MS3 Basic 6000 rpm backfire and check engine light
 

Gents,

I made some changes to my 1999 that warranted a new tune. Everything was going fantastic. I have not had any problems like this in the past. We were nearly done with the WOT tuning when it backfired at 6000 rpm and threw a Code 64 indicating a loss of sync. The Check Engine light came on. It happens at 6000 rpm regardless of throttle position. Otherwise, the car runs great. Starts and runs just fine below 6000, and even runs fine after the backfire. The Check Engine light does not come on if the engine is turned off then restarted, unless I crack 6000 rpm again.

I replaced the crank position sensor, but no change. Any ideas?

Thanks much,

poormxdad

How far out is the crank sensor placed?

This sounds exactly like the issue that we had on Eds car when I forgot to adjust the VR conditioning pots correctly. But from what I remember @Reverant uses a different circuit for the crank input.

He will be able to answer better.

It's your basic hotel room key card, ~0.031".

This has to be a setting. It runs just fine. Pulls really hard, then bang. I haven't tried to keep going over 6000, but I don't think it will. It's like it has hit a rev limiter.

Thanks much.

Same thing with my issue. VR conditioner pots were set incorrectly. It lost sync at 6k rpm. Then wouldn't restart until it went all the way to zero and re-synced.

My Friend, I have no idea what "VR conditioner pots" are.

I changed the pre-Rotrex intake piping by deleting the gutted MAF and adding a sealed cold air shroud. The restriction had been bad enough to require a retune. I gained nearly two pounds of boost, so there were extensive changes to the tune but everything seemed fine. When the backfire and loss of sync happened the second time, the tuner reloaded the previous tune. No change, so maybe it's NOT a setting.

This blows. I'm over my head and supposed to be on track Monday.

poormxdad

If you have to be on the track on monday then set the rev limiter to 5000rpm and go to the track.

Post your tune and I'll take a look.

A loss of sync is not from a tune.

Post a log

I thought about that, but if something gets worse I'm 300 miles from home. I haven't registered yet, so I'm not out any coin, just the joy.

Here's one of each, tune and log.

Thanks,

poormxdad

I see 4 lost syncs. 2 reason 32 and 2 reason 31.

Those are both cam signal issues. Have you tried replacing the cam sensor?

Still waiting on a reply from Reverant.

Taking a composite log of the issue would also help.

No. It's on the agenda for tomorrow if I don't get some kind of other definitive answer.

Thanks,

poormxdad

Take a composite log.

I can't post a log right now but I did swap in a new cam position sensor. No change.

Try this, see if it's fixed.

Can you say what you changed? I am just curious and to lazy to go through both tunes and figure it out myself.

Opps, sorry. Changed noise filtering.

A few things.

First, I can't find an option for Composite log. I thought it would be in Tools.

Second. When I was at VIR a couple of weeks ago, I ran over the gator coming out of the Oak Tree turn. It shook the crap outta the car and I got a check engine light. No drivability issues though. I came into the grid, turned the car off and back on and all was well. No apparent issues till on the dyno a couple of days ago. I'm checking connections now.

I've got Shadow Dash installed on my cell phone. I hooked it up on the way home from work a little while ago and although it went through all the right notifications--it told me it was connected, loading the file and was going online, it never actually went online. Symptom?

Thanks for everyone's help so far. Beers on me if the opportunity arises.

poormxdad

Composite logger is under the Diagnostics and High Speed Loggers tab

Pat,

Your tune made the dropoff and Check Engine light come on at about 4000.

Since I can't seem to figure out how to make a composite log, I made a log with the SD card I have hooked up to the MS. Not sure if it will help. This should be from Pat's tune.

Thanks,

I swear I don't have that tab. MS3 Basic.

Hmm, well that didn't work! Don't use that then, go back to your original tune. I don't use a MS2 anymore, but I remember having to mess with filtering when I had one. I don't know all the differences, but MS3 doesn't have a filtering tab as best as I can tell. I suppose the different input hardware negates the need for it, but not 100% on that.

