3" Gecko with 4s. I have some 1300mAh that I am using at the moment but have ordered some 850mAh to reduce the weight. It’s got a runcam split 3 micro on it as well which is pretty sweet - only problem is that the 1A BEC on the Kakute Mini is not enough for everything, so I am going to have to carefully power the split from the LiPo with the addition of a TVS diode.
All up it’s 334g with the battery. Smaller battery, maybe 300g - GPS puck is contributing to that as well as the longer 4s wires
If you look in the gitter - some guy did a fork with an implementation of a protocol to control split. Would be nice to be able to turn it on/off and to video/photo modes from a taranis switch.
I just looked at split 3 specs - it can feed from up to 20v. Why not to feed it from lipo? Just make sure to solder capacitor on feed line, as usual
It explicitly says not to in the docs, but I have no option. I have a 470uF 25v on the battery and the feed will come from there, so should be ok, but I am going to add a TVS just to be sure.
Yes, given time I will wire up the runcam serial port and get hacking 
i can still trigger the same error with no actual ESC hardware connected and no port defined for BLH telemetry.
I always had splits on lipo feed, but, yes a big enough cap is a must.
Actually I think this is correct - if I graph the RC outputs they mirror the rpm output, so maybe it’s something to do with props off testing
Turns out we are calculating rpm incorrectly. https://github.com/ArduPilot/ardupilot/pull/12544/commits/943314a223496b1f84ecfee1e1a4e06f71db9a1f fixes this
i added the motor pole param and i am under the impression it returns the actual shaft rpm as is. using a 2200kV 14 pole motor on 11,2 volt i do get ~24500 at WOT with no load.
The calculation was about 30% out, so looked roughly right - it just wasn’t exactly right. I think the error is lower for more poles.
ahh, thanks! the difference is 2% for 14 poles, that’s why i didn’t notice. nice find!
hi, i just received that K7 mini - and i see, even after adding NODMA to all UARTs those messages after reboot:
PreArm: Internal errors (0x8000)
PreArm: Compass not healthy
PreArm: Throttle below Failsafe
PreArm: Internal errors (0x8000)
PreArm: Compass not healthy
PreArm: Throttle below Failsafe
Initialising APM
EKF2 IMU0 tilt alignment complete
EKF2 IMU0 forced reset
PreArm: Compass not healthy
PreArm: Throttle below Failsafe
EKF2 IMU0 tilt alignment complete
Frame: QUAD
gecko3k7 00360032 31385113 38343530
ChibiOS: 0997003f
ArduCopter V4.0.0-dev (9961f4bc)
Barometer 1 calibration complete
Initialising APM
Calibrating barometer
AK09916: Reset Failed
Prepared log system
So, the
PreArm: Internal errors (0x8000)
and
AK09916: Reset Failed
are those messages expected to be shown from the current master code? Is it something new recently added there, or an indication of some issues?
You are seeing the same two errors that I was!
hmm, i can check again, but i think i synced with master on yesterday`s evening or so. may try to do it again.
i never had errors on kakute in relation to pass-through - why would it error now? i simply used same exact config i had on previous models, kakute and omnibus nano - never had errors on them.
it throws this error 8000 on every reboot.
You are using the new KakuteF7Mini hwdef I created right?
Right, only added nodma to all uarts
FWIW I am using UART3 with DMA on GPS and it is working fine so far
i also got it all working, got esc telem working and showing current/voltage from esc, just made a mistake and bent props - and now need to wait until spare props will come from amazon. but, all worked, s.port feed into taranis worked.
nice little controller.
Cool! If you have telemetry working on 4.0, any interest in trying https://github.com/ArduPilot/ardupilot/pull/12544 to give me another datapoint? I haven’t flown this beyond a hover because of work/weather but initial indications are amazing.
What version of BLHeli are you using? Do you have auto-telemetry turned on? What board are you using?