So a lot of you will remember my thread from last year that ran for many many pages, trying to get my '74 2.0(56) running again after literal years of turmoil and confusion.

After much investigation and many suggestions, I was able to isolate the issue to the MPS, a rebuilt 0280100043 unit, remanufactured by Fuel Injection Corp USA and sold by Rich Bontempi's High Performance House back in 2020. Last fall, I was able to get the car running, idling great and driving spottily by simply using the Tangerine Racing MPS tuning kit to adjust the inner idle screw richer, but was never able to achieve full throttle. I ended up cranking the inner screw up to max, and was able to get about 60% throttle before the engine would die. Although idle became excessively rich, this made the car driveable enough to get it hot. Especially since nearing the cut-out point, it seemed to get very lean.

With the warmer weather back, I've been working on trying to tune the car in better, but have been met without much success. But I've dialed in with growing specificity toward what seems to be an MPS transformer problem.

I started by doing a full intake system rebuild/re-seal, just to be certain I wasn't fighting a vacuum leak. Slowly and carefully cleaning and re-seating all mating surfaces and using gasket dressing on gaskets. (I also 3D scanned and modeled the intake system, but that's a topic for another thread.) I'm now up to 100% confident that the car has no vacuum leaks, up from maybe 85-90% before.

Then I set about adjusting the tuning. I got a proper vacuum pump that could pull up to 25+ inHg and bench tested my MPS. Like last time, the 8-10 coil (secondary) curve came in FAR lower than both the graph value and chart values given on Brad Ander's MPS page, at least on the high-power end of the scale, with the primary coil being around 1/2 the secondary. [Graphs and chart shown below.]

With the inner screw maxed out, the 25inHg primary inductance secondary 0.67H and the primary 0.32H. At 0inHg, the secondary approaches .93H and the primary, .49H.

Now, I had posted this very low inductance curve in my last thread, and the community consensus was that the shape of the curve was more important than the actual values, but I question this, and wanted to test where the cutout point on my engine's throttle response was, so I hooked a vacuum gauge up to a long hose, plugged it into the manifold port where the decel valve usually goes, and went for a drive.

After letting the engine get hot, I did a few pulls, where I gradually increased my throttle position and watched the vacuum.

Sure enough, the engine would cut out each time the manifold vacuum fell below about 7-8inHg on the gauge, which is just past the intercept point between my MPS's secondary curve, and the 043 curve Brad has posted on his site.

This aligned well enough with my suspicions that I was willing to spend the money to test out another unit. I went out looking, and settled on a 0280100037 unit from EBS Racing, since I couldn't find any verified '043s, and '037s are compatible with my car/ECU - considered to be an "upgrade" due to their higher stock mixture setting.

When this one arrived, I found it had also been remanufactured by Fuel Injection Corp. And upon testing its coils, while the curve's shape looked much healthier than my '043's, it was similarly very low. And sure enough, the car wouldn't start on it. What's more, I found that on this unit, the inner adjustment screw was nearly maxed out already, only about a quarter turn from its limit.

At this point, I'm scratching my head a bit. Could I really have gotten 2 bad rebuilds in a row? Does the inductance value really not matter, and there's something else wrong with my car? Why are my values both so different than the nominal ones Brad measured?

I checked that my LCR meter isn't bad with a 2.2H coil from DigiKey, and it read that more or less dead on. I also pulled the top off my '043 and checked to see that the transformer actuation is free and clear. Even with the core pulled to its limit, the secondary won't go much above 1.0H. The resistances in the coils are correct, so they're not shorted. All I can think is maybe the cores have been replaced with something non-suitable?

I know Jeff drove his car with my components last summer, which is further confusing, though I don't know if he tried full throttle or not, and I know my 2056 pulls a little less vacuum than an engine with a stock cam.



Values:

Brad Ander's 043 (based on 2 rebuilt and 1 NOS unit)
0 inHg = 1.39H
4 inHg = 1.18H
15 inHg = 0.71H

My Rebuilt 043
0 inHg = 0.929H
4 inHg = 0.915H
15 inHg = 0.809H

Brad Ander's 037 (based on 1 rebuilt unit)
0 inHg = 1.44H
4 inHg = 1.26H
15 inHg = 0.72H

My Rebuilt 043
0 inHg = 0.943H
4 inHg = 0.883H
15 inHg = 0.601H

So I guess, I'd like to ask the forum a favor. Anyone who has an LCR meter and a WORKING D-Jetronic car, if you're willing, do me a favor and check the inductance across your 8-10 coil with no vacuum, and post it here. If you feel like graphing your response curve, all the better!

