I guess this is going to be my EFI blog for awhile. It is going to seem a little scattered, and maybe not focused enough, but all the issues are interrelated.

If I just start the car up, and jump on the highway, it runs pretty well,with decent mileage in the low 20's. If I try driving to work, or around town, it is kind of a mess, with unreliable idle. I want to fix this so I can drive it to work with un-due drama. I drove it to work Thursday and it died at a light, and I had to wave other cars around. Embarrassing, annoying, and inconvenient to say the least. It started after a few moments,and I went on to work, but determined to make drivability THE issue. I do NOT care how fast it is. I do NOT intend to race it. I just want to drive it on a regular basis, and do NOT feel comfortable doing that now. Here are the theories as to what is going on:

1. I think the electric fan is raising hell with the EFI. I am pretty sure the around town issues are related to it. I think it died on the way to work when the fan came on and spikes from it were picked up as rpm counts, flooding it. I am pretty sure it was flooded because it would not start right back up. I had to wait a few moments. When I drive on the highway, I think the fan just never comes on.

2. The large injectors required for the Turbo are very sensitive to idle variances. A smaller injector would have a larger open-to-closed relationship at idle, so small changes might not effect the idle so much.

3. The VR pickup is still not stable enough. It is too prone to interference.

Here are the fixes floating around in my mind now.

1. Try to filter the fan. Tell me what to do. I know, I know, you have told me before.

2. Get rid of the electric fan altogether. The BMW has a honking big clutched fan as primary cooling. The electric pusher in front of the radiator only comes on when idling for 5 minutes or longer on the hottest,100 degree days.

3. Get rid of the turbo charger and intercooler, and replace injectors with smaller ones.

4. Investigate feasibility of adding a filtered, regulated,isolated power supply for the EFI system. This should not be hard to do, and could be put under the dash out of the way of all the EMI noise under hood.

5. Finish the trigger wheel pick up. I have it in, but not wired. It would require a re-config of the MS ECU, which I am not smart enough to do. I did not build this myself,so am kind of dumb about its configuration internally. I am told this would be more stable and less prone to interference from EMI. Right now it has a runout of .020". I am also told I need to improve that, which would require taking the radiator out again.

6. Go back to fuel only on EFI and set up a distributor with advance, or an MSD advance box as Lou has done.

7. Put a two barrel carb on it.

So here are the first questions, and least invasive things to try:

1.Filter the fan motor.
I know I have been told this many times, but what size capacitor should I use to filter the fan, and how do I wire it? Do I place the capacitor between ground and the 12+ that feeds the fan? And does it matter where along the wire I place it? And which way to I orient it? I intend to try putting a ferrite choke on it also. But, as I learned from moving the HEI ground to the ECU and off the engine block, some interference can come back through the ground plane.So am wondering if I should put on the ground wires as well. You know every computer peripheral has one on its cord.

2. Provide a filtered power supply. I would guess the designers of this thought they had that in their design,but it is clearly not up to the task. Any ideas about that? I might ask Peter to weigh in here.

I really think most of the problems here are electronic noise related. And I do not know how to deal with them. This is just not my thing. The BMW is a silky smooth slant six that is so unruffled and reliable that I am tempted to take a serious look at their engine management system. However, I also think that if I took the entire engine and put it into the Dart it would run badly because of the hostile environment the current system operates in.

Any thought on any of this. I am open to any, or all of the changes mentioned above, including just putting a carb back on it. Was the drivability of the carb system really that bad? I remember driving this car all over the place with a two bbl carb, getting excellent gas mileage, and never worrying about whether it would die and not start. Well, that happened once when I tried a Ford carb on it.

Sorry for the length of this post, but the EFI section has needed a little spark of controversy here of late.

Sam

_________________

Start with something simple, like a 2200 mfd capacitor across the plus and minus terminals of the fan. See what happens.

I like your solution of a separate power supply for the fuel injection system. That would keep the power clean and rock stable, no matter what else was going on throughout the rest of the vehicle.

A reasonable test to see if noise from other systems is the issue would be to simply power the entire EFI system from a battery, temporarily mounted somewhere under the hood. If your around town problems magically disappear, then you are on the right track.

Jameco Electronics has DC to DC power supplies at reasonable prices.

- Mac

Sam,

I am a simple minded country boy and your car is maze of electronics and controls and gadgets. I would never be able to begin to start diagnosing a problem with your car. But at the same time it better be something that you "enjoy" working on and diagnosing. If that is not something you like, then I would simplify everything and make it dependable to drive and enjoy, even if that meat removing the turbo or electronics. If it is a constant "work" project that you don't want to work on, then that robs the joy out of driving the car and it will end up setting in the garage

I really feel sorry for you sometimes as you continue to have problems. That would drive me crazy. Hope you get it all worked out. It really is a sweet ride.
Rick

_________________
2 Mopars come with Spark plug tubes. One is a world class, racing machine. The other is a 426 CI. boat anchor!

