EFI conversion - VIN1049

[opethmike]

I am not about to retype the whole thing, so here is a link to the thread on the other forum:

http://www.dmchelp.com/showthread.php?2056-EFI-conversion-VIN1049

I'll be updating both sites with the same info each time I work on the car, in case people go to one but not the other.

Lots of good work done today:

Made a bracket and mounted the fuel pressure regulator. A little unnerving drilling into my car!
Mounted the vacuum advance solenoid, and run the vacuum lines and harness connector to it. Had to widen the mounting hole in, and used an M8 to bolt it to the manifold.
Mounted the idle speed motor and hooked up the air lines and harness connector
Mounted the relay board behind the rear parcel shelf
Mounted the ECU in the ECU area
Connected the ECU and relay board

Was going to drill & tap for the coolant temperature sensor, and begin wiring, but I ended up hanging out with my buddy Steve and driving his Viper for a few hours.

[opethmike]

Lots of good stuff done today:

Ran live 12v bulkhead terminal on passenger side
Ran switched 12v from white bulkhead connector for frequency valve
Ran ground to metal panel in ECU area
Ran wire from FIdle terminal output to post-diode of idle circuit wire; for controlling ISM/vacuum solenoid
Ran wire from coil (-) for tach signal
Wired intake air temperature sensor

One thing I did to avoid future headaches was both solder and crimp any connectors I used to the wiring.

[Farrar]

Looks great, Michael! I'm following this project with great anticipation.

Farrar

[Bitsyncmaster]

Are you using the FV with EFI?

[opethmike]

No, the frequency valve has been completely removed from the car. I am just using its old 12v wire.

[opethmike]

Swung by my mother's place (I keep the car there) on my way to the gym today. I took a few minutes, and played around with the mount for the throttle spool.

No reasonable way I'll be able to use the old one. The new one doesn't just drop right in, but I should be able to make it fit without too much hassle.

I find it interesting that so far its the small stuff like this that are the most aggravating/confusing/demanding parts of the project. I tell you what, if you want to learn how to improvise, convert your car to EFI! :)

[Farrar]

Have your silicone plugs arrived yet?

Farrar

[DCUK Martin]

No reasonable way I'll be able to use the old one. The new one doesn't just drop right in, but I should be able to make it fit without too much hassle.



With a little plate to mate the double holes to the single one on the top of the water pump, it'll fit fine. The main problem is of course that it rotates the other way. You need the throttle cable to enter from the rear/right meaning the cable needs to be longer. You will also need to bend up your throttle pedal *slightly* to give you full throw.

As custom work goes, it's pretty minor ;)

[opethmike]

Have your silicone plugs arrived yet?

Farrar

Supposedly a box from McMaster showed up at my office today, but I'm not at work, so I'll have to get them tomorrow.


With a little plate to mate the double holes to the single one on the top of the water pump, it'll fit fine. The main problem is of course that it rotates the other way. You need the throttle cable to enter from the rear/right meaning the cable needs to be longer. You will also need to bend up your throttle pedal *slightly* to give you full throw.

As custom work goes, it's pretty minor ;)

Cool, thanks Martin. Sounds like I just need to go buy a longer cable then.

[opethmike]

Martin, do you use just a longer inner cable, or do you use a longer outer cable as well? Do you have any pictures of the setup?


As far as progress goes, today:

I wired up the coolant temperature sensor
I wired up one bank of injectors

[DCUK Martin]

What I do with the cable is to have a little "joiner" made. It's nothing more than a piece of 1/4" circular stock, say 60mm long. Drill a 5mm hole into each end to a depth of 28mm, then drill a 2mm hole through the 4mm centre that's left. Then you can add a length of outer sheath and replace the centre cable entirely with a longer one. Much easier than making an entirely new assembly. I use a length of electrical heat shrink to seal up the join which ends up neatly behind the coolant bottle.

[Bitsyncmaster]

Model airplanes have a lot of flex cables that you can buy at a hobby store or on line. Lots of little fittings also to connect them.

[opethmike]

Thanks for the pointers, Dave & Martin. The throttle setup has been one of the more challenging parts for me.

Addressing the throttle will be one of the last things I take care of, but I will be sure to let you know how it turns out when I get there.

[Farrar]

Michael, you're doing exactly what I would (will?) do -- mounting all of the electrics inside the car. I did this with my new AT computer. Isn't it nice how much space there is back there? :)

FYI, those plugs go in easier than they come out. When I first pushed one of them in I thought "uh oh, this seems loose." However, the slight taper seems to be enough to hold a vacuum-tight seal, and the silicone doesn't get squishy when it heats up. (Hooray for science!)

Looking forward to seeing what you can knock out this weekend!

Farrar

[opethmike]

Accomplished tonight:

Wired second bank of injectors
Wired, calibrated, and installed wide-band oxygen sensor + controller

Hoping to get done this weekend:

Run MAP vacuum line
Plug old injector bungs
Make bracket and install TPS
Figure out TPS wires
Power on MS, and make sure that it sees all sensors
Calibrate TPS, and program FIdle output, make sure it works
Install Megatune & get communication with ECU
Perform initial configuration
Get car to start

Once I can get the car to start and idle reasonably well, I still will need to:

Wrap up wiring
Install throttle spool
Install air intake tube + air filter
Put the interior of the car back together
Fix the stripped body bolt nut behind the rear driver's side spring/shock assembly
Drive, tune, drive, tune, drive, tune

[opethmike]

Today:

Ran MAP vacuum line
Made bracket and installed TPS
Got Megatune communicating with ECU
Set FIdle spare output to provide ground to ISM & vacuum advance solenoid when throttle is closed. Verified it works.

Did basic configurations in Megatune. Was able to get car to start and run (somewhat rough) for about 15-20 seconds, then the car stalled. Further attempts to start didn't work - noticed first that I wasn't seeing any RPM in Megatune, and was confused. Then, noticed that my battery had run down too low - 11.8 resting, 11.1 during cranking.

Going to swap out the battery from my Saturn into the car tomorrow, temporarily. Once I can get the car to start and idle, and once I've charged the correct battery, I'll put the correct battery back in.

[opethmike]

No luck starting so far - datalogging shows that I'm not getting a tach signal (0 the whole way), but I do see sensor recordings, so I do have constant 12v. (I also checked voltage between ground and my 12v hot feed during cranking, its constant).

I have a shielded wire running from the tach terminal strip port on the MS ECU Relay board, to the "1" post on the ignition coil. I thought I should be able to get tach from there?

I should also note that the dash tachometer does show RPM while it is cranking.

[opethmike]

Datalog is attached. Pretty clear that I'm not getting tach! :angry: Also can see that the CLT and MAT are show readings during cranking, so I am indeed getting voltage to MS while cranking.

Rename the .txt in the zip file to .xls to view in Excel.

[opethmike]

Further diagnostics - I verified using my multimeter (reads tach) that I am using the correct post of the coil.

I also verified that I have continuity from one end of the wire I am using to the other.

While cranking, I hooked my multimeter to ground and to the tach port on the relay board after plugging the wire back in. The meter showed ~200 RPM while cranking, so RPM signal IS getting to the relay board.

However, datalogging shows only 0 or 1 for RPM. Aggravating.

I am running fuel control only, FYI.

Can't attach the data log for some reason. PM your e-mail address and I will send it to anyone interested.

[DMCMW Dave]

Get your tach pulse from the line inside the car, a handy place is the idle ECU wiring. It's the white/slate wire on the connector toward the inside of the car. This is probably close to where you have your MS ECU anyway.

[opethmike]

Thanks Dave, I gave that a shot. Unfortunately, still no luck. When I turn the car to run, the RPM gauge in MegaTune will show '1' briefly, then go to zero. Same thing during cranking.

I am now starting to post on the MS forums; will hopefully get some further information.

[DCUK Martin]

I have had the Adaptronic triggering directly off the distributor rather than the coil -ve.

