Seemingly random Sync loss
 

After getting car (99) back together, I am re-tuning the VE table. However, I am getting what seems to be random sync loss.

Any temp, any load, any RPM.

Cranks fine.

I have never used the high-speed logging as car ran for 18 months with no real issues. Only time I lost sync was when I got a back-fire, before I got the starting tuned properly.

MS3-Basic.

Things I have done: Added COP's including re-wiring the Ignition and Injection leads so they don't go to the front of the engine and then back and then forward. They just come through the firewall, through connectors, and then forward to either COP's or injectors.

To that end, the grounds are now at the back of the head, rather than the front.

How should I proceed to diagnose this?

https://cimg6.ibsrv.net/gimg/www.mia...4d9320c53d.png

Hmmm, wonder if the glitch in the TPS at the same time as the killed RPM is a hint?

Bad ground connection- all your sensors go haywire at that point.

Update:

More logging and I saw that I was sometimes getting erroneous TPS and MAP signals (greater than 100%, and more boost than possible, respectively). This solidified the ground connection.

I had the sensor ground at the engine under a VC bolt. So, I moved it to the intake manifold, as I should have originally.

I have attached a truncated log, taken after the ground wires move. It shows a sync loss at the beginning and at the end. I am no longer seeing the drastic sensor changes with the sync losses, but the sync losses remain.

My next step will be to open up the DB37 harness and look for issues. However, before I do that, is there any way to look into the logs and determine if I have cam sensor or crank sensor issue. Or perhaps take a cranking high speed log and look at the signals.

Regards,

General rule of thumb -- if the car was working fine, you changed something, and now it's not, then the problem is the thing that you changed. It doesn't matter how wacky, off-the-wall, and unrelated it seems, it's almost always the thing you changed. :)

When you rewired it, did you touch the wires for the cam or crank sensors? They're shielded wire from the factory, right? Did you perhaps disturb the shields?

If you turn on the composite logger, it turns the MS3 into a logic analyzer for those two signals. You should three lines, one crank, one cam, one error. There should be four pulses on the crank, followed by one on the cam, followed by 4 more on the crank, followed by 2 on the cam, and then that pattern should repeat while the car is running. The MS3 will trigger sync loss if it gets missing or extra crank pulses, or missing cam pulses (It doesn't care about extra cam pulses). When it declares sync loss there will be a red "pulse" on the third line on the chart.

--Ian

Good rule of thumb.

I did not purposefully touch either one. None of the sensor grounds were touched either, on the factory harness side as far as cutting and splicing goes.

Here is my sync log. Could not find a comparison, though I know I have in the past. What is the red stuff at the bottom?

Is it generally correct?

https://cimg5.ibsrv.net/gimg/www.mia...1fcf6e5ff1.png

Generally correct. The red bits are sync errors. There should not be that many, but at least they vanish.

Could you post a longer log with fixed grounds please. Also it looks like you don't have your gearing set up correctly in TS....

I did check my Crank Sensor clearance. It's fine.

I have not set up gears in MS3, as I have no need yet.

Attached is longer log from this afternoon.

You want to post the composite log csv file, not the msl.

In particular, it would be interesting to see the bit of the composite log where the red bits *start*. The bit you displayed above is when the sync error ends. Green is cam, blue is crank, and you can see the 4-1-4-2 pattern I described above.

--Ian

You're getting cam sensor errors.
Check how you adjusted the resistor for CAM sensor input. I had the same thing. One more turn on the pot and it never happened again.

Ian: I only have this composite log, which is during cranking. The pic I posted was at the beginning. I just turned off Cranking Pulse and turned it over.

You can find the attached CSV. In the morning I suppose I can run a high speed while the car is operating. Was not sure that was possible.

Stefanst: I did not build or set this MS up. Not sure what pot to turn. This is Reverant built unit. Definitely I will trouble-shoot my harness first as taking the ECU out is hard on this old man.

To that point, why are you saying it is losing Cam sync?

I've been having a similar issue, replaced the crank position sensor and even replaced the cam position connector with a Deutsch connector.

OK, so there's no sync at the beginning of that log, presumably because the engine has just been started. To learn why you're getting sync errors you need to have the composite logger running *while* the sync error happens and capture it.

How old is the Reverant-built unit? I have a very early MS3 that he did which had a bunch of random sync error problems, but everyone with later units has reported great success with them. I dunno why mine is different, perhaps it's just a bad component.

