Thanks for the speedy response guys, much appreciated. I will read the notes but it looks like Craig has delivered the crucial news. Anytime in the next half hour would be cool ! No pressure.
I cannot get my red safety switch to go solid red under any circumstances at the moment, i don’t know if i have a procedure error on my part or a sick pixhawk. Does it know when the Esc’s are not correctly calibrated ?
I do not get any Rc failsafe messages except i have once, and once only this evening seen a “Safety switch” red text message appear in the heads up display.
Is there a way to dump out a status of the firmware that i can try to diagnose please ?
Hello all,
I confirm, I have exactly the same configuration has previously :
-TX: Frsky Taranis (with latest fw)
-RX: Frsky D8R
-ESC: EMAX 4in1 25Ax4
-Pixhawk FC with GPS (ArduCopter 3.2.1)
-3S 3000 mAh battery
I can indeed calibrate the ESCs when I connect them directly to the radio.
And the Pixhawk doesn’t seem to communicate with the ESCs. Somethings happen when I plug some servos to port 5 and 6 (for gimbal), but still, not motor rotation…
Expecting to have this work soon, would be very nice…
I’ve actually been debugging this. I’m going to throw it on an oscilloscope later to verify, but the problem seems to be that the Emax 4in1 is pulling down the input voltage. From what I can tell, it is drawing 5.31 mA of current from the Pixhawk when HIGH, based on the 220 ohm resistors on the servo outputs. My Pixhawk is outputting 3.27 V at HIGH. With the Emax connected, the HIGH is pulled down to 2.1 V. It is a bit unreliable, as changes to the GND or power input of the Pixhawk can cause the output to go high enough to “wake up” the ESCs. I’ve even had them all come alive once or twice when plugging in the main battery, but they go into a programming mode.
I tested the Emax 4in1 on a signal generator with a servo position control, and I varied the input amplitude. The Emax becomes unresponsive below 2.8 V input, and I would say it is unreliable below 3.0 V. The signal generator was able to drive the current needed to maintain 3.3 V on the input. The Pixhawk doesn’t seem to be capable of this. It is possible the APM was able to supply enough current, the outputs are directly connected to the Atmel via a 220 ohm resistor. On the Pixhawk, there is a level translator on the output, and it looks like that level translator cannot keep up with the Emax’s power draw. If I had to lay blame, I would say the input impedance isn’t high enough on the Emax. Does anyone else see a fix here?
Even with this solved, the Emax would still have a problem with the Pixhawk because the Pixhawk sets the Signal wires to contact HIGH when in “safe” mode, no safety switch pressed. This will cause the Emax ESCs to go into a “panic” mode, where they pulse on and off every 2 seconds or so. Is there some reason the Pixhawk holds this input HIGH instead of LOW?
I took the problem the other way. I asked EMAX why it doesn’t work. I wrote them an email and here is their answer:
“The current 4 in 1 ESC 25A doesn’t support Pixhawk, we had noticed that and are currently modifying the firmware and doing test so that the new firmware will be compatible with Pixhawk flight controller. We will finish test in around a week. I will keep you posted about it.”
So, hopefully, we should have some news next week.
I am having a similar problem with Pixhawk and Turnigy Plush 25A ESCs.
The ESCs worked with APM 2.6 hardware
They worked at first with the Pixhawk, but then I believe the Pixhawk firmware was updated.
I can now go through calibration successfully, but when I try to ARM, the ESCs complain that the throttle signal is too high. Same symptom whether I calibrate 4 at once or do the manual ESC calibration.
When I use MP to look at throttle input and output, there seems to be a higher throttle output on RC1,RCS,RC3,RC4 (ouputs) than the throttle input from my Spectrum receiver, which is suspicious. The RC outputs do follow the input (go up when throttle goes up, do down when throttle stick goes down)
Tried 3 different receivers from two different manufacturers, get same problem
I can calibrate and get the ESCs to ARM with this work around:
Follow calibration procedure but do not bring the throttle sick right down to the bottom, keep it about 15% above low throttle
Calibrate ESCs this way so they think that low thottle is when the throttle stick is at 15% or so
When I ARM, keep the throttle stick right down on the bottom below 15%
Is the Pixhawk keeping throttle output on RC1-4 too high when arming? If yes, what can I do to fix this, I cannot find a parameter to tweak (I have tried several)?
