Megasquirt and Variants Running Thread

[cordes]

Maybe the car is simply cursed. I say replace every piece of copper on that car and run dedicated grounds from each system to the battery! Are there any rust issues causing continuity issues with the chassis?

Jeff

It could be cursed. Frankly, that's becoming more likely than not as we progress through this.

In regard to the grounds, the negative battery terminal grounds to the head via a 2ga mil-spec wire. The ECU, fan, alt, and just about anything else I ran grounds at that point also. From there, a 2ga mil-spec wire heads to the stock ground location behind the battery. I have made a 1/4" hole there and I use an M6 bolt and washer with wide flanges to hold the large ground, and the few smaller grounds from relays to the body. I don't think I should have any problems with ground loops, etc. The area was thoroughly cleaned at multiple points during this ordeal.

I should be able to get the new crank sensor in on Saturday, and the MS arrived safely back at the house today. We'll see how it all goes.

[Ondonti]

It could be cursed. Frankly, that's becoming more likely than not as we progress through this.

In regard to the grounds, the negative battery terminal grounds to the head via a 2ga mil-spec wire. The ECU, fan, alt, and just about anything else I ran grounds at that point also. From there, a 2ga mil-spec wire heads to the stock ground location behind the battery. I have made a 1/4" hole there and I use an M6 bolt and washer with wide flanges to hold the large ground, and the few smaller grounds from relays to the body. I don't think I should have any problems with ground loops, etc. The area was thoroughly cleaned at multiple points during this ordeal.

I should be able to get the new crank sensor in on Saturday, and the MS arrived safely back at the house today. We'll see how it all goes.

Are you grounding all that stuff to the head location or back to the battery? While I do some "remote" star point grounds sometimes, the engine bay stuff I wouldn't shortcut anything. Unique grounds back to the battery. Then, work to keep trigger wires away from noisy items. The crank sensor and ECU should have its own dedicated power source/relay that comes directly from the battery. NOT from a branch. My P body is MS2 powered and I did a lot of wiring wrong on that, but the distributor signal is a light diode so the output is very robust. I still would redo things on the next adventure similar to my last wiring project. My last project, the trigger signal runs through a sensor branch that runs the front of the car instead of across the firewall so nothing exciting is going on. Its miles away from coils and injectors and alternator. My biggest problem with that car has been the alternator charging. Can't tell if its the silly mopar control box, connector, or junk Bosch alternator. My P body never had that problem, was probably wired/grounded wrong, but has a nice Nippo unit.

[cordes]

Are you grounding all that stuff to the head location or back to the battery? While I do some "remote" star point grounds sometimes, the engine bay stuff I wouldn't shortcut anything. Unique grounds back to the battery. Then, work to keep trigger wires away from noisy items. The crank sensor and ECU should have its own dedicated power source/relay that comes directly from the battery. NOT from a branch. My P body is MS2 powered and I did a lot of wiring wrong on that, but the distributor signal is a light diode so the output is very robust. I still would redo things on the next adventure similar to my last wiring project. My last project, the trigger signal runs through a sensor branch that runs the front of the car instead of across the firewall so nothing exciting is going on. Its miles away from coils and injectors and alternator. My biggest problem with that car has been the alternator charging. Can't tell if its the silly mopar control box, connector, or junk Bosch alternator. My P body never had that problem, was probably wired/grounded wrong, but has a nice Nippo unit.

It all grounds at the head where the stock ground for the alternator was. The battery negative terminal has one large wire which grounds there also. From there, it goes via a 2ga wire to the body. There are a few relays and front end lighting grounds which come in at the same spot.

The car fired right up and ran fine today with the exception of the O2 showing full rich the whole time and the alt wouldn't charge. At one point, the alt attempted to charge, but it essentially went full field so I turned the car off after about 2-3 seconds of that.

I plan on buying another ECU on Black Friday to have something I can test against. To me, everything seems to point at the ECU, but they say it's fine. I just don't see how I could have such intermittent problems with multiple inputs and outputs when at least the crank signal seems to be rock solid when view with an oscilloscope. It's all just baffling to me, but then I'm not an electrical engineer.

Here is a link to some files from today in case anyone would like to view them.
https://drive.google.com/drive/folde..._K?usp=sharing

[wallace]

Are you using the new sensor?

[tryingbe]

It all grounds at the head where the stock ground for the alternator was.

What? Are you NOT grounding the sensors to sensor grounds provided by Megasquirt but grounding all sensors and everything else on the head?

https://www.diyautotune.com/tech-tip...ng-strategies/

[cordes]

Are you using the new sensor?

Yes. The newer red one which they currently sell on their site.

What? Are you NOT grounding the sensors to sensor grounds provided by Megasquirt but grounding all sensors and everything else on the head?

https://www.diyautotune.com/tech-tip...ng-strategies/

Sorry. All sensors, to include the crank and cam sensors are grounded via the sensor signal ground. All power grounds go to the head and then to the body behind the battery with the exception of some relays located where the PM was. Those ground straight to the body where everything else meets.


ETA: I attempted to test the output for the alternator tonight, and I could detect no difference in anything regardless of what state I set it to in the test mode. I plugged everything back in and attempted to start the car. It cranked with no signal detected for too long, but eventually saw a signal and fired up. What's more, the alternator charged at a rock solid 14.2 volts. I let it run for a couple minutes to charge the battery a bit. I turned the car off, and attempted to restart it after waiting a few seconds. Of course, there was no crank signal again. This thing is driving me mad.

