Idle Speed Motor Problems: I made it worse by cleaning it?????

[RamblinDMC]

Today, I took out my idle speed motor because I thought it was causing my idle problems. Turns out it was fine and moved freely, but I decided to clean out the built up gunk with carb cleaner. Oddly, this made it seize up and it won't turn freely now. Anything else I can try? Has anyone had luck disassembling the motor. I noticed the casing is crimped together, but it could be pried open.

[Bitsyncmaster]

Don't bother to take it apart. There ar no brushes to clean so only thing is the bearings. Just put some thin oil into the ends. You can oil the tip end if you take that cap off that has a paint "lock" on it but probably not necessary.

[RamblinDMC]

I did try oiling it, but it didn't help. Should I put some solvent in it?

[mluder]

Correct me if I'm wrong, but I believe you are not supposed to clean the idle speed motor with brake or carb cleaner as it removes the lubrication inside and can cause the bearings to seize. The correct cleaner is WD40 - It has enough solvents that it can remove the gunk without taking away the lubricants.

Cheers
Steven

[Bitsyncmaster]

I have used a solvent to clean it and then "3 in 1" oil. I use plenty of oil and just let it all drip out. I only clean and oil the valve side of the motor.

[RamblinDMC]

I'm soaking some pb blaster in it right now to try to free it up. It looks like Mluder is correct because it siezed up when I put carb cleaner in it.

Sent from my LGMS631 using Tapatalk

[Gfrank]

I've had to pop the bottom off and clean the bushings to free em up in the past. Not going to lie, I wrecked the first one I took apart, learned quick

[nkemp]

First, do you have 12Vdc at the center pin of the connector? If not, then that needs to be resolved first.

Second ...Have you made a harness to cycle the motor? It is a three wire connector and by alternating the two outer pins/wires against ground, you can drive it full one way then the other. The center is positive and the two sides are negative. One side drives it open the other side drives it closed. This may break through the gum-up.

Don't ground both at the same time!!!! Don't wire it any other way. Failure can occur.

I have a theory that poor idle (hunting) occurs when the idle motor rotates into the gummy zone and gets a bit hung up (responds a bit slow). The controller senses that the idle is not where it should be and adds more control signal. The valve responds (breaks loose) and then overshoots. This repeats and causes hunting. But it is only a theory.

[RamblinDMC]

Gfrank. How would you recommend taking it apart?


I have tried powering it with a bench power supply. It won't turn. Yes my car itself is outputting the correct voltage.

Sent from my LGMS631 using Tapatalk

[Bitsyncmaster]

First, do you have 12Vdc at the center pin of the connector? If not, then that needs to be resolved first.

Second ...Have you made a harness to cycle the motor? It is a three wire connector and by alternating the two outer pins/wires against ground, you can drive it full one way then the other. The center is positive and the two sides are negative. One side drives it open the other side drives it closed. This may break through the gum-up.

Don't ground both at the same time!!!! Don't wire it any other way. Failure can occur.

I have a theory that poor idle (hunting) occurs when the idle motor rotates into the gummy zone and gets a bit hung up (responds a bit slow). The controller senses that the idle is not where it should be and adds more control signal. The valve responds (breaks loose) and then overshoots. This repeats and causes hunting. But it is only a theory.

Correct for powering the motor.

There are thin copper legs that make contact with the armature. So you have to be really carful if you open it that the armature stays put inside the housing.

Have you tried backing off that top "bearing" screw on the connector end, that sets the end play. Maybe yours is just to tight.

[nkemp]

When you are soaking it with PB, did you plug it and prop it so that the PB can pool around the sticky parts? And let it sit for a few days? Since the carb cleaner seemed to cause it to get stuck, maybe more carb (or brake or manifold) cleaner would free it up. Plug the end and let it soak like above. It may be preferred to try that before disassembly.

Once it breaks loose, then soak it more in PB, a light oil or similar (not WD40). Exercise the motor on the bench to help break loose the crud and give it full range again.

While you are waiting for the solvent to break loose the crud, here is some reading on the subject:
- Read post #12 here: http://www.peachparts.com/shopforum/sl-forum/243090-idle-speed-air-control-slide-valve-confidence-check.html
- Step 10 in this is pertinent: http://www.hiperformancestore.com/Motronic.htm

But as always, YMMV.

I wouldn't jump to the conclusion that the cleaner caused the problem other than it only partially cleaned the valve.

Or ... you may be better off at the local scrap yards. Bosch made a number of designs that will work and it has been argued that the later models are better. If you get a junkyard one, you should clean it before usee (back into the frying pan? :-)

Here is the interchange info that I have (note: the 516 will not work)

Type of idle air motor: Cars (Volvo's and maybe others)on which it can be found:
Bosch 500 ....240 ’81-’82 and 240 Turbo ’81-’83.
Bosch 501 ....240/740 ’83 through April,’87, also 240/740 Turbo ’84 through April,’87.
Bosch 520 ....240/700 May,’87 through all of 1988.

Grab the connector and wire while you are at it ... just to have and use if you don't already have one (doesn't every D owner have an extra ISM wire connector/harness).

EDIT: That's weird. I pasted a table in above when I first created. But it did not display when posted yet does when I opened the edit window to fix it. So I deleted it and pasted plain text that I created from the table in Word. Hmmmm.

[RamblinDMC]

There is an adjustment screw? Is it under the paint in that little well next to the plug? I should be able to remove the valve end without too much problems because there shouldn't be any motor stator wires there. Right? I may give it a try today.

[Bitsyncmaster]

There is an adjustment screw? Is it under the paint in that little well next to the plug? I should be able to remove the valve end without too much problems because there shouldn't be any motor stator wires there. Right? I may give it a try today.

Yes, under the paint is a "screw".

The wires are in the top where the connector is. But when you remove the bottom, nothing to hold the armature in place but gravity.

[RamblinDMC]

So if I un crimp both the top and bottom. I should be able to slide the stator out? What is involved in removing the rotor?



Sent from my LGMS631 using Tapatalk

[Bitsyncmaster]

So if I un crimp both the top and bottom. I should be able to slide the stator out? What is involved in removing the rotor?



Sent from my LGMS631 using Tapatalk

You would have to pull the "stator" out with the top side still on the end of the shaft. The thin flat copper connections are attached so you don't want to damage them.

[nkemp]

Dave ..Getting motors apart is usually easier than getting them back together. Is that the case with this motor? Will he have to retain the brushes in order to reassemble?

[Gfrank]

So if I un crimp both the top and bottom. I should be able to slide the stator out? What is involved in removing the rotor?



Sent from my LGMS631 using Tapatalk

I peel back the crimps on both ends, then tap/pry the bottom end off, then push the top off from the bottom, stator and all.
I've got one on the bench at home apart I can take a photo of it tonight

[Bitsyncmaster]

Dave ..Getting motors apart is usually easier than getting them back together. Is that the case with this motor? Will he have to retain the brushes in order to reassemble?

There are no brushes in this motor. Since it only rotates about 90 deg, they use a flexible copper to contact the armature.

You just tap the crimps back to assemble. There are O-ring that seal the ends to the case.

The fact there are no brushes lets you use a lot of oil since it can't contaminate contacts.

[DMCVegas]

I have a theory that poor idle (hunting) occurs when the idle motor rotates into the gummy zone and gets a bit hung up (responds a bit slow). The controller senses that the idle is not where it should be and adds more control signal. The valve responds (breaks loose) and then overshoots. This repeats and causes hunting. But it is only a theory.

Not quite, no. A hunting Idle is the result of vacuum leaks.

The first thing to understand is that both the Idle Speed Motor and the LAMBDA systems function by running on a loop. Both have target values that they must shoot for, and under normal operating conditions, they hit their targets very quickly. However, you must understand that they are modular additions that function independently of one another and never communicate. Neither with one another, nor the "dumb" mechanical K-Jetronic system. So they've no idea when there is a problem since they only have a single source of data to analyze.

How The Systems Work
So, the way it works is like this: The ISM is fed the engine RPM, and will increase and decrease speeds by rotating left and right until it hits the predetermined speed as set by the ECU. 750± for example when warm (that number isn't specific because it may still cycle a few RPMs per se, but you never notice it. Even on modern cars this happens).

Now the Air Flow Meter of course measures the amount of air moving past the plate to determine how much fuel to meter out in order to determine as optimal of an Air/Fuel ratio as possible. This side is all mechanical and does a great job. But sometimes it isn't perfect for ALL conditions. That is where the LAMBDA system comes into play. The Oxygen, or O2 Sensor monitors the oxygen content of the exhaust to see if the optimal AFR is being delivered to the engine, and will tweak the mixture accordingly to perfect it. This helps performance a bit, but more importantly reduces emissions. How it does this is a whole other story, but it introduces a variable on the Primary Pressure Regulator. Having said that, the ISM only feeds the engine air from after to meter. So all it does is bypass the throttle plates, and still uses metered air.

The only thing you have to remember is that when there is too little oxygen in the exhaust, LAMBDA leans out the mixture. Too much oxygen, it enriches it. And just like the idle speed, on a proper running system you hardly ever notice the variation because it becomes so tiny.


How these two systems cause Idle Surging
Now that you know how both systems work, here is how they fail: With a vacuum leak, additional air is introduced into the intake AFTER the air meter. The additional oxygen in the exhaust is detected by the O2 Sensor. As a result LAMBDA will enrich the AFR. Feeding additional oxygen & fuel into an engine is what you do when you press on the gas pedal, so it causes the RPMs to jump! The Idle Speed ECU sees the high RPMs and closes the ISM to slow the engine speed down. This reduction of air into the engine now collides with LAMBDA's rich mixture. So once the RPMs go down, suddenly the engine is running too rich. That causes the RPMs to plummet too far below the target RPM speed, so the ISM opens back up. As the RPMs now increase, the additional air causes a lean condition, so LAMBDA adds more fuel which will enrich the AFR. Feeding additional oxygen & fuel into an engine is what you do when you press on the gas pedal...

Repeat forever. *Some* cars will see the idle speed hunt and eventually stabilize, depending upon the leak.


How do you test your Idle Speed Motor?

Ever seen the Karate Kid II? Remember that scene where everyone has the drums as some way of cheering on Daniel-San. That is going to be you.

https://i.makeagif.com/media/11-13-2015/7tqPb2.gif

Remove the ISM and hold it vertically in your hand, and spin it back and forth like those drums. If it goes click-clack-click-clack-click-clack-click-clack-click-clack... then your Idle Speed Motor is most likely just fine. It's when your car CAN'T idle at all, or is stuck at a specific RPM that you have a problem with either it, or continuity on the wires.


Why don't modern cars do this?
Because everything is integrated together, and they have a TON of sensors. If the O2 Sensor detects higher oxygen than the MAF sees (minus what is consumed during combustion), or a cylinder bank is out of control when the other is not, then it partially disables emissions and just runs on default signals for a "limp home" mode to avoid this same idle hunting problem.

[David T]

You can never eliminate the hunting. As you point out there are multiple systems running in control loops and they are not coordinated or synchronized. To make things even worse, they attempt to control things that have already occurred so they overshoot and delay. The only thing that keeps things working as well as they do is the inertia of the system, since it is mechanical it doesn't change quickly. Therefore an "out of bounds" condition won't happen too fast for the control systems to recognize and react to it. To minimize hunting you MUST try to balance each of the 6 cylinders so they all fire as evenly as possible to reduce the amount the control systems must adjust for. That includes any vacuum leaks since they can affect the cylinder nearest to the leak more than others, affecting the balance. Keep in mind the K-Jet system is basically a mechanical fuel injection system with an electronic control system "grafted" onto it to make fine adjustments. For it's time it was way ahead of the other major system being used, carburetors. While today it looks antiquated and overly complex, it was State-of-the-Art back in the day. Balancing the cylinders starts with a compression test, a leak-down test, adjusting the valves, replacing the spark plugs, ignition wires, anything that can make the cylinders uneven.

[Bitsyncmaster]

I found three OEM idle motors tend to have dead spots where the valve seem sticky. My idle ECU has a test mode that slowly runs the idle motor back and forth. Just watching it do that test shows how sticky they are. Now run that test with the smaller Volvo idle motor and it looks smooth throughout the range. The dead spots are close to half open so it would probably only cause hunting on a very cold engine.

Usually if you run the lambda open loop, that will stop idle hunting. That just proves the hunting is caused by the two closed loop systems fighting each other.

Another common hunting fix is to install a new O2 sensor. Maybe the dirty sensor slows down the lambda loop.

Another problem is our O2 sensor is not heated. If the idle speed drops to low, the sensor output drops because it's to cool.

