I haven’t gotten far enough down those experiments. so I don’t know.
report your findings.
Depinning just the temp sensor disables the temperature sensor but does not disrupt your hall sensors.
FOC can run hall sensorless. so perhaps you mean you don’t want to lose the hall sensor smooth starts functionality. (HFI is a new option to explore… but new… )
2 Likes
Yep thts what i wanna say i dont want to miss the smooth strat
Don’t worry, you’ll keep that.
Ok guys just wanna let you know i disconnected the temp wire as a solution
Did this because i have really less space in my cable channels of my diy integrated deck so i wasnt able to fiddle the cat cable through
Foc mode works temp readings are not schown any more in the ackmaniac app so this is solved for me
Thx for your help!!
3 Likes
I set fault stop time down from 3s to 50ms triggered several faults in a row and didn’t even notice. ie the cutout didn’t send me wobbling.
I advise depining/disconnecting a noisy temp sensor if you can’t figure out how to reroute, shield it, make it not noisy. and/or get better software filtering from some 5.02 or higher firmware.
I haven’t followed my own advice yet.
Today I was doing a temperature testing run. trying to see how fast my ESCs would heat up and get good metr logs to try to look over.
turns out I triggered several temp faults on that motor in a few seconds. and I didn’t even notice. because “fault stop time” set to 50ms. something I saw @b264 advise somewhere.
so that’s a pretty interesting “safety” feature. would have avoided the original Crash this thread talks about.
The following faults were registered since start:
Fault : FAULT_CODE_OVER_TEMP_MOTOR
Current : 0.0
Current filtered : 0.0
Voltage : 37.85
Duty : 0.835
RPM : 37644.9
Tacho : 1621362
Cycles running : 0
TIM duty : 7010
TIM val samp : 2
TIM current samp : 4200
TIM top : 8400
Comm step : 0
Temperature : 84.86
Fault : FAULT_CODE_OVER_TEMP_MOTOR
Current : 0.0
Current filtered : 0.0
Voltage : 37.85
Duty : 0.848
RPM : 37185.5
Tacho : 1622003
Cycles running : 0
TIM duty : 7126
TIM val samp : 2
TIM current samp : 4200
TIM top : 8400
Comm step : 0
Temperature : 84.37
Fault : FAULT_CODE_OVER_TEMP_MOTOR
Current : 0.0
Current filtered : 0.0
Voltage : 37.88
Duty : 0.828
RPM : 37175.8
Tacho : 1623035
Cycles running : 0
TIM duty : 6954
TIM val samp : 2
TIM current samp : 4200
TIM top : 8400
Comm step : 0
Temperature : 84.86
Fault : FAULT_CODE_OVER_TEMP_MOTOR
Current : 0.0
Current filtered : 0.0
Voltage : 37.91
Duty : 0.811
RPM : 36794.7
Tacho : 1623996
Cycles running : 0
TIM duty : 6813
TIM val samp : 2
TIM current samp : 4200
TIM top : 8400
Comm step : 0
Temperature : 85.02
Fault : FAULT_CODE_OVER_TEMP_MOTOR
Current : 0.0
Current filtered : 0.0
Voltage : 37.93
Duty : 0.808
RPM : 36291.4
Tacho : 1624919
Cycles running : 0
TIM duty : 6788
TIM val samp : 2
Fault : FAULT_CODE_OVER_TEMP_MOTOR
Current : 0.0
Current filtered : 0.0
Voltage : 38.00
Duty : 0.791
RPM : 35318.0
Tacho : 1629455
1 Like
@fessyfoo I found the same thing 100%. I got to thank @b264 in person for that tip. Has saved me for sure.
1 Like
Just so you are aware setting fault time that low can cause your ESC to blow more easily when other faults are triggered (overcurrent etc). Played around with this a bit when I was debugging something else and found that setting the fault timeout super low would blow FETs much more easily.
Vedder is planning to make a few extra settings for this motor temp stuff next release.
2 Likes
Jeff
do you recommend the NO_limits fork for those that want temp. sensors… but they set the fault temp way higher than possible/likely?
funny you should say that. I did blow up a fet on my other board recently.
do you have guidance on how long it should be? I think default is 500ms
2 Likes
There is also a fine line between “save my ESC, or save my ass” 
In most cases, I’ll choose my ass over the ESC
If I hit a rock or something in traffic while braking heavily and my ESC turns off for 500ms due to an overcurrent fault, it’s almost certainly going to mean getting run over by a car when I fall off.
If anything, I’d rather it gradually turn off so I don’t lose my balance. Then it could stay off for 2500ms
To me, it seems 50ms is a nice amount where an unexpected temporary hard cutout won’t usually cause me to loose my balance
2 Likes
link to that convo in the PR because it took me a moment to find.
3 Likes
Yeah but if the ESC pops you are in the same place as a cutout and a bonus of dead ESC, but I guess there are the times when it is a non critical fault. I’d probably give a healthy dose of caution before advising more people to do this.
I think this is the best path for the time being while we get this setting in place. In practice motors seem to be able to handle a lot more heat than we think.
4 Likes
I’m never going to get back to that experiment of trying run a shielded sensor wire setup.
I finally followed that advice and depinned the temp sensor on the noisy motor.
straight pin to help push out the connector. tiny screw driver to pry up the clip.
2 Likes
Diff board. Unity and lacroix/maytech 6380 motors.
Temp sensors went nuts. no faults because unity firmware has a setting to disable them.
3 Likes
interesting example from a third board. (proto#2) with new lacroix/maytech 6374 motors. dual focbox 1.6
The motor temperature noise kinda bursts onto both ESCs at the same time at the end as heat of motors gets high in general
someone did do the cable shielding thing I never got to… cool!
@Deodand i noticed you’re active again.
Whatever happened with this idea? I still see “cut the temp wire” advice out there. and have seen motors that continue to have this problem. so your proposed solution seems better than that technique.