Just my $.02, from my experience: The LCR meter readings are interesting and good guidelines, but to set the MPS correctly, once the outer screw is in the correct ballpark, you really only need to be able to read the AFR for a given vacuum to adjust things further. Even with the exact LCR meter that Anders used, my car wouldn't run correctly when I mirrored his settings. You'll get different readings with different brand LCRs. Yet another reason relying solely on an LCR isn't recommended.

Also, you shouldn't be using the inner screw to dial in WOT. Use the full load stop screw for that.

My order of operations:
Get the outer screw set first on the bench. I do this by putting the MPS in a vice with the adjustment screws facing up, then insert the adjustment tool in the MPS and rest it upright in place. Then using a vacuum gauge start pulling a vacuum. Adjust the outer screw so you see the tool stop moving down somewhere at around 5-6 in Hg. (reflected in Anders curve you posted above).

Then install the MPS in your car and adjust the inner and full load stop screws (in that order) using an AFR wideband sensor to guide adjustments.

Just to make sure we're all on the same page regarding which screw is which- I'm using Anders definitions:
Click to view attachment

Yeah, what I really should do is take the thing to a dyno to at least get an AFR reading on it. But I'm disinclined to spend a bunch more money without checking off that I'm not overlooking something as simple as a bad unit.

You're also correct that the outer screw is the correct adjustment for WOT mixture, but at this point, I'm just running with it uninstalled. I backed it out 'til it was no longer doing anything, and for occasional test runs, it shouldn't hurt the diaphragm to stretch to its limit. The only reason the inner screw is maxed out is that it gives me enough useful throttle range to drive the car for testing. With the idle tuned more correctly that intercept falls well into the teens, and the moment I touch the throttle the car just shuts off.

How close were your numbers to Anders'? Were you off by a similar ~30%? I know his guide gives a tolerance of .01H!

I did not say the outer screw is for adjusting WOT. The full load stop screw is for that. Running the car without the full load stop screw will shorten the life of the diaphragm. Not a good idea and not necessary if you use the replacement full load stop screw that's provided with the kit since you can adjust the inner screw and outer screws with the full load stop crew in place.

You can't adjust the MPS correctly without a wideband AFR sensor.
Forget the LCR numbers. You're chasing your tail with those at this point. That said, a number of people, including myself, posted their numbers on an earlier thread. (The #s I posted in the thread aren't current anymore as I've made some adjustments since then).
MPS Tuning Analysis Thread

Click to view attachment

Yeah, apologies. I meant the full load stop when I said "outer screw". I've been referring to the 3 adjustments as inner, mid, and outer to myself.

Regardless, the full load stop is out, not stopping anything at any point. The outer adjustment on the diaphragm hasn't been touched.

I'll definitely be looking through that thread though.

Getting the outer screw set correctly first is crucial (and easy to do using the steps I outlined above). I learned that the hard way. Good luck!

My recollection from the original thread was that there were still basic assumptions that hadn’t been resolved such as low vacuum at idle and/or specific cam grind and potentially cam timing at assembly was an unknown.

Am I recalling that correctly?

D-jet just isn’t very tolerant of changes that would affect volumetric efficiency.

After my little rebuild, it still idles at 15inHg at around 1100-1200rpm. Some people said that's low, but it fits comfortably in the idle zone from Brad's page. To quote, "At idle, the manifold vacuum is about 10 to 15 in. Hg".

I don't know what's happened to Mark DeBernardi, but I haven't been able to reach him about the cam spec - that said, and as I've said before - this car ran with this ECU and this cam/engine hardware for several years. Only new fuel control parts are the injectors and MPS, and I have bench tested the injectors to be almost exactly to spec and all identical. The fact that the MPS is very definitely not in spec is the only thing I've found that's off about this entire engine, and it happens to perfectly match the symptoms.





In other news, I know you guys are anti-induction-curve, but in that other thread, while there is variance, I don't see anybody with a full load data point below 1.15H, and most are in the 1.3-1.5H range. What's more, is where Not_A_Six was experimenting with the inner screw in that thread, it was translating his curve uniformly across the board, while mine stays within a couple dozen mH on the full load end, and only adjusts on the idle/overrun end.