The fact it made it worse is odd, maybe try less filtering?

Again, I don't know just something to try. I highly agree that logging it via composite logger, which is inside TunerStudio, is the best approach.

I'm a big dumbhead...

Jeez... I pulled my head out of my ass and here's the file

I thought I'd recap.

No tuning problems but I need a retune due to changes to the intake piping.
Retune going great, then a backfire at 6000 rpm and check engine light.
Next run up on the dyno produces the same backfire and check engine light.
Tuner reloads the tune I drove in with, and the next run has same backfire and light.
We both assume hardware, like crank/cam position sensor.
Car runs great otherwise.
I replace crank position sensor. No change.
I replace cam position sensor. No change.
Pat's tune moves the drop off down to about 4000 rpm.
We confirm I don't really know what I'm doing.
There's not enough beer...

Firmware version?

MS3 Release 1.2.0 2012228 23:46 GMT

Is it possible there is something in play now due to the increase in boost that was in already in the tune that wasn't in play before?

are you making more than 11psi of boost?

sounds like youre just simply hitting your boost cut limiter at 6K.

Max boost went from 7 to 9 psi, and I'm cracking through 7 at 6000 rpm now.

This is what the Rev Limiter page looks like.

your boost cut is set to 180kPa. Increase that to 200 and see what happens. Since you make boost linearly, the MS might think youre going to pass 180kPa and cut you, makes sense given your rotrex and consistent rpm where it cuts.

Yes. Isn't that ~26 psi.

no. 180kpa is 11psi.

This might help you for future reference; MegaSquirt^® and Turbocharging

I set 200 kPa. No change.

Given the change in Pat's tune, I turned off Noise Filtering. She rev'ed to ~6800, then did the same thing. I took a composite log.

Very nice. Thanks.

Is that a composite log of the issue happening? We need a composite log of the rev and shutdown.

Is it not? I pulled over, started the log, did the pull to where the dropoff occurs, then pulled over and stopped the log.

The car doesn't completely shut off. It acts like it hits a rev limiter, but the check engine light comes on. The previous composite log was done the same way, but the dropoff occurred at 6000 instead of 6800.

Nevermind. You did it correctly.

Interesting. The only thing wrong with it is there is an extra crank pulse. No idea why, but I'm guessing that it caused the megasquirt to lose sync and basically shit the bed until you restarted.

https://www.miataturbo.net/attachmen...ine=1460219139

I installed a new crank position sensor and no change. It's an Intermotor unit from CARQUEST. I have a new Mazda unit on hand I could try. What say ye?

I had a feeling that had something to do with it.

Based on the pic of the composite log, go space your crank sensor further away from the wheel. I bet that solves your problem.

But how? I'm not saying you're wrong, I'm just having trouble getting my noggin' around it.

Sooooooooooo, everything is going fine and I have a crank angle sensor fail. Shit happens. It's like a bad ballast on a fluorescent light. It's not completely dead, but not functioning 100%. I install a new replacement, but it behaves exactly the same way. How can the system that's my engine and electronics suddenly need an adjustment?

I've already made the adjustment, but I haven't taken her out yet.

Thanks much,

The crank sensor on a 99+ miata ouputs a square wave digital signal. But internally, the actual element that senses the changing magnetic field (caused by the teeth on your crank wheel) is a VR sensor. The signal from the VR sensor is a sinewave that varies in amplitude depending on RPM. The sensor itself has the circuitry inside it to take a sine wave of varying amplitude, and convert it to a constant amplitude square wave signal. Great design, it minimizes noise by putting the signal conversion hardware right next to the sensing element.

Ideally the sensor is spaced at such a distance that the sensor can pickup an actual pulse at say, 100 RPMs on the starter with a dead battery, and at redline, without the internal circuitry failing to count a real tooth, or, picking up a noise pulse and thinking that is a pulse.

