Microsquirt loosing synch

Printable View

10-07-2018, 08:43 AM
cineman

Microsquirt loosing synch

Hello!
I did a Microsquirt convertion last year on a delorean with stock intake, using Ed ( Seo Motorsport ) hardware kit. http://dmctalk.org/showthread.php?10...I-hardware-kit
The kit is very helpfull, and i made also a "fake" fuel distributor to obtain a stock looking engine.

The actual problem is we have lot of sync loss problems, and we tried lot of different solutions with no improvement now.

We configured it to run on Fuel only management, using original coil signal on OPTO input.
It worked for some time, but always had to made software filter to improve the reading of the signal.
We then decided to mount a 36-1 wheel, with a hall sensor, and got the reading in the VR input.
Again, it worked but struggled after around 2000rpm.
So I thought it was a better idea to improve the signal with an LM1815 amplifier, and again entering the OPTO input.
Again, while i could see a perfect square signal on the output of LM1815 entering the OPTO input, the Microsquirt was loosing quite entire the signals after 2000rpms.
Being the signal good, I think now it could be a firmware problem, or an impedance problem.
I remember also Lazzabby had a similar problem with the SMD version 3.57 ; while we use the Microsquirt maybe that problem was brought to this version too.

Has anyone used the MicroSquirt and has suggest ?
Thank you!
10-07-2018, 09:30 AM
FABombjoy

Which hall sensor? How is it wired? Do you have a sync error log?

IIRC most issues with the 3.57 were when using the VR pickup in the distributor.

On my setup I'm running a geartooth hall sensor (55505) on a 36-1 wheel, and a modified distributor w/ half moon wheel (also a 55505). I have a v3 PCB with Megasquirt 3X. Crank is on the VR input with the pots set per hardware manual for geartooth sensor. I had 0 sync loss for half a year, but made a change at some point that did induce some noise (false tooth reading), at which point I enabled and tuned the the software filters, and added an additional shield to the crank sensor wire. Have had 0 problems since.

The software filter settings are not covered very well in the manual, but they're a feature of the microcontroller and you can learn more from the MCU datasheets.
10-07-2018, 09:47 AM
Bitsyncmaster

Your problem is probably noise getting into the circuit you use to input the signal to read RPM and cam location.

Make sure your signal grounds and shields don't conduct current. So those grounds are run separate to the engine block and not driving any power to your ECU.

When I designed my idle ECU I would get an occasional ignition miss. I ended up using a filter circuit driving my optical isolator to fix that problem.
10-08-2018, 09:02 PM
FABombjoy

Looking at the microsquirt schematics, there's really not much to it compared to current megasquirts.

Having problems at the same RPM point based on 2 different input methods may seem odd, but the output of the VR conditioner and the optoisolator are tied together AND to the same input pin on the MCU. If you post a sync loss log / tooth log / datalog I'll bet we can figure it out fairly quickly.
10-14-2018, 07:33 AM
cineman

Quote:

Originally Posted by FABombjoy

Looking at the microsquirt schematics, there's really not much to it compared to current megasquirts.

Having problems at the same RPM point based on 2 different input methods may seem odd, but the output of the VR conditioner and the optoisolator are tied together AND to the same input pin on the MCU. If you post a sync loss log / tooth log / datalog I'll bet we can figure it out fairly quickly.

Thanks for the suggests. I’ll make a log and update you all. I saw the schematics and realized it was the same ecu input, that’s why i tought of a firmware problem first; but quite sure is a problem with noise, i will make also a test with a signal generator.

Inviato dal mio iPhone utilizzando Tapatalk