Continuous ECU Resets (The Death)
 

So after having the car running on MS for a little over a month now I am all of a sudden having problems with resets. I took the car out for the first time in a week and while datalogging I noticed that i was getting resets on and off. By the end of my 15 minute drive I had about 32 resets total.
I just want to make sure there is nothing in the settings if MS that will cause resets. I'm correct in that assumption right?
My first course of action was to clean the stock ECU ground wire at the block because that is the same wire that MS is grounded to as well as my wideband. That didn't solve my problem.
So next is to take the wiring apart again and redo my grounds. The plan is to run a few more stranded ground wires from the engine block, bring them to a terminal block and attach all my grounds to that. That will for sure guarantee me that I have good grounds on that end.
Now I have read that there are other things that could cause resets, such as a noisy alternator, but I don't think that's the problem seeing how NONE of my past datalogs show a single reset ever.
Anyone have any other ideas on things I should do while i'm in there doing that?

You're starting at the right place-- it's usually either a ground issue, or interference from sources such as plug wires/coils or even fans-- I'd make sure all MS wiring is clear of any of these....

Is the ms has been constructed by you? if so, have you built the flyback circuit for low impedance injector? in the manual it says you can eiter run high impedance injector with the flyback circuit. I experienced it, runnig low impedance injector at first, but after a couples week, I retro-fitted high impedance injector and I started getting reset. My problem solved when I removed the flyback components off the board.

anyway, I would recomend to use shielded wires for all inputs and the sparks outputs. Also, add a 4.7uf capacitor across the tach input line. Add another 25uf cap, across the 12v feed line of your coils if the're not OEM.

Do you have a datalog of your ms shutting off? that would help to understand why.

Seeing how I ran the car the way it was for a month without resets I'm not sure I need those sort of things but if it comes to that I don't mind tackling those mods. First thing i need to look at is the physical wiring mess I have at the ECU.
Here's that datalog.
Added the MSQ as well just incase.

Here is what my datalogs used to look like. My fuel table was all messed up and EGO correction was all catywompus but you can see that there are no tack dip-spikes like I have now.

cracked solder joint?

That's what I'm thinking. I need time away from school to take my wiring apart.

This is the reason why I'm putting together a patch harness. That way i can leave the stock one alone for good.

Ever get this figured out? I'm having the same problems but right from the start =/

Check your grounds...

Run minimum 5-6 separate wires, 18-20ga, each from one of pins 7-19, doesn't matter which, all the way back to the same point on your block. That's usually it...

Jerry,
Is it imperative to run a whole bunch of small gauge wires? What about running just a couple 12 ga wires?

Also, must it be grounded to the block? I was planning on running my MS in parallel with my stock 97 ECU and mounting right there with it, grounding it to the chassis itself. It seems that running wires from behind the passenger seat to the engine for a ground seems a little unnecessary.

in your car it would be easier to run grounds straight to the battery.

It is better to run individual wires rather than one larger gauge wire (this is how the factory does it too), and the block would be the best source for your grounds.... many people have gotten away with less... I just always try to tell people what the absolute best option would be.

On the 1990-1993 pins 2A thru 2D are grounds if you're using the factory harness.

many in the dsm community were getting the same issue, figured out it was the ms case. Those with the issue were bolting it down therefore grounding it (?). Isolating it with rome rubber "feet" never gave me a problem and has fixed the issues for others since.

Problem solved :) it was the grounds. My WB02 is MUCH smoother now too.

I cleaned up the two that go to the block from the stock ecu, even though they looked pretty clean. I then added 3 more 20g wires to the MS harness and bridged them to the 6 remaining pins that were not already grounded so now I'm using them all. Hooked those up to the nuts that hold down the bracket for the ECU. My guess is that the weak point may be the pins on the harness and using all of them takes care of that because like I said the stock ECU wires looked just fine (also tested them and it showed 0 ohms).

Yay, now I can move on to spark too =)

Awesome-- glad to hear it man!

Well i thought I had all my ground problems tackled and then I took the car for a drive and had the g/f hold the laptop. Autotune was working great and all was dandy. Then i had her close the autotune screen and what do you know, 26 resets were shown in the bottom right corner. The weird thing is I'm not feeling these resets like a misfire. Here is what I have setup right now
I'm using 6 gnd pins on the DB37, 2 16 gauge wires go to 2 pins each then 2 18 ga wires go to a single pin each. Pins 13-19. They then go to a junction block with spades that has 4 wires coming from the stock ECU grounding point on the block. I cleaned the battery gnd, where the cable mounts to the body in the trunk, the gnd connection in the PPF. Then in the engine bay I cleaned the block to chassis gnd cable but could only do so where it connected to the body. So I added another heavy gnd wire from the chassis to the intake mani close to where the stock ECU gnds out just to make sure all was well.
I'm still having reset problems it seems. We were just cruising on the interstate trying to make sure all the cruise cells were good and Bam the resets of death come back.
Can weak coils cause resets by chance, I don't see how they could considering they don't provide power, though I do see how they could cause a misfire. Heck maybe they are tach spikes but i'm not sure. The CAS is grounded at the same spot on the intake as everything else as well and gets it's 12v from the fuse box.
I'll add my current MSQ and a datalog here quick.

