I would like to raise a concern I have about the 3DR power module. I have completed my hex (900mm) setup and have done around 10 flights. I am running 2x 4S 5000 mAh in parallel with 2814 770kv Tmotors 13x4 props APM 2.6 with V3.1.5 FW and no servos connected. I managed to do 2 successful autotune but I am not happy with the performance. I noticed when I go to auto and start going through my waypoints there is a very big altitude change/drop when arriving at the waypoint and setting up the heading to the next. When the altitude drop happens the motors spin up quit hard and it is a bit scary to see this happen. I was expecting a more fluent flight behaviour. Unfortunately I erased the logs so I can’t ok back at the amp draw and how much the altitude drop is but I will make sure I get the logs this weekend.
When I landed the last 3 flights I noticed that my receiver (XPS 8 channel) shows an orange LED instead of the expected Green. Based on the Received manual it means that the voltage for the received has dropped below 4.4V I can’t tell how much below 4.4v, how long and how many times… I hope I can find more accurate details this weekend since I am going to install a TattleTale and compare them to the APM logs.
The TattleTale is a tiny and lightweight device that plugs into a empty receiver channel (or through a Y-cable). This device measures the input voltage and displays the lowest voltage threshold that was captured after being powered on. With a programmable acquisition time of 1ms to 100ms, even the shortest voltage sags can be detected!
The tattletale has helped me in many occasions when I am facing an abnormal behaviour in my Heli FBL or Fixed wing power setup. I strongly recommended this product from XPS.
My question is in regards to the power module performance:
- Shouldn’t the 3DR power module be able to power my APM 2.6 in my current config? I am surprised that the voltage of the power module dropped below 4.4V.
- Can a voltage drop to or below 4.4V that was registered by my receiver cause the behaviour that I am seeing on my Hex?
- I was planning to cut the red wire from the power module going to the PM port and connect it to an external BEC so that the APM + Received get powered from the BEC instead of the power module. Does that make any sense?
- should I get a new power module and try that out?
All suggestions are welcome…
Thanks a lot,
Replies
Hi Mike,
I do not think that the voltage drop is causing the altitude change, but more likely the other way around. Your logs show a very clear correlation between throttle output (and thus current) and Vcc drop. Every time the motors go to high power, the Vcc drops. Yours is very substantial.
On your heavily loaded setup, the current draw can reach 80A, which is fairly high. On the lightly loaded setup, it only hits 45A.
My personal opinion is that the PM should not be used on copters that draw more than 50A max. It would be interesting to see what happens if you bypass the current measuring section of the PM.
Hi Rob,
Thank you for looking at at the logs and your advice.
I will remove the PM as you suggested and power the APM and my RX from an external BEC and see how it goes.
I will see how it goes and post the update after that.
I wouldn't remove it just yet. It's still useful as a power supply. Just connect the ESC power wires directly to the input side, and that will take the current measurement part out of the loop.
Thanks Rob,
I got it.
Want to try something interesting? Should be totally safe.
Check out your PM. You will see that the two negative wires, the one for the ESC and the one for the battery, are not directly touching. They are soldered to either end of a heavy trace on the PM board.
Cut the shrink wrap off, and either solder a short, heavy piece of wire between the two butt ends of the negative wires. Or you could just add more solder to the trace. Or unsolder once of the wires, strip more jacket, and then resolder the wire such that it's almost touching the other. Basically, I'm trying to decrease the resistance on the negative battery side.
Then obviously, put some new shrink wrap on it.
You could also jumper out the positive side current sense resistor, but soldering wire across that resistor. That is one way to get rid of the current sensor portion, while still using the voltage sensing and power supply portions.
Hi Rob,
I did the modifications but no luck, I still have the same behavior. I have attached the logs of today's 4 test flights. I did an autotune with the 2 batteries but no real improvement.
flight 1: 2 batteries trying auto
flight 2: 1 battery trying auto was successful
flight 3: 2 batteries trying auto ended up doing autotune
flight 4: 2 batteries trying auto with new parameter obtained from autotune in previous flight.
Can you see anything else from the logs?
Thanks a lot.
2014-11-10 12-24-10 2x 5000 mAh.log
2014-11-10 12-33-27 1x 4200 mAh.log
2014-11-10 12-42-39 2x5000 mAh autotune.log
Yeah, I actually spent a while playing around with this myself on Sunday. I made a 20 Amp load bank so I could check the voltage readings. I wish I could get 80A load, but that's a lot.
Anyway, I checked a number of places and, I just can't find the problem. Interestingly, the negative battery side has almost as much voltage drop as the current sense resistor. But that's still not where the problem is coming from.
I dunno.
I appreciate your help and effort trying troubleshooting this.
I have some ideas but not sure if any of them will solve my issue. Anyway here is what I am going to do changing 1 at the time:
- I just ordered 6x X-Rotor 40A opto ESC and swap my maytech 30A ESC's out.
- replace the PM
- I will try to power my RX and APM via a BEC instead of the PM and see how it goes
I will keep updating my findings here.
Yes, the Opto ESC's are an interesting step in the right direction, especially if they are truly Opto. Check for continuity between their signal ground and battery ground wires. There should not be continuity, the two sides should be completely separated.
Something else interesting to try, would be to power the Pixhawk completely separately from the motor battery. And especially if you do this with Opto ESC's. If the problem still exists then... I got nothing.
log of 4th flight and new parameter settings
2014-11-10 12-58-21 2x5000 mah.log
141110 FY-680 autotune with 2 batteries .param
-
1
-
2
of 2 Next