[opethmike]

I ran K-Jet for 5 years on my car, and never once had any hunting at startup, hot idle, etc. I've also driven and seen numerous DMCs without any idle hunting. It is entirely achievable.

[DMCVegas]

You can never eliminate the hunting.

It is true that you will not eliminate the hunting entirely, no. Because ALL vehicles, even brand new ones will still have a bit of hunt to them at idle. That is why the manuals always list the idle speeds with the"±" symbol at the end to indicate the over/under on them.

But if your DeLorean is going vroooom!-stall-vroooom!-stall-vroooom!-stall-vroooom!-stall-vroooom!-stall with a massive idle hunt of a hundred RPMs or more, you have a problem that you can, and should repair. It's one thing to have a small fluctuation of RPMs, but if you have a hunting idle caused by sudden RPM accelerations with violent drops from stalling, that is indicative of a vacuum leak.

The other thing I have found as well is to extend the striker screw on the Idle Speed Micro Switch, and to then bend the bracket up so that it engages the Idle Speed Circuit as quickly as possible. Bend it up to where once the striker is fully extended, the switch is then fully depressed. That helps stabilize idle.

From what I have read online, the Microswitch engages the ISM circuit when closed. When open, it is supposed to crack the ISM full open as well to feed the engine. I found on my own car, even when working properly, the car was dropping down to 600 or so RPMs. That would cause a mini-hunt as the ISM overcompensated by going WOT and overshooting 750±. Sometimes the engine would even die. But it would eventually settle. The Micro Switch wouldn't close until the throttle was in the fully closed position. Once I bent the bracket up and adjusted the micro switch, the engine behavior was different. When closing the butterflies, the idle then gently decelerated and quickly settled in proper like a more modern car. Then when taking off, I could get a little extra "pop" out of the engine for quick surge to help prevent lugging when accelerating from a dead stop. Now in theory, this shouldn't be happening. I should have had the results completely backwards. But I don't. Only thing I can figure is that there is a small delay from the time the ECU is triggered to "wake-up" to begin it's loop function. Possibly from it resetting itself from Wide Open to Fully Closed, and then it becomes reactionary to RPMs.

Granted now, there are some other factors at play here:

Volvo 760 GLE ISM (Shouldn't matter though).
Non-original Idle Speed ECU (supposed to be DMC, might be 760 GLE).
Throttle Linkage tightened to eliminate factory slack.

[nkemp]

I agree with what was said. It may be the solution to many cars. And I am adding this theory ...

Idle speed, in and by itself, is a closed loop system. And yes, it is affected by other control systems in the car (not an optimal design). The ISM controller applies correction in response to the input(RPM). The ISM control circuit has to respond to the operational characteristics of the ISM. If an ISM develops "stiction" then it will not respond as the control circuit is expecting. The controller may apply more control signal possibly overshooting when the ISM responds.

Has anyone tried disconnecting the FV while hunting? If it is a battle of the "minds" then taking the FV out of the loop should make a big impact.

If there are air leaks, the ISM control circuit should automatically compensate assuming the air leak is fairly constant (The basic design, using the bypass screws, was a deliberate leak).

Why develop stiction at the extremes of valve operation? The valve likely operates in a narrow band for the most part. As such, gunk gets pushed out only so far in both directions. It doesn't get cleared of the system. The stiction is not uniform across the full arc. Or maybe the opposite is the case where only part of the valve is exposed to the air stream and it is that part that develops stiction (not the extremes of the arc)

As I said, this is a theory and as such may be right or wrong.

[DMCVegas]

Idle speed, in and by itself, is a closed loop system. And yes, it is affected by other control systems in the car (not an optimal design). The ISM controller applies correction in response to the input(RPM). If an ISM develops "stiction" then it will not respond as the control circuit is expecting. The controller may apply more control signal possibly overshooting when the ISM responds.

I believe that is a negative. The ECU would not up the voltage or amperage beyond it's normal value in order to "overcome" the valve sticking. The unit just has a 12V feed. Think of it as a pinball on a track between two magnets. You turn on the magnet to the left, the ball rolls left. Turn on the one to the right, it turns right. Turn them both on, and it locks in place.



Has anyone tried disconnecting the FV while hunting? If it is a battle of the "minds" then taking the FV out of the loop should make a big impact.

No need for all of that. As Dave pointed out, running LAMBDA in open loop will cause the unit to default to a static setting and will then eliminate the idle hunt. HOWEVER, you still have a lean condition present. The size of the leak can fluctuate causing other mixture problems. Worst of all is that you're now running your engine lean. Best case scenario is poor performance and you WILL fail emissions testing for high NOx. Worst case is you risk damaging the engine. Be it from those high temperatures, or you start sucking in sand, leaves, and other debris.



If there are air leaks, the ISM control circuit should automatically compensate assuming the air leak is fairly constant (The basic design, using the bypass screws, was a deliberate leak).

Nope! There is a difference between a leak and a "bypass". I pointed this out earlier, but I'll reiterate it again:
Anytime you have an idle speed circuit on a gasoline engine, it is metered air. It is not a leak. It might be idle speed screws, it might be an Auxillary Air Valve, an Idle Speed Motor, or any other variable valve that EFI uses like an IAC pintle valve. In all of these cases, the air is metered. Be it by the venturi on a carburetor, the Mass Air Flow sensor on EFI, or the Metering Plate on K-Jetronic; ALL of these items meter the air delivered to the Idle Speed Circuits. This ensures that the proper amount of fuel is delivered to match the air being consumed.

Now these idle speed systems will bypass the throttle plate(s) that go into the engine. That way they take over the responsibility of engine speed at idle. But they do not "leak" unfiltered air from the outside, no.



Why develop stiction at the extremes of valve operation? The valve likely operates in a narrow band for the most part. As such, gunk gets pushed out only so far in both directions. It doesn't get cleared of the system. The stiction is not uniform across the full arc. Or maybe the opposite is the case where only part of the valve is exposed to the air stream and it is that part that develops stiction (not the extremes of the arc)

Remember how I spoke about leaks letting dirt and debris into the system? The ISM shouldn't have any gunk inside of it. If it does, you can just spray the stuff out to clean it. The valve does operate in a narrow field of movement. As pointed out it only rotates 90°. It's a wedge looking piece that takes up a quarter of the circular chamber inside, with a metal rod that goes into a niche below that limits it's movement. That is what the indentation on the bottom is.



As I said, this is a theory and as such may be right or wrong.

Hey, you're learning. We've all been there. Hell, the first time I had a problem with my ISM, I wound up replacing the ECU and tearing my motor open to try and fix it. Turned out it was just a bad wire that had to be replaced. I certainly don't know everything, but I can tell you one thing that will aid with learning: Never twist facts to suit a theory. Always twist your theory to suit your facts.

[David T]

An old, tired O2 sensor will have a diminished output and it will change voltage slower so the Lambda system will respond even slower making any hunting exaggerated. Hunting will be most noticeable when cold, the mixture is less well tuned because the Lambda system is in "Open Loop" waiting for the O2 sensor and the motor to warm up. Combustion is not optimal, all of the clearances are not right because the motor isn't at operating temperature and the combustion chambers are still cold. A heated O2 sensor just shortens the warm-up cycle reducing emissions. On a Delorean the Lambda system is waiting for the whole motor to warm up before it will listen to the O2 sensor. Hunting cannot be eliminated but Yes, it can be minimized to the point where you can just hear it because the frequency valve is modulating. That is what you hear mostly, the buzzing from the frequency valve. If you can see the RPM's changing enough to show up on the Tach, you have a problem.

[nkemp]

I believe that is a negative. The ECU would not up the voltage or amperage beyond it's normal value in order to "overcome" the valve sticking. The unit just has a 12V feed. Think of it as a pinball on a track between two magnets. You turn on the magnet to the left, the ball rolls left. Turn on the one to the right, it turns right. Turn them both on, and it locks in place.

Dave, you can probably give the best reply here as to how the ISM is controlled.

My suspicion is that the controller modulates the signal (modulates the 12Vdc) to the motor. It either is a fixed width pulse and varies frequency or it uses Pulse Width Modulation. In both cases the controller can hold the valve open x% if it meets the idle thresholds. It moves the % open to bring the idle into compliance.

Note please that I'm not arguing that vacuum leaks, bad O2 and various other problems cause hunting. My point is not that every hunting engine has a sticky ISM. And hunting is a problem worth fixing. In those cases we are in "violent agreement" :smile::smile: ( I love that phrase:smile:) My point is that a sticky sluggish valve can impact idle.

[Elvis]

To all these facts and also some misinformation I just add (again) some facts:

When I bought 6548 in 2001 - it was not hunting - until I cleand and lubed
the completely frozen ISM. Then suddenly it started to "breathe" when cold.
Setting everything up solved/improved it. Eventually I got a "bad" idle ECU
and took it apart and found the basic PID circuits inside...

the microswitch does not turn anything on or off !
it just changes the set speed from 775rpm to something higher !
(don't we have an FAQ for it ?)

I am happy to see that others repeat me - how to test the ISM electrically :-)
I remember how I got flamed for suggesting this some years ago :-P


PID ! PID Controller ! the ECU is a PID control circuit.
Any electronics student should know PID control loops.
I found it - read here:

https://en.wikipedia.org/wiki/PID_controller

(don't get confused by the mathematics, just read the text)


Slow moving or sticky idle motor - yes ! it will probably be hunting.

rich mixture - engine reacts slower to changes of the ISM - YESSS - hunting !
-> try to control a closed loop system with a dead time - not an
"invention" of me, just read the basics about control loops and find the facts.

At the SDSU they explained us how to set up/calculate a stable control loop.

by changing the PID parameters the hunting can be reduced. (not completely
eliminated)
Well, also a suggestion I got flamed for...I still have my test ECU with the
potentiometers installed...if I remember correctly - reducing "P" worked pretty good.

[nkemp]

the ECU is a PID control circuit.

Good sluething. Good to know. I often wondered (having nothing else to think about:) ) if it was PID or simply PI control. Regardless, good actuators and inputs are need for good control.

Do you have a link to your posts with more info?

[Bitsyncmaster]

Dave, you can probably give the best reply here as to how the ISM is controlled.

My suspicion is that the controller modulates the signal (modulates the 12Vdc) to the motor. It either is a fixed width pulse and varies frequency or it uses Pulse Width Modulation. In both cases the controller can hold the valve open x% if it meets the idle thresholds. It moves the % open to bring the idle into compliance.

Note please that I'm not arguing that vacuum leaks, bad O2 and various other problems cause hunting. My point is not that every hunting engine has a sticky ISM. And hunting is a problem worth fixing. In those cases we are in "violent agreement" :smile::smile: ( I love that phrase:smile:) My point is that a sticky sluggish valve can impact idle.

The stock ECU drives one side of the idle motor to ground then off and drives the other side to ground than off. It does that a 100 times a second which is why you hear and feel a buzz. So the valve is actually swinging back and forth but the mass of the armature limits how much it swings. If you look closely you see the valve is moving very quickly. The duty cycle of one side is changed. One side gets a shorter pulse and the other side gets a longer pulse. That is how the valve makes it's movement.

[Elvis]

...and that's why the motor sees different voltages in average (positive and negative)
and therefore gets more power to overcome a sticky position - after a time delay
which then results in a dead time -> hunting !

[DMCVegas]

The stock ECU drives one side of the idle motor to ground then off and drives the other side to ground than off. It does that a 100 times a second which is why you hear and feel a buzz. So the valve is actually swinging back and forth but the mass of the armature limits how much it swings. If you look closely you see the valve is moving very quickly. The duty cycle of one side is changed. One side gets a shorter pulse and the other side gets a longer pulse. That is how the valve makes it's movement.

Exactly! That's why I use the pinball & magnet analogy to start building that understanding. In relation to the air flow direction, there are the 3 contacts on top of the motor. Center is 12V feed, left rotates open, right rotates closed. The ECU as pointed out controls the ground to each to close the circuits to force the valve to rotate in that particular direction.

The function of the duty cycles is exactly as Dave describes it.



...and that's why the motor sees different voltages in average (positive and negative)
and therefore gets more power to overcome a sticky position - after a time delay
which then results in a dead time -> hunting !

I would have to respectfully disagree on that. The reason that the ECU runs constant pulses is to avoid wild swings of the internal armature. If it was physically sticking, we wouldn't see wild swings from over compensation. We would instead see it just get stuck on a particular RPM speed for any extended period of time. The rapid back and forth polarity that occurs inside of the motor also acts as a brake of sorts to help stop this very thing from happening.

