Is this a bad ecu?

[Jason]

My car has developed a problem where when it's started from cold it will run fine for the first minute or so, then start running rough and with a lack of power. Feels almost like a cylinder or two isn't firing. I have a wide band O2 sensor on the right cylinder bank, and when the car is running rough it reads richer than normal. After anywhere between two and twenty minutes the car will suddenly "wake up" and start running normally. The problem has been slowly getting worse, where initially it would happen in maybe 1 in 20 cold starts, and now it happens almost every cold start.

So I went to check the mixture by looking at the frequency value duty cycle at the diagnostic port, and found that the signal was cutting in and out.


https://www.youtube.com/watch?v=kRydNKBea5M

This is after about three minutes of idling from a cold start. I'm guessing this isn't normal, and that I should be seeing a constant signal, correct? Also the duty cycle looks wrong to me. Shouldn't it be around 50% during warm up? For the first minute or so the duty cycle was 40%, though still cutting in and out as above. Does this mean the ECU is bad, or can something else cause this?

Thanks

- Jason

[Bitsyncmaster]

Your scope is looking at noise since it's on the 5 nano/second scale. The frequency of the FV is 70 Hz so your scale would be around 10 milliseconds.

[DMCVegas]

The FV also has a nasty bend in it's wiring right before the BOSCH connector. I had a failure with my Idle Speed Motor's wiring that resulted in a bad connection that I had to graft a replacement wire in to rememdy.

Aside from adjusting your scope like Bitsyncmaster said, also physically check your wiring for continuity.

[Jason]

Your scope is looking at noise since it's on the 5 nano/second scale. The frequency of the FV is 70 Hz so your scale would be around 10 milliseconds.

The scope was on auto, so when there is no signal the time scale goes small while it's searching for something. When it sees a signal the time scale switches to 5ms.

[Bitsyncmaster]

The scope was on auto, so when there is no signal the time scale goes small while it's searching for something. When it sees a signal the time scale switches to 5ms.

Can you switch it off auto? You don't really know what your looking at when it's switching the timebase. You should also limit the bandwidth of your input if your scope can do that. Car circuits are full of high frequency noises.

[Jonathan]

I don't know enough about determining whether ECUs have gone bad to comment, but one thing that did come to mind when you mentioned it being intermittent and related to the O2 sensor, was possibly the O2 sensor ground has gotten corroded (and making a poor connection).

It is located on the passenger side of the intake manifold, at the end nearest the firewall (very close to where the frequency valve is mounted actually). You can see the screw used for that ground on this diagram (http://store.delorean.com/c-334-1-4-1-air-induction-system.aspx), and it is #45 that mounts together with the bracket #44.

The ground wire has a quick connector (which was black on my car) and the short end of the wire between that connector and the ground screw is RED in colour. Should be easy to spot. You might get lucky and a quick check of the screw and connector could be all it was. Go easy on the screw as if it is corroded, it may snap off trying to get it free.

Here is a photo showing the red wire in the top right hand corner (pic was from Dana showing the throttle cable clip install).

38994

[Bitsyncmaster]

I watched your video again and I see the 67 Hz shown in the display but that waveform is nowhere near the 49% the other display shows. Maybe your trigger threshold is set to high causing it to blank out.

[Jason]

So, yes, it was scope incorrectly switching to a different time scale that was making the signal "disappear". Doh! I feel stupid.

But I still think something funny is going on. Or I don't understand how its supposed to work.

I let the car cool down for 2 hours (outside temp was about 40) and started it again. 30 seconds after starting I got..


https://www.youtube.com/watch?v=87cmOIMAUic

Which I guess is what it should look like running open loop.

[Jason]

About a minute later it changed too


https://www.youtube.com/watch?v=2arb_p3nQvs

Isn't it too soon for the system to go out of open loop mode?

[Jason]

So I drove the car 10 minutes to the gas station, filled the tank, and drove 10 minutes back to get it completely warmed up. It ran rough for the first minute, and then normally. When I got back I looked at the ECU signal. It appeared to be back in open loop mode, except at and 50% duty cycle instead of 40%. Is that supposed to happen? I was reving the engine to see if I could get the ECU signal to change at all. For the first few revs nothing, and then it seemed to suddenly switch back to closed loop operation.


https://www.youtube.com/watch?v=UE2bgS_m-ps

[Bitsyncmaster]

Now your scope looks good. I will have to look if that 40% is really inverted. I have not hooked my scope to an ECU yet, I use my Snap-on meter to read the duty cycle. With the Snap-on It reads correctly with 50% at start up and it would read 60% pressing the wide open throttle switch. Or if your engine temp is below 59 deg. F than you should see 60%. But when I use my mutimeter to read duty cycle it reads backwards. I wonder if the car meters all reversed the readings.