Also, can't help but notice Chris Foley trash talking Fuel Injection Corp's refurb quality at the end, though he was disappointed with diaphragms rather than transformers.




Before posting this, I had an idea, so I actually just tried out popping the disassembled MPS into the car, with no cells or diaphragm, just the core extended out to its absolute limit, which is further than it probably should be, since this MPS seems to be missing the little tabs that engage the core stops.

The primary secondary coil reads 1.066H in this configuration. The car starts up immediately, pretty stinking rich but not too smoky, and responds very well to throttle input. But full throttle - even ~80% throttle - still kills it. Like, this MPS CANNOT PHYSICALLY provide any more pulse width, and it's still not enough fuel.

I agree with superhawk. My build actually produces more vacuum than the mps expects, which I believe was/is causing my cruise leanness (I think I finally have that fixed after my last trip to the Dyno). On my car there is a noticeable transition from 12inhg to 9inhg which matches nicely with dipping into the throttle from cruise. If d-jet is expecting a particular demand based on that you might be having that occur in the wrong place.

What does your car idle at for vacuum? What does it cruise at? Have you tried playing with the timing and watching the vacuum? The old way to time cars was max vacuum then bring it back a little. Maybe a degree. Getting the car in the ballpark vacuum can only help. As best you can with that cam.


And this is d-jet. Vacuum leaks don't matter.

As stated above, idle is 15inHg on the dot. Right after startup it's a bit lower - like 10-13inHg - but stabilizes once the oil pressure and rpms are stable.

Agreed - other than the small trimming of pulsewidth by the TPS, the D-Jet system will just establish a higher idle if there's a leak. Heck, that's what the AAR does (and it's clamped off at the moment).

From the test drive I did, cruise is around in the 10-15 inHg range, light acceleration is like 8-12in Hg. I didn't see more than maybe 21-22inHg on overrun, but obviously I couldn't do a WOT pull. The MAP numbers seemed pretty reasonable for the car.

But again, it seems like I'm just not getting a steep enough EFI response curve. I set it super rich at idle, and it's leaned to death at WOT.

Just checked my spare 043 MPS with the Amprobe LCR set on 20

pins 8 and 10

This has been adjusted to the below


0 inHg -- 1.49

4 inHg -- 1.30

15 inHg -- 0.81


I am running a 2056 with a Raby 9590cam and a 037 ECU so my MPS is an 043 tuned to richer than 037 MPS specs.

This is my spare so you would be welcome to try it on your engine.




and just for reference sake here is a graph I posted in the above mentioned thread.

Click to view attachment

The numbers for the one on this graph were.

0 inHg - 1.52
4 inHg - 1.34
15 inHg - 0.83


But have been adjusted away from that, Graph just posted to give an idea of slope.


I know some people say you can't mix these parts but I have over 20k miles on this combo that says different.


PM sent.

Innovate AFR up the tail pipe
No dyno needed
Drive down the road. Pull over. Slight tweek and test drive some more

Takes up to a dozen tests to get right
Then a drop of epoxy on the screws that nice you have it correct

It may be my simple mind, but it seems that something that is already relatively complicated is being over-complicated.

While it’s true that the curve is important, it seems to me that it is already predetermined by the design of the MPS - the design of the variable inductor and the aneroid cell. While we can set various points along the curve, I and not aware of any way to control the curve between those points.

I built a 2056 about 3 years ago and took great care to ensure that the MPS was functioning properly. Verified that the inductor core moved freely and that the aneroid cell was intact. I then adjusted it per Ander's spec for a 2.0 using a LCR 55a as a starting point.

I then installed an O2 and vacuum sensor so that I could read both while driving.

I drove it and noted that AFR at 15 inHg, 4 inHg and 0 inHg and adjusted the AFR targeting 12, but setting leaner toward high vacuum and richer toward lower vacuum resulting in:
15 inHg - 80h - 14.5 AFR (inner screw)
4 inHg - 128h - 13.5 AFR (outer screw)
0 inHg - 150h - 12.5 AFR (plug screw)

This was a long process:
1. Drive and monitor
2. Pull the MPS
3. Adjust
4. Repeat

What’s the curve – who knows?