[opethmike]

MS actually recommends tach signal from coil if doing fuel only control.

I don't think the source of the signal is the problem - using my multimeter, I can read tach signal from the relay board's terminal strip port for the tach signal. The relay board is a separate unit from the ECU; it centralizes all of the wiring.

So, the relay board IS getting tach signal, but it seems like it isn't reaching the ECU. Odd, very odd.

[dmc6960]

Trace the ignition circuit on the Megasquirt board. Perhaps you have a bad solder joint. Particularly pay attention to whatever jumper you used to identify tach select.

[opethmike]

Its a pre-assembled board, so I didn't build it. According to diyautotune.com's website, it comes jumpered for coil (-).

I currently have a stimulator on order to test with. It has to be something with the ECU because I can read the tach signal at the relay board's tach port on the terminal strip.

[AdmiralSenn]

How do you have your tach trigger set up? There are separate jumpers for the VR vs optical circuits, VR is for coming off the distributor and optical is for the coil terminal. Your unit is prebuilt so it may be set to VR as I think that's a lot more common. Open the case and check it out.

Not sure which board you have so the jumpers may not be where I'd expect, but the boards should be VERY clearly labeled.

From the manual:

52. Select the tach input circuit with jumpers:

For the VR sensor:

Jumper VRIN to TACHSELECT on the bottom side of the PCB (near the DB37, opposite the heat sink.)
Jumper TSEL to VROUT (Or VROUTINV if you want the VR input to be inverted) on the bottom side of the PCB, near the center.


OR (Do NOT install both sets of jumpers, chose one set or the other!)
For the Hall sensor, optical sensor, coil negative terminal or points:

Jumper XG1 to XG2 on the bottom side of the PCB, near the 40 pin socket,
Jumper OPTOIN to TACHSELECT on the bottom side of the PCB, near the DB37 connector, opposite the heat sink.
Jumper TSEL to OPTOOUT on the bottom side of the PCB, near the center.

[opethmike]

From the diyautotune.com website for the pre-built unit:

Hall Sensor/Coil(-) and VR Sensor Ignition Trigger Components (jumpered for Hall sensor, optical sensor, points, or fuel only using negative coil input- you can reconfigure for any input you like)

I opened the case, and I saw a jumper on XG1/XG2.

I couldn't find the optoin or tsel parts of the board.

[dmc6960]

Do you have a 3.0 board or a 3.57 board?

[AdmiralSenn]

From the diyautotune.com website for the pre-built unit:

Hall Sensor/Coil(-) and VR Sensor Ignition Trigger Components (jumpered for Hall sensor, optical sensor, points, or fuel only using negative coil input- you can reconfigure for any input you like)

I opened the case, and I saw a jumper on XG1/XG2.

I couldn't find the optoin or tsel parts of the board.

They may be on the underside.

Hmm, that is weird. Try setting your software trigger to basic trigger instead of fuel only? I had this problem once and that fixed it.

[opethmike]

Do you have a 3.0 board or a 3.57 board?

I have a 3.57 board.


They may be on the underside.

Hmm, that is weird. Try setting your software trigger to basic trigger instead of fuel only? I had this problem once and that fixed it.

I haven't tried basic trigger yet. If that works, I will be very happy. Thanks for the suggestion, I will try it after work today.

[opethmike]

They may be on the underside.

Hmm, that is weird. Try setting your software trigger to basic trigger instead of fuel only? I had this problem once and that fixed it.

I found them. All the jumpers are set correctly.

[dmc6960]

You didn't happen to be logging when you had the engine running the one time, did you?

[opethmike]

No, I didn't think to. But I think the only reason it ran then is because of fuel from the priming pulse; as it ran VERY briefly.

[opethmike]

I haven't tried basic trigger yet. If that works, I will be very happy. Thanks for the suggestion, I will try it after work today.

Matt Cramer from diyautotune.com has stated the following to me in an e-mail, so I am feeling optimistic about the basic trigger:

"I believe there's a bit of an odd code quirk with odd fire engines - you
have to put it in Basic Trigger mode when running fuel only, instead of
the usual Fuel Only mode. Let me know if this gets it synced - if not,
there's some input tests I can walk you through."

Can't wait to get done with work today so I can try it out.

[opethmike]

Went over to the car during my lunch break. Unfortunately, setting the ignition type to 'basic triggering' made no difference whatsoever. :(

Log and msq attached.

[AdmiralSenn]

Your CLT and MAT sensors are really noisy looking.. and yep, no RPM signal.

That's really weird that the relay board sees it but the ECU doesn't. I never used the relay board; how does the signal get from one to the other?

[opethmike]

What can I do to reduce the noise on those sensors?

The relay board and ECU are attached via relay cable -

http://www.diyautotune.com/catalog/megasquirt-relay-cable-p-47.html

[AdmiralSenn]

Where are they grounded? The grounds for those two and the TPS should be at the special sensor ground (pin 2?).

I'd honestly say you should just use a DVM and check continuity between the relay board and the ECU. Maybe just a bad solder joint or something. Or something off on the board itself - are you using the relay board to run the fuel pump, or anything else RPM based?

[opethmike]

Where are they grounded? The grounds for those two and the TPS should be at the special sensor ground (pin 2?).

I'd honestly say you should just use a DVM and check continuity between the relay board and the ECU. Maybe just a bad solder joint or something. Or something off on the board itself - are you using the relay board to run the fuel pump, or anything else RPM based?

There are ground ports for each individual sensor on the terminal strip of the relay board. There is one ground that comes off of the relay board, that I have grounded to with a screw through the metal in the ECU area of the car.

I have a stimulator coming in the mail tomorrow. My plan of attack for further diagnosis is:

Test ECU on stimulator
See if continuity is found in relay cable on pin 24
See if tach out pin 24 of relay board is giving RPM

[Bitsyncmaster]

There are ground ports for each individual sensor on the terminal strip of the relay board. There is one ground that comes off of the relay board, that I have grounded to with a screw through the metal in the ECU area of the car.

I have a stimulator coming in the mail tomorrow. My plan of attack for further diagnosis is:

Test ECU on stimulator
See if continuity is found in relay cable on pin 24
See if tach out pin 24 of relay board is giving RPM

I would not consider the metal plate in the ECU area to be a good ground unless you added a good ground to it.

I would recommend a 14 AWG wire from the metal plate to the grounds your using in the engine compartment (Engine block?).

[opethmike]

I would not consider the metal plate in the ECU area to be a good ground unless you added a good ground to it.

I would recommend a 14 AWG wire from the metal plate to the grounds your using in the engine compartment (Engine block?).

Good idea, I'll do that.

[AdmiralSenn]

Yeah, those sensor grounds are supposed to be through the ECU or to a REALLY good ground, like the block or battery.

See pin 19 here: http://www.megamanual.com/ms2/v3ext_wiring.gif

Grounding things improperly is one of the easiest ways to mess up an installation.

I look forward to your stim test results. I can't really think of anything else offhand.

[opethmike]

One of the first things I will be doing after work is relocate the ground to the engine block. I'll let everyone know if it makes any difference.

[opethmike]

Would it matter at all that I am using a non-stock ignition coil?

[dmc6960]

Did you change it at the same time?

[opethmike]

No, I've been running this same coil for almost four years now. The exact coil I am using is:

http://specialtauto.com/delorean-parts/images/ign-coil.alone.jpg

I am using it with a single ballast resistor.

[dmc6960]

Good for you that you've been able to run it, but I had NOTHING BUT PROBLEMS when trying to use the exact same coil. I dont honestly think its related to your issue, but I'll always have that nagging feeling when anybody uses non-stock ignition components.

[opethmike]

I think I still have the original components lying around in a box in the attic. I'll see if I can dig them up.

[opethmike]

Just received the stimulator in the mail. Hopefully this will help narrow things down.

[Bitsyncmaster]

Would it matter at all that I am using a non-stock ignition coil?