Anyway, you can check the reading by pressing the wide open throttle switch If you still have the stock setup.

It looks like your car went closed loop but the mixture is not adjusted. But on your return trip it when to normal open loop at 50% which it should not have done.

[Bitsyncmaster]

I just bench checked a lambda ECU with my scope and the orange wire does have a backwards signal. When I set the WOT signal on the waveform showed 40% at 12 volts and 60% at ground.

So it seems like the automotive meters do reverse the electronic definition of duty cycle.

So it looks like your cold start was open loop at 60% which would indicate your engine was below 59 deg. F or the WOT switch was stuck on.

[Jason]

Now your scope looks good. I will have to look if that 40% is really inverted. I have not hooked my scope to an ECU yet, I use my Snap-on meter to read the duty cycle. With the Snap-on It reads correctly with 50% at start up and it would read 60% pressing the wide open throttle switch. Or if your engine temp is below 59 deg. F than you should see 60%. But when I use my mutimeter to read duty cycle it reads backwards. I wonder if the car meters all reversed the readings.

Anyway, you can check the reading by pressing the wide open throttle switch If you still have the stock setup.

It looks like your car went closed loop but the mixture is not adjusted. But on your return trip it when to normal open loop at 50% which it should not have done.

When it was "open loop" after warming up when I pressed the WOT switch the duty cycle went from 50% to 40%.

I've adjust the mixure, though I did it at normal idle, not the 950 RPM mentioned in the manual.


https://www.youtube.com/watch?v=CFQaCuE7DlE

I'm going to let it cool down for a few hours and then see what it does.

[Bitsyncmaster]

If your using the scope to adjust the mixture, you would want to adjust closed loop for 56%. That would really set you at 44% or 40 deg. dwell if you were using a dwell meter.

It's not really important to idle at 950 RPM. That may let the o2 sensor warm up more and get a more stable reading since our O2 sensor is not heated.

[David T]

Instead of getting out your scope you should be hooking up a pressure gauge and see what the control pressure is doing. My guess is the CPR is bad and your control pressure is too high when cold. Besides, your O2 sensor won't read accurately when the sensor is cold.

[Jason]

Instead of getting out your scope you should be hooking up a pressure gauge and see what the control pressure is doing. My guess is the CPR is bad and your control pressure is too high when cold. Besides, your O2 sensor won't read accurately when the sensor is cold.

I was thinking that, but I don't have a fuel pressure gauge, so I figured I'd check the mixture setting before buying more tools. I did find some crud in the screen on the CPR, but after cleaning it the problem remained.

[DMCVegas]

I was thinking that, but I don't have a fuel pressure gauge, so I figured I'd check the mixture setting before buying more tools. I did find some crud in the screen on the CPR, but after cleaning it the problem remained.

K-Jetronic operates off of hydraulic pressures, and incorrect ones can absolutely screw with the A/F mixtures, which leads the LAMBDA system go through a larger duty cycle.

If nothing else, run the tank down, throw some Techron in, and then give it the old Italian Tune-up. Then see how she goes.

[Jason]

Well, I'm a little surprised, but it looks like just adjusting the mixture fixed the problem.

[refugeefromcalif]

throw some Techron in, and then give it the old Italian Tune-up.

Forgive me but, What the heck is an Italian Tune Up? Just go as fast as you can and pray it works?

George

[DMCVegas]

Forgive me but, What the heck is an Italian Tune Up? Just go as fast as you can and pray it works?

George

That's a great question. And nope, no prayer involved at all. Here's a quick rundown of an Italian Tune-Up:

What is it?
It's a way to de-carbonize the Intake Valves and Spark Plugs inside of an engine. Engines function best under an optimal engine temperature, because this ensures a proper, complete burn of the fuel inside. If the engine isn't ran at temperature allot, it is at risk of building carbon.

How is it performed?
The idea is to run the engine under load, at near red line. Best accomplished by driving on the highway without engaging your overdrive gears. So on a manual, you want to keep it in 4th Gear, and 2nd on your automatic.

What is supposed to happen inside of the engine?
As stated before (and as common engine knowledge), engines run best under a specific temperature where heat is involved to ensure a clean burn, which helps eliminate carbon. By placing the engine under load, we increase and maintain the temperature on the cylinder heads. The other part of this is that gasoline is a natural solvent. So as it gets sprayed down onto the valves, it will wash them clean, and this can help the combustion chamber too. When we have an added solvent such as Techron, it hastens the cleaning process.

