Its very inconsistent and pretty rare but some times when I boot up my heli 9plug in a battery) the servos all go to the extremes and bend things or snap rod ends.
H743-Mini latest firmware. Align RCE-BL25A(Hobbywing Platinum PRO v4-A) Servos are powered off the ESC built in bec with a capacitor added on to that power dist rail. Servos are Align DS450 and DS455 Digital.
I use an ELRS receiver and ESP8266 WiFi in case it matters what sort of command and control there are that turns on with the heli.
It doesnāt seem to matter if it starts with the transmitter on or off.
Any hints on what params to check/investigate/look in to?
Is there another section that would better suite this post? Its not really heli specific or related to helis other than its happening on a heli but when iv looked around before on this it seemed mostly other heli people have it and the rare plane people that say anything the response is to adjust linkage to stop it binding but thatās not really an options on my heli.
Iām not sure I can help you much, but Iāve just realised that I may have had a similar issue with H743-MINI. It was installed in my TRex 500 heli and after a few successful Arducopter flights I was making adjustments on the workbench and noticed that my rear collective servo was behaving strangely. It wasnāt moving smoothly, and would often stop at the top of its travel. Just to be sure, I replaced all three collective servos.
I then found that, even though it had flown okay before, as soon as I armed the system it would immediately disarm. They guys here couldnāt see anything in my log files that would explain it, and I never did solve it. Iāve today just finished setting it up with an H742-WING board instead.
That does not really sound like what mine is doing at all. Itās only on start up. Once it gets to the servo linkage check phase the control surfaces are all solid and work as expected.
Maybe not the same, though I donāt know when my servo was damaged, or how it was damaged. At the time I was thinking it had maybe been over-driven by the FC at some point, which is why I replaced the other two as a precaution.
Admittedly this is little speculative, but I think this issue is more specific to the way some (possibly cheaper) servos respond to spikes in the signal line before the signal is consistent. I donāt think any particular autopilot is necessarily to blame. Another user reported similar behavior independent of both the autopilot AND the RC hardware, which leads me to this conclusion.
To fully mitigate it, simply put a switch or circuit breaker on the servo rail power source. Power the autopilot first, wait for it to begin producing output, and then switch the servo rail to hot.
spova, I havenāt flown this particular heli with the H743WNG yet, due to weather and other things; but Iāve got an H743WING in another Arducopter heli with no issues so far.
Yuri_Rage, you may be right in your speculation, though Iāve never encountered the issue myself in 30+ years of RC flying with both cheap and āqualityā servos. I donāt like the idea of a switch in the servo power source though, for thatās just another potential failure point. Switching Tx on before powering Rx should be enough to ensure that the servos get a valid, usually mid-point, signal when theyāre powered on. All bets are off if the Rx is powered before the Tx.
I understand your trepidation with installing such a switch for sure. Remember, though, this is NOT pure RC flying. You have an autopilot in the middle which needs to boot and configure itself according to the userās parameter set. As pins get attached to internal resistor networks and get initialized, there may be brief non-PWM output that is inconsistent with the transmitter. Now, I agree that servos should not respond to non-PWM output, which is why I speculate rather than state anything conclusively here.
I understand what youāre saying. If thereās a possibility of it being a problem, would a choke in the +ve supply be the answer ā or the ferrite core that often installed on ESC Rx connections (which Iāve always removed to save weight since I donāt use 35MHz any more)?
Edit: Thinking about it a bit more, a choke in the +ve line is unlikely to do any good
iāve been monitoring this thread. The only thing I can add is to provide some insight on the control system. For the swashplate servo, the limits are hardcoded to be 1000 and 2000 PWM. It is done this way so that the swashplate movement can be consistent and standardized for any set up. Thus it is possible that the flight controller could command the swashplate servo to limits that would exceed mechanical limits in the users heli. I donāt see how the tail servo could go beyond the the user defined limits in the servo set up (servo parameters). I would say that that would have to be a hardware issue either in the flight controller or some other hardware. Exceeding the swashplate servo mechanical limits could happen through the software, but it is unlikely.