But just like you said, it's a PID control circuit. So it has to compensate for the variables. The variable here causing wild swings in RPM hunting isn't caused by the intertia of the valve. When the engine is revving up and down, that is something different. That is, in fact a vacuum leak. The leak causes problems with the oxygen content in the exhaust, the LAMBDA steps in to correct it. Additional fuel causes the RPMs to shoot up along with the additional oxygen because it's feeding the engine. That then becomes the variable that the ECU is trying to compensate against.

In doing so, the ECU then creates a variable that LAMBDA has to fight against because of the additional un-metered air being sucked in that further throws the AFR off. But that is why once you disconnect the O2 sensor, the problem goes away. LAMBDA then stops trying to correct the mixture and simply sends a default duty cycle to the Frequency Valve. With that variable removed, the ECU is then able to hit it's target RPM range. Granted now the fuel mixture is screwed up and can cause other engine and emission problems, but the idle is stabilized.

Now there may definitely be a delay with the ECU activating from the time the Micro Switch is tripped. Because as soon as I forced mine to activate sooner, my RPMs would float down at land smoothly at their target speed. Waiting until the last minute always caused too great of a dip in speeds, and THAT would cause a mini-cycle of idle hunting. But activating sooner that eliminated it.

I can also assure you that it's not a variable in voltage. I had a knackered wire on the open side that wasn't passing full ground to the ECU. The close side never compensated by lowering it's voltage to compensate, therefore the valve would never open up. Once I grafted in a new wire, the valve fired right up and operated as normal.

[nkemp]

What Dave describes is PWM, Pulse Width Modulation. That is where the time duration, per a very short period, of the 12Vdc signal is varied depending on the amount of drive needed.


The reason that the ECU runs constant pulses is to avoid wild swings of the internal armature.

The only reason the ECU provides a constant pulse in terms of width and time is that the input is where the controller wants it to be. Once the input varies, it will change the signal to compensate.


If it was physically sticking, we wouldn't see wild swings from over compensation. We would instead see it just get stuck on a particular RPM speed for any extended period of time. The rapid back and forth polarity that occurs inside of the motor also acts as a brake of sorts to help stop this very thing from happening.

Try this metaphor... You want to change the throttle position because the load had changed (maybe the road is sloping up). You push on the pedal and you do not get the response you are looking for (maybe there is something physically blocking the foot pedal). So you increase the pressure (In PID terms you increase the D factor). But low and behold, the thing blocking your pedal breaks loose and now the car lurches forward. You you modify pressure again by backing off, way off because the think blocking is in play again (More D in the opposite direction) and you slow down too much once it breaks loose again. This continues because PID control does not do well with a variable and undefined output response. It assumes the controller will respond in line with the signal. If the thing blocking continues to block, then the oscillations can continue.


I can also assure you that it's not a variable in voltage.

Yes and no. A pulsed 12Vdc signal (pulse width or frequency) can be measured by a meter (digital or analog) as a variable voltage. The longer the pulses or the more frequent the pulses the more the "measurement bucket" fills towards the 12Vdc signal (or whatever it is). Try it with a speed sensor (VSS) hooked to a drill and monitored by a meter. They produce pulses. In this case, the more pulses, the more the voltage. As I recall, Ford VSS based speedometers (maybe others) measure the VSS by not counting pulses but by measuring the resultant voltage regardless of the fact that you are feeding it with a fixed voltage. So yes, the individual height of a pulse is 12Vdc (This you would see on an oscilloscope). But on a "meter" you may not see the full 12Vdc but rather the integrated value.

Note that that is more of a "hacker's" interpretation. The experts can tweak it.

And I must reiterate, because there is a vacuum leak does not mean the ISM isn't hunting partially because of poor ISM performance. There are 3 separate scenarios:
- There is hunting due to a vacuum leak
- There is hunting due to poor ISM response
- There is hunting affected by vacuum leaks and poor ISM response.

[nkemp]

It may be valuable to compare a 2 wire ISM to a 3 wire ISM.

The 2 wire ISM controller provides a control signal to open a needed amount against a spring. The spring pulls it closed, the signal pulls it open. At a specific signal strength (pulse width as I recall), the electrical pull equals the spring pull and it hovers there until it needs more or less air.

The 3 wire version drives the opening either more open or closed by competing signals (but not concurrent competing signals. I think that burns the valve out). AFAIK, when the throttle is opened, the ISM quits getting signals and it stays where it was last positioned (subject to any mechanical movement/vibrations). It is kinda like a water valve where you can turn it one way to increase and the other to reduce water flow and it stays where you leave it.

And like the throttle metaphor above, I do have a valve on the house that sticks. No matter how much i turn it it does not open then all of a sudden it goes full blast. If I over compensate, I can turn it full off and it sticks again when I open it repeating the process. It is a freeze proof valve that is in the garage... I really should fix it.

[Bitsyncmaster]

I like how a lot of owners are understanding the operation of the DeLorean idle control.:)

Yes I would also call the motor control a PWM signal. But one of the two sides of the motor are always powered (grounded). If one side is powered for 1 ms. then the other side is powered for 9 ms. At 100 Hz rate total time is 10 ms.

The idle motor has the best power (prevent sticking) when at 50% (each side is 5 ms powered). On a warmed up engine the idle motor is running almost closed which has very little power to overcome a sticking valve. Hence most idle hunting problems due to a dirty idle motor are with the warmed up engine.

[nkemp]

Dave, do new clean ISM's behave the same as older sticky ISM's in terms of the dead spot and current draw?

[Bitsyncmaster]

Dave, do new clean ISM's behave the same as older sticky ISM's in terms of the dead spot and current draw?

I have never tested a new ISM. I've only had three OEM used ISMs that I tested (one I cleaned) and they all had those "dead" spots. I bought a used Volvo ISM (got it for $25) on ebay. It looked new but it was frozen. The cleaning with solvent and oil fixed it. The OEM units draw 800 ma. and the Volvo draws 400 ma. The Volvo unit is much smaller so I don't even use a bracket to hold it (just the hose holds it). I think the EU guys sell the Volvo unit new.

I thought I had a demo on youtube but I don't see it anymore. Maybe I can do another video.

[Elvis]

Again - it does NOT shut off ! (only when you turn or pull the key)

go out to your car and touch the ISM while playing with the throttle. Does it stop vibrating ?
NO !

Some more details to hopefully improve the understanding:

50/50 - ISM stays in the current position.
90/10 - open fast
10/90 - close fast
50/50 - stay in that position ! fully open or fully closed or somewhere in between - just stay there !



It may be valuable to compare a 2 wire ISM to a 3 wire ISM.

...

The 3 wire version drives the opening either more open or closed by competing signals (but not concurrent competing signals. I think that burns the valve out). AFAIK, when the throttle is opened, the ISM quits getting signals and it stays where it was last positioned (subject to any mechanical movement/vibrations). It is kinda like a water valve where you can turn it one way to increase and the other to reduce water flow and it stays where you leave it.

...

[Elvis]

Too much to write right now - but I also respectfully disagree with your theories.

Robert - a little question for you:

why do many cars (even without vacuum leaks) hunt badly when cold and suddenly
stop hunting and run smoothly ?
because the vacuum leak disappears ?
or maybe because a temperatur switch changes something in the engine control ?



AND - PWM IS varying the voltage ! I do this for living.

One more question for you to think about:

Let's assume the ISM is sticking, Engine speed is dropping.

The ECU will change the PWM from 50/50 (stay exactly there) to 60/40 up to 90/10.
(P and I are doing this )
where do you think it will break free and what will happen to the engine speed ?


OK, gotta go to work, a motor with PWM control is waiting...



...

I would have to respectfully disagree on that. The reason that the ECU runs constant pulses is to avoid wild swings of the internal armature. If it was physically sticking, we wouldn't see wild swings from over compensation. We would instead see it just get stuck on a particular RPM speed for any extended period of time. The rapid back and forth polarity that occurs inside of the motor also acts as a brake of sorts to help stop this very thing from happening.

But just like you said, it's a PID control circuit. So it has to compensate for the variables. The variable here causing wild swings in RPM hunting isn't caused by the intertia of the valve. When the engine is revving up and down, that is something different. That is, in fact a vacuum leak. The leak causes problems with the oxygen content in the exhaust, the LAMBDA steps in to correct it. Additional fuel causes the RPMs to shoot up along with the additional oxygen because it's feeding the engine. That then becomes the variable that the ECU is trying to compensate against.

In doing so, the ECU then creates a variable that LAMBDA has to fight against because of the additional un-metered air being sucked in that further throws the AFR off. But that is why once you disconnect the O2 sensor, the problem goes away. LAMBDA then stops trying to correct the mixture and simply sends a default duty cycle to the Frequency Valve. With that variable removed, the ECU is then able to hit it's target RPM range. Granted now the fuel mixture is screwed up and can cause other engine and emission problems, but the idle is stabilized.

Now there may definitely be a delay with the ECU activating from the time the Micro Switch is tripped. Because as soon as I forced mine to activate sooner, my RPMs would float down at land smoothly at their target speed. Waiting until the last minute always caused too great of a dip in speeds, and THAT would cause a mini-cycle of idle hunting. But activating sooner that eliminated it.

I can also assure you that it's not a variable in voltage. I had a knackered wire on the open side that wasn't passing full ground to the ECU. The close side never compensated by lowering it's voltage to compensate, therefore the valve would never open up. Once I grafted in a new wire, the valve fired right up and operated as normal.

[nkemp]

Again - it does NOT shut off ! (only when you turn or pull the key)

go out to your car and touch the ISM while playing with the throttle. Does it stop vibrating ?
NO !

Some more details to hopefully improve the understanding:

50/50 - ISM stays in the current position.
90/10 - open fast
10/90 - close fast
50/50 - stay in that position ! fully open or fully closed or somewhere in between - just stay there !

Good input. Thanks for clarifying.

When the throttle is depressed:
- does it just send the 50/50 signal leaving it where it was just before throttle?
or
- does it go full closed in a dim-witted attempt to bring the RPM's back to 750'ish?

I presume it is the former.

[Elvis]

idle switch pressed -> control to 775rpm

engine very cold or idle switch not pressed >- control the speed to something higher than 775rpm !

[nkemp]

This thread has taken a bit of a scenic side road. RamblinDMC, how goes your project that started this thread?

[DMCVegas]

Good input. Thanks for clarifying.

When the throttle is depressed:
- does it just send the 50/50 signal leaving it where it was just before throttle?
or
- does it go full closed in a dim-witted attempt to bring the RPM's back to 750'ish?

I presume it is the former.

The first thing is the Temperature Switch. That will set the target RPM at either 775± when warm (I just verified this number in the Worshop Manual), or something like 1,250± when cold. That determines the target range.

Now then, as I understand it, the function of the Idle Speed Circuit is as follows:

Micro Switch Closed = ECU adjusting Idle Speed Motor/Regulator activates to try and adjust engine RPM speed.
Micro Switch Open = ECU sends a wide-open signal to the motor to feed the engine.


So under those two conditions, you have the ISM behaving in the following way:

Throttle Open = Wide Open ISM.
Throttle Closed w/Engine @ Idle Speed = ISM fluctuating to control Idle RPM Speed.
Throttle Closed w/Engine Decelerating = ISM closed because actual RPM speed is still too high. But once the RPMs drop, it will revert to behavior #2.


And of course the "closed" & "open" values here will be based off of the duty cycle variations.

But that is going to be the really odd thing about the duty cycle in this application. There is a little bit different of a function here than a regular one than say what we have on the Frequency Valve. So check out this diagram:

http://wiki.bildr.org/images/thumb/3/35/Duty_cycle.png/300px-Duty_cycle.png

Take a look and you will see a 10%, 50%, and 90% duty cycle. Complete square waves here because it's a digital signal of course, and the percentages represent how often the signals are applied. So in the case of something with a 100 Hz rate like the FV for instance, a 10% duty cycle means that the valve is only open 10% of the time during a 1 second interval. And a 90% duty cycle means it would only be open for 90% of a single second of that interval, right?

Now, take a look at the ISM. The unit has a constant 12V being applied to it, and the ECU simply closes a path to ground on each side of the motor when it wants it to change direction. So the duty cycle doesn't represent when 12V is being applied. It merely represents when 12V is able to reach ground when the ECU completes a circuit. And what the ECU itself does it is switches back and forth between the two negative sides. So what you have is a completely inverse duty cycle on the opposite side that looks something like this:

https://build.openmodelica.org/Documentation/Modelica%203.2.2/Resources/Images/Electrical/PowerConverters/dutyCycle.png

So if we were to reference our Duty Cycle percentages as a representation of the ISM rotating Clockwise for instance, a 60% duty cycle @ 100 Hz would mean that within a 1 second window, the motor spent 60% of it's time rotating clockwise, and another 40% rotating counterclockwise. Overall that would represent a 20% deviation from the mean with the ISM valve now rotated 20% farther away in the clockwise direction from where it started at the beginning of our sample window in relationship to the distance that it could have traveled. Now, what actual distance that represents to show a percentage between fully closed and fully open, that I have no idea of. I haven't got the slightest clue on what the ratio of duty cycle to valve position/rotational speed is. But that's OK because it doesn't matter. The ECU can, and will continue to modify this duty cycle as needed to reach the target RPM range.