What else does it do?
It also cleans up the fuel system itself. The final part is that we're also getting rid of any old gas by burning that first tank off, and then running a fresh tank through as well. The fresh fuel and the solvents help dissolve any potential varnish in the system, and help get it cleaned out. No matter if it is a carburetor, or a CIS system, gasoline comes into contact with moving parts inside of the fuel system that can stick, and you want to make sure they're able to move freely. This is particularly true with our fuel injectors as we increase line pressure and clean out any gum or varnish that could be causing other problems. Either a bad spray pattern, or preventing the injector from closing that causes a Hot Start issue.

Does it actually work?
Yes. A prime example are the two-piece spark plugs that Ford installed in many of their Triton engines (5.4L and all 3Vs). Carbon would build up between the plug and the cylinder head to get it stuck in place, and it would break off during removal. Performing this procedure before hand helps break down the carbon for easier plug removal.

Why is it called an "Italian Tune-Up"?
There are varying accounts, but it's believed that the name came about a long time ago because of Ferrari owners who did not drive their cars very often. They would start the cars up, but never get them up to temperature. Aside from accumulating moisture inside of the oil before it could be boiled-off, this would never heat the cylinder heads up properly, so the valves would stay relatively cold, and would build carbon up because it wouldn't fully burn. So as part of the maintenance procedure, Ferrari mechanics would run the engines at load for prolonged periods of time in order to burn the excess carbon off, and out of the engine.

Why Techron?
Techron seems to do a really great job on fuel systems, and especially those injectors. Most importantly, I've found it to be very gentle on the rubber components of the DeLorean's fuel system (I've used Berryman's B12 Chemtool before, and it did some damage). Allot of people will argue Sea Foam is better, but I wouldn't. I would instead say that it's more like a different tool for a different job. Techron for the fuel system, and Sea Foam more for the intake and combustion chambers (and Chemtool as a parts dip). But with regular use, and proper top-tier fuels, you shouldn't need to quite go that far.

[Jason]

The problem came back again.

Below is frequency valve duty cycle and control pressure during warm up from an over night cold start. The control pressure looks reasonable to me but the CPR response to vacuum changes seemed a little strange. When I revved the engine (starting at 5:20) the control pressure would stay high, drop when I let off the throttle, and the recover in a funny bouncy way. This is the first time I've ever looked at the fuel system pressure so I don't really know what's normal. Primary pressure was 70 psi.

You probably want to turn your volume down before playing the video. The engine sound was too much for the mic, so the audio is loud and distorted.


https://www.youtube.com/watch?v=zrE0jmxTC-Y

Right after taking the video above, without shutting off the engine in between, I took the car for a quick drive and had the low power / rough running symptoms, so I'm reasonably confident that the measurements were taken while things were in the failure state. It would be nice to be able to record control pressure while driving.

[Bitsyncmaster]

I think it took way to long to turn off the engine temp sensor (thermostat) which would release the open loop 60% duty to open loop 50% duty. So your wide open throttle switch, the temp sensor or the ECU must be bad. You can cut the light green wire going to the lambda ECU to find out if the ECU is bad.

It also took to long to go closed loop (although it was pretty cold this morning). That would suggest the O2 sensor, plugged exhaust (CAT) or ECU is bad.

I have a few spare ECUs which I could loan or sell if you decide the ECU is bad. Shipping would take a medium flat rate box.

[Jason]

I think it took way to long to turn off the engine temp sensor (thermostat) which would release the open loop 60% duty to open loop 50% duty. So your wide open throttle switch, the temp sensor or the ECU must be bad. You can cut the light green wire going to the lambda ECU to find out if the ECU is bad.

It also took to long to go closed loop (although it was pretty cold this morning). That would suggest the O2 sensor, plugged exhaust (CAT) or ECU is bad.

I have a few spare ECUs which I could loan or sell if you decide the ECU is bad. Shipping would take a medium flat rate box.

Would it tell me anything useful if I tried driving with the O2 sensor unplugged so the system stayed open loop?

[Bitsyncmaster]

Would it tell me anything useful if I tried driving with the O2 sensor unplugged so the system stayed open loop?

That should hold 50% duty when your light green wire is not grounded. If the ECU when to the changing duty cycle it would indicate the ECU is bad,

[David T]

That should hold 50% duty when your light green wire is not grounded. If the ECU when to the changing duty cycle it would indicate the ECU is bad,

You don't need a scope to test the Lambda ECU. Refer to D:04:15 in your Workshop Manual. It gives you the proper readings for:
O2 sensor disconnected
O2 sensor grounded
O2 sensor disconnected with 1.5 volts applied ECU. If you don't get the proper readings you check the wiring BEFORE deciding the ECU is the problem.
As for fuel system pressures, refer to D:02:01 for the Primary pressure and the chart for the Control pressures. If you are going to get out diagnostic equipment you should know what you are looking for and at. The most important piece of diagnostic equipment you can have is a brain to interpret what you are seeing. Put your scope away. It will only serve to confuse you.