I also am not a fan on intentionally installing something that cuts power to the servos.
Iām inclined to think its something to do with the FC (soft or hardware) since if I keep the FC powered off and just put power to the servos they donāt move.
As for making sure the TX is on first. I generally do this and it doesnāt seem to matter as it does the servo over travel randomly. In addition it being an autopilot I donāt want to even have to use my TX every time. I ultimately would like to be able to fly using only the GCS sometimes.
I have the swash travel limited in the heli settings and usually the FC obeys they set limits but not always. Iv read about plane guys solving this by using another safety switch that also disables servos but that this setting interferes with things on heli and should not be used here.
Does anyone know of some other servo parameters that dictate startup movement or positions?
There are no such parameters. Because it happens sporadically and only with some servos, it leads me to the speculation above.
Another option that might work is to power the autopilot via USB, and allow it to boot that way prior to applying full battery voltage, then remove USB power after the battery is connected. You could use a USB battery bank in the field for such a procedure.
I have a spare Matek board and various servos on the bench, but Iāll be unable to test for a couple of weeks. My plan is to try a handful of servos on different output pins with various firmware variants to see if thereās a way to replicate results as described here. I may monitor pin output with a scope, as well. Iād love to hear results if someone beats me to it.
I thought about going that route (powerbank to start the FC first). Its a bit more stuff to fiddle with but starting a heli is already sorta fiddly with putting the canopy on with them powered up and stuff.
I unfortunately canāt reliably replicate the issue myself to really report when it happens. It does seem more common when the heli has been recently powered already and I do use caps on both the servo power rail and on the FC power.
This does not happen with any normal heli FCs I have used but all of those have the power run through them too.
Its not unique to helis. Iv read of the same issue on planes with ardupilot as well. It seems more detrimental on helis than other things though.
Those of you using the H743-wing, are you powering the servos through the FC or separately? That could possibly be a factor I suppose.
I think initializing on USB power is the safest route until we have a more definitive answer. At least it doesnāt require additional hardware to be installed onboard.
I was slightly mistaken about āno parameters.ā I think the servo trim parameters are the initialization positions.
Iām mitigating a similar issue on my Rovers, where GPIO pin state changes are unavoidable during ChibiOS boot with a PWM-driven relay between my relay board and the autopilot. Not ideal, but the safest solution I can achieve reasonably.
spova, with my H743Wings Iām powering from the 5v output of its built-in PDB. With my H743Mini thereās no such facility, so I was powering from an external CC BEC.
(BTW, how do you show the other personās text that youāre responding to?)
I recently bought my first matek board (H743 wing) and have had the same issues. I concur with Yuri that powering USB first seems as the best option (read least ugly fix). I have considered soldering in a little switch and boot battery that gives 5v to the USB and then stays charged when when the main battery is plugged in. Just have not figured out yet what battery to use.
You can hit the reply button on a specific post and it will try to quote that post in your reply or you can highlight specific text then mouse over it and you should get a āquoteā option that when you click on it will start a reply with the quoted text.
OK, got it thank you. Iāve often used the āReplyā button on the post but it never copies over the post, but highlighting the text and hitting the qote-marks does the job for me.
The on post reply button is hit and miss for me. Some times it copies only part of the text or some times non. This post was supposed to be a quoted reply too and as you can see, it didnāt quoteā¦
I moved this our of trad heli and in to the general 4.3 area. In further messing around Iām convinced its a problem in the way the FC boots up. There are numerous posts going back through previous versions with no clear resolution. This is an aircraft breaking issue the is part of the most basic operation as you have to turn things on to fly them. I get that many or maybe most users are just not using servos because they are using multi rotors. If there was just a basic failsafe or startup servo position setting that would be great but I cant find it.