Too much to write right now - but I also respectfully disagree with your theories.

Not a problem at all. This is how debates are supposed to work. We disagree, and eventually come to a common understanding.



Robert - a little question for you:

why do many cars (even without vacuum leaks) hunt badly when cold and suddenly
stop hunting and run smoothly ?
because the vacuum leak disappears ?
or maybe because a temperatur switch changes something in the engine control ?

It is hard to say since I don't know all of the control conditions involved. 59°F/15°C is supposed to be the magic number the coolant reaches for when all of the systems in the DeLorean are then ready to operate in a proper, warmed up mode. With 3 separate, independently functioning systems, 2 of which are competing to modify the the fuel mixture, who knows. We have the Idle Speed Circuit, LAMBDA, and the Warm-Up Regulator. Then we have potential variables in the probes that feed back information to those systems, such as the Temperature Switches, the Oxygen Sensor, and even heaters inside of the Warm Up Regulator. Which is then another problem since LAMBDA is digital with a static duty cycle when cold, but the WUR is analog as the bi-metal arm moves to slowly raise control pressure for the fuel injectors. That would defeintely cause a rough idle speed cycling. The WUR thinks the engine is cold, so it runs rich. LAMBDA thinks the engine is warm, so it attempts to correct. Instead of a leak with unmetered air, it would then be a system conflict injecting unaccounted fuel. Or even too lean an amount if the WUR deactivated ahead of time, and LAMBDA things the mixture is supposed to be richer, and then compensates for the fluctuating engine speed based upon the changing AFR.

There is a good chance that the sole reason that the problems finally go away once the engine is warmed up is simply because the systems are then able to overcome their variances and can then operate as intended.

Now there might just be another factor here at play with thermal expansion, and I will get to that in a minute.



AND - PWM IS varying the voltage ! I do this for living.

One more question for you to think about:

Let's assume the ISM is sticking, Engine speed is dropping.

The ECU will change the PWM from 50/50 (stay exactly there) to 60/40 up to 90/10.
(P and I are doing this )
where do you think it will break free and what will happen to the engine speed ?


OK, gotta go to work, a motor with PWM control is waiting...

I'm not denying at all that this is a PWM system. But here is my idea:

This is part of the consideration of thermal expansion. Perhaps a cold valve due to age and/or a lack of lubrication or whatever, might be more difficult to move. But once it warms up, then the expanding metal case and/or valve housing allows it to more more freely. (themal expansion could also account for poor connections on other temperature sensors and probes too, although we'd probably see more problems in those cases).

Honestly, I believe you are correct, but respectfully, that you are correct for the wrong reason. Let me explain:

I don't think this is a problem with overshooting. What I do think though is that there is a delay on a sticking valve that causes a greater fluctuation in that particular direction the RPMs are heading.

The idea over-correction being presented, as I understand it, would mean that let us say the engine is holding at 720 RPMs for example. As the ISM opens, it sticks. So the duty cycle increases it percentages until it overcomes the blockage, and then opens too far, thus increasing the RPMs. Is that what I am understanding?

My theory on a sticky ISM by chance would instead mean that the motor sticks and simply moves slower. So let us look at this same scenario. The problem isn't that the engine is holding @ 720 RPMs, it would be that the engine speed is continuing to fall. Before the ISM finally opens up, then engine begins to stall. So what in reality we are looking at isn't a an over-correction, it is instead a delayed correction. The delay causes the extreme deviation, and simply becomes mistaken for being an over-correction.

What I am willing to bet is that integrated into the Idle Speed ECU, the Duty Cycles are dictated by the deviation from the mean target speed.
For instance (and I am completely making up these numbers only to set a theoretical example), let us say that that within deviation of 25± RPMs, we're probably looking at a 50% Duty Cycle.
For 26-50 RPMs too high, it would be a 60% Duty Cycle.
For 26-50 RPMs too low, it would be a 40% Duty Cycle.
51-150 RPMs too high= 75% Duty Cycle.
51-150 RPMs too low= 25% Duty Cycle.
151+ RPMs too high = 90% Duty Cycle.
151+ RPMs too low = 10% Duty Cycle.

On top of all of this, there seems to be another odd behavior of the ECU. It seems like as soon as the Micro Switch is closed, the ISM snaps shut, and then has a full second delay where it wants to sample the RPMs and the calculate the Duty Cycle to send out. The reason I hypothesis this is because in the factory default position, I would ALWAYS get the engine stalling and dipping too low into the upper 600's. But once I readjusted the Micro Switch Position to hit sooner, that problem went away with no other modifications or adjustments to the engine. Then it would never go below 740 or so.

And here is an important thing to consider: The Duty Cycle does NOT represent the position of the Valve in relation to fully open or fully closed. It merely represents how far it will move from its current position. So a 50% duty cycle might keep the valve from moving, BUT the valve may still be in too far open or closed of a position, thus it hits another minuscule duty cycle adjustment again to set it right back on track when the RPMs eventually wander too far away from target.

That is why I don't see the example of an engine throttle being pressed to overcome a hill as vaild. A throttle is an analog example. To apply such a theory to the operation would mean an example of where you can only press the throttle to go WOT, or you lock up the brakes. So you would slam the gas pedal down to reach a target speed of 55 MPH. Then as soon as you hit 56 MPH, you would slam the brake pedal. Then once you hit 54 MPH, you'd slam the gas again to the floor and do this indefinitely as a means of speed control. Which if the idea of which seems almost violent, now you know why your ISM buzzes!

Which is to also say, if the ECU could actually accomplish such a task to increase torque to overcoming sticking, it would need to fluctuate the central output voltage constantly, which instead of a simple duty cycle would become a Modified Sine Wave. Which given that it would then have to have an opposite fluctuation of voltage akin to the switching duty cycle, I am highly skeptical of. But that then asks the question: Has anyone witnessed a modified sine wave on the output, or it is a true digital duty cycle?

[nkemp]

So the throttle example didn't work for you. Here is another. Have you ever put a ratchet on a nut, applied a bit of force and it didn't move, then applied more and more, until it breaks loose and you bang your knuckles on the first part in your hands way? That is overshoot. That can happen in a control system.

Or imagine a control system where the controller is sensing something and controlling to that value. It signals the actuator to open or close to control it. Let's imagine that the actuator is a valve and it is disconnected. The controller does not know it is disconnected. It only knows of the value it senses and controls to. When it does not see the response expected, it says open more and more until it is full open output mode. In the case of a PWM signal, the full output is 100% on. In the Idle ECU it might be as much as 100/0 or maybe it is less such as 99/1 or 98/2.... whatever Bosch designed it to.

Now imagine plugging in the actuator. What happens... it goes full open and overshoots.

We are proposing that a sticky valve can cause overshooting and loss of control possibly including hunting.

[Elvis]

what a long post, I'll need some time to go thru it.

Hope we can make a FAQ out of it afterwards.

But for now - my Wurlitzer is waiting for TLC...:-)

[David T]

what a long post, I'll need some time to go thru it.

Hope we can make a FAQ out of it afterwards.

But for now - my Wurlitzer is waiting for TLC...:-)

Overshoot is only part of the problem. All of these observations assume each cylinder fires equally. Often they do not and if one cylinder has just a little less or more power the motor will speed up or slow down. The idle ECU senses the change and sends a signal to the idle motor to change it's position to compensate. By then another cylinder is firing and may cause the control system to compensate the other way. Point is the control system is always "behind the curve" and is adjusting for what already happened, not what is going to happen or is happening now. To complicate things further other control systems are adjusting and compensating so they superimpose their cycles onto the motor too. Since hunting cannot be eliminated the goal is to reduce it to as minimal a level as possible. That starts with balancing the mechanical system that the control systems are controlling so there is a smaller band they actually have to control. The wider the control band the harder it will be for the control systems and they will be less powerful. Another factor at play (at least for the idle control system) is vacuum leaks. If they are occurring after the air sensor plate they reduce the power of the idle motor to actually control the motor. They can also have the insidious affect of unbalancing the motor too if the leak occurs close to one cylinder affecting it more than all of the others. Hunting can be induced in many ways. On one car the impulse coil was not evenly adjusted and the gaps were not all the same. It caused the ignition system to be unbalanced and therefore all of the cylinders were not firing evenly. Bottom line, to reduce hunting each and every cylinder should be as close to all of the others so the motor can run smoothly.

[DMCVegas]

So the throttle example didn't work for you. Here is another. Have you ever put a ratchet on a nut, applied a bit of force and it didn't move, then applied more and more, until it breaks loose and you bang your knuckles on the first part in your hands way? That is overshoot. That can happen in a control system.

Or imagine a control system where the controller is sensing something and controlling to that value. It signals the actuator to open or close to control it. Let's imagine that the actuator is a valve and it is disconnected. The controller does not know it is disconnected. It only knows of the value it senses and controls to. When it does not see the response expected, it says open more and more until it is full open output mode. In the case of a PWM signal, the full output is 100% on. In the Idle ECU it might be as much as 100/0 or maybe it is less such as 99/1 or 98/2.... whatever Bosch designed it to.

Now imagine plugging in the actuator. What happens... it goes full open and overshoots.

We are proposing that a sticky valve can cause overshooting and loss of control possibly including hunting.

No, I totally get what you're saying. But this is the reason why I don't believe the comparisons items apply here. The first reason is that with the example of the throttle or the ratchet with the stuck bolt, you are increasing the amount of applied torque to remedy a non-movement situation. With the Idle Speed Motor, there is no increase in torque because it is not capable of such a feat. If it could not overcome an obstruction at all, it simply will not move. A variation in duty cycle wouldn't have any affect upon the torque. The other problem is that with both scenarios, we're only applying force in a single direction. With the Idle Speed Motor, we have a constantly changing velocity as the internal motor rotates back and forth between two directions.

More to the point though is that it isn't overshooting like you say. It's just not being able to correct itself in time, which gives a dip in the opposite direction.

Let me use the pinball analogy again here. So imagine you have a pinball, and it sits on a track between two electromagnets on each opposite end. The track itself is a full 1000 cm long (I'm going to go metric with large values here to use round numbers instead of imperial fractions to keep things simple). If we were to apply a full magnet on on side for 1 full second, That would be a 100% duty cycle. And during that time, the ball would roll 10 cm away from it's starting point 1 second prior.

So 100% DC = a 100:0 ratio. That results in 10 cm of movement.

Now we apply a 50% duty cycle. The ball might rotate a single cm back and forth, but at the end of a 1 second time frame, the ball will remain back in the same exact location it was in.

Here though is where things get interesting: While there are two separate Duty Cycles here for the left and right magnets they both are fed from the same individual power source, which is always live. They do not activate, however, until their ground circuits are complete. And control of that goes through a single switch which can only physically activate a single ground at a time, and must always have at least one of those grounds activated. So, that means that if we have a 60% duty cycle to activate the right magnet, the other 40% of the time that magnet is disconnected, the left magnet will then instead be activated on the opposite side. Does that make sense?

So keeping our duty cycle% number as the representation of the right magnet that moves the ball right, subtract that number from 100 to see what the representation of the opposite left magnet would be to see how far left it would have moved. So, a 60% duty cycle would indicate that the ball would have moved over 60 cm to the right, but back left 40 cm. 60-40=20. Thus in the end the metal ball moved 20 cm to the right.

But lets say we have a 40% duty cycle. 40-60=-20. So a negative number of course means that we now move left 20 cm from the original spot within the time frame of a single second. It's just like using a number line in algebra.
http://image.tutorvista.com/cms/images/38/positive-and-negative-number-line.jpg
Now, here is where things get REALLY interesting. Take the 100 cm track which the ball sits on, and at absolute zero in the center of the track you're going to place the track onto a fulcrum so that it now tilts like a seesaw. The tilting of the track has now become representative of the variable RPMs of the engine. When perfectly balanced, that represents the engine running at target RPM, like 775±.

The angle of the track represents engine speed. When tilted fully to the left, the ball rolls off the track and the engine has died. That of course requires a reset, or restart. When tilted fully to the right, the ball can roll until it hits a stopper that represents the max 2,000 RPMs the ISM valve can achieve.