Depends how good the MS input circuit was designed.

If you have higher current than your low voltage ("points closed") may be above the MS designed threshold. Even with the stock setup that voltage is between 1 to 2 volts. That is due to the wiring and the ignition output transistor voltage drops.

[Farrar]

Just received the stimulator in the mail. Hopefully this will help narrow things down.

That's what she said.

(Well, somebody had to say it!)

Farrar

[opethmike]

Yep, I giggle a little every time I read the name of the thing. It gets even better, the specific product name is "MegaStim".

[Farrar]

Yep, I giggle a little every time I read the name of the thing. It gets even better, the specific product name is "MegaStim".

First, the Mega Stim ... then, the Mega Squirt.

Methinks whoever designed the system had a naughty streak.

Farrar

[opethmike]

Depends how good the MS input circuit was designed.

If you have higher current than your low voltage ("points closed") may be above the MS designed threshold. Even with the stock setup that voltage is between 1 to 2 volts. That is due to the wiring and the ignition output transistor voltage drops.

Interesting point. This is the sort of thing I wouldn't think of, as electronics are my automotive weak point. Would reading voltage during cranking be helpful at all?

[Bitsyncmaster]

Interesting point. This is the sort of thing I wouldn't think of, as electronics are my automotive weak point. Would reading voltage during cranking be helpful at all?

Actually with the ignition on and the engine not running then the coil is energized (points closed). So you can read the voltage on the (-) side of your coil to the ground your using. Normally you would read 1 to 2 volts. If your reading 3 or 4 then you would want to ask on the MS forum if that would be a problem.

Edit:
Probably better if you read it at your RPM input terminals since that is where you you sample the signal.

[opethmike]

With the new coil, I have ~1.6 volts. With the original coil + ballast resistor, I have ~1.8.

I moved the ground for the relay board to the engine block.

I am getting VERY confounded at this point. Today I checked the following:

The relay board has a good ground on the engine block.

The RPM gauge shows input when on the stimulator, but ONLY if I use "fuel only". With "basic triggering", it is either zero, or when I turn the knob, all the way up with very erratic readings.

Opened the case and saw that XG1/XG2 are jumpered to each other. JP1 is 2/3 jumpered. J1 is 1/2 jumpered.

The continuity on pin 24 of the relay cable is good.

It seems like everything check out at this point?

[dmc6960]

New thought this morning....

You said your multimeter says ~200 while cranking? Well the default cranking speed in MS is 300. Try changing it to 200.

[opethmike]

New thought this morning....

You said your multimeter says ~200 while cranking? Well the default cranking speed in MS is 300. Try changing it to 200.

Good thought, I will definitely give that a try.

[opethmike]

Jim & AdmiralSenn,

Where are you guys getting your tach signals from?

-Mike

[dmc6960]

RPM relay socket.

[AdmiralSenn]

Tapped from the sensor in the distributor right next to the ignition computer. Toying with switching to coil negative from the same area just to see if it's better.

[opethmike]

Good news! It starts and runs.... sort of.

This morning, there was a new firmware release. So I upgraded to that. Took a little bit of fiddling, as the new release REQUIRES TunerStudio, not Megatune.

The very first thing I did today, for shits and giggles, was to change the wire I'm running to coil negative to an unshielded wire from a shielded one.

Amazingly enough, after some cranking, the car started and ran quite roughly for a bit. I'd say less than 30 seconds or so. I sat there stupefied that it was actually running, and it stalled before I had time to go back to the engine to play with the throttle. (The pedal is currently disconnected.)

Then, I realized that in the tune I was using, that I hadn't configured the FIdle to provide ground to the idle speed motor and vacuum advance solenoid yet. So I addressed that, so that the ISM would open like it should.

Went to start the car again, and it started up a little more quickly this time. It kept running, still rough, but not nearly as bad as before. Presumably because it was getting air from the ISM.

Opening the throttle quickly would cause the car to want to stall. If opened it slowly, it would rev nicely until about 4k RPM or so when it would want to stall again. After running for several, I'd say 5-10, minutes, I shut the car off.

I couldn't get the car to start again at the point. I checked the battery with my multimeter, and it was only 11.2 volts! So, I had run it down too far to continue. I am charging it over night and will get back to it tomorrow or Friday.

I took data logs of two starts and runs, and saved my tune as well. For anyone interested, they are attached in a zip file.

[opethmike]

You'll probably notice in the logs that the RPMs are, um, a bit nutty. The dash tachometer showed reasonably steady readings while it was running.

[DCUK Martin]

Engine RPM/position is THE key signal for any ECU. It will act according to what it's seeing, and if what it's seeing is screwy...

I wish you'd decided to go with the Adaptronic, you'd be up and running by now :D

[Bitsyncmaster]

I could see where a shielded wire on the coil "-" could cause a lot of problems. Even though the frequency is pretty low at a few hundred HZ the edges (rise and fall times) are pretty quick. Also I've measured the inductive spike (which you must have to run) is about 130 volts.

Shielded wire adds a lot of capacitance and slow down the rise time on that inductive spike.

[opethmike]

Engine RPM/position is THE key signal for any ECU. It will act according to what it's seeing, and if what it's seeing is screwy...

I wish you'd decided to go with the Adaptronic, you'd be up and running by now :D

Heh, I hadn't even heard of it before I had already purchased my MS :-)

Oh yeah, I understand that correct RPM is important, but that's the next issue to resolve. The really infuriating issue of the car not even starting at all has been seemingly resolved, so I consider today a victory.

Dave - Yeah, I am not much of an electronics guy, but I figured, what the heck, a different wire was just about the only thing I had tried yet.

[opethmike]

Just looked up the Adaptronic - holy balls, its expensive. Though, I suppose you get what you pay for! 8)

I'm going to stick with Megasquirt for now, but if I ever decided to change ECU, its certainly an alternative to consider.

[20098]

I've been watching and reading this thread with great anticipation of you getting the car running - congratulations!

I did the Eagle 3.0 conversion and went with the MegaSquirt 1 with the 3.57 board. I've been where you are with wiring issues and the tuning aspect. With great patience and the help of Gregg Zilb, I have a well tuned and running 3.0 DeLorean.

It is a great feeling knowing that you were the one to get your project through to completion. Good luck with this. Looking forward to reading more about this project.

[opethmike]

Thanks for the encouragement Josh, I appreciate it. Though the car isn't close to drivable yet, it is a HUGE relief that I was able to get it to start several times today. That means progress, and progress is good.

[dmc6960]

Congrats! I just spent this whole evening building a brand new tune from scratch as I decided to try different fueling settings. Computer hasn't quit on me yet since my repair, have a good 3-4 hours on the engine now, so it's ready to drive to work again. Having a good stable working wideband allowed me to tune as much in this one evening than I did manually for a whole month last year when I first got it up and running. Best of luck with everything moving forward.

[DCUK Martin]

Just looked up the Adaptronic - holy balls, its expensive.

It is? Compared to something like a TEC3, it's peanuts (or it was when I originally researched it!). I think it's the other way around.... MS is very inexpensive

Edit: their range has changed now, the e420c has evolved into the e420d for a bit more money, but advanced features such as an internal map sensor

[opethmike]

I think it's the other way around.... MS is very inexpensive


You know, I think that's a better way to put it!

[opethmike]

Congrats! I just spent this whole evening building a brand new tune from scratch as I decided to try different fueling settings. Computer hasn't quit on me yet since my repair, have a good 3-4 hours on the engine now, so it's ready to drive to work again. Having a good stable working wideband allowed me to tune as much in this one evening than I did manually for a whole month last year when I first got it up and running. Best of luck with everything moving forward.

Are you using VEAnalyze/Autotune?

