I never quite finished it, but for my racecar, I designed, purchased, and almost assembled a system to deliver exactly 17.0V to run the ignition system on my racecar independent of battery voltage. For now those parts run on the 16V battery, which actually works pretty well. But I decided to hardware-fix the low voltage problems.

To the OP, I didn't see if you're running batch or sequential, but depending on the LS coil, as sonoftheirun11slikeitsnothinghill mentioned, a lot of LS coils have discharge problems with high dwell especially when batch driven.

 

Isn't 3.7 ms dwell pretty reasonable for LS Coils? Not saying that couldn't be the issue but just referencing what FlowForce recommends that seems like it should be pretty safe, no?

 

Alright, so I took some of these suggestions and these were the results. It should be noted that there were maybe one or two 2nd gear pulls that felt pretty good with possibly no issues. But the rest of the pulls had breakup. I feel that the issue is coming on strictly with boost pressure. In 4th gear I encounter it somewhere around 3500-4000rpms, and in 2nd gear I might not encounter it until around 6000rpms, it all depends on how quick the turbo spools.

@Simba, You are correct it was tuned in the fall. Thank you for posting screen shots so patsmx5 could view.

@patsmx5, I did not try pulling 10% fuel and adding 4 degrees of timing yet because I took another log after the dwell changes and the results were weird. And I am currently wired for batch firing.

Attached you will find the log with dwell table enabled. I originally set it to Flowforces recommendation but noticed that the dwell time wasn't any lower than my nominal 3.5ms with voltage correction to 3.7ms. So after a few attempts at their recommendation I lowered it another .5ms. No improvement.
Attached are the screenshots from two data logs taken last night and I was not low on my AFR for any of last nights pulls so I didn't want to try less fuel. I am not sure if this was a one time occurrence for the data log I gave you. If you still think I should try less fuel, I will..
In addition, I made a new coil harness just because I thought it looked sketchy before. So now I have dedicated ground and 12v from the battery to a new coil harness. The dedicated wires were already there for a sound system that I confiscated for the meantime.

I'll ask this with the risk of looking dumb lol, Has anyone heard of similar issues from turbo oil blow by? My EFR has always had excessive, in my opinion, blow by. I have seen coating the inside of my intake and charge pipes. The drain line is the proper angle and 10AM, I'm using a good flange that has a good ID size. I have dual 10AN lines on my valve cover to a catch can for crankcase pressure. But I do have a high volume oil pump so I wonder if my upper RPMs are causing blow by that is oil misting the air causing a bad ignition/combustion?

 

I should have included that I held the throttle for a little longer on these logs while it was breaking up. You can see that my AFR was 12.3 and the MAP was 250KPA, and then the AFRs go up to about 13 and my MAP starts going down, all before I let off the throttle.

 

Do you run a cat? Is it possible you have some sort of exhaust restriction?
I will say that both Pat and myself have researched coils and high boost ignition problems a great deal. I do think going over all your wiring is a good idea.
We also run LS style coils, but ours are more substantial. I am able to run my dwell at 5.5ms on MS2, but my coils can be dwelled excessively and are not true LS coils.
If wiring doesn't fix it, I might be inclined to try stock coils dwelled at 5ms. I think I remember AndyFloyd figuring out that we could dwell them pretty fat, and having good luck with them.

 

 

I do not have a cat. Allergic lol, metaphorically and actually. I am running krakens low mount cast manifold, and their full 3" with resonator and muffler.
I will try to add new wires from the ms3x to the coil harness tonight to completely bypass any original vehicle wiring. Maybe one of the solder joints in the db37 cracked or a wire rubbed somewhere.
I have been seeing your names pop up everywhere with conversations regarding coils, I appreciate your insights.
I am not sure if I have the factory coil connectors laying around anymore to fab something up to the current wiring but I will check.
I know used OEM LS coils are often favored over amazon/auto parts store new coils. Is that your contention as well? Or is there something new out there that's good? I am running the Flowforce LS coil bracket

 

Are you still on ms2 or if not what ecu?

 

I am on megasquirt 3 with the expansion pack.
MS3X

 

Try gapping the plugs down more, or lowering the boost. Looks like a misfire, probably from spark blowout. More timing, less fuel, less gap, lower boost, will all help get it to light off.

I've never messed with LS coils, but I have read there are good ones, china crap ones, etc. Do you have good ones? Are you sure?

I run some high quality IGN1A coils, new/custom wires/brackets, and I can fire off methanol at 8,000 RPM no problem with 30+ lbs of boost. These coils were 80 each I think. So not cheap, but they WORK. If you have a healthy stock coil, try that with 5ms dwell. A healthy stock coil (and new stock NGK wires) are actually pretty powerful. The coils/wires just don't last all that long.

Last but important, I would not run those coils batch.

 

I run the IGN1A coils, and am happy with them. I use the factory wiring to trigger my coils and to power the relay that uses 10ga wire to supply power to my coils. The IGN1A's have 3 grounds, all 10ga except for trigger/signal wires. Good crimps are extremely important.
I don't really remember all that much about the true LS coils, except don't overdwell them.

 

I agree with Pat about not running batch if you have MS3.

 

Patsmx5, I am gapped at .016 right now but I'll try a few thou lower for testing. And as far as lowering boost.. NO 😂 I need all the PSI's, I love them too much. You make good points and I'll see where I am at after this weekends tests.

Soneofthehill, that is some pretty serious wiring. But those are some serious coils.