The farther the track tilts to create a steep incline, the greater the duty cycle variable to move the ball faster in order to quickly balance the track again. A small tilt may only trigger a 60% or 40% duty cycle. But a more severe tilt will cause an 80% or 20% or greater one. It all just depends upon how fast it wants to move the ball based upon how severely it wants to correct things. But don't forget, those magnets are powerful, and they move that ball at lightning speed up and down the track.

The problem with a sticking ISM would be like having glue or gunk on the track. It's not that the magnets are going to get any more powerful to overcome the obstruction, it's just that the pinball then moves much slower. That means that it can't counter the continuing tilt of the track as fast. Which means either rapidly tilting in one direction or the other. I.e. rapidly falling or rising idle RPMs. That delay is what you are seeing and thinking is overshooting. That's the point I'm trying to make. There might be a bit of additional correction needed on the opposite end, but not that much. However, once that pinball rolls back into the sticky area, the cycle would repeat itself AGAIN, and the engine RPMs never get the chance to settle down properly.

Now imagine that the track is clean and working fine, but then this blind man comes up and start smacking the ends of the track up & down and doesn't tell you why. All you know is some spoiled brat of a kid keeps farting and some maid keeps spraying air freshener. The blind guy he responds by futzing with your track each time he smells something different! As a result the system then has a hard time balancing the track because of this annoying interference. The blind guy is your Air/Fuel Ratio, the maid is the LAMBDA, and the kid is a vacuum leak.

Certainly the idle speed will never perfectly balance out. Not when even at a 50% duty cycle, there is still a vibration of the ball. Let alone when it's on an opposite side of the fulcrum, and the track slowly tilts, which will eventually require a variation in the duty cycle to bring about a new balancing act. And it still even happens on new cars. You'll still see your tach jump a tiny bit, and even when you plug a computer into the car, you'll see the RPMs bounce ever so slightly. And that's OK. We just want to eliminate the terrible hunting problems the cause noticeable engine revving.

Sometimes it helps to visualize things, and in a case like this, I can now definitely say that I too am a now a firm believer after this thread that a sticky, or gunked up ISM can certainly cause a nasty surging idle problem.

[nkemp]

I can now definitely say that I too am a now a firm believer after this thread that a sticky, or gunked up ISM can certainly cause a nasty surging idle problem.

The phrase "violent agreement" comes to mind again.:) Or should I say "enthusiastic agreement"?



The first reason is that with the example of the throttle or the ratchet with the stuck bolt, you are increasing the amount of applied torque to remedy a non-movement situation. With the Idle Speed Motor, there is no increase in torque because it is not capable of such a feat.

Correct. PWM does not increase torque but it can change power. The longer the PWM "on" periods, the more power (50% PWM is more power than 20%) assuming some work occurs.

For those who are reading this thread looking for more background, and likely not the contributors, here is some background:
- Torque is not power (torque is not horsepower)
- Torque is expressed in terms of pressure.
- Torque is part of the horsepower equation
- Power refers to work done over time
- When torque is applied to something over time and causes some action then work is done
- Watts is a measure of power
- Watts = Volts * Amps
- If Volts is pressure then think of Amps as the volume (number of electrons in electricity (gallons in a water metaphor))
- In our case torque refers to the voltage (12Vdc pressure) which is colloquially referred to as "12Vdc" in the car
- PWM does not affect volts delivered against the "job to be done". You can see this on an oscilloscope where the PWM peaks (the height of the square-waves (pulses)) always shows voltage as 12Vdc.
- PWM affects the number of amps, number of electrons, delivered against the job
- 20% duty cycle allows 20% of the electrons to flow. 50% allows 50% of the electrons to flow... or there about

And therein is why a smart controller, including a PID controller, will increase the PWM in order to cause a change. When a smart controller does not see the change it needs in the time it needs, it will attempt to cause the actuator to try harder. Using the ball on the sticky track, and adding a smart controller and sensors, when the controller does not see via the sensors the action desired in the time desired, it will increase the effort (PWM), and this can cause overshoot if the control apparatus (sensors, controller and actuator) cannot control within the required time when the sluggishness is overcome (a non-linear actuator response). In our case, the idle ECU, assuming PID control, sends a signal to affect the idle, it monitors the feedback, if the idle does not move towards the desired fast enough, it increases the PWM on one side and decreases it against the other side until it gets the response needed and then it moves back to 50:50 (side to side). (That is my assumption of how it plays out)

BTW ... something similar (but not 2-sided like the idle control) happens with the Lambda control and FV as it attempts to modulate the fuel distributor to affect fuel flow modulation. In this case the fuel pressure is "volts" and fuel volume is "amps". Assuming some sort of smart control like PID control.

In the case of bloodying your knuckles breaking a nut loose, the actuator is the hand, sensors includes eyes and muscle feedback and the controller is the brain. In this case we modified torque ( we can't apply more arms ... but maybe we engaged more muscle cells... Hmmm) to affect more ability to do work and once the threshold is breached, the brain, sensor and actuators didn't react fast enough to prevent the pain that is soon to happen. A really fast system would apply counter control to keep it inbounds.

SO ... so much of this discussion also really depends on what the Idle ECU does. What kind of control does it perform? PID would be preferred for the most part since it is best able to maintain tight even control. And I have to go on the input of others that it is in fact PID control. This really is worth the read: https://en.wikipedia.org/wiki/PID_controller

BUT ... If it is PID control but the sensor is sluggish (a buildup of material slowing its response to temperature change) or the actuator is sluggish (sticky), then the PID control cannot do its job as well as designed.

For further control process thought ... think of temperature control and your furnace:
- Using only threshold control (really dumb control regardless of the smartness of the electronics controlling it), if the furnace comes on at 68 and off at 70 trying to control to 69, there would be overshoot of some amount over 69 and undershoot of some amount of 68 (and this amount is a function of the external temperature and the furnace capacity) ... Hunting
- The "simple" mechanical T87 round Honeywell thermostat improved control by adding an "anticipator" to avoid the overshoot. (it was a tiny heater that tried to warm the bimetal spring detecting temperature to avoid overshoot ... and it was implemented to work both ways) In our example we might assume it tried to stay between 68 & 70,
- Really smart thermostats (not all smart t'stats are really smart) look at the rate of temperature change (when the equipment is off as well as when it is on) and determine when to start, how long to run and when to stop based on the rate of change to flatten the swings (minimizing hunting)

You might say all three perform PWM on the furnace. But it is not as simple as pulsing a bunch to flatten the temperature change (because too many pulses reduce furnace efficiency with too much heat going out the exhaust compared to transferred to the space). Or as simple as one long pulse that is the most energy efficient but has the greatest over and undershooting (hunting in our example)

But I digress.

The point I'm making is that the control algorithm is key to the hunting problem as are the sensors, actuator performance and all the interdependencies of other control systems. And of course leaks. It is the "system view".

And this probably is more "violent agreement". :) But I tend to think of it as enthusiastic. :)

Technical experts: please feel free to correct my comments as needed rather than to perpetuate an error caused by me.

[Bitsyncmaster]

What also hurts the ability of idle control is the delay. The RPM value you read is two ignition pulses so your value is already 51 ms late. We use two pulses because of the un-even fire engine so the timing for RPM is the same. I ran some test software using one pulse and two tables in software and did not notice an improvement so I went back to using two pulses like the OEM ECU does.

Then you have the delay of the ISM PWM to finish the current value and complete the new value. That adds another 20 ms. delay.

Then you have the delay of the ISM inertia. Delay unknown but most likely not important since it already is moving both directions.

Then you have the engine inertia delay (unknown).

[RamblinDMC]

Wow, I've been inactive for a few days. There's a ton of great info here.
I've been unable to free up my original ISM, so I ordered one of the newer replacements with the small diameter housing (0280140501). It was considerably more affordable than an OEM style. It should get here soon.
In the mean time, I've disconnected the harness from my existing ISM, and used a screwdriver to set it to about 1/3 of the way open. I'm still able to drive my car, although it is hard to start and idles high at 1,200 RPM or so.

The new ISM should be here Tuesday.

[nkemp]

Wow, I've been inactive for a few days. There's a ton of great info here.

As I said, this thread took a scenic detour. :)

I run the smaller ISM. Seems to run as well as the original and from what others have said, less burden on the system. I'd be interested to hear if you ever get the original loosened up.

[RamblinDMC]

As I said, this thread took a scenic detour. :)

I run the smaller ISM. Seems to run as well as the original and from what others have said, less burden on the system. I'd be interested to hear if you ever get the original loosened up.
I'm planning on taking it apart once the new one gets here.

Sent from my LGMS631 using Tapatalk

[RamblinDMC]

So update on the new ISM.

I've installed it only to find out that it isn't working. I've confirmed that the harness from the idle ECU is good, the throttle microswitch is good. The ECU is outputting voltage.

From what I can tell, it appears that the plug on the 0280140501 ISM is wired backwards. Where the center pin is a constant 12V and the motor is actuated back and forth by grounding the outer two pins.

A bit confused here.

[NckT]

So update on the new ISM.

I've installed it only to find out that it isn't working. I've confirmed that the harness from the idle ECU is good, the throttle microswitch is good. The ECU is outputting voltage.

From what I can tell, it appears that the plug on the 0280140501 ISM is wired backwards. Where the center pin is a constant 12V and the motor is actuated back and forth by grounding the outer two pins.

A bit confused here.

That's how it's supposed to work !

The switching signal from the wire from the ecu to the throttle switch is not continuous to earth (0 volts battery negative reference), it goes via a anti feedback diode as the throttle switch earths other circuits, such as the distributor (spark) advance solenoid.

I've read this thread with interest, but you can adjust the throttle position and settings to get rid of any idle fluctuations, especially when coming off throttle, they are just not to the manual that's all.

[nkemp]

... but you can adjust the throttle position and settings to get rid of any idle fluctuations, especially when coming off throttle, they are just not to the manual that's all.

Further input?????

[DMCVegas]

What also hurts the ability of idle control is the delay. The RPM value you read is two ignition pulses so your value is already 51 ms late. We use two pulses because of the un-even fire engine so the timing for RPM is the same. I ran some test software using one pulse and two tables in software and did not notice an improvement so I went back to using two pulses like the OEM ECU does.

Then you have the delay of the ISM PWM to finish the current value and complete the new value. That adds another 20 ms. delay.

Then you have the delay of the ISM inertia. Delay unknown but most likely not important since it already is moving both directions.

Then you have the engine inertia delay (unknown).

Even at a total of 70-80ms, that isn't enough of a delay to cause a wild swing in RPMs that would lead to surging. Absolutely it prevents the engine speed from locking to exactly 775 RPMs, that is true. Unless you are setting the engine speed via set screws, with or without an ancillary device to increase Speeds where both are able to introduce a relatively static amount of airflow (i.e idle solenoids for carbs), no vehicle with microprocessor control ever seems to lock in a specific idle speed. There are always subtle fluctuations around the target speed in the tens of RPMs, but nothing where it causes wild swings. Otherwise, the system wouldn't even work at all.

Likewise, even on a car where we do not have vacuum leaks and LAMBDA is functioning properly, we do in fact see that the RPMs are able to settle down. That means that the microprocessor is indeed fast enough to calculate & implement its airflow changes to counter even the inertia of the engine speeds so as to correct the speeds. Otherwise, even on a functioning engine, it wouldn't work at all, let alone one with vacuum leaks.




The phrase "violent agreement" comes to mind again.:) Or should I say "enthusiastic agreement"?

I believe we certainly are in the ballpark of one another! And I also have to admit, this has been a quite fun argument so far!



Correct. PWM does not increase torque but it can change power. The longer the PWM "on" periods, the more power (50% PWM is more power than 20%) assuming some work occurs.

That is the thing though, work is ALWAYS occurring here. It's 12V DC, as we have established. BUT, we have a single, positive power source, and then two individual negative grounds to complete the separate circuits. Both of which are indicative of what direction the ISM rotates.

I know for a fact that if one side doesn't have a proper ground connection, the entire valve will snap fully closed, or fully open, and will not move at all in the other direction. And the reason for that is because as stated, work is always occurring. Without the pull in the opposite direction, the valve will simply continue in one direction indefinitely until it stops at either fully open, or fully closed.

The negative sides of the ISM circuits are completely switched back and forth. It has to ground either one side or the other. This results in a constant "push/pull" action on the motor. So rather than just being a single duty cycle, there are two. HOWEVER, they are NOT independent of one another. The second duty cycle is always inverse from the other. So of it's 60% open, that means it's 40% closed. If it's 10% open, it's 90% closed.