[dmc6960]

I could never get autotune to work well at all. I'm logging everything for 5-10 minutes, then doing VE Analyze in MegaLogViewer. Drive a little, retune, drive a little, retune, etc and so on for the whole evening. Then I keep doing it on my drives to and from work. You still need to manually manipulate your VE table enough to keep the mixture from going too lean/rich in most of the low load powerbands. Then you gradually expand the envelope of variables for the engine. A good starting point for your A/F mixture is to set the whole band to 13.5 or so. Autotune may actually work for that, dunno. Then again, your running MSII with newer software than I have. When you get the engine running pretty solidy at the one mixture ratio, go back and edit your mixture table to lean out the medium/low loads to 14.5 and 15.0 respectively, and your really high loads to about 13.0 or so. You want to avoid quick swings in the set mixture ratio in ranges close together. When crusing at 70mph I was swinging between 15afr and 13afr because I had my VE table switching right around that RPM/kPa level. End result would be some bouncy operation. On my new tune I'm changing it much more gradually.

[opethmike]

Well, interestingly enough, without having changed anything whatsoever, other than plugging the MAP hose back in, the car went from starting twice yesterday to never starting today. WTF?

Some interesting observations:

While cranking, the live gauge viewer under the tuning menu doesn't show anything reading; including my infamous RPMs not showing up. If I crank for long enough, it goes 'offline' - I thought that maybe I didn't have a great 12v source (post behind the parcel shelf), so I hooked it straight to the battery. Same behavior.

If I use some start fluid, and get it running for a few seconds, everything including RPMs show up correctly. So I'm not sure what to think right now?

For shits and giggles, in case my problems are a settings issue, could I possibly ask someone to send me their .msq?

[opethmike]

So this Saturday, the car will have been down for a month. If I can't get it to consistently start by mid-next week, I'll be reverting to K-Jet for the rest of the summer so that I can drive the car, then trying the conversion again over the winter.

[DMCMW Dave]

IIRC his original reason for doing all this was not because his K-jet had any problems, more like "It's a science experiment".

[opethmike]

IIRC his original reason for doing all this was not because his K-jet had any problems, more like "It's a science experiment".

That is correct.

[Farrar]

I think the fact that Mike is documenting everything here in great detail will be, when complete, the closest thing to a "how to" for EFI conversion. I for one am taking notes. :) Keep up the good work, Mike! I look forward to a successful conclusion!

Farrar

[DCUK Martin]

One thing about the approach Mike's taking currently is that in using the coil -ve as the ECU's trigger, there's no provision for controlling the ignition from the ECU somewhere down the line without re-working the trigger. For batch-fired injection, all the ECU needs to know is "how many cylinders". I know Mike's taken this approach with eyes wide open, so I'm not criticising him but as a general "how to", it's a limited approach.

Personally I fit a later bellhousing and flywheel with integral trigger - it's just a way of avoiding a trigger wheel which always scream "aftermarket" to me. However a flywheel-based trigger alone won't give the ECU enough information to run ignition on an odd-fire. My Adaptronic setup still uses the stock ignition system, feeding directly into the original distributor connection.

If I were to do a EFI odd-fire, I'd probably adapt the distributor into a trigger wheel OR put a trigger wheel on the end of the camshaft (there's a seal available from Renault that was used on the cars which have a distributor next to #1)

[opethmike]

Seriously, can you guys get your EFI/Carb debate crap out of here? This thread is for MY SPECIFIC CONVERSION.

If this EFI/Carb debate crap continues in MY THREAD FOR MY SPECIFIC CAR, I will stop updating this thread and forget that it exists.

On that note, I will add that I very much appreciate Martin's continued efforts to try to help me trouble shoot my EFI setup.

[sean]

Seriously, can you guys get your EFI/Carb debate crap out of here? This thread is for MY SPECIFIC CONVERSION.

Just did some house cleaning. I guess I need to follow this thread better. Keep this on topic, which is Mike's conversion.

[opethmike]

Thanks Sean.

[dmc6960]

Ok, here is my MSQ. Couple things to note...

This is the tune I have been running on the car most of the year so far. It is not the new one I just started a few days ago. It runs MY car quite well. Starting and warmup are not perfect, but work sufficiently to start and drive the car above 50º. I typically start my car by turning the key on, then letting it prime for 1 second then continuing the key to crank. Assuming your RPM relay primes with ignition, this should be the same for you.

This file is from MS 1 Extra, NOT MS 2/Extra. I do not know if it will load without problems on a MS 2 system.

This file is setup for PWM on my low impedance injectors, in theory they should be identical to the injectors you have as well.

Squirt method is 6/alternating. It has worked well for me, but many other tuners strongly suggest 2/alternating, which is what I started to setup with the new tune.

The VE table should be nowhere near what you need, since I have different compression, camshafts, and exhaust characteristics. Also my idle kPa is around 68-70, MUCH higher than normally cammed engines. I do not know how exactly it will idle for you.

If you are seeing fluctuating RPM readings when your finally running, that's because your looking at the "high res" RPM gauge, which cannot do Odd-Fire averaging. You need to view the low res, RPM/100 gauge. (I think the gauge is simply called "Engine Speed").

This file also has cranking speed set at the default 300rpm. You should change that to 200 since that's what you observed with your meter.

Over the weekend I may be checking online less frequently, since I only have my phone to do so. Worst case scenario I wont see your progress reports until Monday morning. However tomorrow my parents are taking both my kids, and my wife is working a garage sale at her mom's. I THINK that is going to leave me alone to do whatever I want (which will either be yard work or DeLorean play). If you would like to call for advice, my area code is 763, prefix is 639, and the last four are my VIN number.

Good Luck.

890

[opethmike]

Same post is going up on the other forum.

The following are the facts put together today:

The injectors squirt for the prime pulse. I pulled the rails and put the injectors into bottles. They 100% for certain spray during the prime pulse.
The injectors DO NOT squirt during cranking.
Spark is good. Car runs for a few seconds on starter fluid.
During cranking, I can use my multimeter to read RPM directly off of the TSEL jumper on the ECU board.
During cranking, RPMs do not register in the live gauge monitor, and noticeably, nor does anything else.
When the car runs for a few seconds on start fluid, RPMs (and all other info) shows up in live gauge monitor. However, the injectors still aren't firing at this point as the car stalls as soon as the starter fluid runs out.

The relay board has three pads related to power on it, that I have run as follows:

Ground - I run ground to the engine block - passenger side valve cover with a bolt.
12V battery - I run this to the positive battery post behind the parcel shelf; on the passenger side. Currently, I have it straight to battery.
12V switched - I ran a wire to the bulkhead connector area in the engine bay, and tapped into the green/brown wire off the red connector.

[AdmiralSenn]

Hmmm. I didn't realize your tach wire was unshielded to start with - that would definitely hurt things.

If it's still not firing injectors I can pretty much guarantee that it's the RPM signal. I just don't know what's wrong with it. Your "shorter log of running" shows no RPM signal. Your "start with tach wire" log shows that MS THINKS it's opening the injectors - you have a non-zero pulsewidth and RPM signal. In that case it might just need an adjustment to your basic fuel settings - check out Jim's post and copy the basics (Req_fuel and such) and try it out.

I wonder if you got a bad board or something... I hope not.

[dmc6960]

Have you changed the cranking speed yet?

[opethmike]

Yes, that did nothing.

[opethmike]

Hmmm. I didn't realize your tach wire was unshielded to start with - that would definitely hurt things.

If it's still not firing injectors I can pretty much guarantee that it's the RPM signal. I just don't know what's wrong with it. Your "shorter log of running" shows no RPM signal. Your "start with tach wire" log shows that MS THINKS it's opening the injectors - you have a non-zero pulsewidth and RPM signal. In that case it might just need an adjustment to your basic fuel settings - check out Jim's post and copy the basics (Req_fuel and such) and try it out.

I wonder if you got a bad board or something... I hope not.

I've tried both unshielded and shielded tach wires.

The tach signal absolutely gets to the board; I can read it from the TSEL jumper during cranking. But for some reason, it doesn't get logged during cranking. When the car does run for a few seconds on starter fluid, RPMs do show up in the live gauge viewer.

