Plug gap and MS resets
 

I had some time to kill today so I went for a random datalogging drive.

Three times, when I accelerated from zero directly into boost, I got resets! And that's with a .025 gap on new-ish plugs, new-ish 8.5mm wires, and working-but-generally-suck-ass stock coils.

I have an unopened set of FM's extra-long plugs. My idea is to try those with a .020 gap.

Thoughts?

(Please keep any technical feedback on the simple side. I'm still a lowly student of Prof. Paul :) and genuinely trying to expand my comprehension of MS fundamentals, without getting over my head or breaking anything - this car is my DD.)

Two questions:

1- What makes you think that plug gap is causing this?

2- Are these in fact processor resets (marked by a red line in the log) or just misfire events? A processor reset, when it happens, is a fairly violent event.

Let's see the log.

EK, check your boost cut off settings.

Log is attached.

When Paul initially installed the MS a few months ago, we had the same problem. We were using a gap of .03x and we kept getting resets under boost which (I think) he said were caused by spark plug "blow-by" (is that the correct term?) .... eventually we lowered the gap to .025 thanks to suggestions found here on MT and all was well. So, since I was in the car when those resets happened the first time, I recognized the same reset tonight. When I opened the log in MLV, sure enough, there were red marks and 3 resets listed. Artie -- no, it wasn't boost cut; this happened at around 8 or 9 psi if I recall correctly. My boost cut is much higher than that.

Last night I was searching around the forum and saw that some folks use a gap as low as .020, so that's how I got this idea today.

I don't know if using the longer plugs has any impact on the required gap.

Yep, those are definitely bona-fide resets- you can see the barometer go apeshit when it happens. Under "More Settings" go to "Barometric Correction" and set the Min to 90 and Max to 100.

As to the root cause... There's certainly no harm in your gapping the plugs down a bit, I'm just having a hard time imagining how plug gap is causing your problems unless there's something else very wrong with your ignition system- insulation missing from wires, degraded coil windings causing an internal arcover from secondary to primary, that sort of thing.

Is your MS set to activate any external solenoids or relays when in boost? EBC for instance, or some other custom gizmo? WI?

Will do ... hey, I checked just now, it's already at min 95 and max 105 ... pretty close to what you suggested (and presumably what Paul put in) ... hmmm....

Yes, there is a GM boost solenoid.

Not sure I can explain that. Looking at the log you posted, baro starts at 100, then after the first reset at 226.500 it goes to 155 (which is about what MAP was at the time) and then at 228.480 on the second reset it goes to 116.


I'm just trying to think of things that could be creating electrical noise within the MS, a common cause of processor resets. Do we have the flyback diode installed across the EBC solenoid?

The unit was built by DIY so I imagine they did if it should be there.

No, I'm talking about an external diode 1N400x), placed across the two terminals of the EBC solenoid in a reverse-biased configuration. IOW, the striped end connected to the +12v lead, and the other end connected to the lead which connects to the MS.

Give it a shot and see if it has any effect.

Probably wrong, but...

It could be noise from the ignition system interfering with it. I'm running EDIS on my 99' and I made some fancy MSD low ohm sparkplug wires. Well, I've had an undiagnosable miss the entire time. Replacing the ignition system (except for the sparkplug wires) doesn't fix it. I have noise like crazy and it gets worse with RPM (imagine that...). I'm 99% sure it's the low ohm wires causing the problem. I don't suppose you're running low ohm wires or Champion plugs are you?


EDIT: And you'r point???? My point was I get resets too. And I get crazy ass numbers in datalogs on almost every sensor sometimes. Moving grounds around moves the problems around, but it never fixes it. Seperating every ground helped the most. Then I put them all on a ground bar like what you would see in a breaker box. I have noise bad. I even added ground and cleaned ground and tried my damnedest to avoid ground loops. It's gotta be the ignition wires.

OK. will try that.

It better NOT be the frigging spark plug wires. I spent $83.95 + s/h on FM's Magnecor 8.5-mm wires because they're supposed to be good.

Are you getting the resets consistently now or are they just a fluke? If consistently happenening at 8-9psi I suggest unplugging the boost control solenoid to see if that solves the problem. You can build that much boost without it, right? Remember when we first were having the issue I suspected that damn thing but abandoned the idea after the regapping seemingly solved the issue.

