Base Ignition Settings

[hufferfuely]

yes...cylinders are 180 degrees apart.

The physical tooth spacing is spaced evenly... but the tooth spacing shown by Tachref isn't even. It is all over the place.

I was trying to use the second crank sensor on my setup....that was the yellow crank pulse.

[24c]

So you have two teeth, 180 deg apart, two cylinders 180 deg apart and it's a two stroke.

You need to fire twice per revolution, that's every time you pass the tooth, you can't use the missing teeth option AFAIK.

You might be able to use "independent dual ignition", if you move the second VR sensor to 180 degrees (opposite the VR1), but I expect grippo might have something better to suggest.

[grippo]

So you have two teeth, 180 deg apart, two cylinders 180 deg apart and it's a two stroke.

You need to fire twice per revolution, that's every time you pass the tooth, you can't use the missing teeth option AFAIK.

You might be able to use "independent dual ignition", if you move the second VR sensor to 180 degrees (opposite the VR1), but I expect grippo might have something better to suggest.

I think this sums it up perfectly. You have to get a setup that has 1 sensor and produces 2 evenly spaced pulses per rev. In this case use NON-dual spark, 0 wheel teeth. This would be equivalent to a standard distributor/ reluctor setup. If you use 2 sensors, 180 deg apart, then you would have to grind off one tooth, and then get only 1 pulse per sensor per rev. But this dual spark mode is meant for separate coils since you would have 2 outputs. To tie them together you would have to use an OR logic chip on a perf board. But if one of the sensors gave you one pulse, then leaving the other tooth on should give you 2 pulses and you wouldn't need the second sensor.

On your last tachref plot, the yellow trace looks promising since it looks evenly spaced. What is the time between the yellow pulses and does this time correspond to 1/2 a revolution. You would need an estimate of cranking rpm for this. If so disconnect the blue sensor and you should have the correct first setup above.

[hufferfuely]

Grippo,

I tried to run each sensor along independently. No matter which one I run, If I run one of them, I get the three tooth plot, If I run the other I get the four tooth plot. It is only if I run them together, that i get the good single plot with one pulse every time I need an ignition event. The problem must be in the way the processor recognizes the signals.

Is there any way to tell the processor to use the signal from the second sensor as both the tachometer signal and tach event, so it will have the correct rpm and fire twice per revolution?

-Hufferfuely

[24c]

The problem must be in the way the processor recognizes the signals.

The problem is your signals are garbage

You cannot run two VR sensors together into a shared ground apparently, something the MicroSquirt asks you to do. I placed some Zeners on mine to isolate the shared grounds, but now I just use a 1N007 on the cam sensor side. You don't need to with your set up as you_should_be using one VR sensor. Try it using grippos suggested "wheel" settings and try grab a datalog, at least you'll see some trigger±, resets or such.

The other thing the VR inputs are sensitive to polarity, so when you were running TachRef the polarity was affecting things.
I would disconnect one of the VR sensors, and remove it from the bike. I would then use TachRef on the one VR sensor and see what you get, you should get a toothed pattern from one VR sensor with your 2 flywheel teeth.

It might be worth your while to make a simple rig using your flywheel and a VR sensor of the engine, at least you can tweak easier.

[hufferfuely]

I think there is something wrong with my processor.
The stock ECU used these same sensors with no problem.

I just rigged up a 68-2 missing tooth setup on a non used starter ring gear on my sled. Tachref picks up the pulses fine, but i can't get the unit to recognize rpm. i will post a datalog of this along with one per Grippo's suggestion.

-hufferfuelly

[24c]

I think there is something wrong with my processor.The stock ECU used these same sensors with no problem.

OK, but just because a stock sensor triggers a stock ECU, unless the trigger pattern/processing info is in the MicroSquirt, it will not work, as they have to be compatible...eg I have a Yamaha Thunderace YZF1000, there is 8-1 teeth, but the missing tooth is not a full tooth...so it doesn't work with a Microsquirt, but if there was a software revision to take this long tooth into consideration, it would.

You can test this out by putting something in a cordless drill or router, using a recognised wheel pattern and VR sensor, and you'd see cranking rpms etc.

[hufferfuely]

Here are the .msq files and datalogs for the two methods.

[hufferfuely]

Here is the last one.

[hufferfuely]

I am afraid the only good fix is to filter the signal from one VR sensor so that the processor recognizes it correctly. Unless someone could modify the code so it utilizes only the first and fourth pulses (see attachment) as generated by my sensor.

Any ideas?