Then you break these all down by treating the "duty cycle" numbers as a math problem. 50/50 means "50-50=0", which results in no change in the ISM's valve position. 60/40 means "60-40=20", which results in a opening of 20%. 40/60 by contrast means "40-60=(-20)" which then means that the valve then closed 20% that time around. Now keep in mind that while we know the function of the duty cycles, we have no specific measurements for the results of when the are applied. I.e. the specific positioning of the valve or CFM variations. It does feel a bit frustrating in that regard, but on the other hand they are not needed here. That is because we're not concerned about CFM, but rather the main function of the switch and it's impact.

The other thing to keep in mind is that while it looks like a PWM, it's not. PWM is to send data. It's actually a square wave with the inverse simply applying to the opposite circuit that is being grounded. There are no harmonics at work here. It may look like a PWM as the amplitude of the wave increases or decreases, but there is no information being sent or received.



For those who are reading this thread looking for more background, and likely not the contributors, here is some background:
- Torque is not power (torque is not horsepower)
- Torque is expressed in terms of pressure.
- Torque is part of the horsepower equation
- Power refers to work done over time
- When torque is applied to something over time and causes some action then work is done
- Watts is a measure of power
- Watts = Volts * Amps
- If Volts is pressure then think of Amps as the volume (number of electrons in electricity (gallons in a water metaphor))
- In our case torque refers to the voltage (12Vdc pressure) which is colloquially referred to as "12Vdc" in the car
- PWM does not affect volts delivered against the "job to be done". You can see this on an oscilloscope where the PWM peaks (the height of the square-waves (pulses)) always shows voltage as 12Vdc.
- PWM affects the number of amps, number of electrons, delivered against the job
- 20% duty cycle allows 20% of the electrons to flow. 50% allows 50% of the electrons to flow... or there about

Good info, but not applicable to the duty cycles we are speaking of here.



And therein is why a smart controller, including a PID controller, will increase the PWM in order to cause a change. When a smart controller does not see the change it needs in the time it needs, it will attempt to cause the actuator to try harder. Using the ball on the sticky track, and adding a smart controller and sensors, when the controller does not see via the sensors the action desired in the time desired, it will increase the effort (PWM), and this can cause overshoot if the control apparatus (sensors, controller and actuator) cannot control within the required time when the sluggishness is overcome (a non-linear actuator response). In our case, the idle ECU, assuming PID control, sends a signal to affect the idle, it monitors the feedback, if the idle does not move towards the desired fast enough, it increases the PWM on one side and decreases it against the other side until it gets the response needed and then it moves back to 50:50 (side to side). (That is my assumption of how it plays out)

BTW ... something similar (but not 2-sided like the idle control) happens with the Lambda control and FV as it attempts to modulate the fuel distributor to affect fuel flow modulation. In this case the fuel pressure is "volts" and fuel volume is "amps". Assuming some sort of smart control like PID control.

BINGO! The ECU doesn't know what the valve's position is. Once activated, the only feedback it receives is what the RPMs are. Then it will adjust the ISM Valve by opening it and closing it in reaction to that RPM value. Now the higher a deviation from the mean target the ECU sees may affect the duty cycles, but that's it. There are no variations is applied power, so there are no variations in applied torque. It's not as through the ISM needs to be dead-center in order to achieve it's target RPM speed. Otherwise there would be some sort of feedback ability from the ISM for valve position. As stands, all that the ECU does is send open and closed signals to the ISM and nothing more.

But keep in mind that while the applied duty cycles can change, that does not affect the amount of power being exerted on the sticky area. More on that in a second though...



In the case of bloodying your knuckles breaking a nut loose, the actuator is the hand, sensors includes eyes and muscle feedback and the controller is the brain. In this case we modified torque ( we can't apply more arms ... but maybe we engaged more muscle cells... Hmmm) to affect more ability to do work and once the threshold is breached, the brain, sensor and actuators didn't react fast enough to prevent the pain that is soon to happen. A really fast system would apply counter control to keep it inbounds.

Again, other than the tach feed for the RPMs, there are no other feedback loops. Using your example, it's more like you're a leper than has lost all sense of touch, and then you pick up a hot skillet. You can't feel anything, and then rely upon me to tell you that you're burning your hand, please put the skillet down.

But as for the torque thing, no additional torque is supplied. We only give additional time to move, but there is no increase pressure applied. Otherwise the ISM will never open the valve and the system is stuck. So that is not the reason for the surging. Again, more on that in a second.



SO ... so much of this discussion also really depends on what the Idle ECU does. What kind of control does it perform? PID would be preferred for the most part since it is best able to maintain tight even control. And I have to go on the input of others that it is in fact PID control. This really is worth the read: https://en.wikipedia.org/wiki/PID_controller

Oh, I can explain that right now. The ECU is a very simplistic PID Controller. When the RPMs are too high, it tells the ISM to close it's valve. When the RPMs are too low, it then tells the ISM to open it's valve. And it does this by modifying the primary and inverse duty cycles that are applied to the ISM.

There are two RPM targets: The default 775± one, and the latter 1,200± (or something like that) one which is activated by the thermal probe in the coolant Y-Pipe. Then you have the micro switch on the throttle that tells the ECU to only activate when the throttle plates are closed. Otherwise it leaves the ISM's Valve fully open. That's the entire function of the ECU.



BUT ... If it is PID control but the sensor is sluggish (a buildup of material slowing its response to temperature change) or the actuator is sluggish (sticky), then the PID control cannot do its job as well as designed.

Totally agreed there. And that is what I was talking about in my last post about how we don't know exactly how far the valve moves with each final percentage of the duty cycles. Remember, it's a constant push/pull situation with the ISM. If there is a "sticky" area caused either by grime or thermal constriction, then yes, we could absolutely see a problem with overshoot, but not because of a variation of power. Let us explore a hypothetical here:

So remember when I spoke about how to read the duty cycles to determine the final variations? Check this out: Let us say that we have a 60/40 split. However, we have a sticky spot in the opening range of perhaps a 10% value, so it's like it is robbing us of advancing the valve because the thing can't move fast enough to get out of the sticky spot. So let us say that that the sticky spot reduces the rotational speed of the valve by a ½. Now the ECU applies a 60/40 duty cycle, with the number on the left representing opening the ISM valve. It it worked properly, we'd have 60-40= 20 for a 20% opening value. BUT let's say that with being sticky that 60% gets cut in half, so now we only have 30/40 resulting in 30-40=(-10). HOWEVER, this value really only equals a zero. And the reason for that is because while it cannot advance beyond the sticky area, the closing cycle keeps pulling it back and resetting it back to zero. That we could also check both because the final sum has an opposite value of what it should have been, but also because 30+40 =70 and not 100%. Those two conditions mean 0 gain.

Now keep in mind, because we have not been able to increase the amount of airflow during this full duty cycle, the RPMs are still falling. As a result the ECU keeps ramping up the duty cycles as the deviation value from the target RPM speed increases. So then lets say it gets to 70/30. Remember, the sticky spot decreases speed by half. So now we have 35-30=5. Congratulations! We now finally has a positive gain, but it is only 5% which doesn't exceed the break-even value. Next go round can push it further back. So then the duty cycle switches to 80/20. Reduce to 40-20=20 (remember, our value remains the same as before). Because that value is greater than the 10% sticky spot value, it means we have gotten out of the sticky spot! Now though while we are starting to feed additional air into the engine, we still will have high values because the RPMs have dropped too far down. So we'll have a surge back up.

Now comes our new problem: The RPMs will eventually climb too high. Because remember, the ECU doesn't monitor the position of the ISM, it only tells it what to do. So then the ECU starts sending 40/60 cycles to close the valve. That then pushes us back past the sticky spot, and our problems repeat themselves. In fact, if the sticky spot goes both ways, it can even cause the same problem, but by letting the RPMs go too high instead (which in that case our sums should be negative, and if they are not after the reduction and have a positive value, the final gain is 0). In either case, that is a bad ISM that needs to be replaced. You could try repairing it, but it's really not serviceable.

Still yet though, this is a good example of how a sticking ISM can absolutely cause cycling RPMs. It should also be noted that if the sticky spot is large enough, the ISM will not be able to correct in time and the engine will either lock at a high speed, or will simply stall out. There is no way for it to apply more power by increasing voltage or amperage. That isn't what the ISM does.

Remember, it's not that we're surging because we overshot our target (like your bloody knuckles example), we're surging because we didn't correct the whole entire thing in time! Which is more like the nut is spinning the the opposite direction that you need to counter.



For further control process thought ... think of temperature control and your furnace:
- Using only threshold control (really dumb control regardless of the smartness of the electronics controlling it), if the furnace comes on at 68 and off at 70 trying to control to 69, there would be overshoot of some amount over 69 and undershoot of some amount of 68 (and this amount is a function of the external temperature and the furnace capacity) ... Hunting
- The "simple" mechanical T87 round Honeywell thermostat improved control by adding an "anticipator" to avoid the overshoot. (it was a tiny heater that tried to warm the bimetal spring detecting temperature to avoid overshoot ... and it was implemented to work both ways) In our example we might assume it tried to stay between 68 & 70,
- Really smart thermostats (not all smart t'stats are really smart) look at the rate of temperature change (when the equipment is off as well as when it is on) and determine when to start, how long to run and when to stop based on the rate of change to flatten the swings (minimizing hunting)

You might say all three perform PWM on the furnace. But it is not as simple as pulsing a bunch to flatten the temperature change (because too many pulses reduce furnace efficiency with too much heat going out the exhaust compared to transferred to the space). Or as simple as one long pulse that is the most energy efficient but has the greatest over and undershooting (hunting in our example)

But I digress.

The furnace example is apples to oranges, but not because of the variations in the control system. Rather it's because you have non-applicable accessories being introduced that don't really apply to the Idle Speed Circuit. A couple of better examples would be poor insulation around the thermostat, or even a vent that was aimed directly at it. Room stays cold, and every time the heat kicks on, the vent bathes the thermostat in warm air that causes it to prematurely kick off.



The point I'm making is that the control algorithm is key to the hunting problem as are the sensors, actuator performance and all the interdependencies of other control systems. And of course leaks. It is the "system view".

That's the problem though. Unlike a modern car with a communications bus and integrated modules, these are all separate ones! There are not interdependancies; everything operates independently without sharing any common algorithms. Modern cars with Idle Air Control Valves have static duty cycles they can apply be default, and will when they recognize such malfunctions as the wild swings in idle speeds. With the older modular systems like we're dealing with, noting communicates with one another to either identify a problem, let alone is able to throw an error code to help identify WHAT the cause of the problem is.


And this probably is more "violent agreement". :) But I tend to think of it as enthusiastic. :)

Technical experts: please feel free to correct my comments as needed rather than to perpetuate an error caused by me.

Same here of course too on both counts. These are the kinds of technical discussions I enjoy the most on here.

[NckT]

Further input?????

As others have said, the idle speed ecu is just a fixed PID control valve positioning system that uses the rpm as process variable (PV) fixed setpoint (SP) of 775 rpm and the output (OP) is the valve position utilising the earthing of the motor outer pins as a very fast "nudge wait" type algorithm.

PID utilises control function of
P: proportional, or Gain of response
I: Integral
D: Derivative.
In the ecu, those parameters are fixed and cannot be adjusted (unless the components are modified)

Now, if they were adjustable Is be inclined to reduce the gain and increase derivative values, but we're stuck.

However, the Output is the full scale deflection of the Valve, ie fully closed = 0rpm fully open = approx 2000-2500 rpm

The small incremental deflection of the valve causes a fairly large change of the engine rpm, and this assumed that the motor is free, had no sticking wear characteristics etc.

The control function of this can be adjusted however by changing the characteristics of the control system so that the full scale deflection of the valve only affects a small proportion of the idle revs, and to do this is to increase air bypass around the throttle plate and adjust to include air damping.

In simple terms this is how I've set my engine (which has full working lambda control also) and is NOT to 'the book': (I don't have pictures to hand as I'm on my mobile device)

Ensure engine is mechanically fine such as piston compression and valve clearances are correct
Ensure that the vacuum pipes are on and in the correct order and actually work
Ensure fuel pressures are correct both primary and control pressure and re-shim fuel pressure and recalibrate control pressure regulator (wur)
Ensure injectors are spraying correct and not leaking
Ensure ignition timing is correct and the distributor advance solenoid works
Ensure there is a bung or the anti tamper plug is in place where you adjust the fuel ratio with the long 3mm Allen key
Ensure lambda system is working and the frequency valve works and oxygen sensor operates and dwell feedback is within spec (as this can cause an idle surge in itself) this will assume pre catalytic conveyer emissions is correct.
Ensure deceleration valve springs operate to fully close the poppet valves
And, of course that the idle motor does not have sticking areas and the idle system engages with the microswitch ie the in line anti feedback diode works

To begin:

Undo throttle link rod
Wind back adjustment of hammer on throttle stop microswitch
Wind back adjustment of lower throttle stop screw so it's fully backed off
Wind in lower stop throttle screw until it just touches the throttle stop. Wind in 2 full turns and lock the nut
Open the throttle slightly and place a metric feeler gauge with '30' on it on the throttle stop and close the throttle so the feeler gauge opens the throttle slightly.
Wind in the microswitch hammer slowly so that it just clicks the microswitch. Tighten this lock nut.
(The above ensures that the Idle ecu operates at the correct time as too early or too late will cause incorrect valve positioning causing surge on idle)

Ok so far ?

