Synch Loss Issues - Solved.

[Tennesseejed]

Project specs: BMW R100GS motorcycle, two cylinder, 1000cc, 180 degree crank, microsquirt (code v. 3.3.3), 36-1 tone wheel w/VR sensor, MSQ and MSL attached.

Issue: Massive synch loss preventing startup.

My original install is detailed is this multi-page thread of woe and victory (http://www.msextra.com/forums/viewtopic ... 01&t=57652).

After the original fuel system proved inadequate I am resinstalling MS this Winter using a walboro automotive fuel pump and all of the original sensors.

Everything is wired. Have spark, have fuel, injectors are not firing.

Composite log reflects substantial synch loss. RPM values in MSL log corroborate.



Scoped the VR sensor and polarity is correct. Signal is also non-interrupted, i.e., constant, although the amplitude varies a bit, probably on account of minor wheel migration while turning.

Can anyone point me in the right direction to understand the reason for synch loss? I gather that the reason for snych loss is data logged in MSL but I'm not able to understand what is the reason for my snych loss. That should be a big clue.

During my initial install in I ran into some unusual electrial conditions. In particular, MS would not fire injectors if the injector power supply was fed via a relay, but would fire if the injector power supply was fed via a simple toggle switch (http://www.msextra.com/forums/viewtopic ... 00#p436686).

As before I currently have the fuel pump and wideband O2 powered through a relay. MS and the injectors are toggle switched directly from battery positive to the fuse block. This may not be related to my current synch loss issues since they did not occur previously.

The 36-1 tone wheel is mounted on the alternator as before. Same VR sensor, sensor location, and bracket.

I haven't tried noise filtering yet but the previsouly install ran fine without it.

Are there tools in tuner studio that would assist in diagnosing this issue? Would a synch log file be helpful?

Hopefully this is something obvious in which case please feel free to call me a dummy head.

Thanks in advance.

Jed

[hybrid]

Can you post a regular tooth log?

[Tennesseejed]

Files attached as well as pics.

The slightly shorter line immediately in front of the tallest one in one pic/file is triggering (no pun) a memory of something that is commonly wrong. I'm going to look into that right now.

There was also much more data shown during cranking than is represented in the pics/files. Not sure what setting I can change which would assist with that.

[Tennesseejed]

Tuner Studio manual (p. 191) suggests that two long teeth means the polarity is wrong.

New .cvs attached after leads reversed. Still having two teeth.



Edit: link to tuner studio manual didn't work properly.

[Matt Cramer]

This log looks more like it's failing to see real teeth. Try adjusting the VR trim pots for more sensitivity.

[Tennesseejed]

This log looks more like it's failing to see real teeth. Try adjusting the VR trim pots for more sensitivity.

Apologies, I should have specified that I'm working with Microquirt v.3, assembled.

I took a quick measurement of the sensor gap before leaving the house today. It was .4mm. Manual suggests .75 to 1.0mm.

I'll adjust to 1mm this evening and do some more testing.

[R100RT]

Wow Jed, should have just left the Jim Stim with you
What type of VR sensor did you end up with? Bike close enough to run a test with tone wheel off crank nose and spin in drill press with VR to side? I've found some of the VR sensors I played with didn't work properly, some do. If you run a log with plugs out and ignition feed fuse pulled (or leads grounded) do you note a difference? Your wiring with toggle switch on MS and injectors seems unusual, care to send me a sketch of schematic? Also, if your tone wheel has run out that may affect sync. Have you run a dial indicator against it?
Regards,
Lorne

[Tennesseejed]

She lives! Short startup video below.

Unfortunately, I can't explain what the problem was. For posterity here was my troubleshooting method which I may add to my earlier thread so all my BMW stuff is available to others in the same thread.

First I adjusted the VR sensor gap to 1mm and attempted to start/log. No joy. Reversed VR sensor leads, same result.

Followed Lorne's suggestion and confirmed the Microquirt box was still working by putting the tone wheel in a vice/drill. No triggering. Reversed the VR sensor leads and achieved reliable triggering.

Reinstalled the tone wheel and VR sensor on the bike and it started. *shrug.gif*

While I can't identify what changed that lead to a successful outcome, this is eerily similar to the first time around. First install wouldn't work, borrowed Lorne's Jim Stim, achieved drill assisted triggering, reinstalled everything and, poof, it worked.

Anyway, I'm grateful to have this awesome forum to help with questions and troubleshooting. Thanks.

https://vimeo.com/200917551

[LAV1000]

Congratilations
Keep a close eye to the microsquirt connector, might be some lose connection pin/socket over there.

[R100RT]

I think I might have a thought - is it the "Alaskan Death Watch Beetle of Connectivity Interruption". Seen this little devil before scurrying undercover as serious troubleshooting takes place

[rickb794]

Sounds fishy, hope your troubles are over.

[Tennesseejed]

Hi LAV1000, how are you? Last I recall you were putting Japanese throttle bodies on an Italian three cylinder. Do you have a build thread somewhere? I'd like to see/hear how that turned out.

R100RT, . I'm just happy that this time around it didn't take three months to slay the gremlin. I was losing my mind the first time.

Rickb794, it really is a weird situation. Both the tone wheel and sensor/mount are physically attached to the alternator. My best guess is somehow the either operation of the alternator is affecting the sensor operation or the sensor/mount are somehow affecting the alternator circuit. I'll pull the schematic and see if I can't figure something out. Strange, though, that the problem would be intermittent with the same sensor located in the same manner and then work one time but not the next.

[rickb794]

That i what I found fishy, my spidey senses think it might come back.

[LAV1000]

Hi LAV1000, how are you? Last I recall you were putting Japanese throttle bodies on an Italian three cylinder. Do you have a build thread somewhere? I'd like to see/hear how that turned out.

Turned out very well.

Didn't make a build threat.
Those body's are on it for 5 or 6 years now with major and minor improvements.
Started with oem low z 2.5 bar injectors.

Switched to 3.0 bar Magneti Marelli IPW injectors, had to use something Italian
They are small stainless steel units and used on a lot of european cars.
This way I have a bigger choice in injector sizes.

[Tennesseejed]

That's a clean, beautiful install LAV. Thanks for sharing a pic. Love the sound of three cylinders when you wind the engine.

That i what I found fishy, my spidey senses think it might come back.

Turns out one of my sensor bracket bolts was shunting the alternator coil to ground through the bracket and VR sensor. Oops.

On the previous install I had insulated the connection and on the re-install neglected to do the same.

Should be clear sailing going forward.