The result is a 914 that runs better than any 914 I’ve ever driven, and I’ve owned 3 and driven many. Smooth cruising, smooth acceleration, rock solid idle at 800 RPM and 31 MPG on the highway. (I also have cruise control that helps with the MPG) Note that a 123 dizzy also improved the build significantly.

I’ve helped many with their MPS and found that almost every MPS has problems. I’ve even seen one, brand new, purchased from a Porsche dealer that leaked and had a crack in the diaphragm that was almost impossible to see. I’ve also seen those that were run with a cracked diaphragm for some time, allowing moisture to corrode the inductor core so that it did not move freely.

I also built this web page to help with adjusting for altitude: https://mps.dolimpios.com/
Note that if your system has GPS capability, it will automatically determine your altitude.

The meter gives you a curve. For a hypothetical motor

The AFR gives allows you to tune your actual motor. Even if it has some wear. Even if it’s fresh/tired/etc.

I wonder what material was used for the diaphragm

Beryllium copper
Lots of old threads about this
Doubt there will ever be new ones again beyond CFR

Fuel inject corp we’re ….. back in the day….. using a different allloy
Which also had different characteristics and did not work as well as original

Exactly my point

The values DO matter as well as the curve HOWEVER your values will most likely not match Brads as the readings are affected by altitude and atmospheric pressure. I would assume when anyone says the values dont matter that is only in the context of "I tested my stock MPS and it doesn't match Brads values, is it bad?" For tuning an MPS there is absolutely only one set of values that will be correct, but again, they wont match what anyone else posts.

I think I have at least one of every 914 MPS part number here, I could give you readings from any of them but it would do you no good.

The way I do it (in my opinion the only reliable way to bench tune a MPS) involves cloning the settings of a known good matching MPS side by side by way of a T'd vacuum setup and switch on the LCR meter between the two MPSs. This way you ensure the exact same vacuum reading produces the same inductance across the entire curve.

Unless you are using an identically calibrated vacuum gauge and the same LCR meter under absolutely identical atmospheric pressure, tuning from someone else readings will at best only get you in the ballpark, and often wont even do that.

There is also a secondary factory when dealing with the 914 2.0 MPS that have been rebuilt. There was an additional spacer ring around the diaphragm on the 2.0 units that was not used in the majority of d-jet MPS. If for some reason during the rebuild process that ring was omitted it could impact your ability to dial it in to the necessary settings. Then there is also the question already mentioned around the material and quality of the diaphragms FIC uses. I have a pretty large stash of MPSs, at least one of which is an FIC rebuild from an unknown date, and the diaphragm absolutely responds differently though the midrange.

Those vacuum numbers look good to me. I pull 20 at idle and 24 max on overrun but the rest are about right. I would definitely try the mps on offer.

The curve is modifiable, as in you can ruin it, and shift it with adjusting the inner and outer screws, but in my experience it was best to establish the curve as it is supposed to look on the bench (starting with Chris's tips on preset) and then make sure it responds/ ramps correctly. After that, in the car mess with the inner and full stop as you tune to taste with the AFR. At the Dyno I had to bring in the full load stop, I was too rich at wot across the board.

You did check your injectors and you did move to a 123ignition right?

I have examples of FIC diaphragms in my office. They are hand made from brass sheet, soldered to trimmed used hubs. They don't appear to have good flexural properties and also look like they are difficult to seal.
My diaphragms are precise reproductions of what Bosch used, but I have to be very careful pressing the hub onto the paper thin disc or I risk stretching the center, which changes how they respond to vacuum signals.
I recently ordered more hubs but I will need to order another batch of diaphragms later this year. It takes a long time to use up one order and sales so far this year is lagging previous years. It will probably be my final order of diaphragm pressings.

Thanks for the input, all!

Olympic, I'll definitely take you up on your offer - it should provide a definitive answer as to whether I'm dealing with bad MPSs or not. That curve looks very healthy. You've also reminded me that I have the same cam profile as you! WebCam 9590 - you mentioning the number made me remember it shared its number with the processor my PC was running at the time lol

Adolimpio, that's a handy calculator you've made there - I'll be sure to use it if I can get to that point. Although, I'm only about 600ft ASL where I live, so not a ton of altitude compensation for me.

Emery, I did check my injectors. I'm still running my stock dizzy though, since it seems perfectly healthy.

