Pixhawk 2.4.8 (clone) and problems with copter firmware above 3.2.1

Hi,

I’ve added a new PDB but I’m not noticing any difference at all. However, in my haste to calibrate the ESCs, I forgot to power on/power off, so the sequence was THR max > power on > yellow LED > throttle down > musical beeps and then the motors spin up and down correctly with the throttle, but no change when restarted and running through the Pix.

Running calibration in auto and semi auto is exactly the same as before.

I’m beginning to think that somewhere in the software is a change that’s not compatible with the current batch of clone boards out there and the Chinese haven’t updated them to work with the new firmware. My Pix is probably about a year old, but I’ve only just started on the build so I’ve no idea whether this has been an issue from the start.

I buy clones when times are tight and genuine when I’m more flush, but I’m loath to spend three times as much as a clone on a genuine Pix. I can’t find anywhere in the UK that sells the boards without the power module, buzzer, button etc. I’ve got all those bits in abundance and I’d rather save a few quid and not buy any more.

This has now turned into a bit of a rant so I’ll stop here :angry:

The only thing is really strange that there is only two of us, I got new different unit and still same problem!
I starting to think the problem is somewhere else!
I have taranis x9e and I see you also have taranis
I have afro 30 opto with afro_nfet firmware - I have noticed when I uploaded a blheli to them they are going crazy as hell but no (1s) beeping , anyway I am back now to afro_nfet.hex for tests
I have also Quanum MT4010 475KV I don’t know them at all! never used with anything else!
what is rest of your setup ?
thanks

Greg, I think we may be the only ones, or everyone else is keeping quite about it. I did wonder if it could be caused by interference from my WIFI router, but that’s in a different room so should be far enough away. In any case, I figure if it’s 2.4GHz interference causing issues with TX/RX communications then I’d see it affecting other models. Also it would affect TX/RX comms when in passthrough, but it doesn’t. It only show when the Pix is involved.

I’ve the X9D Plus Taranis with X8R RX connected through SBUS, HK F30A ESCs flashed with SimonK bs_nfet.hex and Turnigy Multistar 4220-880Kv motors.

It seems I can run 3.3.3 without any issues, but I’m stuck in setup because of this problem so haven’t flown it yet with either FW versions. Who knows, there may still be issues with 3.3.3 that only show up when it’s hovering at 100m and I flick the RTL switch!

My clone pixhawk about a year & half old, if installed any copter firmware newer than 3.2.1, the heading will drift both sides, I calibrated the compasses a hundred times doesn’t help, but it work perfectly on 3.2.1.

1 Like

I think I may have a solution to your “crash” problem. There’s a parameter labeled crash detection or something similar. When testing on the bench, you are inputting attitude commands, but the accelerometers are not detecting any movement. The pixhawk thinks it has crashed so it disarms to prevent motor damage. I disabled mine for testing, plus if I do some rough flying, I don’t want it to disarm at 300ft.

1 Like

Hi

Have the same problem after upgrade from V3.3.3 to V3.4.5.

We are three now.

I’ll look into that crash detection setting. Before I do though, I’ve an idea what might be to blame for my problem. I built up a second mini quad at the weekend and after calibrating the ESCs in BLHeliSuite, the motors spun up to 100% through the software, but when spun with the throttle on the TX they stopped responding to input after 50% on the stick. This worried me as it could mean my TX is faulty, but i took it outside for a test hover and it flew great with full throttle punch outs. So on my mini quad, the FC is receiving instructions on the throttle channel, but the accelerometers are not detecting movement so the output throttle is being limited as a kind of failsafe.

But with the Pixhawk, if it is the crash detection setting, then everyone will surely notice this behaviour and not just us few here. Unless everyone sees this as normal and never thought to question this behaviour because the machines fly fine.

My next test is to spin the motors up in acro mode to remove any compensation from the accelerometers.

Ok, so I get the same behaviour when throttling up in acro mode, so no fix there. I was really hoping that would be it!

I seems the only thing to do is settle on using 3.2.1. It’s the only version where the ESCs aren’t constantly beeping and the throttle acts normally.

I’ve just been reading over at DIYDrones of people having similar problems with SimonK and Pix, so I’m going to flash BLHeli and see what happens. The posts over there are a year or two old but there’s not been an update to bs_nfet since the version I have. I’m waiting on a USB Linker from HK to arrive as I don’t want to have to dismantle the hex again to flash the ESCs with the HK’s flashing tool I already have.

That this problem only exists when connected to the Pix, and that the motors spin up fine when directly connected to the RX, suggests that this really isn’t a problem. It’s what would be expected when the PID loop does what it should, altering the output to the motors in accordance with the input it is getting from the sensors, which in this case is no movement. I’ve almost convinced myself of this, but I can only be sure of by taking it for a test flight and I’m really nervous about trying that!

Other than that, I’m close to being out of ideas.

My final update on my experience with this issue.

