V Twin Ignition Woes

[hwright]

Hi,

I am installing Microsquirt as ignition only on my 1989 Honda Spada. it is a 90deg V-twin 4 stroke.
It has a 8-1 Crank Trigger, and dual coils. The stock setup uses 2 VR sensors off the same wheel, however I am using the one for cyl 1 (rear cylinder) and have an offset for the 2nd cyl.
I am not sure exactly on the relationship of the trigger to TDC, but I thought I could get it idling and then adjust it with a timing light. The stock ignition idles at 13Deg BTDC.

Basically I have installed Microsquirt, got commumicating. Put in everything I think it needs to run.

I get a steady RPM signal during cranking. I don't have a scope but I do have LED's on the coils to show me when they are triggering and I can see them flashing. The Trigger+- thing shows no dropped pulses.

However I can't get it to fire at all when cranking, I have been trying every combination of tooth delay and inital advance value I can, but no joy. Sometimes I get a few pops when cranking but nothing continous.

Attached is my MSQ and a datalog taken with these settings. Is there something I have wrong or anything else I could try?

[SQLGUY]

If so, you should be able to use that to verify your trigger offset and advanced offset. On my bike, the timing light didn't always fire reliably during cranking. I was able to get a good signal out of it by wrapping a piece of primary wire half-a-dozen times around the plug lead and using that instead of the inductive pickup.

What you should do is, first of all, set your ignition map to either all zeroes, or at lease zeroes around the "cranking corner" (low RPM, high kPa). Set your cranking trigger to "calculated". Charge the battery up well and disconnect the injectors so you can crank without worrying about actually starting/backfiring. You may want to remove the plugs for easier cranking.

Check first that you are getting TDC firing on cylinder one, and adjust the trigger offset in base ignition options until this is correct. Then check that you're getting 90 degrees off TDC firing on cylinder 2, and adjust the advanced offset in advanced ignition options if you need to move this. If the advanced offset doesn't work with -90 or -270, try changing the polarity of your ignition capture (base ignition options rising edge or falling edge).

Once you're getting TDC firing correct for both channels (cylinder 1 and cylinder 2), your spark should be OK, and you can try to start the bike. From my experience, you'll probably want to keep zeroes in two or three cranking squares of your ignition map so that you don't get kickback while trying to start. The rest of the map you can set back to the advance needed by the engine.

[hwright]

Thanks for the tip.
It was a struggle but got the timing light to flash enough times to set the timing.

Ended up with -220Deg on the inital advance and +90 deg on the offset. I know it's supposed to be -90 but it was 180deg out of phase when that happened. Will it cause any other problems or should I try -270 deg?

Anyway I was doing ok when I had the plugs out and cranking that way, but when I connected everything back together my RPM signal is dropping out. I think I will try a capacitor on the coil. Is this the best way to reduce noise?
I am using the bike's stock wiring plugged into the standard ECU connector, it takes ignition and VR lines right next to each other with what looks like no shielding. I really don't want to chop up the standard loom as its very neat at the moment and can easily be swapped between stock and Microsquirt. Also the stock ignition has no problems, how do they do it?
Will keep you posted.

[SQLGUY]

Looking in the code, I didn't see a clear reason why -270 would work any differently than +90 for the advanced offset. Of course, that may change with future versions of the firmware. I think I'd try -270 just to be consistent with documentation... it should be the same anyway.

For my setup I replaced the ignition drive wiring with coax grounded at the coils. Capacitors should help too, though. The reason the stock systems are more immune to switching noise is that they don't have CPU's to be disrupted.

Anyway, glad to hear you're making progress.

Cheers,
Paul

[hwright]

Well I put 0.022uF caps from the switching side of the coils to ground and ran coax (actually shielded speaker wire) cable to each of the coils. And.....

..... I get no more tach problems!

However I still can't get the bike to start. Have double checked that its firing on TDC on cyl 1. The timing light doesn't pick up on cyl2 for some reason however I can see my LED on the coil flashing so it's firing.

2 other notes. When I first powered up the coils the Mirosquirt fuse blew. Hmmmm. Replaced it and everything works fine and didn't blow again. Also sometimes I get a reset on the first time I power the Microsquirt on but not after that or when running. These things did not happen before. I have 2x checked for shorts etc.