Chris, I may send my FIC MPSs to you for check over/repair, regardless of outcome with Olympic's unit - I could sell them to recoup some costs if nothing else.


And for everyone, I wanted to highlight this experiment I did yesterday, since the thread is moving around a lot. I don't believe this is a tuning issue. Based on this behavior, there are 2 options as I see them: Either (1) the MPS transformer is damaged/broken to the point that it cannot provide pulse width necessary for WOT, or (2) something else in the EFI system is causing inadequate fuel delivery. At the moment I'm leaning toward Option 1, since I've double, triple, and quadruple checked everything except the MPS, which is displaying odd traits, and is missing at least one original component.

Mr. Cessnaporsche01 & 914ers.
If interested I'll send you my OEM (new) MPS to test your 914 motor activities, but you must send it back to me with no adjustments or modifications. (see Photo)
This would be helpful details for the 914 guys on this list
Send me a PM with your ship address if interested. Just promise to send back.
Tom in Michigan .

Don't discount those trigger points. That could easily be causing your lean condition. The 123 really is the most fabulous thing...

Don't tease me like that - i'm back to my 123 install in 10 days...

Has anyone considered checking the TPS?

If not working properly the ECU may think the engine is idling and will not consider the MPS input properly.

Yes, a good idea. You can check that without removing it.

Here is a dumb question... Have you checked your fuel pressure? You seem to be in the range of Andres settings and have made other corrections suggested. If your fuel pressure is off or fluctuating or dropping you could have the same condition at WOT.

TJB, thanks for the offer. I'm going with Olympic's MPS for testing, but your offer is most appreciated.

Emery, I tested my trigger points with some LEDs in circuit with the injectors, and I'm quite comfortable that they're not skipping. That said, the 123 is still on the horizon for when the car is back on its proverbial feet.

Adolimpio, TPS is good. It's got the 914Rubber board, but I tested and calibrated it while messing with the intake system, and it does what it's supposed to.

And Rick, yeah, I have a pressure gauge in line in the injector loop. I have it set a little high, around 32psi, since that's where I used to run it, but it's rock steady.


I should be getting Olympic's MPS in the mail on Friday. If it gets me WOT, I'll know what's been plaguing me. If not, welp, another run down the old Dr-DJet checklist like a madman, I guess lol. I'll report back with results.

Well dang. I can rule out the MPS as the root cause. The car goes right back to its start-no-run behavior with Olympic's installed.

What this has taught me is that my LCR meter cannot measure the coils in the MPS - it reads his at .927H @ 0inHg, when it's actually 1.49H. So that explains the incorrect MPS readings. It's weird that the primary coil reads proportionally, when I know it can measure a sub-100-Ohm inductor accurately, but maybe that's to do with it being in a transformer and not free standing? Regardless, behavior is unchanged between 3 MPSs, so I'll call them all innocent.


I popped the 043 back in, and added a 370 Ohm resistor to the CHT, just to crank the mixture right up there and went for a drive so I could get video of the behavior. The resistor allowed me to get more throttle in than ever before, like, close to ~90% WOT, but it still sputters wide open, and of course, the idle mixture is pretty filthy rich - mild amounts of brown smoke out the tailpipe while idling. Engine is at temperature in the video.

The engine sputters whenever my foot is on the floor, but it recovers instantly when I let up. I should have thought to do it in the video, but the engine will redline, just at lower throttle settings, so I'm pretty sure the injector points aren't the culprit. I'm thinking is has to be pulse-width related, but I'm coming up blank on what could be causing the pulses not to scale proportionally with the MPS signal. Maybe shitty injectors?? And that doesn't answer why the engine doesn't want to idle below a fairly high MPS
setting - although that may just be tuning, since it will idle at a full inner screw rotation below max.

Anyone heard one of these engines sound like this before?

https://youtu.be/WSK4shNqZiE

It sure does seem happy in the part-throttle regime.

Did this ever get resolved?

Not sure if there's anything to be resolved there. I think the 3500 Ohms (which I measured on the cold engine again today, with an ambient of 55F) was just because the block was colder than the ambient air. I went and checked since you asked about it, and it measured 1500 Ohms with the engine still noticeably warm from having been out on the road a few hours ago. Chassis ground at battery to harness ground on block has only 1.2 Ohms of resistance, and the CHT connection at the harness to the harness ground on the block measures the same as to the battery ground, so there's not a big resistance anywhere that there shouldn't be.

