Hi, we (my collegue David Kellermann and i) build up an ocotocopter based on the cinestar 8 with hoverfly pro controller. When "starting", the octocopter tilts to one side very fast an rolls over the ground like an angry beast. Breaking props and anything in its way. We checked rotation direction of motors and and mouting of props accordingly, everything is correct as it seems. ESCs are calibrated and configured, gain in hoverfly is set to 25%. One reason could be, that we had to install the hoverfly board heads first, so that to bottom of the pcb is looking upwards. Still in the software sky is shown upside, so we thought, that so software somehow detected the upside-down mounting. Has anybody experience with this kind of mounting? How can one test without risking an other crash? Kind regards Janosch Scharlipp
Hi Janosch: I'm sorry, but I don't know about the Hoverfly Pro controller. Hopefully there are other forum members who do and who can help. Andy.
Janosch. The HF Board needs to be installed with the USB connector on top, facing the sky. If you are using a CS8 with the QC motors raise the gain to about 70 as a starting point. I have mine flying at about 100-110 on my MX-20. Different transmitters will react differently. Make sure you have done all of the calibrations to the HF board that are in the manual before flying. (Temperature, Accelerometer, Compass if you have the GPS.) Check while connected to the HF board through HoverFly setup and power to the ship with your transmitter turned on that your switch inputs are working correctly. I found setting these in the transmitter a bit tricky. If they are reversed then you will run into problems. Hope this helps.
Hi Gary, thanks for your reply, do you know for sure that the installation upside down for the controller board is not possible? Since in the hoverfly software everything looks fine. (sky up, earth down) Switches and sticks seem to be ok (in hoverfly software correctly reflecting input at the transmitter). We did skip the calibration routines for GPS, since we have no such module installed. Can a low gain seting lead to such an effect? I thought, that low gain means slow and gentle reaction to tilt, not such an agressive chaos as we can overserve it. How did you get started, did you setup everthing and just press play and the thing started to fly (somehow without gain optimization etc.) or did you somehow fix it at the ground to overserve if it works? I had one more idea this night, maybe, if one installs the hoverfly board upside down, one has to mirror motor-connections from left side and right side? If i understand corectly, the board detects rotations around three axes with gyro sensors. So looking from back to front, a clockwise roll movement, would mean that the motors on the right hand side of the board must get more power. But if i mount the board upside down (that is rotated by 180° around the roll axis), then the same movement will give the same measurements in the board, but this time the motors on the left hand side of the board have to get more power. Anyone who nows how this thing works internally? I wish there was a way to test these things without such a high risk ...
Ok, we have good news and bad news First of all, David turned around the hoverfly board provisionally, and now it did in fact fly! (not very stable yet, but i guess thats because of the low gain value) On the other hand, i think i made an error in my reasoning in my last post, the way it looks now to me, flipping the hoverfly board shouldn't change anything at all ... which it obviously does. Hope those guys from hoverfly tech can explain that, but for now we will go on with the properly installed hoverfly board. Thanks so far Janosch
Janosch the accelerometer chip needs to know which side is up other wise disaster waits. When you do the Accelerometer calibration the software has you arrange the board through the six steps of each side edge pointing down, then the bottom and then the top. If you started with the gains at 25 you probably have the same control issues I had. Start at 70 and work up and down from there. Currently I am set at 100 on my MX-20. That shows up as something different on the readings in the HFSetup software but since it is controlled from your transmitter that's what I go by. Install the board with the USB port and the arrow on the top of the board facing forward and then work on the gains.
If the accelerometer is like the ones I work with, it will be reading a constant 9.48066 meters/second squared down through the z-axis when it's installed correctly. That's 1G. (Although G varies depending on where you are on the planet.) If you have it inverted, then it's going to reverse the sign of the z-axis and, as Gary says, that's A Very Bad Thing. The ACC will react as though it's flying upside down and the flight control system will attempt to flip the copter over, right side up. Which is where this story started, I believe? Hope that explains things. Andy.