Re attach link rod and ensure that closing the throttle using the accelerator pedal has repeatable operation to fully close the throttle at the stop, slightly shorten the link rod so that the main throttle spool spring assists in closing the throttle repeatable.

The following I do to add additional air damping around the throttle plate:

Gently screw all of the 3 brass screws in so they are seated.
Unwind the 2 brass screws nearest to you by 2 full turns each
Run the engine to temperature
Use a multimeter on the coil to measure engine rpm and note the reading
Unscrew the larger brass screw with the slot slowly out until the engine revs read slightly higher, then screw back in quarter of a turn.
Re check dwell reading from lambda computer and that at 3000 rpm the oxygen cycles from 0.2v to 0.8v at least twice a second to check fueling is correct
Blip the throttle open and let go to then check idle surge has been reduced

Check idle by putting lights on, air conditioning, rear heater etc and check idle soured remains

I have done the above for a number of cars and had been repeatable.

In addition, even with the correct thermistor fitted (there's 2 types), the fast idle only operates with an additional 10 k ohm resistor in series with the thermistor (coolant temp sensor that varies its resistance with temperature)

I hope this helps,

Nick

Btw if the ecu fails it's usually one of the earth switching output transistors first followed by the operational amplifiers then the remaining logic chips..

[NckT]

In addition, when you turn the ignition key to position 2, the ecu will position the valve fully open (assuming no mechanical issues with the idle speed valve of course)
This aids starting so you don't need to put your foot on the accelerator pedal.

The ecu will only position the valve to engine idle with the micro switch engaged.

If the micro switch is not engaged then the valve will fully open and explains why engine revs are high if the throttle does not shut fully and the micro switch is not operated.

In addition, if the micro switch is activated too soon then the throttle is still open and the engine revs are high so the valve closes prematurely then the engine falls below spec then increases up then the throttle is shut then the Idle valve attempts to open, all underdamped if you view it as a control feedback response. The microswitch switch activation at the right time is important.

Anyway, too much theory, I thought I'd just write down what I do to help.

Nick

[DMCVegas]

To begin:

Undo throttle link rod
Wind back adjustment of hammer on throttle stop microswitch
Wind back adjustment of lower throttle stop screw so it's fully backed off
Wind in lower stop throttle screw until it just touches the throttle stop. Wind in 2 full turns and lock the nut
Open the throttle slightly and place a metric feeler gauge with '30' on it on the throttle stop and close the throttle so the feeler gauge opens the throttle slightly.
Wind in the microswitch hammer slowly so that it just clicks the microswitch. Tighten this lock nut.
(The above ensures that the Idle ecu operates at the correct time as too early or too late will cause incorrect valve positioning causing surge on idle)

That is exactly what I've been saying too. While there is naturally a slight delay in the ECU responding to the current RPMs to adjust the Valve, there is by contrast a MASSIVE one when clicking the Idle Micro Switch! If you adjust the striker, and even bend up the bracket if need be, you can get that ECU to activate sooner to avoid the idle speed surging! I've seen this on my own car as well! I, however, prefer to go ahead and cause the switch to click early. That way the ECU takes over, and the RPMs even out more gently. When this happens, the Deceleration Valves on the Throttle Plates take over and help feed the engine to prevent a lean/stall condition.

It will also give you a little bit of a boost in power when taking off because you're feeding the engine off of throttle plates, and then it kicks up suddenly to avoid engine lugging.



Re attach link rod and ensure that closing the throttle using the accelerator pedal has repeatable operation to fully close the throttle at the stop, slightly shorten the link rod so that the main throttle spool spring assists in closing the throttle repeatable.

YES! YES! YES! This cannot be stressed enough! This is important! When the engines were assembled in Douvrin, the linkage between the Throttle Spool and the Throttle plates was never properly adjusted! So even through you press the pedal to the floor, you never fully open up your throttle plates! So you still have untapped power in the engine you never use. I can't say that this is true for every single DeLorean because of Renault's manufacturing, but I HAVE seen it on ever car I've ever worked on.



The following I do to add additional air damping around the throttle plate:

Gently screw all of the 3 brass screws in so they are seated.
Unwind the 2 brass screws nearest to you by 2 full turns each
Run the engine to temperature
Use a multimeter on the coil to measure engine rpm and note the reading
Unscrew the larger brass screw with the slot slowly out until the engine revs read slightly higher, then screw back in quarter of a turn.
Re check dwell reading from lambda computer and that at 3000 rpm the oxygen cycles from 0.2v to 0.8v at least twice a second to check fueling is correct
Blip the throttle open and let go to then check idle surge has been reduced

Check idle by putting lights on, air conditioning, rear heater etc and check idle soured remains

I have done the above for a number of cars and had been repeatable.

That isn't exactly necessary, and not something I personally recommend. Two reasons for that:

1. You really only ever need to use those screws when adjusting the CO Screw to set the RPMs to 950. Otherwise the modification to the Micro Switch takes over nicely.
2. Take a look at the earlier K-Jetronic system with the previous Auxiliary Air Valve (for cold starting & A/C) that used those screws to set the idle speed:

https://drive-my.com/images/renault30/Renault-30-TX-04.jpg

Disregarding the breather hose, you can see a long tube connecting between the left and right cylinder banks that has been plumbed into the Throttle Tubes (part 102279). That tube is important because it allows for ΔP to occur. That way the airflow between both cylinder banks equalizes and doesn't cause one side to run too rich or too lean.

[NckT]

I've done the adjustments going on for a few Deloreans now and all have had the same positive results now.

I understand about the auxiluary air valve and the adjustment of the brass screws on an oem setup 'by the book', and the adjustment of them in respect to balancing the airflow between each cylinder head by the split air path but I'm applying practise based on experience of dampened response by practical experiments on the vast from when I first got it when it ran like crap.

For explanation for others reading this, the auxiliary air valve is a device that allows additional air around the throttle plate by a smaller valve that when open causes a faster idle. It slowly closes because it's connected to a bi metallic strip with a heater coil around it. With residual engine heat and electrical supply to heat the coil in addition, the valve will close fully so when the engine is warmed up the idle reduces to normal idle back to the plate position and screws. In essence, the Idle motor just replaces this in the same air path.

In addition, what is not often checked is that the stroke of the accelerator pedal fully going down actually fully opens the throttle plates in their entirety. Adjustment is done by screwing up the bolt headed accelerator pedal so and adjustment of the cable to the throttle spool. You'll find then that the pedal is slightly higher than the brake pedal bit the full stroke of the pedal does then fully open the throttle plates.

The pressure differential pipe you're referring to is part of the crankcase ventilation system, you can see it connected to the oil filler cap.

I don't mind technical discussions, but I do apply in practice first rather than copy/ paste what others may have previously said before.

I think I've met you btw at euro fest as I helped out on the parts stall in the hotel

[nkemp]

I thought I'd just write down what I do to help.

Nick

And I appreciate the details!
Nick ... the other Nick

[NckT]

And I appreciate the details!
Nick ... the other Nick

Alright there other Nick, no problem

[DMCVegas]

In addition, what is not often checked is that the stroke of the accelerator pedal fully going down actually fully opens the throttle plates in their entirety. Adjustment is done by screwing up the bolt headed accelerator pedal so and adjustment of the cable to the throttle spool. You'll find then that the pedal is slightly higher than the brake pedal bit the full stroke of the pedal does then fully open the throttle plates.

Fascinating. I've definitely seen many cases of too much slack in the linkage rod between the Throttle Spool and the Butterflies. That results in a terrible delay where the spool rotates about a cm or so before finally cracking the throttle open. So you pretty much end up with dead travel in the accelerator pedal. Not to mention that it never fully goes full WOT. Slack in the cable down at the the Pedal Box is something that should be inspected too. Thank you for bringing that up.



The pressure differential pipe you're referring to is part of the crankcase ventilation system, you can see it connected to the oil filler cap.

That is correct. But while it is correct, that doesn't mean that it doesn't also have a secondary purpose. And that purpose would be to allow a ΔP function in order to let the manifold vacuum pressures between the left and right cylinder banks balance out. That is why there are 3 individual screws for the idle system (and two areas for infrared probes to balance CO). So if say the right cylinder bank was being starved of air, it could then instead pull the additional air from the crankcase or Auxillary Air Valve to automatically balance itself out to avoid a lean condition.

With the next generation K-Jet system that introduced CIS to replace the Auxiliary Air Valve, the screws were no longer needed except for adjusting the fuel mixture. That now meant that since the ISM's variable valve function was going to handle ALL of the airflow, that pipe was no longer needed for ΔP since the remainder of the intake manifold could now take care if it itself. So naturally the crank case ventilation then just plugged into the Cold Start Tube that the ISM was using and was done with it.



I don't mind technical discussions, but I do apply in practice first rather than copy/ paste what others may have previously said before.

I think I've met you btw at euro fest as I helped out on the parts stall in the hotel

Technical discussions are great. It's nice to see these come together.

If you're talking about me at Eurofest though, nope, twas not I.

[RamblinDMC]

Update:

I have confirmed that the newer style Idle Speed Motor is in fact backwards (although not in the way I originally thought) from the OEM style. In that the outside two pins are swapped. I was able to make an adapter so that it can be used. I am now unsure if my idle ECU is functioning properly. I have tried unplugging the harness closest to the center of the car to observe if the RPMs jump. No response.
The idle motor does not move at all, however I can manually actuate it by applying 12V across the appropriate pins.

[NckT]

Update:

I have confirmed that the newer style Idle Speed Motor is in fact backwards (although not in the way I originally thought) from the OEM style. In that the outside two pins are swapped. I was able to make an adapter so that it can be used. I am now unsure if my idle ECU is functioning properly. I have tried unplugging the harness closest to the center of the car to observe if the RPMs jump. No response.
The idle motor does not move at all, however I can manually actuate it by applying 12V across the appropriate pins.

If you still have your old idle speed motor then measure the resistance between the centre pin to the outer pub then do the range to the other outer pin. If one reads short circuit then chances are you've blown the switching transistor in the ecu. You could try the test on your old motor by applying 12v to the centre pin then momentarily earthing the outer pin to see if of moves. In one instance I've had to fix a couple of ecu s due to this where the fault was with the idle motor.

I have one of the newer idle motors on my car and is not backwards wired, unless you have a bad one for some reason

[Bitsyncmaster]

Update:

I have confirmed that the newer style Idle Speed Motor is in fact backwards (although not in the way I originally thought) from the OEM style. In that the outside two pins are swapped. I was able to make an adapter so that it can be used. I am now unsure if my idle ECU is functioning properly. I have tried unplugging the harness closest to the center of the car to observe if the RPMs jump. No response.
The idle motor does not move at all, however I can manually actuate it by applying 12V across the appropriate pins.

I use the smaller Volvo idle motor and did not need to swap the driver wires. Maybe your ECU is bad.

[RamblinDMC]

What was the part number on it?
My ECU is outputting voltage but maybe it cannot meet the current demands.



Sent from my LGMS631 using Tapatalk

[Bitsyncmaster]

What was the part number on it?
My ECU is outputting voltage but maybe it cannot meet the current demands.



Sent from my LGMS631 using Tapatalk

I don't know the part number and I painted it black. The smaller motor takes half the current of the OEM motor.

[nkemp]

I have a "501" installed and it works as does the DeLorean's original. For them to be interchangeable, they should have the same pin configuration. What made you think that it was wired backwards?

[nkemp]

Fascinating. I've definitely seen many cases of too much slack in the linkage rod between the Throttle Spool and the Butterflies. That results in a terrible delay where the spool rotates about a cm or so before finally cracking the throttle open. So you pretty much end up with dead travel in the accelerator pedal.

Maybe we should spin this off as a new thread. but ...

I think part of this problem is that as the accelerator pedal is depressed, the initial rod action at the spool is side to side. Being that the closed throttle relaxed position is the "6 o'clock" position (or very close), it moves towards the side rather than inline with the linkage.. albeit a very short amount. I believe that full throttle is the "9 o'clock" position.

Has anyone tried repositioning the closed throttle at something closer to the "7 o'clock" position? That implies repositioning the throttle spool ... not simply shortening the rod to a 7 o'clock start.