1.2 ohms of resistance isn’t insignificant from one ground reference to another.

Is this resistance in your meter leads? What does your meter read when you simply short the leads, one to another?

Resistance between battery negative and engine should be zero (after accounting for resistance in your test leads).

Not saying this is the root cause but issues like this need to be resolved. The ECU depends on having proper ground reference for input signals such as MPS, CHT, TPS. Etc.

If I found a 1.2 ohm ground plane difference in the metal substructure I'd run dedicated cables to specific ground points throughout the car.

I did this once to a 79 Fiat X19 (in the top ten list for wonky electronics).

Wife drove it for 80K without a single electrical gremlin.

I'd fix the grounds FIRST!

More than 50% of the time electrical gremlins are ground related.

Which makes sense when you realize the electrons actually flow from negative to positive...

Listening to your car run.... have you changed out the Throttle Position Sensor? When I first got my car running I had issues until I installed a new board. Just a thought.

I did some testing for resistance around the car this morning. Superhawk, good call on checking my meter probes, as they were measuring 0.2-0.5 Ohms shorted to each other. I scrubbed them with scotchbrite and was able to get the short resistance down to 0.1-0.2 Ohms. Getting a good dig into the metal on the case and ground bolts saw resistances under 1 Ohm. Most of the resistance seems to be in the block and transmission case, since the transmission ground strap reads 0.1 Ohm, and the ground strap mount to the battery ground is only 0.2 Ohms, but from any of the engine bolts to the strap it's more like 0.5-2 Ohms, with the fan cowl resisting at a massive 15 Ohms, not that any electrical components are attached to that. All the metal is quite clean, so I don't know if there's anything I could do to reduce resistance further, other than ground the engine loom straight to the battery like Ninja suggests.

Ishley, yes, I did replace the TPS board with a new 914Rubber one. It's calibrated correctly, and I don't feel any of that rough bouncing between throttle positions that indicates a bad board.

You really should be measuring 0 ohms after compensating for what ever your test leads add.

Personally I wouldn’t accept any more than 0.1-0.2 ohms between battery neg and engine ground lug that FI is grounding to. Battery cables (either pos or neg) can appear good but be corroded internally adding resistance. The ground weld stud at the body (under trunk) may be getting dicey 50 years in. Modern cars have begun to move away from spot welded ground studs due to corrosion and increasing resistance as they age causing CAN bus issues.

As suggested by Ninja, I would use a temporary jumper to get a 0 ohm ground over at the FI ground lug. This could be as simple as a temporary jumper cable for troubleshooting on a drive. Get any possibility of poor grounds out of the equation. If it doesn’t improve anything it’s simple to remove.

As an aside, listening to your video I’m not convinced that you don’t have an ignition related misfire going on.

Ignition and fuel issues are so commonly mistaken one for the other.

Very hard to relate to the video since I don’t have a good sense of speed, load on the engine, and how the throttle is being modulated.

But can hear that it does seem to run better at what is probably lighter load and less throttle, less RPM.

Did ignition timing ever get set properly? At one point you said you couldn’t even get enough revs to verify ignition advance.

Fair enough on the resistance thing. I doubt a few dozen milliohms is the issue, but it's not much trouble to run a temporary wire to battery ground.

And the ignition theory is interesting. Although, wouldn't that be more RPM dependent than throttle position dependent? Like not enough dwell to charge the coil? I guess the advance is moving up a few degrees when manifold pressure goes to 0, so I could test with the vacuum retard hose disconnected and see if anything changes.

But otherwise, the timing is set to the notch at 3000rpm - once I got the mixture richened up on the MPS it gave me the full rev range so I was able to tune it no problem.

And the holes flow from positive to negative.

Bad common ground = floating voltages = bad behavior.

During this troubleshooting I would disconnect both advance and retard on the dizzy if you haven't already. Plug the ports on the TB of course.

Let there be static advance + centrifugal advance only without a vacuum canister compounding things based on vacuum and changes in throttle position.

The thing is that your engine ground also is the path by which the coil draws current.

I’m going to use an extreme example - let’s say your initial statement of 1.2 ohms between batt negative and engine case was true.

The Bosch OEM coil has an internal primary resistance of 3-4 ohms.

