Hello everyone,
I'm running a core8 board and a rotax v990 engine used by Aprilia in the first decade of the years 2k.
The tune is loaded in the http://www.rusefi.com/online page and there are some related logs too (not to that tune directly, but same configuration).
The engine provides its trigger signals thru a 36/2 flywheel read by a VR sensor for crank angle and notch on the intake camshaft read by a hall sensor for phase.
The engine and the ecu run well on speeduino firmware, just to avoid looking for hardware/cabling/conditioning issues.
I've been trying for a couple of weeks without success. Matthew gave some hints on tuning the trigger decoder to suit my situation (as you can see in the msq the trigger decoder is using tuned parameters).
Being on a motorcycle engine makes things harder as i can only use the timing light after replacing the stator cover with a cut open version of it with the VR sensor installed in a temporary position.
The sparks are in the right places during cranking.
I cant run the engine with this cut open stator cover, I risk to damage the engine, spill oil and blew everything.
Additionally every unsuccesful attemp ends with some very harsh kickbacks and i've already replaced a freewheel.
Seems like that the spark timing becomes erroneous whenever the engine fires succesfully, if you look at engine phase in the log looks like it gets corrupted after succesful firing attemp.
The tune is loaded in the http://www.rusefi.com/online page and there are some related logs too (not to that tune directly, but same configuration).
Thank you! Those are important 60%
The other 40% would be providing direct links to best tune and best logs. I have a very cute 4 y.o. kid to watch, I do not have time to search for those, sorry.
Re: 36/2 trigger problems on a big V twin
Posted: Wed Mar 23, 2022 12:16 am
by mck1117
Can you try to run it as a single cylinder? Unplug the injector for whichever cylinder is configured with a non-zero offset and see if it starts without kicking.
Seems like that the spark timing becomes erroneous whenever the engine fires succesfully, if you look at engine phase in the log looks like it gets corrupted after succesful firing attemp.
Thank you.
sounds like 'triggerSyncGapRatio' and 'trg err' are the entries to inspect closely
Re: 36/2 trigger problems on a big V twin
Posted: Wed Mar 23, 2022 9:06 am
by deffie83
This is a relevant log entry related to the previously posted tune:
I am failing to understand the problem statement. "36/2 trigger" sounds like a conclusion and I wonder what makes you think that you have a trigger problem. Are you referring to some private conversation with someone or are you mixing your conclusions into the problem statement?
I see reliable RPM reading, I see very off falling "fuel: running" amount and I see a strange/broken TPS reading.
image.png (131.9 KiB) Viewed 4270 times
I do not see a complete link to the tune.
A link to a helpful youtube video of your problem could help. Please do not attach a video file to the forum.
Re: 36/2 trigger problems on a big V twin
Posted: Wed Mar 23, 2022 8:57 pm
by deffie83
This is a little video of a cranking attemp, I think there I were retarding ignition a little to avoid kickbacks:
I followed the problem classification as a trigger problem as that was the area where Matthew Kennedy put his efforts in helping me while discussing the problem on slack's public channels #support_tuning and #support_member_vehicles.
As initially reported I'm using tuned trigger parameters proposed by Matthew (included in the tune I posted) which indeed attenuated the kickbacks.
Please note the "trigger ok" indicator going red on succesful firing.
The initial post contains a link to the current msq. I've now uploaded my entire rusefi project folder, including all the past logs, if it can be of any utility:
This is a spreadsheet populated with a couple of toothlog dumps where I tried to recreate trigger decoder logic:
It is not making much sense, sure I'm missing much of its basic working principles, was just a try.
I believe that you have never provided an exact URL of the exact tune in question. Overall it looks like you and I have a major communication issue
I believe that you might have had trigger decoding issues but at the moment I see no indication of trigger decoding issue - it looks to me that "trg err" is not incrementing. By the way I should rename that "trg err" into something more readable.
Do you have problems positioning TDC? To investigate that rusEFI console engine sniffer tab could be the best tool - it would show you the TDC as rusEFI assumes it to be. If that's in fact a TDC location issue one of the ways to confirm would be a high data rate MAP log file that's if you have a MAP sensor.
I believe that you have never provided an exact URL of the exact tune in question. Overall it looks like you and I have a major communication issue
You are right I haven't provided the EXACT URL of the EXACT tune in question.
However, the tune is the first one of the list in the page i mentioned, close to this same nickname.
Are we robots ?
However I've linked the whole folder, can we please stop quibbling ?
I believe that you might have had trigger decoding issues but at the moment I see no indication of trigger decoding issue - it looks to me that "trg err" is not incrementing. By the way I should rename that "trg err" into something more readable.
What about trigger ok going red in the video ? Isn't that meaning trigger issues ?
Do you have problems positioning TDC? To investigate that rusEFI console engine sniffer tab could be the best tool - it would show you the TDC as rusEFI assumes it to be. If that's in fact a TDC location issue one of the ways to confirm would be a high data rate MAP log file that's if you have a MAP sensor.
Comms issues again ? I think I talked about cut open flywheel covers, timing lights flashing on TDCs, broken freewheels...etcetera.
Sorry, I do not have infants to watch for, but I value my time too, and that was a couple of weeks every day after work trying to solve this problem, asking for help on your messaging platform, where you didn't arise at all. Not a few minutes reading a forum post or fractions of a second trying to find out a tune in the beginning of a list.
Can you try to run it as a single cylinder? Unplug the injector for whichever cylinder is configured with a non-zero offset and see if it starts without kicking.
However I've linked the whole folder, can we please stop quibbling ?
The whole folder is not useful. It's much more useful for us if you link tune on rusefi.com/online where it can be viewed without TunerStudio (on your phone while away from a pc!).
I haven't provided the EXACT URL of the EXACT tune in question.