[DMCVegas]

I have a "501" installed and it works as does the DeLorean's original. For them to be interchangeable, they should have the same pin configuration. What made you think that it was wired backwards?

I beleive that is the same unit I have as well. Gold in color from the 760 GLE. Pretty sure it is the same size as the stock unit though. Which are these smaller ones everyone else is using, and what do they come off of?



Maybe we should spin this off as a new thread. but ...

I think part of this problem is that as the accelerator pedal is depressed, the initial rod action at the spool is side to side. Being that the closed throttle relaxed position is the "6 o'clock" position (or very close), it moves towards the side rather than inline with the linkage.. albeit a very short amount. I believe that full throttle is the "9 o'clock" position.

Has anyone tried repositioning the closed throttle at something closer to the "7 o'clock" position?

Indeed this is begining to sound like an entirely different thread topic.

It's been a while since I've been around my car, but what you're describing sounds like what I'm talking about as well. It appears as though as the spool begins to rotate, it merely turns where there is a delay on pulling on the lever for the throttle plates. The best way I found to adjust that is to disconnect the linkage below, then rotate each end fitting by 180°. If you do 360° on a single one, it will excessively pre-load the linkage, and the throttle won't fully close. So you've got to do both ends.

This is also dealing with the later style of throttle spool. That version is the open top with an adjustable throttle cable. The earlier style is closed on top, and the cable uses a barrel connector. So you have to adjust everything at the accelerator pedal. When I get back, I can take some photos and measurements.

[NckT]

That's going back to what I was on about with the full stroke of the accelerator pedal.

Disconnect the link rod and remove the cable from the spool. The throttle should spin round and hit the stop.

Reconnect the cable to the spool and press down on the accelerator pedal fully. The spool should fully open so it hits the full open stop and operate the full load microswitch. If it doesn't then wind the bolt in where the accelerator pedal rests (about 3 full turns usually). Undo the cable from the spool to let the spool hit the closed stop then re attach cable and repeat the test.

Set the throttle itself as I described in an earlier post.

Set the link rod so the spool is only just off the stop. This way the spring in the spool ensures that the throttle is closed repeatedly and takes into account any slight wear in the system.

Enjoy

Nick

[RamblinDMC]

I have a "501" installed and it works as does the DeLorean's original. For them to be interchangeable, they should have the same pin configuration. What made you think that it was wired backwards?

When I ground the same pin on both motors, one opens and the other one closes and vice versa.

[Bitsyncmaster]

When I ground the same pin on both motors, one opens and the other one closes and vice versa.

That would prove it's backwards. Are you using the connector "key" as your reference? Or the original is backwards.

[nkemp]

RamblinDMC ... what ISM are you using?

My original was a 500 and the junk yard "upgrade" was a 501. Maybe someone else knows if the DeLorean ever came standard with a 501. Note: I have read (Here: http://www.k-jet.org/articles/information/k-jet-in-detail/#CIS) that the 516 is not interchangeable with the 500, 501 & 520 but it did not say why.

[RamblinDMC]

That would prove it's backwards. Are you using the connector "key" as your reference? Or the original is backwards.
Yes, I'm using the key of the connector as reference.

Regardless, the motor still does not move under power of the ECU.

Sent from my LGMS631 using Tapatalk

[RamblinDMC]

RamblinDMC ... what ISM are you using?

My original was a 500 and the junk yard "upgrade" was a 501. Maybe someone else knows if the DeLorean ever came standard with a 501. Note: I have read (Here: http://www.k-jet.org/articles/information/k-jet-in-detail/#CIS) that the 516 is not interchangeable with the 500, 501 & 520 but it did not say why.
I am using a 501

Sent from my LGMS631 using Tapatalk

[nkemp]

Bosch idle air motor:

Volvo's used:



500

240 81-82
240 Turbo 81-82



501

240/740 83 - 4/87
240/740 Turbo 84 - 4/87



520

240/700 5/87 - 12/88.
Many late model turbo vehicles use the 520.



516

Most late model 240, 700, 900 vehicles LH 2.4 & 3.1 Fuel Injection (1989+).



Just as reference ... I believe this to be accurate but one should check/verify for themselves.

[Bitsyncmaster]

I found the part number of the Volvo idle motor I'm using.

1983-92 Volvo 240 740 ICV Idle Control Valve 0280140520

[DMCVegas]

RamblinDMC ... what ISM are you using?

Note: I have read (Here: http://www.k-jet.org/articles/information/k-jet-in-detail/#CIS) that the 516 is not interchangeable with the 500, 501 & 520 but it did not say why.

I can't speak to every individual application of L/LH-Jetronic and what it looks like in relation to the engine manifold. For instance with K-Jet, BMW keeps their air mixture unit with the distributor off to the side, away from the engine. Which results in some strange plumbing. But, some of the LH ISMs don't have parallel inlets and outlets, and instead have one that is 90°. So physically, that would definitely not work on a PRV application.

https://www.volvoclub.org.uk/faq/Images-Parts/IdleAirValve.gif

But then again as well, those later systems starting changing and evolving too. I think I've even seen some later Jetronic systems that would use both the Auxillary Air Valve (for fast idle) and an ISM for normal idle control. So who knows what the CFM ratings are, and how that could affect performance. That's why when I was in a pinch with the D being my only car back when I was a bachelor, I would only opt for '83 Volvo GLE V6 w/B28F parts. Otherwise, it's OEM all the way to avoid compatibility problems like this.

[RamblinDMC]

I can't speak to every individual application of L/LH-Jetronic and what it looks like in relation to the engine manifold. For instance with K-Jet, BMW keeps their air mixture unit with the distributor off to the side, away from the engine. Which results in some strange plumbing. But, some of the LH ISMs don't have parallel inlets and outlets, and instead have one that is 90°. So physically, that would definitely not work on a PRV application.

https://www.volvoclub.org.uk/faq/Images-Parts/IdleAirValve.gif

But then again as well, those later systems starting changing and evolving too. I think I've even seen some later Jetronic systems that would use both the Auxillary Air Valve (for fast idle) and an ISM for normal idle control. So who knows what the CFM ratings are, and how that could affect performance. That's why when I was in a pinch with the D being my only car back when I was a bachelor, I would only opt for '83 Volvo GLE V6 w/B28F parts. Otherwise, it's OEM all the way to avoid compatibility problems like this.


As a broke engineering student I'm trying to avoid paying the high prices for many of the OEM parts. At this point I do believe my Idle ECU is bad. However I cannot confirm this as it is outputting voltage but it won't turn the motor.

[nkemp]

However I cannot confirm this as it is outputting voltage but it won't turn the motor.

As I recall, the circuit switches the ground side. As such, the fact that there is an output voltage doesn't tell much. If you can solder, you may be able to fix it for a few dollars.

Go to post #11 here: http://dmctalk.org/showthread.php?9761-Remove-Deceleration-valves>

Just replace all the parts listed. No guarantee that it will fix the problem, but the cost is minimal in the scope of thing.

To test if the switching is working, I think (someone verify this please) you can put a meter, set to Vdc closest and over 12Vdc between the center tap and the side taps, one at a time. Run the car with the connector disconnected. You should see a voltage on each side. If no voltage I suspect a bad solid state switch on that side. I wouldn't expect to see 12Vdc (or whatever your car's voltage is). It should be relative to the pulse width on that side (50% would be about 6Vdc).

Now I open this to the electrical experts to clean up the mess I made with that suggestion :smile:

[Elvis]

Send me the ECU, I will repair and test it for a fair price.

[RamblinDMC]

Send me the ECU, I will repair and test it for a fair price.

What do you charge? Where are you located?

Thanks

[Elvis]

Germany

[RamblinDMC]

Ok,
What do you typically charge?

[Elvis]

Shipping costs are the problem - you check what it costst to ship, and if you are still interested - then we can talk.

But I won't post prices here !

I am not a company, I do this as hobby, for fun

[DMCVegas]

As a broke engineering student I'm trying to avoid paying the high prices for many of the OEM parts. At this point I do believe my Idle ECU is bad. However I cannot confirm this as it is outputting voltage but it won't turn the motor.

No shame in that at all. I know exactly what that is like, and don't fault anyone for being in that situation. The important thing though is always that you MUST ensure that the repairs are completed properly the first time around, otherwise you will end up spending more money in the long run. Not to mention encounter allot more frustration. Now there are ways to temporarily band-aid things to keep the car going. You could either block-off the hoses and then just use the Idle Speed Screws in the manifold temporarily, or just connect a manual valve in between the hoses and set your idle speed that way. Cheap, works temporarily, fully reversible, and most of all keeps the car running until you can implement a proper repair. Which is fine so long as you actually DO go back and properly complete the repair later on.

Like I say, I've been in a very similar situation as you. In fact, with the Idle Speed Circuit itself. And the best thing I can recommend to you is to learn from my mistakes.

First thing you need to do is to check for continuity and resistance in the wiring that runs from the ECU to the ISM.

http://www.k-jet.org/img/articles/diagrams/cis_system1.jpg

Page D:05:03 of the Workshop Manual has a schematic of the wiring that you can follow. Which is easier to understand because it better lays out the numbers on the plugs (1, 2, & 3 the top, 4 & 5 bottom). The wires you're concerned about would be Black/Red, Black/Pink, & Black Green. This is why I advise to always buy a copy of the manual from a vendor. Don't just Google it for a bootleg copy, because that is wrong, OK? Don't download it by searching for it on Google.

Anyhow, what you'll need to do then is connect up a multimeter and start testing those wires. Check to ensure your voltage from Black/Pink is the same coming out of pin #4 on the ECU as it is when it reaches the ISM. Then you will need to check for resistance on the Black/Red & Black/Green wires. You'll need leads to do this in order to reach into the passenger cabin. And you'll need to test the wires end-to-end for a measurement of inherent resistance to create a baseline for your final measurement. See the video below:

https://youtu.be/InJhgwmj2So?t=20m48s

Now if you have proper continuity on your lines, then you will move onto the ECU. The repairs myself I am not familiar with, but soldering components is not hard at all if you do want to go nuclear on the whole thing. But first start with your wiring, THEN move on to your other components.

[RamblinDMC]

Update:
These are the following things I've checked:
Continuity from idle ECU plug to ISM plug in all three wires.
No resistance between any of the three wires.
Idle ECU has power and a good ground.
Good connection to terminal on igniton coil.
Idle speed switch is good.
My new ISM shows 20-something ohms on either side.
ISM can be actuated back and forth freely when using a bench power supply.

A 6-8V pwm signal can be seen on the ISM plug with it unplugged. HOWEVER it will not actuate the ISM. I then checked the voltage at the ISM while plugged in and it slumps down to less than one volt. Seems like the idle ECU cannot meet the current demand of the ISM which causes the voltage to plummet. It looks like I have a few fried components.
Now to order some parts and get to soldering!


Sent from my LGMS631 using Tapatalk

[RamblinDMC]

Shipping costs are the problem - you check what it costst to ship, and if you are still interested - then we can talk.

But I won't post prices here !

I am not a company, I do this as hobby, for fun
Hi Elvis,
Do you mind sharing the part numbers that you use?



Sent from my LGMS631 using Tapatalk

[DMCVegas]

That is great progress!

Now the next question is where is your original ISM? Is it still intact, and does it bench test fine as well? If so, you could return the new unit to get your money back.

[RamblinDMC]

That is great progress!

Now the next question is where is your original ISM? Is it still intact, and does it bench test fine as well? If so, you could return the new unit to get your money back.

No, it was seized and no amount of carb cleaner would free it, I disassembled it to attempt to fix it, but ended up screwing it up. I didn't pay much for the 501.

[nkemp]

A 6-8V pwm signal can be seen on the ISM plug with it unplugged.

Just to be clear... you are measuring between center and the outside pins. And you are seeing 6-8V on both sides?

[RamblinDMC]

Just to be clear... you are measuring between center and the outside pins. And you are seeing 6-8V on both sides?

Yes that's correct. But once it's put under load, the voltage drops away.

Thanks

[RamblinDMC]

I'm currently placing an order on Mouser to replace the diodes, ICs, transistors, and electrolytic capacitors. Having a tough time identifying the diodes. Is there a schematic and parts list available?

[RamblinDMC]

Update:

I was able to trade some machining/fabrication to another DeLorean owner in return for a good used idle ECU. Everything on my car works as intended now! Turns out the 501 ISM isn't wired backwards. Not sure what I was thinking there! :p

I'm still waiting on the parts to repair my old ECU, I'll still try to fix it just to see if I can do it.

[Elvis]

80% are the power stages. done.

how to test a transistor is most likely all you need to know.