MegaSquirt losing crank signal

[Austrianvespaguy]

I HAVE SOLVED THIS FOR EVERYONE!!!
But the solution varies depending on exactly which Megasquirt setup/VR condition circuit you have, so make sure you understand this! @Malcolm206, as I understand it you're using MICROsquirt v3, and if that's correct, you are not done yet! So here's the rundown:

1.) If you are using an LM1815 chip for as your VR conditioner (found as an optional extra on some PCBs such as the DIYPNP), then you want the 18k resistor in series with the VR+ sensor wire (to limit the current that the IC needs to sink).

2.) If you are using MEGAsquirt 3.0 or 3.57, then you want the 10k resistor in series with the VR+ sensor wire (to filter out noise at higher RPMs).

3.) If you are using MICROsquirt, then you want a 10k SHUNT resistor that goes BETWEEN the VR+ and VR- sensor wires! (<- This really solved my problem and is the proper setup for the MAX9926 chip on the Microsquirt board)! See page 47 of the Microsquirt Maunal here!

http://www.msextra.com/doc/pdf/Microsquirt_Hardware-3.3.pdf

Malcolm and DigitalWave, I'd suggest you double check which VR condition circuit you are using; if you have the white Mirosquirt daughter card that looks like this:

http://www.microsquirtmodule.com/micromodule_V2.2_front.jpg

THEN YOU NEED THE SHUNT RESISTOR IN SOLUTION 3!!!

[Austrianvespaguy]

Well a little more info for the general knowledge-base here; I'm using the DIYPNP board w/Microsquirt card, and it has TWO different VR signal conditioners on it. The Microsquirt uses a MAX9926 chip, and the DIYPNP board has an extra optional LM1815 chip. The MAX9926 wants the ~10k resistor on the input for noise reduction, and the LM1815 wants the ~18k ohm resistor on the input to lower the current that it needs to sink. I made the small mistake of installing the 18k resistor with the MAX9926 circuit, and I think it was blocking too much of the signal. Just switched down to the more proper 10k, and things are certainly improved. I'm not closing the book on this one just yet, only tested briefly and think I still dropped once or twice at redline, so might go with an even lower resistance, but looks like I'm on the path to improvement now!

[Austrianvespaguy]

Ah, I didn't realize it was mounted to the rubber part of the pulley, good call! I have checked it's resistance (575 ohms at the sensor connector) but I'll verify at the 55 pin connector also. My gap is pretty small, about 0.5mm and no I don't really see much runout, but I do see what I think is a very small amount in the tooth logger.

[Digitalwave]

Here's my tooth log. You can see the extra long gap in the middle that's then only followed by 57 instead of 58 teeth. Also, this problem seems to worsen as engine temperature increases, which I really can't explain. Revs fine to redline dead cold (though I don't like to do that!) Here it's about 190 and cutting out around 6000 RPM, and fully warm around 210 it cuts out at 4500-5000 RPM.
Hmm, attachment not working, so here's a link to the screenshot:

https://drive.google.com/open?id=1uC1cpuL1KncqjAmVwZYogu-rwSgjyIWz

On a final note, is the small wave along the normal teeth just a bit of runout on the wheel itself?

Since yours is temperature dependent I would consider that it could be the toothed wheel itself. It is mounted on a rubber damper that degrades over time. When your car is running does it look like the toothed wheel has a lot of run out? Have you tried other/new CPS's? Have you ohm tested yours both at the sensor and at the 55 pin connector?

[Austrianvespaguy]

Good stuff! Glad we were able to get you're solved! Now if I can just figure out my issue. . .

[Malcolm206]

I was under the impression that the resistor on the board was 25k, but after reviewing the schematics realized it was actually only 5k. I added a 10k resistor and that seems to have fixed the issue with the CPS.


Sent from my iPhone using Tapatalk

[Austrianvespaguy]

Here's my tooth log. You can see the extra long gap in the middle that's then only followed by 57 instead of 58 teeth. Also, this problem seems to worsen as engine temperature increases, which I really can't explain. Revs fine to redline dead cold (though I don't like to do that!) Here it's about 190 and cutting out around 6000 RPM, and fully warm around 210 it cuts out at 4500-5000 RPM.
Hmm, attachment not working, so here's a link to the screenshot:

https://drive.google.com/open?id=1uC1cpuL1KncqjAmVwZYogu-rwSgjyIWz

On a final note, is the small wave along the normal teeth just a bit of runout on the wheel itself?

[Austrianvespaguy]

Oh and to the OP, your graphs certainly look like your problem is not seeing the missing teeth gap properly (sorta opposite my problem), but for this the 18k (or 10k) resistor in series with the VR+ wire should be the first fix you try if you haven't already!

[Austrianvespaguy]

My screenshot is of the composite log instead of the tooth log, so that's why it looks a little different. The red graph along the bottom is the de-sync counter. Next time I fire up the car I'll grab a tooth log for comparison, but basically it'll show the missing teeth section as being too tall of a bar and then only one too few short bars in between.

[Digitalwave]

^ Is your crank wheel set up for rising edge or falling edge? The tooth log should be looking like this:

[Austrianvespaguy]

Interestingly I'm currently struggling with the exact same problem on my DIYPNP setup (which includes an 18k resistor in-line with VR+). From the composite log file, I can see that the problem stems from the system only seeing 57 instead of 58 teeth and then getting thrown off, but I can't for the life of my figure out WHY/HOW it's missing a tooth, as all the timings look pretty good! Also strange is that sometimes, it all works fine up to my 7000 RPM redline limit, but others (like now) the cutout happens right around 5000 RPM. I *think* it might be a bit temperature dependent, but not entirely sure on that one yet.

[justinpirie]

I had to add a resistor in line to get mine to work. From memory I think it was 10k.


Sent from my iPhone using Tapatalk

[Digitalwave]

I tried a bunch of different resistors and didn't find much if any change. You do need a 10-20k resistor based on most of my research. You can read my experience here: https://www.r3vlimited.com/board/sho...d.php?t=418713

Have you tested the resistance of the CPS @ the 55 pin connector inside the car? It should be 540 ohms +/- 10%.

[Malcolm206]

First try different resistors (you did already). Next replace the CPS with a new OEM metal CPS and set the air gap properly (1.0mm). MS is picky about how good the signal is, much more than Motronic. If those both fail to fix the issue, start tracing and testing your CPS wires in the engine harness.

I’ve tried three different used CPS units, and the system still cuts out at three similar but slightly different RPMs. I’m currently waiting for my new CPS unit.

Has anyone had luck adding a 10k, 17k, 20k, etc. resistor in line for the sensor, or is this issue most often stemming from MegaSquirt being more particular about a properly operating CPS sensor?


Sent from my iPhone using Tapatalk

[Malcolm206]

Odd, perhaps the bracket is bent/damaged. Of all the m20's we have worked on the bracket adjustment is enough to make the CPS actually touch the relluctor wheel.

My mistake, when I was tightening the bolts on the bracket I wasn’t holding it firmly enough and it would rotate away from the wheel. I now have it set to 0.88 mm away from the wheel, which didn’t change the issue.


Sent from my iPhone using Tapatalk