Do you have noise filtering enabled in the Megasquirt? Look under "Ignition settings" and "Noise Filtering".

The Megasquirt generates a "sync loss reason" code, which you can read after a sync loss has happened. With the Miata it's one of two values, either claiming cam or crank problems, but IME it's actually not very useful because I've looked at the code, and it can't tell the difference between a missing cam pulse or an extra crank pulse that showed up before the cam pulse did. With noise on the input lines you can get either symptom, so that value is pretty useless IMHO.

NB1 cam sensors are a fairly common failure point, especially with heat. They're also relatively cheap and easy to swap (one bolt and it's on the front of the valve cover, so it's probably the single easiest thing to swap in the entire engine bay). It's probably worth just buying one and swapping it out to see if it fixes the problem. If not, put it in your spares box, because you'll need it eventually anyway.

In my case, the problem was that the hardware filtering that was happening inside my MS3 was not doing a good-enough job of taking out the noise. If I turned off noise filtering in software, the thing would barely even start because of continual sync errors while cranking. I fixed it by bypassing the filtering circuits that were shipped with it and installing my own RC filters instead.

--Ian

I have ordered a Camshaft Sensor. it still does not make sense to me that I get a very short glitch every now an again. Saturday I will start dismantling my DB37 harness.

Meanwhile ,
https://cimg6.ibsrv.net/gimg/www.mia...1f5c008de8.png
here are my filter settings:

NB1 cam/crank sensors shouldn't need any noise filtering. I have MS3X on NB1 with both turned off and have 0 issues.
However: Reverant units have non-standard hardware and code, so my experience with this is of limited use. Talk to Rev himself, he may be able to tell you, what settings are best to use.
This is also why I can't tell you which potentiometer to turn- different hardware AFAIK.

Edit: Your latest MSL posted shows Sync loss reasons 31 and 32. They are:
31 = Miata 99-00 - 2 cams not seen
32 = Miata 99-00 - 0 cams seen

So looks like cam sensor problem. But if it's the sensor, the hardware it's connected to, or the firmware misinterpreting another problem can't be deduced from the information we have.

The MS code always generates those codes for sync losses on NB Miatas -- those are the only codes it has. It works by counting 4 crank pulses then looking for at least one cam pulse, then 4 more cranks and at least one more cam pulse. It ignores cam pulses that appear in places it's not expecting them, it ignores duplicate cam pulses, and it doesn't even lose sync if it gets a single cam pulse instead of a double (this check was taken out to make it more tolerant of the VVT motor). This means that it can't tell the difference between:

crank - crank - crank - (fake-crank-from-noise) - crank - cam

and

crank - crank - crank - (missed crank from noise) - cam - crank - crank

In the first case it miscounts cranks because of the noise-generated fake pulse, and sync errors when it doesn't see a cam pulse before the 4th (real) crank pulse. In the second case it has missed a crank pulse, ignores the (correct) cam pulse because it's coming between what it thinks are crank pulses 3 and 4 and then generates a sync error because there's no cam pulse after what it thinks is the 4th pulse.

Anyway, the tl;dr of it is that you can't trust these error codes to mean anything. You need to look at the signals, either with a composite log or an oscilloscope.

--Ian

Installed a new cam sensor. Did not solve problem.

I have also determined that the Innovate G5 AFR gauge is toast. The controller is outputting correctly, but bench tested, the gauge powers correctly, lights correctly, but runs to 18 with about 1.0V input, rather than the expected 4V.

Sometimes, when I get the fault, the fuel seems to continue flowing because, upon spark return, I get backfire.

Any thoughts that this could be caused by an intermittent COP? And the sync loss is an effect, rather than a cause?

It could also still be the crank sensor- as Codrus explained MS only has the cam sensor error for the NB Miata. So it will throw a CAM error, even if it's a crank error.
Have you talked to Rev yet if the sensitivity of your MS crank and cam inputs can be adjusted?

I have not contacted him yet.

If this was a low RPM problem, or a high RPM problem, I could see how a sensitivity adjustment could be the root cause, but not so much a random mess-up.. But I have not found anything that seems to correlate as to when this occurs. Correct me if experience has show otherwise.

Sure it could be crank sensor. Throwing parts is at something is a poor Mensa of diagnosis, but I realize sometimes that is how to go.

The only way I know how to diagnose this further without throwing parts at it is by capturing a composite log WHILE the problem occurs. Other than that all you got is to guess at problems and swap parts until it's the right one.