Electric Noise PIC problem

[mr.sneezy]

Hi,
Many many years ago I did the development for a GoKart (the racing kind) datalogger and instrument system, based on a HC11 In the R&D phase we spent more time in EMI suppression than in any other part of the design. We tried dozen of supression thechniques before we had a device that would run for more than half a lap once mounted on a Kart... Spark ignition environments are just EMI hell !

[Tobias]

EMI is the bitch of it all. I have a MSD ignition on my bench with a coil and test plug. The ignition and the coil is just like in the race car. I can fire up the MSD and have the board right next to the coil and all is fine, in the car it fails. The problem I am running into is that the wiring on every car is unique too. Some guys aren't even running a ground wire to the battery...they rely on the chassis to carry the load. I do have hope in that there are other electronics like data acq boxes that do work.

[Jerson]

I do not have too much of exposure to car electronics, but the one time I did work on a car was with the central locking. Rev the engine and all hell broke lose. The fix was to plant a simple half wave diode rectifier followed by a smoothing cap to supply my board followed by the normal regulator stuff. Along with this, there were some kind of coils fitted on the spark plugs that reduced the noise immensely.

Maybe I was plain lucky that my board worked after those changes or maybe it can lead you onto something.

[Dave]

Tobias , Once again, Where are the pullups for the inputs? And what is the purpose of R22 and R25? The schematic looks the same as before...

Dave Purola,
N8NTA

[Tobias]

Dave
I am using the weak pull-ups on the first board/second posted schematic. Maybe you didn't see the second schematic

[mr.sneezy]

Are you actually sensing RPM from the spark source in your project (not sure if your posted project schematic is all of it or just some of it) ?
If so I might be able to help suggest some things. If not, then I'd suggest you try initially powering the project from an independent battery source (dry cells etc) and see if your problems go away. If so it gives a clear indication of whether the source of EMI (if that's the issue) is coming in via the cars own power supply.
Also be aware that any long sensor pickup lines in the car are potential receiving antennas for EMI, and use inline RF chokes before they get too close to the MCU.

[Tobias]

Yes I am sensing RPM on this board. I do have all inputs and outputs optically isolated from the board and I am using a DC/DC converter. The last few days of thinking about this problem, I am wondering if the switch supplying the signal to the WOT is 'bouncing' It is pretty much a generic micro-switch under the throttle pedal. The main reason I am wondering, is that the On/Off input is the same voltage but through a toggle switch. I sent out a board yesterday to test with a 22uf cap on the WOT trace. Hopefully that will fix the problem.