12.70 @ 104.6

Rick I appreciate your concern for me, but I do enjoy this car. I am just honest about my thoughts and feelings in a way I suspect many would not be. It has its highs and its lows. Every "failure" ultimately leads me to a new realization which improves things. I am not an EE or and automotive engineer so the learning comes slowly. But sharing my problems helps others out there, and gives us all something to talk about.

Here is what I have learned in the last 2 days:

I opened up the BMW shop manual and inspected their wiring diagrams, specifically looking to see what they did to avoid the interference problems a hobbyist such as myself and others have experiences. I discovered two significant things:

1. Every single component in the EFI system on the BMW has a "shield" designated on its ground wire. It is represented in the schematic by a cylinder about 1/3 the length of the line representing the ground wire, and parallel to the wire, just slightly off to one side of the line. So now I have more questions to answer.

Is this just shielded wire? And if so, how does that work? How can you shield a ground wire? The strategy with shielded wire is to ground the shield, and carry the EMI to ground. What keep EMI from traveling back up the ground wire itself if the shield is grounded to it. Maybe ground the shield at the component end of the wire and not the ground end. Or does this "shield" represent something else? Can you put a diode in a ground wire and keep ground plane noise from being picked up and carried back to the sensor or ECU? I have posted this question on the BMW forum and maybe someone there will know how to interpret the symbol in the shop manual

2. They have the power distribution divided into three basic blocks. The one closest to the battery powers all EFI components. The rest of the power goes to the far fender where 47 fuses, (yes 47) distribute the power to everything else, including noisemakers such as fans and motors.

So now I have a clue as to how to reconfigure the wiring on the Dart to make it run more like an OEM application.

BTW the BMW has maybe 25 times the wire count of the Dart, and 4 times as many fuses. Of course it is too complicated for me, but complicated can work if you do it right. My first task with the BMW was take the power convertible top stuff out and change it over to a manual top. The power top had almost as many components as the entire Dart, and two of the three systems in it had already failed. So, I am on board with simplify. And if I have to, I may take the turbo out of the Dart, but not until I understand the basic fundamentals of good EFI design and fix the fundamental things that are wrong with the Dart. If it still does not behave, then yes, out will come the turbo, and maybe the electric fan. And maybe the EFI. But not until I have done more learning and fusing. I suspect I can get there.

If nothing else, I now know how to read the wiring schematic for a modern car with EFI.

Sam

_________________

Thanks for your thoughts, Sam. You always think hard about things and are honest with yourself and us, which is great. I like your filtering ideas and moving things close to the battery that are noise sensitive.

I have to say, all of these noise issues make me want to go straight for the MSIII upgrade instead of the MSII. It seems the MSI is much less susceptible to noise and I am a bit scared of upgrading now.

The shields are actually the grounds, and are wrapped around signal-carrying wires (positive voltage or pulses or whatever).

Lou

_________________
Home of Slant6-powered fun machines since 1988

Thanks Lou, I am glad you understand my style of both learning and sharing. Your thoughts have been essential in every single correction that has made in basic design here. I am sure there is some small thing, or a number of small things that are important to trouble free operation of EFI which are not common knowledge yet to the hobbyists who apply them.

What do you think of the idea of going back to a mechanical fan? Dan has a Mercedes clutching fan he has offered. We need to measure and find out if it will pass everything there. Mostly concerned about the AC pulley,the attached trigger wheel, and the pickup which is right down there.

I am eager to find out what the BMW diagrams mean with the "shield" icon on all the EFI ground wires. This has to be important. If I can figure that out, I think I have unlocked a new design "secret" here.

And yes, if your MSI set up is working well, don't fuss with it for now. I would actually investigate other packages such as the FAST system, or other widely used set-ups.

Sam

_________________

Sam,

You say the bmw has 25 times more wires but the last I remember seeing pics of your engine compartment you probably had a mess of wires bigger then the BMW. Was any of that ever cleaned up?

Remind me of what the ignition setup is... which pickup is driving what? What's firing the coil? (Dist to megasquirt, megasquirt to coil?) You mention crank trigger...

What plug wires are you using?

Are you certain its the rpm counts that are jumping around when the fan comes on? Fan current at startup is huge - if idle speed is low enough and alternator is marginally keeping up this inrush of current can cause line voltage to drop to point where MS tweaks and won't function. End result is the same, engine dies, but completely different cause. You can always trigger the fan manually while watching voltage and rpm logs.

Don't make any major changes, will only complicate things more. Capacitors maybe, but separate power supply, getting rid of turbo... your way over thinking this. If you've given up I'll gladly take the MS2 off your hands I really wish I could spend a day with your car.

Me too. That would be fun and productive I am sure.