I tried swapping the coils out for other used OEM coils about a month ago so I have been thinking my coils were fine.. I have access to a literal hundred of used OEM ones so I will try several rounds of swapping and see if my results vary. I like the idea of LS coils on a Miata but honestly there's too many problems with them for me to be a fan boy. If a coil solves my problem I will be going with the IGN1A next time because I've already had one coil failure last fall, which is why testing other coils was the first thing I tried when the breakup came back this year.

I will also wire the car for sequential.

 

I'll share what I went through, I don't know if any of this will apply to you or offer insight or help, but maybe it will. A few years ago I switched to speeduino ecu which I'm sure is more susceptible to electrical noise and missfiring than megasquirt, but some of the same principals may apply.

When I first switched ecus I was getting missfires at high boost and rpm which also presented as severe rpm spikes in logs. I tried everything. Several crank and cam sensors, several different types of coils, Alternator, power capacitors on coil power wires (which did seem to help slightly), multiple different types of plugs at every gap from .014 to .025 (tighter gaps also helped), many different grounding scenarios for ecu and coils. The last coils I tried were the bosch r8 coils, which are still on the car, but they behaved no differently at the time than any other coil I tried. Everything I tried offer no real change in behavior if the plugs had any miles or color to them at all. The only thing that would correct was new plugs and only for a very, very short time. Like 100-200 miles or so, or until plugs were colored. Since car is driven only occasionally, but also taken to the dragstrip, I would be sure and have new plugs in the car at every track outing, which would get me through the day. I just lived with that behavior for a while. I had many false successes as when I would change something I would often put in new plugs at same time. Of course wouldnt ya know as soon as plugs colored it would start acting the fool again LOL.....

Now to what actually and ultimately fixed this problem....
I saw a video on ecu grounding from I believe was original creator of the adaptronic ecu. He made a statement in the video about there being a very complex interaction between the ignition system and the crank sensor that can cause electrical noise interference, but he didn't really elaborate further on that statement other than potential ecu filtering. That statement really stuck in my mind.

So about a year or so ago I made up my mind to correct this somehow, someway. What I did to ultimately cure was to run an msd noise capacitor on the 12v supply to the coils, and make changes to the sensor wiring. What I did on the sensor wiring was to bundle all the low current sensor wires, cam, crank, tps, iat, basically any low current wires sending a signal to ecu, together. I then relocated that bundle to the opposite side of engine bay away from alternator, coils and coil wiring as well as injector wiring. So basically all of the high current wiring is on the passenger side of engine while the sensors and low current stuff is on the drivers side inner fender. Basically I separated the high current and low current wiring about as far as I could.
I also encased that entire bundle of sensor wiring in a steel braided sheild which is grounded on one side near the ecu. The crank and cam sensor wires are also shielded on there own, but still incased in this extra sheild that is over the bundle. These changes absolutely corrected for my issue anyways. Plugs in car are old with many miles and several trips to track with no issues. I have them set at ~.020 but could likely run clean at .025. I run e85 at mid 30ish psi through a gtx3576 sized turbo. I'm not sure exactly what change was the biggest contributor to being the fix, but I will say the msd capacitor does help. I did some work on the car a while back to where I forgot to re-attach the capacitor ground and got a couple missfires at high boost. Came back and attached ground and missfire was gone, still on those old plugs. Hope this offers some potential insight. Maybe it will help, maybe not. every custom build combination is different. What works for one may not work for another. Good luck though, I know its frustrating chasing down electrical gremlins.

 

I was just peeking back at the last logs you posted. I know you were just told to pull some fuel out, but at 5500 RPM and 240 KPA you're at 13 AFR which is way leaner than you want to be.

 

I agree. I am targeting 12.3 and before this issue arose I was achieving 12.3, so I am not pulling any

 

Did I miss the reason for the firing order 1-4-3-2?

I don't think it matters with batch fire, but thought I'd mention it. Dunno what else it might **** up.

 

I am not sure why the firing order was put in like that. It didn't matter however because tunerstudio only uses that section of firing order for when you set up trims.
It will always fire in the ECU's pinout of Ign. A-B-C-D order and it is dependent on the wiring in order to connect A-> cyl.1, B-> cyl.3, C->cyl.4, D->cyl.2.

I wired the car for sequential ignition last night and no luck. The problem is still there. I believe unchanged symptoms to where I have been at.
I was unable to test other coils but am still going to do that.

Take a look at this log from last night, I almost find the RPMs to be weird. I know the roads were a little rough and I may have had possible wheel spin but I'm not sure.

I am going to check ECU grounds to the head tonight. If I have any available, which I believe I do, I will add more. My understanding is all ground pins of the non plug and play megasquirt's should be used and connected to the head.

 

Did you try this?

 

Last night I was able to ground the ECU better -made no difference.

And then gapped down to .010" and I believe the problem WENT AWAY. It was late so I was not able to test thoroughly, but I did 4 pulls and only felt 1 momentary hiccup.

SO, now that I know a tiny gap eradicates my problem.. it has gotta be a coil right? I tested 4 other OEM GM coils at the START of this investigation because I initially had this feeling.

It goes to show how influential cylinder pressure is for the sparking system because I pulled the test coils off a 900hp N/A big block. They work fine in there with a gap 3 times bigger than mine.

I hope to be able to reply soon with a sticker of confidence on this solution.. I just need to find some replacement coils fist