Fluke.

I didn't have ANY resets since we changed the gap to .025. Then, 3 in one trip. None since.

If the solenoid is the problem, what's the solution? Some other brand of solenoid? Or a more traditional EBC?

Does any of this affect the big problem I'm having with engine misfires? Remember I had THAT problem BEFORE we put in the megasquirt ......

Did you have the misfires with those sparkplug wires before MS? If so, I suggest getting a new set of stock wires and seeing if that helps. I can't wait to get rid of my ignition miss... Sucks. I'm looking for COPs.

Actually, the Magnecor wires, new plugs, and new coils were some of the things we tried pre-Megasquirt to get rid the misfires in the first place! Then we thought the misfires were fixed (by replacing the injectors), so we went ahead and 'squirted....

Background: the misfires first started several months before I even considered 'squirting. Came out of a bar one night, started my car (stock ECU, 305 cc injectors), and there was a loud, scary-sounding metallic crunching noise, then my CEL started blinking and the car ran very bad. LOL, no I didn't throw a rod or anything. Stupidly, I limped the car home. Paul figured out that I had an injector go bad and the car was running on 3 cylinders. So we replaced the bad one and all seemed well. A month or two later, the roughness returned. That's when new plugs / wires / coils didn't solve the problem, so we put in a whole different set of injectors (I think 265s or something), instead of just replacing the one possibly bad injector. After that all seemed well again. Then a while later still, I bought the Megasquirt, and we replaced the 265s (or whatever) with 460s. For almost all of the time since then, I haven't NOT had a CEL for misfires. Oddly, in real life, the car runs pretty well as my daily commuter! But it does seem to vibrate a lot.

I have a nagging feeling that something other than an injector went awry in the bar parking lot that night. What it was, I don't know.

Paul: remember when we put in the 460s, we saw a small part broken on the fuel rail .... was that significant?

To answer your question about a solenoid solution Joe gave a possible solution, a diode across the terminals.


Small broken part on the fuel rail? I don't recall that at all unless you are referring to a plastic wire bundle thing.

In my defense Evan never told me about the "loud, scary-sounding metallic crunching noise," until about about a week ago.

Installed the diode today. $1.06 for the package of diodes from RadioSuck. Took less than 5 minutes, going slowly and carefully. Will keep an eye on it.

.1uf cap on js8?

JS8? Huh? Isn't that something on the Megasquirt circuit board?

I was talking about a diode that sends current from the EBC solenoid's output wire back to the input wire, as superbly explained (above) by J.P.

Doesn't DIY do that on all the current versions now?

What exactly is this capacitor specific to? IE, what circuit is it cleaning up?

yea i know, but wondering if the board has the filter for the CAS input.

My own thread is now officially over my head.

The diode across the EBC solenoid is to prevent voltage spikes.

The .1uF cap on JS8 filters out some noise on the CAS input. A MS flashed with the HR code requires this cap.

Brainey- he's getting processor resets, not mis-triggers. You can see the resets in the log, and there are no RPM spikes.

The JS8 cap will have no effect whatsoever upon this.


evank, it's been found some Miatas were getting noise leading to false triggers on the second trigger input. This problem seems to be more pronounced when running HR code. Placing a .01uf capacitor between JS8 and ground on the MS corcuit board seems to cure this. But I don't think this is your problem.

Okay.

For my reference, what would the symptoms be if this ever happens?

misses.

We solved my friends MSPNP reset problem with extra grounds.. About 8 of them. My home built v2.2 MS went ape shit one day last week and reset about every .0125 seconds... Then just, stopped. Hasn't reset since, though, once in a while I'm getting a misfire during cruise which I THOUGHT was resets until I checked logs.. Have a new set of plugs and MORE COPs coming.. I'll get that damn COP crap working if its the last thing I do!!

The diode trick didn't work. I confirmed that the diode is installed in the right directions, etc., but I still get resets at around 14-15 psi. So, I will try going to a smaller plug gap (.020). I happen to have an unopened set of FM's extended-length plugs which are supposed to give a fuller burn. Will try the gap with those instead of with my current plugs.

Made the aforementioned changes today (.020, new plugs) ... went for a ride while datalogging and didn't get any resets while holding 15psi. But the outside temp was only 50s. Won't declare the resets "fixed" until I get the same results in much colder weather.