Losing RPM Signal

[PhilH]

 

Hi all

I've got an issue with the ECU losing signal which is noticeable in the car.  Doesn't appear to be doing it on the dyno, just on the track.  I've re-wired the crank sensor plug, (only have the single sensor) but just want to know if there is anything else I can check before the next race meeting?

As such all the race logs are showing high numbers of restarts, log 136 had 833 starts during a race?

I'm new to the Link software, which was a replacement ECU for a older OMEX system we has been running on the car for 20yrs

Regards Phil.

[Adamw]

Its a trigger error.  If the ecu receives a trigger signal that doesnt match the expected pattern then it has no idea of crank position, so it cuts all fuel and spark, sets RPM to zero, and then attempts to resync.  Effectively the engine has been stopped and restarted.

Attach the log, a copy of the tune, and a trigger scope will be the best place to start.      

[PhilH]

Thanks.  I'm new to Link, so taking a while to find my way around the interface.  Doesn't look like I can trim down an already download log (which is 350mb) so will reconnect and download just a time section and the engine tune.

I've just read the how to guide on how to record a trigger scope log, so will run a trigger scope and upload it as well.

Regards

Phil.

[koracing]

Perform trigger scopes every 1000 rpm from 1000-5000 or whatever you rev out to and save each one separately to get the best trigger settings help.

[PhilH]

Trigger Scopes attached, plus race logs and the ECU map.

Really appreciate the help.   (Hopefully I did the trigger scopes correctly)  

Regards Phil

 

TriggerScope 1000rpm.llgx TriggerScope 2000rpm.llgx TriggerScope 3000rpm.llgx TriggerScope 4000rpm.llgx TriggerScope 4005rpm.llgx TriggerScope 5000rpm.llgx TriggerScope 6000rpm.llgx TriggerScope 7000rpm.llgx MM ECU Log 2023-10-23 3;39;46 pm.zip AlfaMontreal.pclx

Keen to learn the about the Link ECU and software, so happy to have the long explanation and the what and why behind any recommended changes to hopefully solve this rpm cut-out issue.

Cheers Phil

[koracing]

The primary thing I notice is your reluctor signal is very strong on the car - high voltage output and your arming threshold at 1000rpm is 0.5v.  At the voltages you're seeing it's likely a small amount of noise could exceed this threshold easily.  Your 1000rpm nominal peak voltage is around 10v so I would raise your 1000RPM arming threshold to about 5v, 2000 rpm to 6v, and 3000rpm to 7v.  Do a trigger scope while cranking with the injectors disabled and then set the 500rpm point to about half of the peak voltage you record at cranking rpm.  The highest setting in the link is 7V. 

In you log when it stalls it's because the oil pressure is tripping your GP limit 1.  Perhaps you need to change that a little bit for the hot oil idle pressure.  Other than that in your log - it starts with 137 trigger errors and ends with 138 - so not a dramatic change and very normal to pick up a trigger error when the engine actually stalls, but it is also not possible to see where the 137 errors came from prior to the log.

There may be some benefit to increasing the distance between the sensor and the trigger wheel if possible to lower the voltage output of your crank trigger wheel so it falls closer in line with the avaialbe arming thresholds (lowering max at 7000rpm to be only 15V instead of 40+V).  I don't know if that would help for certain, but it may be a possibility.  Perhaps @Adamw can chime in if he has an opinion on these values.

[PhilH]

Thank Kris.  

I'll upload another log or perhaps a link to it, as the forum file limit was 10mb and the log downloads exceeded this, so only the one was uploaded.

Yes I've noticed the low oil pressure engine cut-off, the dry sump pump isn't great at low speed when the oil is hot, so have gone up a oil grade and will monitor at the next race meeting.

I've been trying to get the pickup closer to the timing wheel thinking that would be a better, so will adjust it out a fraction as you mention.   Just trying to fully understand your comments on the arming threshold, "Your 1000rpm nominal peak voltage is around 10v so I would raise your 1000RPM arming threshold to about 5v, 2000 rpm to 6v, and 3000rpm to 7v.  Do a trigger scope while cranking with the injectors disabled and then set the 500rpm point to about half of the peak voltage you record at cranking rpm." and where in the software this parameter lives to make the changes for what you have suggested.

Regards Phil.

 

[koracing]

Triggers -> Trigger 1 --> Trigger 1 Arming Threshold Table

If you are exceeding the attachment size limit - save the files into a dropbox or google drive or onedrive and then post a share link to the file here.

[PhilH]

 

Links to 2 logs files that show lots of restarts

https://1drv.ms/u/s!Al0yFrp3AwTVj3XCF-7nE7-nFXyU?e=HVmwCZ

https://1drv.ms/u/s!Al0yFrp3AwTVj20mYJo97NGbQ5DG?e=IZhefO

Or the full race day log.. 350mb

https://1drv.ms/u/s!Al0yFrp3AwTVj3dwpGAPyjhLXFgC?e=evH6rb

 

 

 

[Adamw]

I dont see anything obvious cause of the trigger errors in any of the logs or scopes yet.  As Kris suggests, I also think it would be a good idea to increase the sensor air gap to get the trigger voltage down a bit.  Although there is no evidence of the high voltage causing a problem in any of the info provided, there are a couple of reasons I would be more comfortable if it were more typical.  1. You can get a situation at high RPM with high voltage VR signals where the signal cant drop fast enough to always cross zero after the gap.  2. We only test the trigger inputs at up to 70V, it looks like you are probably going above that a little (scope can only read to 60V). 

I noticed in a couple of your high RPM scope captures it looks like we possibly have a software "overflow" type issue in the scope capture when the voltage exceeds 60V, so I dont think this is relevant but have asked one of our engineers to confirm if that is what Im seeing. 

In your logs, the vast majority of your trigger errors occur between about 5000 & 6000 which is probably right around the torque peak where the ignition system is working at it's hardest.  Often a trigger error at peak torque would indicate a spark jumping to something other than across the spark plug gap, or the high voltage/high current ignition system is inducing noise into the trigger signal some other way.  I noticed your spark edge is set to rising - I assume this means it has an MSD system on it?  Have you confirmed the distributor rotor is nicely aligned with the cap posts at the ~25-30BTDC that you have when the errors occur?  Is there much mechanical backlash in the distributor and drive?  

[PhilH]

Thanks.

Yes running a MSD unit.  We ran all new shielded cables as we were getting some interference, e.g. system showing 14000rpm on start up and 65000 max rpm.  I've not seen any weird rpm figures since the cable upgrades.

We're running a huge cap that has an inch between posts to try and avoid any spark jump, but will check the rotor phasing tonight just to be sure.  The dizzy is in good shape I believe, but will check for any play as well.  we didn't have this signal loss issue with the old OMEX system (which was using the same inputs and MSD unit) which indicates to me it's setup or that the Link is more sensitive to interference.

 

 

[Adamw]

46 minutes ago, PhilH said:

we didn't have this signal loss issue with the old OMEX system (which was using the same inputs and MSD unit) which indicates to me it's setup or that the Link is more sensitive to interference.

If the G4X receives a trigger signal that doesnt match the expected pattern then it cuts all fuel and spark, sets RPM to zero and re-syncs from scratch.  Many older ecu's including the gems/omex didnt do any position validation, they would simply keep on sparking based on edge count and not knowing any different - just your timing would be out by at least 1 tooth until it corrected itself at the next sync event.