I asked the EMAX team to see what the could do. Apparently, to what I read in previous posts, it is a problem with the Pixhawk, isn’t it ?
I really don’t know anything about programming the pixhawk, but, guys from the dev, could you please do something about it? how can we have the developers team take us seriously bout this issue and reopen the case? How can we send a bug report? CraigElder, if you read me, please, help us, we would really like to fly our drones
Guys, could you help us? I am sure it is nothing and now with the last post from whittonm you’ll have enough information to include a correction in the next software release, no?
I just verified that the fault is in the Pixhawk. I replaced the Pixhawk with an APM 2.5 controller, kept everything else the same, and the problem is fixed. I can calibrate the ESCs, the drone arms normally, flight is stable, connections over the USB are foolproof. I don’t know if I have a bad Pixhawk (it is a genuine model) or if it is an issue with the software. The Pixhawk did work well at the start, but is now unuseable, and the warranty ran out 12 days ago (it is only 3 months). My setup:
This seems like a hardware issue to me, nothing we can do in software. As a couple people have indicated, the input impedence on the Emax ESC is not high enough, so the PWM pulse amplitude seen by the Emax ESC is not high enough. This is made worse on the Pixhawk , since it’s driving the PWM at 3.3V instead of 5V as the APM2.5 does.
This is not the first time that an ESC has had this problem, there are a number of them out there that have it. Unfortunately, if this is the problem, there is really nothing we can do in software.
I suspect the fact that it sort of works maybe kind of sometimes, is an indication that you are on the ragged edge, and the PWM pulse is going over 3.0V, but takes a bit of time to get there, thus there’s this off-set seen between min throttle.
Now, I have to ask, do you have a ground wire between the ESC and the Pixhawk?
This is the part that’s not totally clear to me at this point.
I do not think the kinda/sorta/sometimes it works situation, is an indication that there is a problem with the software, as if there were, then all ESC’s would have the same problem, not just a few.
I can confirm that the HobbyWing Quattro ESC’s have no problem with the Pixhawk and are available for only about $30, so that is one option if you want to stick with 4-in-1 ESC’s. HobbyWing is a very reliable manufacturer of high quality but low-priced ESC’s. I use them almost exclusively.
Hello Bob,
Thank you for your answer.
But how do you explain that the ESC works with other autopilot ?
Are you talking about this ESC: ebay.fr/itm/HOBBYWING-Brushl … 3ce5d93228
Are you sure it will work with the Pixhawk, I would not want to buy another controller that does not work… but at the end of the day, I think this is what I am going tho do…
[quote=“tonypascal”]Hello Bob,
Thank you for your answer.
But how do you explain that the ESC works with other autopilot ?
Are you talking about this ESC: ebay.fr/itm/HOBBYWING-Brushl … 3ce5d93228
Are you sure it will work with the Pixhawk, I would not want to buy another controller that does not work… but at the end of the day, I think this is what I am going tho do…[/quote]
I would not recommend the Turnigy Plush ESC’s as I cannot get them to work with my genuine 3DR Pixhawk but they work fine with multiple China clone APM 2.X controllers.
I can confirm Mikeb’s findings…my EMAX ESCs wouldn’t even acknowledge the Pixhawk.
As mikeb found the Pixhawk couldn’t supply sufficient current, I tried running the PWM signal outputs through a 74HCT244 line-driver buffer, running off the 5V ESC-BEC power supply, and it works fine. Either the next-gen Pixhawk needs to supply more current and/or the EMAX has to reduce their current draw on the signal lines.
A 74HCT244 is only a 35cent part…but can put out 20ma per line.
Hello
Great news. Amazing to see that there is a solution. Could you please post a schematic of the circuit. It would really help !
Thanks for the advice anyway.
Tonypascal