It is possible that your sensor is so close to the crank wheel that a noise pulse (caused by who knows what- crank vibration for example) is enough to trigger the sensor. Moving the sensor further from the wheel will reduce this amplitude, potentially reducing the noise pulse's amplitude enough that it no longer triggers the sensor's circuitry inside to cause it to count that as a tooth.

If moving the sensor further from the wheel solves your problem (if), then you reduced the amplitude of the signal and noise, and the noise reduction stopped the false trigger and hopefully it can still see the actual tooth at low RPMs.

Sir,

Fantastic explanation. I appreciate you taking the time. But you miss my confusification. How could it happen with two different sensors? In other words, what would cause my car to suddenly need a larger crank position sensor gap (if that actually works) than it did before?

I don't have every answer, sorry. If moving the sensor fixes it, then I'd say maybe the old sensor vibrated and got closer, or the crank is now vibrating in a way that caused it to pickup a false tooth, or who knows what? From memory I thought .040" was the proper gap, and you said you used less than that.

I'm having the same issue here. Right at redline WOT pull. Stock ECU was behaving the same way before I installed my MS3. Pulled a code for cam position sensor over retard. Did a new cam position sensor along with a crank position sensor from o'reilys. Might try to go with an OEM crank position sensor and double check the gap. Shouldn't it be fine as long as it's within the FSM specified range for the air gap? I attached a log you will see how RPM drops right at about 6600. Engine doesn't respond to any throttle input until you switch gears and it syncs back up.

Which MS3 ECU?

built it yourself?
MSPNP-PRO?
MS Labs's MS3-Basic?
Mine?


I see I was totally off base with the boost cut once I actually looked at the logs.

Rev built MS3 basic but like I said this issue was there when I had the stock ecu in the car.

well on yours youre losing the cam sync.

What else could it be besides a faulty cam position sensor? I've already replaced the crank position with an aftermarket one and the cam position sensor with a new OEM one. I can try another new one? I hear that crank position sensor is a culprit since they are on the same circuit or whatever.

Alrighty then. Since the car was cold this morning, I swapped in the Mazda-provided, Japanese-built crank position sensor. All is well. No dropoff or CEL. She's a beast since the retune.

I hope CARQUEST will take the new defective CPS back. I had to wait for the Mazda part to arrive. The CARQUEST part was in stock. Seemed like the right thing to do to try the box store part. Lesson learned.

Sooooooooooooo, I have Noise Filtering OFF. Should I turn it back on?

Thanks for everyone's time and advice.

poormxdad

Turn noise filtering back on.

Did you try the carquest sensor with more spacing by chance? What is the new mazda sensor spaced at?

Glad you got it fixed. :likecat:

Pat,

I did not. Swapping that part out is an asspain, so I wanted to do it when it was cold.

I shot for 0.045" and got 0.043" when it was all done and tightened up. I'm thinking it won't matter. The CARQUEST part was defective. Interesting how it had exactly the same faulty behavior as the part that was in there before.

If CARQUEST gives me my money back, I may purchase another Mazda part for the spares box.

Thanks again,

I'm a fan of mazda sensors, I don't blame you there. Still, if the carquest was at .031" as you said, and new is at .043", that could have been the problem with the carquest part, it was too close and that sensor's internal circuitry could not deal with the signal at high RPMs. Goes along with your "interesting" part too. I would guess mazda's can work with the sensor close to the wheel, it would make sense for there sensor to have a "correct" setting, but also be able to work with human and wear induced tolerances.

I'm not taking it out again, dammit.

Gents,

Just had my first chance to drive her with the new sensor after turning Noise Filtering back on. The drop off and CEL is back, but this time at about 6400 rpm.

Are there any negatives to turning Noise Filtering off and keeping it off? I'll probably try a larger gap first, but this really has me shaking my head.

poormxdad

This worries me. I just put a new OEM sensor in and it fixed the issue I had like you are. Really hoping I don't have to be carrying spares around. Fingers crossed.