When I had no fuel tank on, just the vapors in the carbs I nearly got it started. It was popping away and I thought I just needed some more fuel. So I reconnected the tank and tried to start it, dead nothing, not even a pop. I have checked again and it's definately running on TDC on cyl 1.

This is getting very frustrating! Is it possible that I could have some special sort of CDI coil or something? Something that doesn't properly fire? They only have 2 terminals and the stock unit only switches them to ground.

Attached is the latest MSQ and a trace of it almost started and then the latest trace.

[SQLGUY]

I think you need to concentrate on #2 firing and timing. If the timing light triggers OK off of #1 but not on #2, that sounds like a clue. 2 and 3 should be firing off of the same coil and share the same HV ground path, so verify all the wiring and plug conditions for those two cylinders.

I spent a decent amount of time dealing with 2/3 on my bike. In my case the trigger polarity had to be set to falling edge to get the offset to work correctly between 1/4 and 2/3.

Have you pulled the plug from #2 and visually checked for spark?

[hwright]

I re-connected the stock ECU and it wouldn't start either. (by the way I only have 2 cylinders)

Turns out the plugs were oiled up from all the cranking. Cleaned them, put them back and the stock ECU runs the bike fine. Also checked with a timing light and can get it to flash on coil 2 with the stock ECU but only above about 1000RPM and then only intermittantly - a bit of a problem when I can't get it started with Microsquirt.
Warmed the bike up and went back the Microsquirt and it just pops a few times on the start of cranking then nothing.
It's driving me nuts.

The thing is even the stock ECU can start the bike on 1 cylinder. The Mircosquirt is giving it less than that.
I am completely out of Ideas. the timing is on TDC, I have a good tach signal, no +- trigger pulses, around 10V when cranking. Fuel etc. must be ok because the stock ECU runs the bike. It just doesn't want to run.

If I select falling edge, the tach signal just drops out.

Has anyone got any ideas of what to try?

[SQLGUY]

I don't think the stock ECU is running the bike on one cylinder. I think the timing light is not picking up your spark signal well. On my bike I found I could get better triggering of the timing light by wrapping a length of primary wire half a dozen times around the plug wire and using that for the trigger connection instead of the clip-on pickup. If you can run your timing light like that you may be able to get a better indication of what's actually happening with your timing.

Sorry about the mixup on number of cylinders. For some reason I was thinking you had a V4. Can you post a picture of your crank sensor setup?

[newtyres1]

Just a few ideas, or things to doublecheck,

Skip teeth is set to 4, is this OK for this setup being odd-fire? I cannot find documentation on the oddfire settings for skip teeth, it's quite possible 4 is correct for your setup.
Maybe turn baro correction off for now.
Req fuel is 20 which is a big number, but doesn't come into effect until the engine transitions from cranking to running.
You are running alpha-n blend with SD, alpha-n alone is active up to 3000 rpm. I am not sure if your Alpha-n MAP table makes sense up to 3000 rpm.
Have you tried more cranking dwell?
700 is a high cranking speed but may have nothing to do with it not starting.
Can you smell the fuel when cranking and trying to start?
I cannot find the 2-point cranking pulsewidth. Maybe I have the wrong ini file. What are your cranking pulsewidths?
Check your ignition settings against the fine print in here for odd-fire, (if you haven't already)
http://www.microsquirt.info/dualspark.htm
Generally, the ignition setup that is asked for in the docs (trigger angles, oddfire offset, skip teeth, masks and absolutely everything else) must be 100% compliant or else reliable operation of the Megasquirt/Microsquirt is not guaranteed.

Maybe you have done all this, if so then disregard this post. Beware of too much fuel when trying to start the engine. Always better to start with too little and build up, saves cleaning or replacing the plugs.

Another possible idea (if feasible) is let the stock ecu control ignition for now and use the uS for fuel only until you get the fuel right. Then go for the ignition control.

Ian.

[hwright]

Thanks for the ideas, however I am running ignition only on my setup. Fuel is carb.

I am going to borrow a scope and try and run the Microsquirt in parallel with the stock ECU. I should be able to line up the output signals. Should also be able to get dwell and spark map numbers!

Is there any problem connecting Microsquirt and the stock ECU to the same VR sensor?