Something is happening during cranking (I think) to mess me up.

I'll give Jim's injector characteristics a shot.

[AdmiralSenn]

Have you changed the cranking speed yet?

Isn't cranking speed just an upper limit that defines what is and isn't cranking? E.g. if your engine idled at 300 RPM you wouldn't want that as cranking RPM because it'd be firing extra fuel all the time.

Mike, do you have any noise filters enabled in your settings? If so, try turning them off. I'll load up your MSQ on my tuning PC tomorrow or Sunday and see if I can spot anything.

[opethmike]

Nope, no noise filtering.

I'll be honest guys, if I can't at least make some logical progress this weekend, K-Jet is going back on next week so I can drive the car. Then I will try the MS conversion again over the winter.

Either way, once the car runs again, I will have to remove the driver's side rear spring & shock to fix the captive nut for the body bolt back there. Getting tired of working and not driving. ;)

[DCUK Martin]

I've been pondering your problems this morning while trying to fix the window mechanism on our little Smart roadster. (you thought DeLorean window mechanisms are costly/complicated....?!)

Let's look at this logically.

Your ECU needs an RPM signal. The "brain" needs to "see" this signal in order to "do" anything. Adjusting crank setting will matter little if the ECU does not "see" the engine turn in the first place. It doesn't matter that you can see the signal on the board, if the brain doesn't read the signal, it won't work.

Why doesn't the brain see the signal?

I'll disregard the possibility that the board/ECU is faulty

1) You're not giving it a signal it can "see"
2) You have a wiring fault
3) something within the software is checked that prevents it from registering the engine speed.
4) It's a problem *while cranking*

I would do this

1) Can you trigger it directly from the distributor's pickup?
2) I think you've eliminated that
3) You're on coil -ve which means you're getting 3 pulses/rev. Is your software configured for a 6 cylinder/4stroke engine and does it know it's on the coil?
4) What voltage is the ECU "seeing" during cranking, and how is everything fed/grounded. Just for the hell of it, stuck a meter across the power supply to the ECU while cranking - it could be that daft. And I presume the ground is used to fire the injectors so that has to be SOLID.

[20098]

I too had an issue where the RPM signal wasn't being seen. My engine conversion to the Eagle 3.0 uses a trigger wheel built into the distributor with an optical sensor that tells the ECU when to fire. The optical sensor was a 5v sensor, but needed to be modified with an inline resistor (510 ohm, IIRC), to condition the signal, along with another resistor (1K ohm) between the signal-in line and the switched power-in line. Think of the letter "H" where the resistor is between the to vertical portions of the letter. As it turned out, the 510 ohm resistor broke and wouldn't power the sensor. It drove me nuts for weeks, and the sensor is a bitch to remove. I went through three sensors, testing if they worked with straight 5v power, before I realized that the 510 ohm resistor was inline, keeping it from frying it. Eventually, I went with a 12v sensor, ditched the 510 ohm resistor and the MS saw the RPM signal...I digress.

MegaSqurit can be configured (internally) many different ways. I'm not certain how your are triggering the MS from the distributor, but you many need to add the 1k Ohm resistor to the lines, if it wasn't already built into the MS board. My MS was modified with the 1k ohm resistor already, and unknowingly, I added another one to the lines as described above in the "H" form. I was told I could leave it in without harm and did so. So far so good with my sled.

I would check the wiring from the relay board to the MS inputs, since you said you were getting signal to the relay board. I can only imagine that the MS, itself, just isn't seeing the RPM signal because of some internal setup.

This link may be helpful, though I know you're using MS2:
http://www.msextra.com/doc/ms1extra/MS_Extra_Ignition_Hardware_Manual.htm#hallV33

Check out page three, if you're using the 3.57 board: http://www.megamanual.com/357/MS357.pdf

and again, this link for the 3.57 board: http://www.megamanual.com/357/intro.htm

I initially purchased a pre-built (by DIYAutotune) MS2 and upgraded to the Extra code, but never had any luck with it running the Eagle 3.0 all that well. I probably didn't spend enough time figuring it out. Since Gregg Zilb was running the MS1 with the Extra code, and was willing to help me tune the car, I decided to "down-grade" my setup. I'm not certain if MS1 can run odd fire, but I am very pleased I went with that route.

Once you do get it sorted out, you'll be pleased. It takes trial and error. My car was off the road for a bit over 2 years. The process of the conversion took not time at all, it was getting it to run, run right and then become drivable that took most of the time.

Don't give up and go back to K-Jet. You need to spend the time with the system in place to figure out what is going on. I'm not sure where you live, but being in Connecticut, the winter months offered me no time to try tuning.

I'm on the same page as Martin is with your problem.

If I think of anything else, I'll post it.

[opethmike]

Martin and Josh,

I get the signal from coil negative. Wire from the relay board directly to coil negative.

Ground is the passenger side valve cover. Very clean and tight bolt holding it down. The connector to the wire is crimped AND soldered. 12v battery power is currently directly from the battery itself. Switched 12v powered is from the green/brown wire off of the red bulkhead connector.

The RPM signal does indeed reach the board, as I have already noted. I can read it right off of TSEL. If I get the car started with starter fluid, the RPMs display just fine in the tuning software for the few seconds the car runs. It is only during cranking that RPMs do not seem to register in the software. HOWEVER, the RPMs CAN BE read from the board itself with a multimeter during cranking.

The software knows it is a 6 cylinder, 4 stroke, odd fire engine.

Please don't say that I'm 'giving up' by considering TEMPORARILY going back to K-Jet; that's not fair. My number one want out of the car is driving it. The car has been down a month now, in the very short prime driving season here in Rochester. My winter schedule is no issue for working on the car. I have a good garage available that I can put it in for the winter.

Tuning itself can wait until the spring when I can drive on the roads. However, my issue is getting the car to start, not anything beyond that. I have no issue whatsoever spending lots and lots of time tuning the car, but I am not even at that point yet. I have been spending two weeks trying to even get it to start, and everything I am checking out/testing has been turning out to be fine.

Thanks for the assistance guys, I appreciate it.

-Mike

[Bitsyncmaster]

Over on DMChelp I questioned your green yellow wire.

[20098]

Mike,

Red bulkhead connector #12 is the Green/Brown wire for the reverse light switch to reverse lights. Power to this wire comes from fuse #13 via the green wire. This is fused to the accessory relay. Power is cut to this relay when in the start (cranking or #3 position). You want to pull power from either fuse #1,3,4 or 5 (from the main relay).

I made the mistake of using power from the accessory relay and couldn't figure out why I wasn't getting an RPM signal (along with the other resistor issue). I tapped into a power wire from accessory relay and during cranking (start/position #3), there was no power to the line.

This might be the cause.

[20098]

Wiring diagram:
http://www.cortex.biz/schema/schema_V1.31.pdf

[opethmike]

Mike,

Red bulkhead connector #12 is the Green/Brown wire for the reverse light switch to reverse lights. Power to this wire comes from fuse #13 via the green wire. This is fused to the accessory relay. Power is cut to this relay when in the start (cranking or #3 position). You want to pull power from either fuse #1,3,4 or 5 (from the main relay).

I made the mistake of using power from the accessory relay and couldn't figure out why I wasn't getting an RPM signal (along with the other resistor issue). I tapped into a power wire from accessory relay and during cranking (start/position #3), there was no power to the line.

This might be the cause.

Very good tip. I'll give that a shot.

[20098]

Bulkhead connector diagram (by Dave Delman):
http://www.dmc12.info/DeLorean%20Bulkhead%20Connectors.pdf






























or the DMCH version that doesn't give credit to Dave Delman
http://www.delorean.com/pdf/bulkheadconnectors.pdf

[DCUK Martin]

I made the mistake of using power from the accessory relay and couldn't figure out why I wasn't getting an RPM signal (along with the other resistor issue). I tapped into a power wire from accessory relay and during cranking (start/position #3), there was no power to the line.

