12-3 Crank sync issues; Suzuki Bandit 1200

[tmbryhn]

I'm involved in a project where we're struggling to get proper RPM sync.

We're using custom hardware running Discovery F407 firmware and MAX9926 VR encoder.
The crank trigger implementation is similar to OEM Honda CBR900 with a similar trigger offset of 65°.

We're getting nice sync while cranking with spark plugs removed (no compression, constant speed), but as soon as we're installing the spark plugs and introduce speed variations, we're getting no sync.
In the specific attachments, "Invert Primary" was set to "true", which is incorrect for MAX9926, but according to end user it doesn't affect getting good RPM sync while cranking without spark plugs.

The exact same engine with exact same crank trigger starts and runs fine on an MS2-extra ECU.

Triggerhjul 12-3.jpg (1018.9 KiB) Viewed 6355 times

Scop Triggerhjul 12-3.jpg (919.55 KiB) Viewed 6355 times

Trigger Logger - 12-3 inverted.jpg (275.58 KiB) Viewed 6355 times

Trigger Settings.jpg (504.9 KiB) Viewed 6355 times

log.jpg (308.67 KiB) Viewed 6355 times

Any ideas on where to go from here?

[AndreyB]

Your megalogviewer screenshot focuses on all the right data points but without both good and bad logs I am blind.

Sometimes VR wires just need to be physically flipped.

[tmbryhn]

From the scope trace, it's clear that VR polarity is correct.
I can have the end user provide a log of cranking with no spark plugs that'll give perfect sync

[AndreyB]

I can have the end user provide a log of cranking with no spark plugs that'll give perfect sync

Sure you can but those would provide most value if provided together, the perfect one and the broken one.

[tmbryhn]

"Invert Primary" was set to false on this latest log.
Here's a side-by-side comparison of the logs without and with spark plugs. I hope the pics are good enough (I myself find it very hard to view them in the PC web browser, but it works ok on a phone zooming in).
If need be, I'm happy to upload log files, .msq and pics to a shared google drive folder.

Cranking without spark plug gives perfect sync both with and without "Invert primary" enabled.

Trigger log without & with spark plugs.jpg (740.99 KiB) Viewed 6065 times

Data log without & with spark plugs.jpg (777.08 KiB) Viewed 6065 times

I hope we can get some tips as to how we can solve this weird issue, otherwise we'd have to go back to Megasquirt, which would suck...

[AndreyB]

I hope the pics are good enough
If need be, I'm happy to upload log files, .msq and pics to a shared google drive folder.

No comments.

image.png (530.15 KiB) Viewed 6014 times

[tmbryhn]

I can have the end user provide a log of cranking with no spark plugs that'll give perfect sync

Sure you can but those would provide most value if provided together, the perfect one and the broken one.

Here's a dropbox link with the same data. I don't know how to interpret your request above any other way than to give a direct side-by-side comparison like this. Rest assured, both datalogs are attached to the link as well:
https://drive.google.com/drive/folders/1kZEIUEBljSwTtDi-yZfntMsRfqDmvBJj?usp=sharing

[AndreyB]

https://github.com/rusefi/rusefi/wiki/rusEFI-Quick-Start-PDF was created yesterday.

New line just added into it around https://github.com/rusefi/rusefi/wiki/rusEFI-Quick-Start-PDF#log-it

Shaky videos do not do it. Even high-resolution screen shots do not do it: it has to be a proper log file.

[tmbryhn]

https://github.com/rusefi/rusefi/wiki/rusEFI-Quick-Start-PDF was created yesterday.

New line just added into it around https://github.com/rusefi/rusefi/wiki/rusEFI-Quick-Start-PDF#log-it

Shaky videos do not do it. Even high-resolution screen shots do not do it: it has to be a proper log file.

Nice post.
Yup... That's why dealing with third party users can be difficult - they're not necessarily aware of the data rate setting. Personally, I'm always using max rate for logging trigger related, in addition to a custom logging profile that excludes all the unnecessary junk not related to trigger inputs that otherwise has a tendency to make the log size very large.
I'll see if they can make another log with max data rate.

Anyway. Judging by the .CSV trigger logs done in TS, the gap appears as extremely obvious visually speaking, and the amount of rising/falling edges amounts to the same in both logs (with/without spark plugs). Do you have any hypothesis as to why the firmware doesn't manage to establish proper sync? Are there any software tricks like Trigger Gap Override or the likes that can be done to overcome this issue?
Would a possible solution be to write a dedicated "Honda CBR" trigger pattern for the database (12-3)?

[AndreyB]

I am pretty confused now, need your help.

I open "12-3 wihtout spark plugs.mlg" I see one data point 6.001s and next data point at 6.041s would that be a 25Hz rate?

See https://github.com/rusefi/rusefi/wiki/HOWTO-upload-log#data-rate
See https://github.com/rusefi/rusefi/wiki/rusEFI-Quick-Start-PDF#log-it

[AndreyB]

I am pretty confused now, need your help.

As in I was hoping that you would value my time and spell out "I do not have proper logs let me not bother you with unacceptable 25hz logs".

[tmbryhn]

I'll see if they can make another log with max data rate.

This means; I'll see if I can have THE END USERS make another log with max data rate.
I'm very sorry to have wasted your time. The thread will be updated once proper logs have been made. Hopefully this issue can be resolved then.