MicroSquirt.com read only forum (Jan.15,2018)

OK I checked out your msq and found the folowing:

1. In Advanced ignition menu, you have the advance offset = 180 deg. You don't say this is an oddfire engine, so I assumed it wasn't and set it to 0. This represents the extra advance to be added to all the table advances when the second ignition output is fired - it is only non-zero for an odd-fire setup.

2. I set your trigger offset from 180 deg to 0 deg. In dual spark mode you must use a negative offset - but not one that large. It's best to keep it around 0 to -20 deg and use delay_teeth to get you in the ball park.

3. You set injection timing to 180%. I believe you interpreted this as how often to inject. It doesn't mean that - it means when in the tach cycle to start the injectio - for example, if a 180 deg tach cycle, a 50% injection start would start injecting 90 deg into the cycle. (This is not affecting your ignition, but could screw up your fueling - although I believe there is a program limit on it. In any case it should be left at 0 % until everything is running ok, then you can play with it to see if it improves idle.)

With just these changes, everything seemd to work fine on the bench. Outputs were consistent, evenly spaced the right number of teeth apart. The only thing I found wrong was that the rpm was double what it should be - because you are forcing wasted spark because there are only 2 outputs, even though you have a sequential setup. If you have 4 coils,
you would be better off to wait for the sequencer board. Then just change skip_teeth to 6 instead of 3. In the meantime you could run as is but change the rpm in the ini file (scale it by 1/2).

Another thing, it would be best if you use the attached v2.883g. This code cleans up some missed pulses at low rpm and I am going to ask Lance to put this in place of the existing standard code (v2.883b).

Finally, if you are using COP coils tied together for wasted spark, read Bruce's post on this or you will blow out your VB921 drivers. This was posted today under this general info subforum - topic was something like "MS for 75 deg vtwin..."

Thanks, Grippo, for the suggestions. I've implimented them as well as upgraded to 2.883i and gave it a try.

Unfortunatly, i'm still no go. I've asked around and found someone who will loan me a o-scope for a night... so hopefully in a day or two i'll have a capture of my crank and cam timing signals to figure out exactly what is going on.

Attached is my config and a datalog.

Thanks everyone for your help. I'm close, i can feel it... now only if the microsquirt would agree with me.

Scoping will help immensely. When you scope you want to look at 2 things: the raw VR sensor output, which will be a wave function, and then scope at the point where the conversion circuitry in microsquirt converts this to a square wave that is sent to the processor. If the first is bad, you can stop - it's noise on wires or bad sensor or wheel. If only the second signal is bad, then you may need to make some adjustments in the conversion circuit, for which Bruce can help.

If all looks good, do the same thing for the cam signal.

Ok. i got my hands on an ancient 10mhz scope that looks like it's seen better days. I unfortunatly couldn't get a full cycle into one screen as the scope couldn't go that slow... and actually show a usefull image. the trace just wouldn't lag long enough. But, here is what i was able to capture. I think it's mostly useless as you cant count all the pulses. ... also, the pictures where there is noise on the lower (cam trigger) trace are captures when i have the microsquirt plugged in to the harness (but not powered up).

The thing that i noticed, is there looks to be an extra tooth in the crank wheel... you'll notice two pulses right next to each other. ... is this what's messing me up?

Unfortunately, to get physical access to the crank wheel, i have to completely remove the engine and disassemble a large oil reservoir and the oil pumps which is a big pain in the arse.

Let me know what you guys think.

Thanks.

taking a look again, i may not have an extra pulse.... now it just looks like the trace was overlaping itself at that point. but, i'm still confused.

The 3rd plot definitely looked like overlap rather than an extra pulse. But you are in the best postion to judge that - you can watch the pulses go by and look for missing or extra pulses. But what I would do next is to look at the pulses going into the processor - these will be nice square waves and it is much easier to see if something is going wrong.