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

UMainePhaze09 wrote:We changed the trigger offset to -15 and we had consistent spark for slightly longer than before, but we still lose it after a few seconds. We also lose tach at the same time. What could cause our loss of tach?

I really don't know if this helps you, but I had a similar problem with my rig, and I couldn't identify it at first, when I scoped the VR signals, I got the usual sine waves, but when I used the TachRef program, I got square waves with diagonal lines in them, so they were triangular looking.

I adjusted my VR sensor positioning, and it was slightly better. I altered the polarity of the cam sensor and bingo, the diagonals had disappeared, and when I loaded the files back in the MicroSquirt everythin was working, showing cranking rpms etc.

I used a bank of 12V LEDs instead of coils to check everything was OK first, and I was using the (not recommended) coil drivers circuit to trigger these. They flashed nicely and the only loss of tach I got was when the processor got confused with the moving crank/cam sync (my cam wheel is ever so slightly out so it doesn't hold consistent sync).

When I used the standard coil drivers on IGN1 & 2 to drive two pencil/stick coils things got worse, the spark was weak, and loss of tach kept dropping cranked rpms to 0. I tried moving the VR sensor wires, keeping them apart, but in the end it was HT interference through a poor ground circuit. I got the spark stronger (blue) by strapping some monster booster cables to the battery, and my rig, but I still got loss of tach, which never happened if I reverted back to driving the LEDs.

In the end I wired in the eBay ignition module, switched the jumpers to LLO, and tach is stable. The MicroSquirt is about a foot from the four stick coils, runnin wasted spark, and the spark plug noses are within six inches of the VR sensor wires.

We've been messing around with the settings for the last few days and have had some success. We now have consistent spark in both cylinders, the correct timing and spark on the correct stroke. However, as soon as we introduce fuel into the system, the spark becomes inconsistent, firing only every other revolution and we have inconsistent RPM. We used the tachref program, and our file is attached. The crank sensor appears to be inconsistent, with different spacings and noise at two locations between cam syncs. It's hard to see though because of the number of teeth on the crank.

Also, cranking RPM on the sled is above 300, around 350. Do we need to change the code for this to work? If so, how do we go about doing that? We have set the max cranking speed to 500, otherwise we lose spark as soon as RPMs are above the 300 limit (this was our spark problem before, when we only had spark for a few seconds).

It sounds like there are 2 problems here. But the critical one is not the fuel - I have never seen a fuel problem, it always works, just may need tuning. The only ways fuel could knock out tach synch are: adding fuel causes combustion which makes the engine erratic, wspecially ones with 1 or 2 cylinders, and you lose tach synch. From the picture it looks like this is what is happening, the teeth close together after combustion and slow down before combustion due to pumping loss. The other possibility is injector electrical flyback if the problem occurs with the injectors running but no fuel. This might be because the injector wire are mixed with the ignition wires.

Tach synch is lost when the delta_t between teeth is very different from the last delta_t. There are threshholds for the allowable %difference between times during cranking, after start, and running. The default values for these are 50,70 and 25%. These should work but you could try raising the them to 70 and 80 and 40 - I wouldn't go higher. Also try both last interval and alpha-beta-gamma starting. Do a datalog and see if you can correlate the trigger +/- count with the higher rpm on a tach cycle. The present system is not a good data logger for this sort of thing, but at low cranking rpms you might be able to tell something.

What is happening when you go past cranking rpm is that the fuel in cranking is sprayed into each cylinder simultaneously and they get shots 2x per 720 deg of crank rotation based on a cranking pulsewidth that you enter. After cranking rpm, you get fuel based on the VE table once per 720 deg, you use the after start pulse tolerance and, if you were in trigger rise mode, you switch to calculated ignition advance. So, by playing with the variables you can see which of these is having the biggest effect. There is no need to change the code, you could set crank rpm to 10000 rpm if you wanted. (There are min/max limits in the ini file, but these can be changed by editing the file.) However, you then lose the ability to change fueling - which is worse than having a carb.