This might be the cause.

Snap - I made this mistake (and I believe I called it a couple of pages ago :D :devil:)

Edit: That appears to have been in a post deleted due to my love of mud-slinging...... (sorry)

[opethmike]

Heading out the door now.... we'll see guys :-)

[opethmike]

Thumbs up to Josh and Martin; they were right. My choice of source for the 12v switched input was poor - I changed it to a tap into the brown wire of the RPM relay.

I now run the 12v+ to the battery post, and the 12v switched is a tap into the brown wire on the RPM relay. That, and setting the ignition input trigger to 'fuel only' has Megasquirt registering RPMs during cranking.

The car will fire and run for about a second to a second and a half now. And it will do that each and every single time I try. This is huge progress.

Attached are several logs from today.

[20098]

Great news!

[opethmike]

Yes, yes it is :-)

Time to re-read the documentation to see what I can do about getting it to stay running long enough to get the idle settled in.

[20098]

Checking on one of your log files, I don't see any movement on the TPS. Did you calibrate the TPS? It may only run for a few seconds because it is using the fuel from the priming pulse.

On my throttle body, I have an adjustment screw that opens the throttle plate a bit. It registers about 10%. Although the IAC will open enough for the warm up period, the throttle needed to be open to get everything to work properly, or so I believe (regarding my setup). If the throttle plate was fully closed, it would kill the engine once warm because the IAC closed off the air flow.

I'm not sure on the manifold you are using, but there should be some type of adjustment screw or a way to manipulate the throttle to be slightly open. This should allow the injectors to open and fire accordingly. I was under the impression that the injectors would only inject fuel if the MS can sense a TPS signal greater than 0.Though I may be completely off base on this function!

[opethmike]

Checking on one of you log files, I don't see any movement on the TPS. Did you calibrate the TPS? It may only run for a few seconds because it is using the fuel from the priming pulse.

On my throttle body, I have an adjustment screw that opens the throttle plate a bit. It registers about 10%. Although the IAC will open enough for the warm up period, the throttle needed to be open to get everything to work properly, or so I believe (regarding my setup). If the throttle plate was fully closed, it would kill the engine once warm because the IAC closed off the air flow.

I'm not sure on the manifold you are using, but there should be some type of adjustment screw or a way to manipulate the throttle to be slightly open. This should allow the injectors to open and fire accordingly. I was under the impression that the injectors would only inject fuel if the MS can sense a TPS signal greater than 0.Though I may be completely off base on this function!

Thank you for reviewing my logs.

Yes, the TPS is calibrated. I didn't touch the throttle at all during cranking/brief run. I actually don't even have the throttle pedal linked up to the throttle right now. I can try holding the throttle open slightly while I start the car tomorrow.

[20098]

Sounds like a plan.

In either case, even if the throttle cable isn't hooked up, I believe that the throttle plate should be slightly open. In theory, if the TPS shows a reading, beyond closed, it allows the MS to allow the injectors to open. The injectors will also open when the MS senses an ignition event. You don't want to manipulate the throttle while cranking, unless you are trying for "flood clear" mode.

Did you retain the RPM relay, or did you modify it? I have the following modification, using a 4 prong Bosch relay (I used the hot start relay holder and a suitable relay):
Green from RPM relay terminal #15 to terminal #86
Brown from RPM relay terminal #30 to terminal #30
White/Purple from RPM relay terminal #87 to terminal #87
MS DB37 connector #37 (purple wire) to terminal #85

Refer to the fuel injection circuit on page 56.

[opethmike]

The RPM relay is in place. I have simply tapped into the brown wire going into it.

I won't have too much time with the car tomorrow, but I'll report back with whatever I come up with.

[Bitsyncmaster]

The RPM relay is in place. I have simply tapped into the brown wire going into it.

I won't have too much time with the car tomorrow, but I'll report back with whatever I come up with.

The brown wire on the RPM relay is always +12 volts (fuse #7). You would want the green wire for switched power. You may run your battery down depending on how much current the MS draws keeping it powered always.

[opethmike]

Good tip, Dave. I'll be moving the tap to the green wire tomorrow.

Didn't work on the car today; helped my friend re-assemble the cylinder head for his 89 Mercedes 300E.

I am planning on the following tomorrow:

Relocate brown wire tap to green wire tap
Move the live 12v to MS from the battery back to the positive post in the parcel shelf
Change my tachometer signal wire back to shielded
Finish the second half of the bracket for the TPS, pretty important as it can move a little right now

[opethmike]

Yeah, the TPS bracket not being complete is definitely NOT a good thing. I just played back one of my logs - The TPS shows '-1%'; I can't imagine that helping anything! :D

[opethmike]

Today's work:

Wrapped up the wiring in the engine bay nice and neatly. That was really 'fun'.
Installed the air intake tube. Going to get an adapter and a filter tomorrow.
Moved the brown wire tap to the green wire. This is the correct choice for the switched voltage.
Changed the tach wire back to a shielded wire.

Car will crank and fire, but die immediately. Slightly cracking the throttle will make it take much longer to fire, if it does at all, so it is not too much fuel being injected.

I played around a bit with the cranking pulse, ASE and WUE but couldn't get the car to stay on for more than a second or two. So I'd like to ask for input based on my attached msq and log, please.

[Spittybug]

My required fuel is 18.0

What kind of injectors do you have? My settings are very different, so I presume your's must be low impedance? My PWM current limit is 100%. I THINK THIS IS YOUR PROBLEM.

With the exception of the hump, my cranking pulse looks identical to yours.
Likewise for ASE, and taper.

Your warmup is way higher than mine though, I go nearly linear between 117% to 100% from coolant temp of 0 F to 190 F.

Hope it helps. Hang in there.

I just looked through your thread again and couldn't find any reference to your injector impedance, only Jim's comments that lead me to believe his and yours are both "low". Even so, 18% is very low! Try bumping up. The manual says:

If you are running high-impedance injectors (greater than 10 Ohms), then set:

Opening time to 1.0 ms
PWM Time Threshold to 25.5 ms
PWM Current Limit (%) to 100%.
If you have low impedance injectors (less than 4 Ohms), set the:

Opening time to 0.6ms
PWM Time Threshold to 1.0 ms
PWM Current Limit to 30%
(75% if you are running an old style V2.2 mainboard which not advised).


I found getting the injector characteristics just right to be critical in solving my initial start up problems.

[opethmike]

My required fuel is 18.0

What kind of injectors do you have? My settings are very different, so I presume your's must be low impedance? My PWM current limit is 100%. I THINK THIS IS YOUR PROBLEM.

With the exception of the hump, my cranking pulse looks identical to yours.
Likewise for ASE, and taper.

Your warmup is way higher than mine though, I go nearly linear between 117% to 100% from coolant temp of 0 F to 190 F.

Hope it helps. Hang in there.

I just looked through your thread again and couldn't find any reference to your injector impedance, only Jim's comments that lead me to believe his and yours are both "low". Even so, 18% is very low! Try bumping up. The manual says:

If you are running high-impedance injectors (greater than 10 Ohms), then set:

Opening time to 1.0 ms
PWM Time Threshold to 25.5 ms
PWM Current Limit (%) to 100%.
If you have low impedance injectors (less than 4 Ohms), set the:

Opening time to 0.6ms
PWM Time Threshold to 1.0 ms
PWM Current Limit to 30%
(75% if you are running an old style V2.2 mainboard which not advised).


Howdy Owen,

Yes, my injectors are low impedance. The 18% figure I am using is from his .msq, as he and I have the same manifold from Martin, and probably the same injectors. I'll try bumping it up to the documented 30.

My warm-up and ASE are high in the tune I attached from me playing around, suspecting those. Since they were starting to get high, that's when I decided to call it quits and ask for help :-)

I appreciate the continued help. I think I will have time to play with it more after work today. I will let you all know how it goes.

