Thanks, useful link.
Use a 0.15mm MuMetal sheet above all eletronics and power source, and put the gps/compass 3cm above the sheet. It will solve all your problems with EMI in this micro-drone. By the way, my micro-drone flew for 23.3min and weights 208g. I will put some videos here soon.
This is a great flight time. What Li Ion batteries did you use ?
Hi, Hugues. I use Sony VTC6 and Samsung 30Q. Obs.: 23.3min hovering in Loiter.
What are their capacity in mAh per cell ? 3000 ?
Yes⦠nominally 3A. Some say that Sony have a bit more. The only problem with these batteries is that you have to make sure you bought original batteries. I think that the VTC6 is a standard cell made by one factory and used under several brands. Sony is the āofficialā, but I bought some in AliExpress from Dolidada, Delawake and others and they have the same performance as the Sony ones. I use a computer battery analyzer and I got the same results with 3 different VCT6 cells.
Are you using 2s? If so, what ESC?
I have trouble w/ the ESCās cutting out at over 6v, which is no where near a LiIonās low voltage under load. So, my batteries are basically 1/2 full when I have to land.
Here is my PID tuning for the 3" micro Arducopter (2S)
So, from here⦠Full throttle and high wind flightsā¦
I had similar PIDāsā¦
I had to lower mine for full throttle. Iām at about .7 for pitch, and .1 for roll.
But then the wind⦠I flew in some 20-30mph winds and it was almost uncontrollable. I havenāt overcome that yet. The copter can handle that much wind⦠Itās not a power issue or anything. Betaflight handled the same conditions on a copter w/ the same frame and motors, as if there were no wind at all. (other than having to compensate for the wind pushing me)
Itās not windy anymore, so I canāt really tune for it⦠Wish I had a wind tunnel :).
Wind issue is usually managed in a PID system by the I component. Did you try increasing the I component when there is more wind ?
I did have I set to double P⦠(and various in between). I do like a higher I setting.
It seems the filter settings matter more⦠On the completely uncontrollable flights in high wind, I had my gyro filter down to 40⦠I put it to 100 and it was better⦠But then I have other issues like my copter starts to tilt to the side more and more over a flight⦠lower filter settings seem to mitigate the tilting issue.
It flies almost perfectly on calm/low wind days.
I spent the weekend doing more tuningā¦
As you know if youāve read my posts, I am never quite happy w/ the āfeelā of the copter. It always did little uncommanded movements. The movements are in the logs, desired angles match real angles, as if the FC is commanding them⦠Theyāre usually subtle and it was hard to tell exactly what was going on, but even though they were small, they completely killed my confidence in flying⦠I also occasionally had a major uncommanded yaw movement, of 180 degrees or more after fast maneuvers.
So, I finally spent a day trying different settings for the yaw pidāsā¦
I increased the yaw filter to 20. All the other filters need to be increased for these small copters, I donāt know why I never tried increasing this one⦠I increased the yaw PIDās to P=.3, I=.04, D=.0001. AutoTune had my yaw at about .15.
Now itās great! No more twitches, no more uncommanded spin after a full throttle corner. A lot of the shakes on descent are gone.
I am finally happy with it! (For now⦠There has been no wind this weekend, so that may change things later)
If anyone is having any strange yaw movements on their small copters, try increasing your PIDās and filters.
Hi, Wicked1. I use the 4in1 esc hakrc 15A BLHeli_S, but I ordered the 10A because it is 2g lighter.
Correcting a mistake⦠Sony batteries are now made by Murata company, that acquired Sony Energy in 2017. So, there is only one VTC6 original cell and it is made by Murata. I got 6. They are much better than others when needs more current. Another great option is the Sanyo NCR20700B. A bit heavier, but with 4200mAh and 20A of constant discharge rate.
Do you know the lowest voltage youāve had with the 15a one?
The 10a is rated for 1s, so that will definitely use all of the 18650 2s capacity.
The 10A is 1-2S and is 2g. My battery is 2S. I fly until it reach 5.6V. I got 28.5min of hovering flight with the Sanyo battery. If I remove the embedded NanoPi Neo Air, it can fly for more than 30min. Now Iām fighting with the PID and the optical flow sensor Cheerson cx-ofā¦
Thanks for the info! I canāt wait to see pics/video.
PIDās can be tricky. Iām finally getting everything tuned fairly well.
Iāll post my latest parameter file laterā¦
Excelent! Iāll try some tests with PIDs now and tomorrow too. If you could send your file, would help a lot⦠Regards.
Ok, Iāve attached my current parameter file.
1402 8000kv motors
3025 2 blade props
2s 950mah battery
163 grams with the battery.
I enabled the notch filter and did some tuning last weekend, and have only had a couple flights since then. It seems pretty good, but Iām sure Iāll be making changes soon. Iāve been struggling to get good performance in the wind.
My gyro filter is very high (less filtering), my yaw filter is high, and my loop rate is at 800. All of those are experimental.
z3.notch1.param (19.8 KB)
Man, Iām very upset⦠(trying to avoid a bad word!)
Iām projecting this sub 250g long-endurance micro-quad for a long time, and now Iām having a problem with the chosen ESC that is the same you mentioned priorly. The HAKRC 4in1 BLHeli_S ESCs have a Low Voltage Cutt-off (LVC) of 3V per cell!!! What kind of engineer do this? āI will cutt-off the power to preserve the battery! Great idea! If the drone crash, itās okay! The battery will be safe!ā Ooooohhh, man⦠come on!! The BLHeli_S donāt let you change this parameter, only BLHeli32.
The main problem is that HAKRC does not have at least a contact form or e-mail. The HAKRC 4in1 BLHeli_S 1-2S 10A ESC have only 2g, and this is fantastic. But this LVC donāt let you use all the capacity of your Li-Ion cells. Now, I will have to go back to the Lumenier 4in1 20A BLHeli_S with 9.5g, because this is the one that I remember I could use the battery up to 2.5V and get more than 22 minutes of flight with a Samsung 30Q battery and probably more with the original Sony-Murata VTC6 I got now. Iām contacting some ESCs manufacturers right now to tell them about the people using Li-Ion battery packs with LVC of 2.5V. I hope they consider this factā¦
I had this problem yesterday, during PIDs calibration flights. The battery was stopping with 3V and I remembered that you had this problem. So, now I will put the old Lumenier ESC to continue with PID calibration.
Iām sending some pictures of my stuff just for funā¦
Hi-Fidelity 3D Model
Some Assemblies
Lots of Li-Ion Batteries Brand/Models for Testing
Regards!