Here is the outline:
VR pickup drives the HEI module which sends a digital signal to the ECU. From there it comes back to the HEI module as a signal to fire a coil. This triggers the MSD box with the white wire on the MSD. The MSD fires the MSD blaster coil. The spark plug wires are standard issue NAPA, Echlein stuff. Nothing exotic.

Watching the laptop while the fan comes on you see the voltage drop from 12.8 to 11.8 of so. There is a 1 volt drop the stays constant at that point.

The failure last Thursday duplicated itself while testing for fan/voltage effect. At the time of the second failure I had the lap top hooked up. I could see that the RPM count dropped out. It stayed out while trying to crank during which it would not start. AFter a bit the RPM count returned and it started. This happened about 4 times, then it would not start at all. That is pretty much the same pattern as happened this morning, and after a long wait (maybe 2-3 hours) the RPM count returned and it started. I then swapped in another HEI module, and after it warmed up, the result was the same. At first I thought ,"Ah the HEI was bad". But, no dice, eventually it quit again and went through the same drill. I watched the screen as it died each time and saw nothing go wrong with any other readout other than the RPM count that should have killed it. The AF ratio would go lean, and then back, but I cannot think it was lean enough to kill it. And if so, it would have simply fired back up if the RPM count was present. It seems this is a basic ignition configuration problem. All connections between the distributor and the HEI and the ECU seem fine.

So this is more than one thing it seems. The fan is effecting things for sure. But apparently that was not the actual cause of the shut down last week on the way to work. This is new behavior in the last week. Prior to that it was consistent, and never actually failed to run.

I sincerely appreciate your thoughts and advise here. The task it seems now is to figure out why the RPM count goes away. At least that is what the lap top says. I should share that the RPM does not show immediately on the screen even when it does start. It will crank a few times, then the RPM shows and it fires.

Here is a thought. When I was trying to track down the false reads before at 3 K RPM I closed the gap down a fair amount in the distributor. Could this be closing down too far as it heats up? I am not sure what I had it set at, but it was tighter than stock, at somebody's advice here on the forum. That problem was fixed by grounding the HEI module through the ECU instead of to the block. I just thought about the distributor gap. What should it be set to? It has a standard Mopar electronic ignition VR pickup and reluctor wheel.

Sam

_________________

Errr…remind me, Sam, what alternator are you running?

_________________
一期一会
Too many people who were born on third base actually believe they've hit a triple.

It is a GM one wire. To correct an earlier statement, spark plug wires are Belden, 7mm.
Sam

_________________

Have a look at this page and let us know more detail, please.
Almost certainly inadequate for what you're asking them to do. You'd do better with spiral-core wires (such as Magnecor), possibly with EMI/RFI shield sleeves. You're running a distributor? Or a distributorless system?

_________________
一期一会
Too many people who were born on third base actually believe they've hit a triple.

Had Magnacor before. Those were replaced with Belden suppression wire as a possible cure for break down at 3000 rpm. That break down was eliminated by grounding hei through ECU instead of block. Once I get everything working right, that would be the time to go back to Magnacor wires.

Still running distributor. I was on the verge of moving to distributorless ignition when grounding the hei through ECU seemed to cure its problems. Now challenge is to stabilize pick up and rpm read. First must figure out why it is going away.

I will look over your link for more info in the alternator. I think it is rated at 100 amps or so. It looks like the SI but with a fan like the cs144.

Sam

_________________

Are we looking at a heat related issue here? If that's the suspicion, then how about using a thermal chiller spray to cool off certain components (like the pickup) one at a time to see if you can return the system to normal operation?

It's possible you may have a thermally intermittent component. If so, this may help you locate it.

- Mac

Along the lines of grounding issues, try this simple test:

Using a DVM set to test for DC voltage, place one lead onto the negative terminal of your battery, then move the other lead to various ground connections within your system.

In theory, you should always get a reading of 0.0 volts.

If any points give you a significant reading, then you have either a poor ground, or a ground loop, either of which can cause sensitive systems to go bonkers.

"Grounding the HEI through the ECU" sounds like a classic recipe for a ground loop.

- Mac

That is a good thought. Sort a cold can for electronics. I just talked to an auto enthusiast who is an EE. He says wires that are broken under the insulation will act like this. The VR pickup is currently running through a shielded microphone cable. There is a real possibility that this wire, although electrically adequate for the task, was not up underhood temperatures. I have some shielded cable from DIY which is rated at about 200C, so I am going to replace the microphone cable with that sometime this week, and give it another go. This happened with the hood open, and the ambient temp was not as high as when it failed on the way to work.

I am not defeated yet. I have gotten good support from many folks who are smarter than I am about this. They all say this should be easy to work out. The intermittent nature of the problem makes it tougher of course. And it makes you not want to drive it until you are certain you have tracked down the problem and tested it thoroughly.

_________________