Hoping someone here offer advice on getting my GEPRC Phantom HD BNF to fly well on 4S batteries. I’ve had multiple flyaway situations with this drone on stock BF4.1 firmware both on 450mah or 650mah batteries, and the issues seem to apply to both copies of the drone I currently own. I’ve dumped hours into trying to fix this because when the drone seems to fly well, it is almost too good to be true - fast, quiet for parks, great battery life, and feels so fun in the air due to its super low weight. I don't have 3S batteries the right size, and furthermore have heard the quad is a little underpowered on 3S anyway.

The big pain I can’t stop is the constant flyaways. I’ll arm the quad, take off, and the throttle will feel super light and floaty, almost like antigravity is turned way up.. Then I'll few turns (or worst, a bump or even just a leaf passing over the motor) and suddenly the quad is 30-40 feet in the air and I’m hitting disarm and praying that it crashes into the grass. 

Everything hardware about these quads are stock except that at some point in troubleshooting I dropped in some of that Ummagawd sticky stuff to help prevent battery ejections during crashes.

Quad ships with BF4.1, and recommended batteries are 3-4S, 300-450mah. However, [DutchRC Adventures] reported on his youtube channel that he has heard many reports of flyaways on 4S. His fix was to bump down master mix to 0.9, and move the filtering sliders to “less filtering” 1.4 each. With those settings, I was able to get a floaty but stable 4S at 450mah, but immediate flyaways on 650mah. And even on 450mah there is a jitter I can hear on the motors.

My current (probably broken) understanding of the problem is that there is some measured vibration on the quad that is passing through the filters, causing a "dterm windup". This problem presents on both of my copies (including my second one which has had zero hard crashes and the props are entirely new.)

Things I’ve tried:

* Reseting to completely default tune in BF4.17, BF4.2 - no change
* Disabling anti-gravity - sometimes helps a little with landing the quad but still flyaways occur
* Fuddling around with normal quad tuning regime (start with clean PIDs and low filtering, then slowly build up P to bracket vibrations, etc.) No good, flys away even on lower PID gains settings
* Applying URUAV’s community settings for toothpicks. Helped a little bit on 350mah, but still immediate flyaway on 650mah
* Default sliders on PIDs, but lower PD balance to 0.9, and back down on stick gain to 0.9. Raised Gyro and Dterm multiplier to 1.8. Added RPM filtering and JESC. This last one now flies really well on 450 but with immediate flyaway on 650mah.

The most recent settings are actually quite promising but I don’t fully trust them. For one, why does REDUCING the filtering help? Wouldn’t we want MORE filtering if the problem is vibrations? Secondly, the fact that even the small increase in weight from 450->650 causes problems, which tells me the tune is not very robust.

Can someone wiser than I advise on what might be happening here? What the most likely scenario is the root issue? Is this d-term windup? Can the quad even be made to run on 650mah? Why wouldn’t it? 

BTW: Geprc support was completely useless. They claim their quad works perfectly on 4S at 450mah, and I have video evidence from a new in box quad on 4S 450mah that this is absolutely not true. They were very terse in their responses. I'm starting to think its not a great company to buy from.


Attachments: Stock BF config, Config with the DutchRC adventures, and the most recent config that sort of works but I don't trust.

I heard may complains like yours regarding the fly/away issue with 4s but for now, after a couple of flights with original settings, 450mah worked fine even if the throttle is very sensitive.

One thing I would like to try to mitigate this is to install JazzMav or JESC firmware on ESC and activate RPM filtering. This should probably allow to relax the notch filter and the overall filtering rate. But I will probably postpone this to the spring time when I will have the chance to fly more time with good weather and tune everything properly.

Another strange thing I noticed with this quad is that if you update betaflight to 4.2, reset settings to default and try to fly the drone without further tuning it doesn't lift-off but flips on a side.

Got another flyway today. I can't figure it out. Disabling airmode and antigravity seems to have helped somewhat. I've searched in vain for any recent feedback that these modes can cause flyaway conditions but the last serious discussion I can find on this topic dates back to 2016-2017.

When I started testing today, I moved AG and AM to a switch and then after 2 flights couldn't replicate anything, even switching them in and out during flight and before takeoff. Then boom, it happened again when I disabled that change and re-enabled them both permanently as well as enabled RPM filtering. Suddenly the throttle felt out of control - not on takeoff, but when coming back to my home position. I was taking 'diff all' each step along the way and this just blew my mind. It surely can't be that simple. I had to wrap up testing after that happened.

I'm completely stumped on this one. I can't understand why'd this particular build is so cursed. When it flies well though, it does great. Its super quiet, really manueverable and no jitters or anything.

Here's the diff of the diff, for the record. First file was my bewildered checkpoint after failing to trigger a flyway after 2 attempts. And then the second file was after I got an immediate flyway.

Were you able to get a decent reliable pids for 4s on Phantom HD?

Hi Jgilbert20,

I am assuming that you are flying in ACRO mode. Is that correct??
Now, this might sound funny BUT, have you tried flying it in ANGLE (STAB)
mode?? Just to see what happens. It would be interesting to see if the
behavior is the same in all fly modes or specific to a particular mode.

I have my Fly Modes on a 3 position switch so that I can switch modes 
during the flight.