Sounds like its the connection from the megasquirt to the laptop. It would explain why you dont feel the resets. I found once when my gf was holding the laptop for me she didnt pay attention to the usb plug and hit it a little which caused megatune to see a reset.

That's what I'm thinking. I wish I had socketed the chip instead of cutting the traces so i could try a DB9 cable instead of the internal USB adapter i have.
The USB is soo nice and convenient but I think it's giving me false readings. I've gone so overboard on grounds and cleaning them that I was going to kill myself if someone said grounds again. Maybe I'll try a different laptop or something.

I also cut the traces then decided to go back. Just solder on two jumper wires from chip to chip, worked fine for me.

Something worth trying I suppose just to make sure it's the computer. I hate taking the case apart with my EBC mod on the endplate and my USB velcro'd to the bottom plate. Oh well, life of a megasquirt'r.

I took a look at your data log, and I've confirmed those are *not* genuine resets, but rather a communications dropout. The tip-off is that in the event of a true reset, the SecL counter goes to zero and starts counting up. In a communications dropout, it picks up where left off before the glitch.

I'd try increasing the Interwrite Delay setting in MegaTune's configuration files, as described in the Communications Troubleshooting Page of the MegaManual. Let me know if that fixes it.

I'll give that a try and report back, thanks so much.

Hi I'm having the misfire same as described in a few different threads, that have been fixed by fixing the grounding.
Trouble is that mine is a PNP so the grounding goes through the standard loom and I don't really want to cut it up.
Do you have any ways of getting better grounds for the PNP?
Thanks

You run grounds from the MS board to the middle connector, and then from the middle connector to the head.

I believe the most recent PNP's have some additional ground wiring already provisioned internally, so you'd just have to run the lead(s) out from the center connector. Mine was from the very first batch, so it has had an additional internal ground modification performed.

My misfire issue has drastically reduced in tendancy since adding additional grounding between the MS and head. It's gone from misfiring many times each time I drive, to misfiring like once every few weeks. I can only recall one occurance this month, and I've driven it a lot more than I normally do. We also grounded my accessories (gauges and wideband) to the head through their own wire, instead of having all components share a common ground wire. Future plans involve running a direct ground from the battery to the head. I'll try to knock that out this weekend when the car's already up in the air due to a brake job. I think that will totally cure it.

With there being a couple models out now I'm not 100% sure which you're running based on your username here, but it's probably best to shoot us an email and we can get you whatever details are needed. Please describe the symptom as completely as possible (indications of the problem, when it occurs, etc). We'll get back to you really quickly... websales@diyautotune.com

To add to this.
Couple days ago i turned the key to the on position after letting the car sit all weekend, all of a sudden it goes BANG, like a backfire. No idea where it came from, but some smoke came out of the dash. So the smoke part is really weird and unexplainable. So i took part of the dash off and looked for an electric fire, i found nothing. Turned the key on and started the car, all was beautiful. It idled wonderufully with the 550s with the A/C turned on. Soon as i turned the A/C off it started surging with the AFRs bouncing back and forth between 11-dead lean. So i mess with it a bit and lean it out trying to get it to smooth up because the cells were super rich right there. So nothing fixed it.
I take a break and come back later. It does the same weird backfire thing, this time I have the laptop hooked up while i start it. All of a sudden MT starts going crazy with resets, i confirmed they are communications resets. I'm using the DLP USB adapter built into the case. So 25 resets in two minutes and I'm sick of it already. So i tear out my PNP harness setup and MS. Plan is to rejumper the leads i cut for the USB adapter and use an external adapter and DB9 cable. Hopefully that solves my problem. So i'll clean my harness up while i'm at it. Should have it put together by saturday. But what a ordeal, MS is a heartache. I'd be better off with a $10 shitty laptop and power inverter using a straight DB9 cable. Maybe I should try the palm.

Thanks for the help guys . I've sent an email to the address that FoundSoul asked. I will post to let you know what happens and hopefully how to fix the problem.
Col

I have recieved an email including pictures on how to increase the earthing on the PNP , and will have to pull the ecu to fix it. I will let you guys know how it goes when I am able to get the time to do it.

Shouldn't be to hard for you. It's worked for quite a few other people already.