[Bitsyncmaster]

You don't need a scope to test the Lambda ECU. Refer to D:04:15 in your Workshop Manual. It gives you the proper readings for:
O2 sensor disconnected
O2 sensor grounded
O2 sensor disconnected with 1.5 volts applied ECU. If you don't get the proper readings you check the wiring BEFORE deciding the ECU is the problem.
As for fuel system pressures, refer to D:02:01 for the Primary pressure and the chart for the Control pressures. If you are going to get out diagnostic equipment you should know what you are looking for and at. The most important piece of diagnostic equipment you can have is a brain to interpret what you are seeing. Put your scope away. It will only serve to confuse you.

I think you underestimate the OPs knowledge.

[David T]

I think you underestimate the OPs knowledge.

That may be so but I think is is over-complicating the problem with the scope.

[Jason]

With the O2 sensor disconnected and the car warmed up I got 50% duty cycle with the WOT opened, 60% with it closed, and the car was still running badly. So I think that rules out the ECU.

Then I decided to check the ignition timing. I found 26 degrees (or more) at idle. What?! Sticky vacuum or centrifugal advance maybe? I guess I'll try driving the car around until it starts "working" again and recheck the timing.

[Bitsyncmaster]

With the O2 sensor disconnected and the car warmed up I got 50% duty cycle with the WOT opened, 60% with it closed, and the car was still running badly. So I think that rules out the ECU.

Then I decided to check the ignition timing. I found 26 degrees (or more) at idle. What?! Sticky vacuum or centrifugal advance maybe? I guess I'll try driving the car around until it starts "working" again and recheck the timing.

Pull the vacuum hose off the distributor and see if that brings the timing back to 13 deg. If so, your vacuum solenoid may just be unplugged. If that two pin connector falls off it lets the engine get advance at idle with a warm engine.

[Jason]

Pull the vacuum hose off the distributor and see if that brings the timing back to 13 deg. If so, your vacuum solenoid may just be unplugged. If that two pin connector falls off it lets the engine get advance at idle with a warm engine.

That was it. The wire for the idle switch was loose at the vacuum solenoid connector. So my timing is fine but it still doesn't run right, so I'm back to trying to figure out the original problem.

[Bitsyncmaster]

That was it. The wire for the idle switch was loose at the vacuum solenoid connector. So my timing is fine but it still doesn't run right, so I'm back to trying to figure out the original problem.

Yes that would not make any difference with driving RPM but it really helps the idle better. It may help you see a better dwell reading now.

[Jason]

I pulled the spark plugs, and, well, besides "not good" what does this mean?

39194

I put in new plugs and the car runs beautifully, but I assume there must be something else wrong for the plugs to get that way. The plugs are 3 years / 23k miles old.

While the plugs were out I did a compression test and got, going from left to right in the picture (which are cylinders 4,5,6,1,2,3)

120, 120, 135, 110, 110, 100

[sdg3205]

Jason,

It looks like your injectors are all running drastically differently between rich and lean. Have you verified their flow via spray test?

Also, under what conditions and how did you run the compression test?

[Bitsyncmaster]

Those plugs look awful. I would suggest you now adjust the mixture again and check the plugs after driving on a few tanks of gas. You may have to do a long highway trip to clear out your combustion chambers. I wonder what your CAT looks like.

[Jason]

Jason,

It looks like your injectors are all running drastically differently between rich and lean. Have you verified their flow via spray test?
That's next on my list. I need to round up some glass jars. I also ordered an injector tester, so I'm waiting for it to arrive.



Also, under what conditions and how did you run the compression test?


Engine cold, removed all the spark plugs and the rpm relay, disconnected the + and - leads at the coil. Then for each cylinder attached the compression gauge and held the key on start for 3 engine revolutions.

[DMCVegas]

Avoid the platinum plugs in the future like the plague. Go for the BOSCH HR6's.

[David T]

The compression test should also include holding the throttle plates open all the way and "bumping" each cylinder the same # of times, usually 4-5 with a fully charged battery so it is spinning quickly. Do you know what kind of oil is in the motor and when it was last changed? Some of the plugs look like they are burning oil. After doing the compression test repeat it after squirting some oil in each cylinder to see if the rings are leaking. Those plugs look like they were in there a LONG time!