If you add an additional 1.2 ohms in the ground between battery negative and the engine ground, the coil draws much less (~25% less) current and you get a much weaker spark. Could be enough spark to work at lower rpm and proper mixture but when rpm’s rise and mixture is off, you get misfire. Again not saying this is your problem, but it is why good ground matters.

Now if your coil happens to be a VW bus style mount that mounts the coil off the fan shroud housing and that has 15 ohms of resistance from shroud to battery ground, that could be a real problem. EDIT: coil has its own ground (not chassis grounded) - ground path still comes though engine case via points and dizzy to engine ground. So floating ground to shroud wouldn’t be so much of an issue for the coil.

The other thing that grounds through that fan housing is the alternator. Powder coating or painting the fan shroud without cleaning the mounting points to the engine case can cause this type of grounding issue to the fan shroud. Not sure what your situation might be causing that high of a resistance. I would want to understand this a little better to understand why that should is so far from engine ground.

Not trying to run you around endlessly on electrical hunches but they can’t be ignored when you have the unexplained issues you’re having. Even more so knowing that the car was once in a fire and knowing that fire history cars usually have unexplained wiring and ground issues.

When I was troubleshooting my 914 I ran grounds from everywhere to the battery negative. Shroud, block, transmission, etc. After everything was working great I removed them one by one.

Thanks Chris. You have done us a great service. I can imagine diaphragms made from roof flashing, with or without copper, being sold as replacements. You kept that from happening.

Here's the latest news:
I ran out of diaphragms last month. My supplier of Be-Cu stampings isn't going to make them for me anymore. It was 8 years since my last order of 250 pcs.
I may not have any new kits available for the remainder of 2024, until I work out a new plan. Right now I'm looking into an alternate material created to replace Be-Cu with a non-toxic alloy.
The material I found is also a copper alloy, with similar flexural properties after heat treat.

That is too bad. I hope you can find a new alternative.

Rats on the former supplier bagging out.

Anxious to see what you come up with for a replacement.

Bummer. I really appreciate you making these in the first place. I have 2 MPSs rebuilt with your diaphragms- one in the car, and one as a spare. Even though I don't need another diaphragm, I'd happily preorder one or two if it helps make a new batch happen.

Chris, I hope that you can get a new supply of diaphragms made to proper specification.
To OP, You may need to either get new meters or your current meter calibrated. There is little more frustrating than trusting a device that gives erroneous readings. You also need to make a list of your assumptions, and verify that they are true. Best is to have someone else do the verification; nothing worse than making the same mistake over and over, and I have learned that fresh eyes do make a big difference.

So, I did just read through this from start to finish. Your ground readings are a particular concern. You need to short the probes and get a perfect zero reading before starting. If you are tracing grounds all over the car, then you may need to have a 15 foot test lead of 18 AWG wire with a plug end suitable for your meter, and a ring connector for the other that you can clamp onto the battery post connector. I have found that I cannot trust the ground stud on the engine compartment wall above the battery. While a meter test may indicate a low resistance, when starting the car and 200 to 300 amps are flowing through that ground stud the true resistance may be much higher! Remember that the high current paths are the most important; from alternator to starter solenoid to battery positive on the positive side, and alternator through fan shroud, through engine case, through transmission case, through trans ground strap and body stud, through the body and body stud, and through the battery ground cable to the battery. This ground path is complicated, and every transition from one conducting part to the next connecting part has contact resistance in addition to the bulk resistance of each conductor. I rarely find an alternator harness that is healthy; I replace them all. I found on one car that the battery wall stud was not healthy so I connected a new ground cable to the fan shroud where a suitable 8mm hole was located; problem solved with a shorter and more robust ground path. Don't forget the ground stud near the relay board on the left side of the engine compartment; up to 8 ground wires depend on this point. The grounds and positive connections may have nothing to do with your problem, but it may not hurt to eliminate these potential problems to create a proper baseline. Also, try to compare your meter with those of friends. Find some test parts (inductors say, in the range you are trying to measure) and use them to help calibrate your meter.

Great post.

I'd add;

Resistance to ground on all grounding points should be 0 ohms. Those grounding points are on the factory wiring diagrams. Harness condition is key to a well functioning EFI system whether it's D-jet or other EFI systems.

The post demonstrates the value of a good testing procedure(s) to avoid chasing your tail.