-Mike

[opethmike]

Set the the current limit back to 30% from 18%.

The higher I set the req_fuel, the quicker it would start, and the longer it would run. Unfortunately, I would have to set it to some astronomical value (say, 40-50!!!!) with 1 squirt to get it to stay running for more than 2 or 3 seconds.

Playing with WUE and ASE didn't seem to help when I had reasonable req_fuel values (18-22).

I left WUE and ASE alone with the sky high values. It would run, but it would lug pretty heavily, like it had a hot cam, for anywhere from 30 - 90 seconds, then stall.

Log and MSQ attached.

[opethmike]

I'm pretty convinced at this point (thanks Marc!) that is running uber-lean because of a vacuum leak.

Suggestions on tracking leaks down would be appreciated.

[dmc6960]

I didn't get a chance to finish looking at your MSQ or log earlier today, but from what I did notice was that you had the injection set at 2/alternating. But then I noticed that "bank 2" was turned off. I don't have this funtion on my MS 1, but if one bank is not firing it would certainly explain some of the problems. Check both inj1 and inj2 for pulse-width and duty cycle in your log. They should be identical.

[dmc6960]

Another point about the injector PWM settings...

The procedure I read was to START at 30%, then work down one step at a time until the idle falters, then increase it about 3%. When I did this I started noticing a falter at 12%. I had it at 15% for a while then increased it to 18%. I think that's a good number for these injectors, but who knows about the initial opening time.

[opethmike]

Looks shows both 'PW' and 'PW2' firing identically. The bank 2 menu item is listed as 'bank 2 specifc', so I would think it was in case the second bank needed different settings. Seems bizarre to me.

When it did run for a little while, I heard a very noticeable whistling. So that would indicate a nasty vacuum leak, I believe.

[Spittybug]

You are correct on the bank1 and 2 issue, you do NOT need data in bank 2 unless you want it. Giving extra fuel and having it run longer sounds promising and I second the call on a vacuum leak. Do you perhaps have a vacuum port on the side of the throttle body you've mistaken for a water channel or something? If you can, while the car is running, spray some starting ether around the likely areas. When it gets sucked in it will rev up, helping you find the vacuum leak. A blind mechanic taught me that 30 years ago.

Oh, I also had instances where changing back and forth between simultaneous and alternating banks magically cleared out some problems. I think it is just deep seated issues within the code when you are running oddfire. Darcom (Daryl in San Antonio) and I had a bear of a time and with the help of the Megasquirt guy in Birmingham England, helped him get the code fixed. My bet is that it may have gottten messed up again in subsequent releases; they don't thoroughly check that us oddfire guys are fully functional! I'm running MSIIe v.3.1.0 with no issue, you could try loading that if you don't find a vacuum leak.

[dmc6960]

Did you remember to blank off this hole?

1156

[opethmike]

I'm running MSIIe v.3.1.0 with no issue, you could try loading that if you don't find a vacuum leak.

That is indeed the revision of code that I am running.


Did you remember to blank off this hole?

Yes - the hole on top actually came blanked off on my manifold. The big hole on the side I epoxied closed.

Fortunately, there are not a whole lot of places south of the throttle plate on this set up for the vacuum leaks to occur, so it *crosses fingers* hopefully it should not be too difficult to find the leak.

I'm 100% convinced I have a leak at this point - the uber high fuel needed, and the whistling sound that I can hear while the engine does run.

[opethmike]

Found a vacuum leak - the throttle body wasn't bolted down tightly enough.

However, still not luck getting the car to run. :mad2::mad2::mad2:

Starting to feel like this is out of my league, so I'm going to take a break from the car for a few days while I figure out what is next.

[living_the_dream]

Don't give up yet! You're giving guys like me hope for when I try and tackle this.

[opethmike]

Don't give up yet! You're giving guys like me hope for when I try and tackle this.

Not giving up, just taking a break for a few days.

[Bitsyncmaster]

Not giving up, just taking a break for a few days.

That's always a good thing to do. You save yourself doing (breaking) things when you stop and think about the problem.

[Spittybug]

Great that you found "a" leak. There may always be another remember! I would go back and set all of your values where they should be just in case you changed something when trying to compensate for the leak. I also found that manually burning changes to the ECU was more reliable than the autosave types of burns. I know there are still some gremlins lurking in the code, so my advice remains to make changes away from good settings and then back to them to "lock" them in. Oddfire vs. evenfire, simultaneous vs. sequential, etc... Very easy to flood the engine if you've played with the priming pulse for example.

So close!!

[opethmike]

Great that you found "a" leak. There may always be another remember! I would go back and set all of your values where they should be just in case you changed something when trying to compensate for the leak. I also found that manually burning changes to the ECU was more reliable than the autosave types of burns. I know there are still some gremlins lurking in the code, so my advice remains to make changes away from good settings and then back to them to "lock" them in. Oddfire vs. evenfire, simultaneous vs. sequential, etc... Very easy to flood the engine if you've played with the priming pulse for example.

So close!!

I reloaded a 'sane' msq before I started working on getting it running yesterday. Leaving everything else alone, it would only crank, fire, and then stall right away, unless I jacked the req_fuel through the roof. So I think I still have some big vacuum leak somewhere.

Can't imagine where, as I went through the installation very thoroughly yesterday. I'm currently working on getting a hold of a fog machine to turn into a smoke leak detector.

Nonetheless, I'm getting burnt out on working on this car and project right now, so I'm taking a break from it through the weekend. In the mean time, I'll be helping a friend re-assemble the top end of his Mercedes Benz 300E engine.

[opethmike]

Random thought - I pulled a spark plug out of the DeLorean while picking up tools to work on another car today.

I noticed that the two plugs I pulled both had pretty black (and dry) tips. Could that possibly be why the car stalls shortly after starting? I don't know, which is why I ask, but it seems more like it would make the car very hard/impossible to start, as opposed to start, run briefly, stall.

[Bitsyncmaster]

Random thought - I pulled a spark plug out of the DeLorean while picking up tools to work on another car today.

I noticed that the two plugs I pulled both had pretty black (and dry) tips. Could that possibly be why the car stalls shortly after starting? I don't know, which is why I ask, but it seems more like it would make the car very hard/impossible to start, as opposed to start, run briefly, stall.

You probably can't read much from the plugs until you get it running constantly.

[opethmike]

Didn't think so.

[20098]

Mike,
Regarding the spark plugs...it is very easy to foul them up when initially setting up the MS parameters and trying to get things running. I went through 24 plugs before I was able to get things dialed in right. I also found that with my last set, one of the reasons for poor performance was because I incorrectly gapped them! I really felt dumb, but it solved the majority of my problems. You can clean off the plugs with a wire brush, but I was lazy and purchased a couple boxes so I could continue my efforts without the extra hassle.

[opethmike]

Yeah, I think I'll just wire brush them. If anything, its at least a very likely cause for the car running like crap if/when I get it to actually stay running!

[sifu22]

I have been running megasquirt for 5 years now. I'm not sure what version your running but I'm running the early version board. ms1 v3 board. If there is anything i can do to help just let me know. It takes alot of time and effort but keep going and you will be happy you did :) I'm pushing 13 lbs of boost with my turbo conversion and I'm at 300 whp.

[opethmike]

I have been running megasquirt for 5 years now. I'm not sure what version your running but I'm running the early version board. ms1 v3 board. If there is anything i can do to help just let me know. It takes alot of time and effort but keep going and you will be happy you did :) I'm pushing 13 lbs of boost with my turbo conversion and I'm at 300 whp.

Hey Greg (This is Greg, right?),

I am running the MS2 V3.57 board. Right now, my issue is figuring out why the car stalls immediately after firing. The higher I jack up the fuel injected, the longer it will run. Even so, it takes ASTRONOMICAL req_fuel values for it to run for more than a second. Even then, it will only run for 5 - 10 seconds. I can't think of anything other than a gnarly vacuum leak.