DIYautotune is suggesting I might want to install some pullup resistors in the harness. Is that something which others have had to do? I'd hate to hack into the harness just to have this be a wild goose chase.

[tryingbe]

Is MS driving your alternator? If so, I'd disconnect it and see how the rest of the car run.

[cordes]

Are you using the new sensor?

What? Are you NOT grounding the sensors to sensor grounds provided by Megasquirt but grounding all sensors and everything else on the head?

https://www.diyautotune.com/tech-tip...ng-strategies/

Is MS driving your alternator? If so, I'd disconnect it and see how the rest of the car run.

It does. That's not a bad idea. I'll switch it off tomorrow and see what happens.

[cordes]

Unfortunately, I didn't get around to turning off the injector E output tonight. I did however get a good picture of the waveform when it's working properly. I also got a log of the proper operation and one of the sensors being dead. Of note was the fact that when it won't start the car, the sensors both go to 5v and don't get pulled down as I crank. This is quite frustrating, and I'm up for any tests you guys might suggest. I have been emailing back and forth with the guys at DIYautotune, and I should think this new information will help them to guide me also. Frankly, I would think someone really good with electronics would be able to nail this down pretty easily.

https://drive.google.com/file/d/1-Ve...ew?usp=sharing,

https://drive.google.com/file/d/14Hw...ew?usp=sharing,

https://drive.google.com/file/d/1LnX...ew?usp=sharing

https://drive.google.com/file/d/1u3D...ew?usp=sharing

[cordes]

I did some tests today. When the key is flipped forward, the sensors go high, but in the time the fuel pump primes, there's some interesting stuff going on with the cam and crank waves. Why does it get pulled down repeatedly like that? I was able to duplicate it many times by just flipping on the key.

The only change I made today was to take all of the grounds and run them to the head. So now literally everything with the exception of the wideband gets grounded to the head. The wideband is grounded in the kick panel area where the stock ground is located.

The frustration level is not going down. Good times.

[wheming]

I did some tests today. When the key is flipped forward, the sensors go high, but in the time the fuel pump primes, there's some interesting stuff going on with the cam and crank waves. Why does it get pulled down repeatedly like that? I was able to duplicate it many times by just flipping on the key.

The only change I made today was to take all of the grounds and run them to the head. So now literally everything with the exception of the wideband gets grounded to the head. The wideband is grounded in the kick panel area where the stock ground is located.

The frustration level is not going down. Good times.

I know Innovate instructs to run the ground for the wideband back to the battery.
Any other ground point could have some voltage potential that could affect the sensor reading.

Sent from my KB2005 using Tapatalk

[cordes]

I might run that ground out to the head also. Might as well since I've done that with pretty well everything else. Thanks for the suggestion, Wayne.

[wallace]

Are you reusing any wiring or is it all new?
The tooth logs you posted are with key on engine off?

[cordes]

Are you reusing any wiring or is it all new?
The tooth logs you posted are with key on engine off?

It's all new wiring. I didn't spare much in the way of expense when it came to putting it all together. I used mil-spec wire, good crimp terminals, etc. etc.

Those two screen shots I just posted are the result of simply flipping the key forward while having the trigger set on the oscilloscope. It was very repeatable as far as a result, even if the signal does appear to be different each time. I presume it's the computer doing some tests.

[wallace]

You are using cam and crank sensors? After you swapped out the sensors your original problem persists?
If so Can you scope the crank/cam signal at the ms?
what other components share the 5v reference, sensor return circuits?

[cordes]

You are using cam and crank sensors? After you swapped out the sensors your original problem persists?
If so Can you scope the crank/cam signal at the ms?
what other components share the 5v reference, sensor return circuits?

Yes. That previous screen shot is a trace of the crank and cam sensors.

I'm now running them off of a switched 12v signal which also feeds the injectors. I was getting a similar, if not the same problem with the 5v reference signal. I need to email the folks at DIY back regarding the issue. They suggested that the computer should not be pulling the sensors to ground like that when it sees power. We'll see how this goes, but to me everything still points at a bad ECU.

[wallace]

Agree those traces shouldn't do that. A little noise is expected but that's definitely switching. Is this a 3.0 or 3.57 board?

[cordes]

Agree those traces shouldn't do that. A little noise is expected but that's definitely switching. Is this a 3.0 or 3.57 board?

I haven't gotten that deep into it. It's the MS3 pro mini, and they're fairly new. I got this one in June, so I may be a guinea pig on this. I bought another mini today along with the wiring harness, so I'll have another unit to test with sometime next week.

I did email the guy back and gave him some more info he requested. I'll see if he wants more oscilloscope action. This whole thing is a real pain compared to Joy's Daytona which fired right up and ran continually well. I don't have a ton of miles on it, but it has been rock solid every time with no miss ever.

[wallace]

I feel your pain. I've had excellent results with megasquirt. Twice I had problems one almost made me throw in the towel. Both times it was something I had done.

[cordes]

I feel your pain. I've had excellent results with megasquirt. Twice I had problems one almost made me throw in the towel. Both times it was something I had done.

I'm certainly not ruling out user error completely. I just wish the guy could give me some better tests to see if I'm the problem. I've made it abundantly clear that I have a scope with amp clamps, load pro, meter, etc. etc. but he just keeps saying it's most likely a ground. The ground wire is exceptional according to my readings on the Load Pro, so who knows?