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

I am having dropped RPMs on cranking.

Its a 12-0T crank with a 5V cam sync on Vr2.

On my setup I get a tach pulses ranging from 2V to 4V (pk-zero) depending on RPM, according I expect an average RPM of about 360, but with lows of about 240 and highs of 480.

I left my camera at home last night or I would have a scope output.

The very strange thing is that the output has the traditional VR shape, with a negative peak first, sharp zero crossing to high positive peak (thus I am triggering on rising edge, and since its microsquirt the software says falling), and a smooth return to zero, but it overshoots, and has what looks like a underdamped LC circuit. And not on every tooth either. I am also now dropping RPM somewhere around 5000rpm. This is with or without the stock ecu piggybacked.

I have done some probing on the stock ecu signal conditioning and I see a 1.4V open pull up, which drops to 40mV when connected to the VR sensor. The sensor itself it about 200ohms? (I don't have the data in front of me)

The cam sync had soft edge on the 0-5V trigger, but I have solved that for now with a 5k pull up to Vref, but intend to retime my engine on the falling cam sync to remove the need for this.

Config comes later tonight and will have scope pictures.

Thanks
Geoff

As measured at VR1 Pin 32. But I have a couple new symptoms: I can start when the engine is warm and the battery is fully charged, I think I am right on the borderline of starting. Also I loose my rpm again at 4000rpm, but it comes back at 7000rpms and then reads fine until 12krpm.

As I mentioned earlier, I don't like the ripple, but any ideas are appreciated.

Thanks
Geoff

How could I easily modify the trigger voltages of the VR1 input? They're set to 100mV on and 50mV off, correct? I want to double this to stay clear of the ripple as seen in my cranking scope picture. My current theory is that the microsquirt is reading the ripple in the VR as extra teeth thereby throwing a sync error, once it reads the cam sensor again it reads a few pulses, which shows me an RPM for a bit.

I have already isolated my microsquirt as much possible putting it on its own separate battery, only connecting to the tach and cam sensors. I did not ground the various A2Ds, like the tps, temps, or maps.

I have verified again, the polarity shield to be the ground wire as the center of the coax, and the signal to be on the shield. As people have mentioned this seems somewhat atypical. This wire is run to a stock connector where the connector is run over unshielded parallel wires.

Thanks Again,
Geoff

I am really lost here, I don't know what to do next.

Any help would be greatly appreciated.

Thanks Again,
Geoff

What are the 2 traces in the picture - there is a dark one and a light one like a ghost, that sometimes correlates with the dark one and sometimes is offset ? The best way to scope this is to put the VR1 input to the ECU on one channel and on the other channel put the output from the VR processing circuit (which should be a clean 0/ 5V square wave going to the processor) on the other channel. This would show if any pulses are being missed.

What I notice in the scope shots are a compression and then a slowdown typical during cranking when you have combustion and then an exhaust stroke. If the above scope shots show that no pulses are being missed, then you could try unpluggung the injectors and try cranking like this. If you get correct rpm like this, but get good rpm for a brief period, then a 0 rpm when the plugs in the engine are fired, then you could play with pulse tolerances. Also there is a spare3 variable in the latest ini file that will show missing and extra pulses as a count. You can turn it into a gauge and see if it moves off 0 during cranking.

Also, you need to fix the miswiring of the shield by swapping the wires in the Ampseal. Bruce or Lance has recently put out a post on how to do this.

Lastly, did you try changing the tach input polarity in MT, but leave as is electrically to see if it makes things better or worse. Concentrate on cranking - the high rpm misses may be a totally different problem that can be solved later.

Thanks,

The two traces on the scope are actually the same signal, just operator error and the lcd lag from the prior pass. Where is the VR circuit output on the board?

That INI tweak sounds great, will do immediately.

I will correct the ampseal pins, I should have posted something on that when had all the pins removed initially making the harness. But just to make 100% clear, the shield is the VR-/Gnd?

I tried all 4 combinations of software and hardware polarities to no avail, at it reads zero rpms with the occasional 180rpm blip until it the engine catches.

I assume the high rpm issue a pulse tolerance issue of some sort, but I will work on that later.

Thanks for the help,
Geoff

Thank God / Grippo.
I found the pickup point for tachin at Q2 col, and scoped the output. The circuit is indeed picking up the ripple as an extra pulse. The Spare3 +/- teeth display however did not display any of this, but maybe it is not updated during cranking? I will post a scope photo in a little while.

Now the question becomes how to solve this:
I see three methods mechanical, hardware, or software.
1 Mechanical the sensor is reading off machined teeth on a cast flywheel, see phazer's post. The area between teeth could be smoothed and the signal presumably improved. It might require re-balancing my flywheel.
2 Hardware, modify the VR circuit to increase the hysteresis. It could be internal or external as needed. (I have access to a vast array of surface mount components)
3 Software, mask out these undesired pulses
My preference is pretty much reverse order due to the ease.

Thanks again,
Geoff

Scope as promised

Is your flywheel just like the one in phazers post. If so I wonder if you might be picking up a signal from the magnets inside the flywheel.
It would be interesting to see if your extra pulses coincide with where each pole of the magnet passes the pickup.
What kind of engine is this you are working on.
Just a thought, wouldn't it be easier if you removed one tooth and ran it 12-1 with cop wasted spark? That way no cam sensor.

I agree about magnets. I was concerned about that myself, but there's virtually no magnetic field outside of the flywheel bolts don't stick to it at least. I will bring a compass to double check.

There are about 20 magnets if I recall correct, so the magnetic pulse would be about 1.7 times per tooth and would repeat every 3 teeth. This is napkin math btw, so don't hold me to it. My ripple as I am calling it does not appear quite so evenly spaced.

Thanks for the suggestion, I will look into it, the Aprilia guys could be having the same issue.

Geoff