[dmc6960]

Have you done anything with the afterstart and warmup enrichments?

[opethmike]

Yeah, I've played with those to no effect :(

[dmc6960]

Do you have a recent log of whats happening? Perhaps I missed it.

Another thing, I know it can be bad for the sensor, for just a few times I think it would be fine... manually power up the LC-1 before starting the engine, make sure its fully warmed up and sending a signal. This way you can log the AFR during the short run to see what the engine is doing with the fuel.

[opethmike]

I'll collect a new log after work today. I'll let the O2 warm up before logging today, per your suggestion.

Keep in mind that typically the car runs for only 1 or 2 seconds, so its going to be quite tricky to get any sort of decent logs out of it. I will try though!

[living_the_dream]

If you think it's a vacuum leak maybe get video of it and put it on here. Maybe someone can pin point it that way? I've done that before and it helped.

[Bitsyncmaster]

It's not recommended to warm up the LC-1 before engine start. They say it may crack the sensor of any condensation hits it hot.

But many have done it and still working.

[opethmike]

Some minor progress tonight - confirmed no vacuum leaks via a rigged up smoke machine.

Found that one of the hoses connecting the idle speed motor was collapsing under vacuum. That certainly wasn't helping anything.

Cleaned up all of the spark plugs.

Since I think I have all mechanical bases covered now, I'm going to reload the 3.1.0 firmware tomorrow to ensure that I am starting from a clean slate. Not optimistic I can get it to idle tomorrow, but if I don't.....

[AdmiralSenn]

Just FYI, crank the engine a little before you let the O2 sensor warm up. If there's humidity in the exhaust pipes and it hits a hot sensor it can crack and permanently damage it. Just clear the air out a little first if you're going to warm up the sensor before running it.

[opethmike]

Curiosity question - where are you guys grounding? I'm grounding to the passenger side valve cover.

I suspect that either my ground wire isn't crimped & soldered well enough, or not bolted down enough (really doubt it), or its somehow a bad location. At one point yesterday, the car ran for a bit, but during that time, I heard the relays on the relay board clicking, which means it was resetting. Not good!

[living_the_dream]

Man, I'd be so paranoid about something like that happening I'd almost bond all the grounds together that way you know you won't have any grounding problems

[ramblinmike]

I'm watching this thread closely. I bet the problem is one of those silly little things that tend to stay hidden right under your nose. Don't give up, man.

[opethmike]

Man, I'd be so paranoid about something like that happening I'd almost bond all the grounds together that way you know you won't have any grounding problems

With the relay board, there's only one wire for the MS setup. What you mention about bonding grounds is double relevant for me; as I also own a C5 Corvette. Those have oodles of grounding problems, and Chevy resolved them in the C6 by bonding most of the grounds together.


I'm watching this thread closely. I bet the problem is one of those silly little things that tend to stay hidden right under your nose. Don't give up, man.

I bet so too. That said, I do plan on reverting to K-Jet after this weekend if I don't get the EFI to idle reasonably. That is NOT giving up, as it would be a temporary reversion so I can drive the rest of the short driving season I have. EFI would then happen over the winter.

[opethmike]

Well, I added a second ground to the MS ECU. At least no interrupts experienced again. That said, I still can't get the car to run at all unless I jack the fuel WAY WAY WAY up, either via the VE table, or the REQ_FUEL, etc.

Even that, it runs like ass and eventually dies. To confirm the no vacuum leaks result of my recent smoke test, I got the car running crappy with high fuel and used the old carb cleaner test. Still no leaks found. I am profoundly stumped at this point.:rant::rant::confused::swear1:

As one would expect, after it has run for a little bit of time via the jacked up fuel, the O2 sensor shows WAY rich.

[living_the_dream]

I'm assuming you're going back to good ol' mr. reliable....K jet

[opethmike]

After this weekend, yes. The conversion will be tried again over the winter.

[sifu22]

[QUOTE=opethmike;3548]Hey Greg (This is Greg, right?),

I am running the MS2 V3.57 board. Right now, my issue is figuring out why the car stalls immediately after firing. The higher I jack up the fuel injected, the longer it will run. Even so, it takes ASTRONOMICAL req_fuel values for it to run for more than a second. Even then, it will only run for 5 - 10 seconds. I can't think of anything other than a gnarly vacuum leak.[/QUOT

If it only runs for.5 seconds and then stalls I would say it might be the after start enrichments. If it runs a little longer than I would say its the warmup wizard... Too much fuel and it will do that. Too little fuel and it will.do that. I can give u my settings for the after start and the warmup.. Let me know what u want me to.do:-)

[opethmike]

Thanks, I appreciate it. I'll get back to you after the driving season is over. The EFI system has been taken out and I'm currently re-installing K-Jet so I can actually drive for the short driving season I have. MS will be tried again over the winter.

[sifu22]

Well, I added a second ground to the MS ECU. At least no interrupts experienced again. That said, I still can't get the car to run at all unless I jack the fuel WAY WAY WAY up, either via the VE table, or the REQ_FUEL, etc.

Even that, it runs like ass and eventually dies. To confirm the no vacuum leaks result of my recent smoke test, I got the car running crappy with high fuel and used the old carb cleaner test. Still no leaks found. I am profoundly stumped at this point.:rant::rant::confused::swear1:

As one would expect, after it has run for a little bit of time via the jacked up fuel, the O2 sensor shows WAY rich.

Stop touching the req-fuel,and the tables... its probably not the problem... give me a call... 631-872-1185. And we will talk alittle. I will help as much as I can. :-)

[opethmike]

Again, I really appreciate your offer to help, but the EFI system is already off the car. It has been down for six weeks, which is too much for my tastes. I will be trying the conversion again over the winter, and I will certainly make use of your offer to help.

[opethmike]

I also still need to remove the rear driver's side spring & shock to install a new captive nut for a body bolt. So at least I'll be driving again by this weekend hopefully.

I hope I'm not disappointing/letting down people by my choice to revert. The EFI system WILL be successfully installed, just not now. Over the winter when the car isn't driven anyway. :) :) :)

[Malevy]

I'm pretty convinced at this point (thanks Marc!) that is running uber-lean because of a vacuum leak.

Suggestions on tracking leaks down would be appreciated.


I probably could have been more helpful if I was reading this thread!




I suspect that either my ground wire isn't crimped & soldered well enough, or not bolted down enough (really doubt it), or its somehow a bad location. At one point yesterday, the car ran for a bit, but during that time, I heard the relays on the relay board clicking, which means it was resetting. Not good!

Don't rely on anything other than REAL WIRES for electrical ground. The grounds on the DMC are awful to begin with, they make it impossible for any "electronic" anything to work properly. On the VQ conversion, I have ground wires added all over the place.


After this weekend, yes. The conversion will be tried again over the winter.

:sad30:

[opethmike]

I probably could have been more helpful if I was reading this thread!




Don't rely on anything other than REAL WIRES for electrical ground. The grounds on the DMC are awful to begin with, they make it impossible for any "electronic" anything to work properly. On the VQ conversion, I have ground wires added all over the place.



:sad30:

Don't cry, my favorite person named Marc. EFI (and more!!!) will definitely happen over the winter!

[opethmike]

K-Jet is now back on the car, and it runs beautifully. Felt DAMNED good to drive it again. I have made arrangements to have my EFI installed professionally over the winter, along with other go fast goodies.

[sifu22]

K-Jet is now back on the car, and it runs beautifully. Felt DAMNED good to drive it again. I have made arrangements to have my EFI installed professionally over the winter, along with other go fast goodies.

well, Good luck with your project :)
You will be so happy after you do the EFI. :)

[opethmike]

Thanks man.

Yes, yes I will be. While it is getting EFI how I want it (with my manifold, rails, etc), it will also be getting exhaust (trading my K-Jet parts) and aggressive cams. Car will come back much more 'alive'!