MegaSquirt losing crank signal

[Kershaw]

I'll check my dwell settings. I've looked at the tooth logs but I either miss the drop or it's not there. I'll post one.

[varg]

Did you run a tooth log while revving out and look for lost sync/pulse pattern? Just odd that it doesn't miss in neutral but will under load. I don't know if you have a single coil or not, but at 4,500rpm you only have 3.4ms of available dwell with a single coil. From a spreadsheet I made many years ago:

[Kershaw]

I changed the spark plugs to NGK 5122 gapped to .23 and added 42lb injectors. No change to the issue. A new CPS is arriving tomorrow.

[Kershaw]

Still chasing this problem. I can go past 4k. If I am not at full throttle, like 70% throttle, then it'll cut out at 5k. In neutral the rpms will go up to 7k full throttle which is scary because the redline is set at 6300rpm. So I don't know what that's about but I'm guessing it's a separate issue. I bridged VR- in J5 and VR+ in J6 with a 10k resistor as well as gave R9 2 turns counter clockwise. I'm assuming that R9 is the crank sensor potentiometer, even though it is not labeled as such. See attached picture before the shunt was added.



I'm assuming since I get throttle up to 7k in neutral, it is not a timing issue? The AFRs seem very stable at 12.9/13 AFR @ 8 PSI at full load when it cuts so I am assuming it is not fuel related even though I am only running 19lb injectors at this time. Since I am assuming it's not timing, not fuel, I am looking at spark. I changed the dwell setting from 3.7 to 3.2 and instead of a hard cut at 4k it was a short cut and then it kept climbing and cut at 4.5k again before I could back off. So I bought NGK 5122s/BR7Es spark plugs and will install those and see if that changes anything. I will gap them to .23. Of course, with CX Racings beautiful but large turbo manifold, I am assuming I will end up having to remove the manifold to replace the spark plugs. Joy.

Any further suggestions are appreciated.

[Yuri]

Just went to the dyno and unfortunately had the same issue at 4,500 rpm. Glad I found this thread! Sent from my iPhone using Tapatalk

[reelizmpro]

Strange both of ours had dents in the side. New BMW sensor is on the way from ECS. Yeah there's no O ring. That must have been a reference to another engine.

[nando]

I don't think there's any o-ring - there's nothing to seal, the sensor is on the outside of the engine. probably, sensors that come with an o-ring are generic, meant for use in multiple cars where the sensor could be installed in the block.

I've been fortunate to never have this issue, except early on in my ownership of the car when the clips for the wire were missing and it rubbed through the sheathing, shorting out on the water pump pulley. That is probably worth checking, as every E30 has had many water pump/belt changes over the years, and often that clip gets broken or thrown away, people not realizing how important it is.

That also means I bought a new OE sensor when they were "cheap", lol. Going on 10 years with my Megasquirt-3 and it's never given me a single problem!

[Digitalwave]

Barry, the one I had problems with tested fine but also had a dent on the side. A new metal OEM CPS fixed it immediately.

[reelizmpro]

Glad this was bumped. I've been trying to figure out the cause of my drop out a while now. My CPS resistance and gap is spot on but has a dent on the side and the wiring is stiff and old. I installed a plastic one which gave me problems almost immediately. Looks like I'll be ordering a new OE BMW one soon.

I read about an O ring that's supposed to go on the sensor but don't recall ever seeing one.

[Digitalwave]

Air gap should be 1.0mm +/- 0.2mm. You can use a feeler gauge on it easily. Clean it up, gap it, and if it still occurs, buy a new OEM (and only OEM) CPS and I bet it solves the issue.

[ForcedFirebird]

It takes no time. Use a 1/4" 13mm short socket on a 1" long 1/4" extension and it will fit behind everything in situ. Loosen the bottom bolt a turn, and just loosen the top a smidge. The bracket rotates on the top screw, and most likely the sensor will get sucked into the relcutor wheel from the magnet. Then I will slip a business card between a tooth and the sensor, and tighten it back up. If that's not the issue, check to make sure the wire isn't rubbing the water pump, and that it isn't breaking where the wire meets the sensor. We don;t even play on the race cars anymore. We buy a new OEM BMW sensor (up to $200 clams this year), and toss them as soon as the wire starts getting stiff. Nothing worse than chasing issues at the track.

EDIT: And YES, clean that thing. Those ferrous shards can cause issues.

[Kershaw]

I suppose I could try cleaning it a bit. Not sure how much closer I can scoot the sensor but I will try playing with the air gap.

Too bad you're not closer. I really don't want to deal with it. I just want to race it. I don't have the time anymore to chase down things like this with work dominating my life now.

[ForcedFirebird]

I am getting break up around 4500rpm as well and I see everyone mentioning adding a resistor. But I really haven't seen anyone mention how to add the resistor. Above post mentions inline, am I really cutting the cps wire and adding it in there? I'm fine with that, but I want to be sure before I do so.

The above post says "series" and "shunt". Series is cut the wire and add in-line, shunt is cut both wires and bridge bare wires with resistor, but keeping them isolated from each other.

What sensor and what air gap are you using? A lot of the newer plastic sensors have a resting resistance of ~900 ohms, yet still work with the stock ECU just fine. Haven't logged the differences, but often just simply messing with the air gap has solved this for me. Over the years, even in a professional installation setting, I have not installed a single resistor on an m20 with MS.

[Kershaw]

I am getting break up around 4500rpm as well and I see everyone mentioning adding a resistor. But I really haven't seen anyone mention how to add the resistor. Above post mentions inline, am I really cutting the cps wire and adding it in there? I'm fine with that, but I want to be sure before I do so.

[Austrianvespaguy]

So yes there are various combinations that can probably made to work, but I now completely understand these circuits and can assure you that using a single inline resistor with the Microsquirt VR conditioning circuit is just plain WRONG (even if it 'works'). The reason for this is the MAX9926 chip uses differential input, meaning that it's looking at BOTH the VR+ and VR- signals relative to each other and NOT relative to ground (aka single-ended input, which is how the LM1815 works). So putting extra resistance on one signal and not the other will skew the zero-crossing reading, which is bad. The shunt resistor between the two wires keeps both signals 'even' but still drops the peak voltage generated at higher RPM speeds.
Also the MAX9926 chip is really the best of the bunch; so if you set it up correctly with the shunt resistor, it WILL work for everybody with that setup, and it will take out the touchiness of individual sensors, air gaps, etc. It's a really robust circuit; not that you want to bother, but with the shunt resistor I'll bet you it would also work perfectly with your OLD sensor too!

EDIT: And Malcolm was really on the right track at first by bridging the SB2 on the board to use its built-in shunt resistor, but I suspect that 5k resistor is intended for 'other' trigger wheels and the 60-2 wheel needs the 10k resistor as indicated in the Microsquirt manual. Re-reading Malcolms last post, I think he has done things correctly by putting in a 10k shunt resistor in place of the 5k, but hopefully he'll confirm this for us!