ESC32 random disarm

[afernan]

This post is to summarize a rare issue that can happen to any of you using ESC32. This happened to me and some other testers several times with very bad consequences, as you can image.

This problem happens usually with long signal lines (when you put the ESC32 below the motor) and for prop size > 12" and above Throttle 50% (> ~25A). Me personally only have this issue when using CF tube arms, cabling inside. Using ALU arms even longer, cabling outside tube, never saw the problem.

First of all some words from Bill (the ESC32 creator):

"We are not dealing with a complete, final product when building multi rotor's. Each build can be drastically different yet the FC is expected to perform well in every situation. For the most part it does, but you should not be surprised when a particular build runs into a problem. I know how much it hurts when a lot of expensive hardware plows into the ground - probably more than most do. I tend to either fly cheap, easy to replace components and/or spend a lot of time at low altitude with low risk maneuvers with the more expensive hardware until I have the many hours needed to start to trust it. Of course, it's even riskier when you are flying code that you just came up with an hour earlier."

Solution:
- to use shielded PWM cable, and add a 10nF cap to the ESC32 PWM IN to GND

Hope this helps

Angel

[bluuu]

here is more info abaut issue, for all

viewtopic.php?f=27&t=1962&hilit=unexpected#p9486

[sandmen]

@bluu,
there can be several problems why the esc disarm.
Please read out the disarm reason always.
Not all disarm is related to the pwm signal.

[afernan]

[sandmen]

[bn999]

Peter's summary is correct. Bottom line is that each PWM transition is captured and compared for pulse and period width. The pulse characteristics must fall within the configured parameters. If it does, it resets a timeout counter.

If after 0.2 seconds (a very long time) there has not been a single pulse which can be considered valid, the ESC disarms with REASON_PWM_TIMEOUT.

In some cases, the EMI coming from the close proximity of the PWM lines and the high power, fast switching ESC power lines, the noise can be strong enough to cause false transitions, screwing up the timing. If this goes on for more than 0.2 seconds, the ESC gives up instead of blindly running the motor.

I always twist the ESC power lines on the way to the ESC. This is the simplest thing you can do and the most effective.

Another cause might be the FC. If the FC's PWM out port was intermittently not working correctly, the ESC might end up timing out. So, if you find a single ESC that still fails after the CAP fix and power line twisting, try moving it to another FC port.

[teramax]

short question, when PWM lines are max 10cm, does it mater on which side i put the 0,1uF capacitor?
Thinking, soldering it direct on the FC might be easy.

[bn999]

Dunno, I've never needed to add the CAP, but I would suspect it needs to go close to the ESC end. With the digital filtering I turned on, it might not even be needed.

One thing though - the CAP will likely cause trouble for the 1-wire communications - if you are using it.

[teramax]

ok, thanks for reply
So i´ll forget the CAP and try to fix the disarming with tuning the parameters.
As you and Peter already mentioned, just working around the symtoms is not the best solution

[SuperSharpShooter]

I just had a flipover crash on takeoff, and I'm wondering if it was due to an ESC32 disarm. It flipped over towards the rear left as soon as I had it 0.5m off the ground. The flight beforehand was fine, and I flew it again afterwards without issue too. I've had a look through the , but there's nothing obvious I can see in there regarding the cause.

How do we check the disarm code for the ESC? Is there a log for the ESC32, or can you only see the disarm code for the previous disarm?

I only had one cap on the ESC so I'll replace it with two parallel caps. The power and signal lines to the ESCs are quite short - less than 15cm. I'm using PWM in open loop mode.