Everything is fine with my Pixhawk. This seems to be normal behaviour when spinning the motors up on the bench with no props.

I see it like this: the Pix translates all stick inputs to fulfil the pilot’s instruction relative to the platform’s current attitude.

  1. The Pix sees the THR input value increase, but detects no vertical movement
  2. The Pix increases the THR output, but still detects no vertical movement
  3. The Pix increases the THR output, but still detects no vertical movement
  4. And so on until we reach maximum THR output

I’ve updated to the latest FW version and flown with no issues.

@Obecny I hope yours is flying fine now too, and thank you to everyone who chipped in with suggestions.

Hello l3db6h !!!
Thanks for a great news and your updates!
I had a little break from my drones so no progress made since the last time, but now I feel I can start over again, thanks to you! Have a great time Lloyd!

I gave up on the Pix for a few months and then at the weekend I thought, what the hell, it’ll either fly or it’ll crash. Good luck with testing yours.

Hello guys. Any update on situation with your pixhawk’s problems? I am curious due to I had and still have almost the same troubles with my. So, calibration problem was and still is like in yours case, good (full) range of throttle during the calibration process but after, only half throttle and I reach 100% of RPMs. And as well when I change flight mode from stabilize to some other pix is disarming the motors. RTL set on CH7 on other switch than F. modes works only when motors are disarmed. Anyway, with all of this problems I was so anxious to fly my first quadcopter, so I did it. I was flying in stabilize mode. Quad was hovering nicely and steady (but 100% RPM still was on half of throttle range) so I made few rounds around big yard and after 3 or 4 minutes I had a crush :slight_smile:. But not that hard one. It was looking like one motor suddenly has got lower RPM for some reason and whole copter was spinning down, luckily lowering speed wasn’t so high and I didn’t break anything. Immediately I have checked the temperature of motors and ESCs and all were equal and only slightly warm. I decided to try again. The same story after a while. All of my flying end up with some short circuit of the PDB and I burned receiver (FrSky X8R). Pixhawk is working but it was hard landing on the back from 6m height and my drone’s weight is 4kg so it was nice bang. Frame survived as well the props. Only GPS mast broken.
Now parameters of my machine: custom made H shaped frame, 4X Racerstar Racing Edition 3515 BR3515 500KV on 6S lipo (pull 3950 with props 1655); ESC Dys 40A 20-500Hz OPTO (SN series) with SimonK FW ! and of course our Pixhawk 2.4.8 with APM copter FW 3.4.6 (newest available), Radio Taranis X9D Plus and mentioned receiver X8R, other things are not relevant. As you can see almost same parameters as you had and the same problems or similar.
As I was reading some other forums people were suggesting to change parameter ESC_SEED in full parameter list in mission planner (default is 490) and it helped for some folks. Can someone test it? Due to I am out of home for another few weeks I can’t do that now. Any way I have already purchased Hobbywing XRotor 40A APAC Brushless ESC 2-6S which is confirmed to work with our pixhawk with no problems but it is good to check some easiest options possible.
I am waiting for any update guys. Take care.

Hi all,
I bought a pixhawk 2.4.8 clone (not yet tried to fly) and it seems I can only upload AC3.4.6 firmware: when I try to install latest AC firmware (e.g. 3.5.2) mission planner is not able to recognize the board
Any workaround?

thanks

Try to install with QGroundControl and don’t forget to put the sd-card in.

Regards Rolf

Hi Rolf,
I get same problem.
I’ve been able to install firmware with mission planner downloading firmware from repository (here) and than pushing “Load custom firmware”. The problem is that in repository (at least for AC) I can find only

  • 3.4.6 stable
  • 3.5.3-rc1 beta
  • 3.6-dev
    and cannot find 3.5.3 stable that should be the latest stable.
    Any suggestion?

With QGroundControl not able

Hi Folks,
I am new to UAV and bought a 2.4.8 Pixhawk that will not calibrate my ESCs. I bought it on Amazon at Christmas. I just ordered a different one on Amazon, and it is coming from China! It said ‘Prime Free Shipping’ … :frowning:

What is a ‘real’ Pixhawk and where can I get one?

Thanks

you can get it from mro robotics or profi cnc or cuav or drotek ,
others are clones

You do not need your Pixhawk to calibrate your esc. Here is another simple method:

Using a simple servo tester.
(preferably a tester which shows the actual PWM values): Good servo tester

mS = milli second, uS = microseconds.

Minimum = 1mS or 1000 uS correspond to your Tx throttle minimum. (Except the funny Futaba radios) :smile:

Middle = 1.520 mS or 1520 uS.

Maximum = 2 mS or 2000 uS.

There you have it… Can do all ESC setup without any flight controller.
(However remember to remove your props before starting) :slight_smile:

Beside calibrating your ESC you can check at this point the direction of the motors and insure they are correct…

Hope this help.

Cheers
Henri

I am facing the same problem of sudden stop of Motors in Ail & Ele.
Unequal speed of Motors.