Help pls, MS2/Extra strange spark output issue
#1
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
Hi again,
I have learned much since my last post (slightly ashamed at how silly that issue was), but now I'm having a really weird issue.
Have tried a lot of things and am slowly narrowing it down but would really love some assistance from people who know much much more than me.
Car details,
1.6L NA MX5 (1989)
Most recent MS2/Extra firmware, Spark output - going high
Logical spark outputs to stock igniters
MS2 on V3 PCB, with 4g63 mods from MSExtra website, same issue with or without Joe Perez' spark invert mod.
Basically, after about 5 minutes (cold), or less if warm, the engine will start to lose spark for around a second at a time repeating every few seconds.
With a timing light, I see the spark cut out, the led on D14 (spark output 1) stays on the whole time (as in gets stuck high, probably not good for the igniters), however when i checked D16 with an LED it would spark normally.
After having the car off for a short time ~30 seconds, it will go away for a minute or two of being on, then return, if started immediately after turning on it returns very quickly, have tried swapping the transistor on D14 with the one on D15, which did nothing, both could just be busted since I was a bit silly and let the transistor get quite hot desoldering it. But I would expect the transistor to not work at all.
In my JimStim v 1.5 it works perfectly for ever it seems. This leads me to believe it has to be something todo with the wiring in the car and not the megasquirt itself (I sure hope).
My thoughts are,
the solder joints where we spliced into the factory harness could be causing issues, but they looked like quite solid so not sure,
the crank angle sensor signal could be noisy, (this is what a guy at my work who has contributed alot to the VEMS project said was common), however there is no Sync Loss, and the other spark output seems fine.
the igniter could be bad causing the megasquirt logic output to have to supply more current or something causing it to degrade over time (possibly leaving car off for short time lets everything cool down and work again, however no components on the board have gotten even remotely warm.
possibly there is some setting in my tune somehow causing this, i copied everything from the base map i found onto a blank tune, when the spark issue isnt there, it runs really quite well since my brother (whos car it is) has done a fair bit of tuning to it with VEAL.
Have tried moving all the grounds arounds to 3 different spots, currently the ground near where AFM would be, is used as another location for a block of other wires it seems.. However I should note that the electrical ground for my 14point7 SLC Free A wideband controller is grounded to a different spot, being the negative terminal of the passenger side foot light thing, the wideband heater ground still goes to the ground everything else goes to. ( I really hope that having just the wideband electrical ground in a different spot wouldnt cause this, would again feel very silly, but I don't think it is the cause)
Wondering if anyone has any advice or thoughts, it seems to be very similar to this post on the microsquirt forum Spark cutting out: Please help - MicroSquirt.com Microsquirt Forum however I am not using IGBT's to run coils, just logic outputs to the igniters. Also, the tacho on the dash has stopped working intermittently now :'), this kinda makes me think maybe igniter 1 is dying since iirc tach signal on early mx5's comes from the first igniter.
Here are some videos we took of various things, one of the LED for Spark 1 blinking showing it going HIGH for a good second every now and then (in this video we rev it half way through and it looks like its on for a long time but thats just partially the camera refresh rate.
- LED Video
- Various Sensor Graphs
- More Sensor Graphs
- Composite Logger
- Trigger Logger ( i think, could be Tooth Logger)
Will upload the MSQ when i get sent it by my brother.
I think that covers everything, Would really really really appreciated help since this is a real bummer and am a bit lost with how weird the issue is. If there is any information more you need just say and I'll post it ASAP.
Once this is all done, will try to give back to the forum by posting a sort of build log/process of how I did everything along with the issues I ran into etc, and just try to document everything nicely for future megasquirters.
I have learned much since my last post (slightly ashamed at how silly that issue was), but now I'm having a really weird issue.
Have tried a lot of things and am slowly narrowing it down but would really love some assistance from people who know much much more than me.
Car details,
1.6L NA MX5 (1989)
Most recent MS2/Extra firmware, Spark output - going high
Logical spark outputs to stock igniters
MS2 on V3 PCB, with 4g63 mods from MSExtra website, same issue with or without Joe Perez' spark invert mod.
Basically, after about 5 minutes (cold), or less if warm, the engine will start to lose spark for around a second at a time repeating every few seconds.
With a timing light, I see the spark cut out, the led on D14 (spark output 1) stays on the whole time (as in gets stuck high, probably not good for the igniters), however when i checked D16 with an LED it would spark normally.
After having the car off for a short time ~30 seconds, it will go away for a minute or two of being on, then return, if started immediately after turning on it returns very quickly, have tried swapping the transistor on D14 with the one on D15, which did nothing, both could just be busted since I was a bit silly and let the transistor get quite hot desoldering it. But I would expect the transistor to not work at all.
In my JimStim v 1.5 it works perfectly for ever it seems. This leads me to believe it has to be something todo with the wiring in the car and not the megasquirt itself (I sure hope).
My thoughts are,
the solder joints where we spliced into the factory harness could be causing issues, but they looked like quite solid so not sure,
the crank angle sensor signal could be noisy, (this is what a guy at my work who has contributed alot to the VEMS project said was common), however there is no Sync Loss, and the other spark output seems fine.
the igniter could be bad causing the megasquirt logic output to have to supply more current or something causing it to degrade over time (possibly leaving car off for short time lets everything cool down and work again, however no components on the board have gotten even remotely warm.
possibly there is some setting in my tune somehow causing this, i copied everything from the base map i found onto a blank tune, when the spark issue isnt there, it runs really quite well since my brother (whos car it is) has done a fair bit of tuning to it with VEAL.
Have tried moving all the grounds arounds to 3 different spots, currently the ground near where AFM would be, is used as another location for a block of other wires it seems.. However I should note that the electrical ground for my 14point7 SLC Free A wideband controller is grounded to a different spot, being the negative terminal of the passenger side foot light thing, the wideband heater ground still goes to the ground everything else goes to. ( I really hope that having just the wideband electrical ground in a different spot wouldnt cause this, would again feel very silly, but I don't think it is the cause)
Wondering if anyone has any advice or thoughts, it seems to be very similar to this post on the microsquirt forum Spark cutting out: Please help - MicroSquirt.com Microsquirt Forum however I am not using IGBT's to run coils, just logic outputs to the igniters. Also, the tacho on the dash has stopped working intermittently now :'), this kinda makes me think maybe igniter 1 is dying since iirc tach signal on early mx5's comes from the first igniter.
Here are some videos we took of various things, one of the LED for Spark 1 blinking showing it going HIGH for a good second every now and then (in this video we rev it half way through and it looks like its on for a long time but thats just partially the camera refresh rate.
Will upload the MSQ when i get sent it by my brother.
I think that covers everything, Would really really really appreciated help since this is a real bummer and am a bit lost with how weird the issue is. If there is any information more you need just say and I'll post it ASAP.
Once this is all done, will try to give back to the forum by posting a sort of build log/process of how I did everything along with the issues I ran into etc, and just try to document everything nicely for future megasquirters.
Last edited by venetox; 12-01-2017 at 06:11 PM. Reason: Uploaded Tune File