Josh: I've seen that AH drift-downwards-when-translating problem before. It seems to be a product of slipstream over the pressure transducer as the copter flies sideways: the air pressure decreases due to the effect of the motion (maybe the "venturi effect?"), and that decrease in pressure is sensed by the FC as a gain of altitude, thus the FC throttles back and, of course, that then really does produce a loss in altitude. Not saying that's what's going on here, but the symptoms are suspiciously similar... Andy.
Josh's copter is producing a pretty substantial weather front when it moves with all that thrust, so it's not a stretch to say that a barometric-based AH sensor just might not work. I think we should see if Holger could augment the FC's barometric data with GPS data from the Navi board..
It's certainly true that the copter would be flying into it's own prop-wash, but I would think that with some mechanical (e.g. foam) "averaging" and/or software averaging, it can be compensated for. I would fret that the horizontal toilet bowling we see on position hold would not work well if translated into the vertical plane by using the GPS signal for AH...you're feeding in a second erroneous signal. Wouldn't that make things worse? Andy
I had the same problem about AH. The problem turned out to be from soft rubber dampenings that the FC was sitting on. I changed them with short and stiff ones, and the problem gone.
I tried the same setup just an hour ago: 5215XF-435KV motors 18" props Herkules III MK FC I also have exactly the same AH problem Josh described. So I take back my previous statement about dumpers.
It's something in the X8 configuration. MK electronics are not setup for X8 IMO. At least from my experience I prefer the Flat 8.
That's not true in my experience. Recently I flew a land survey with my X8 where I had to keep the copter perfectly level and straight 500m out and 500m back, and my AH was dead on the whole way. Didn't touch the elevator the whole way out and back. So it can work. Just not for some. I still maintain that it has something to do with the heavier lift machines (bigger props, motors and payload than my copter).
Somebody using a Pixhawk decided to test what might be happening with their altitude hold (which sounds very much like the MK system). It's a good read over on DIYDrones: http://diydrones.com/profiles/blogs/loss-of-altitude-after-horizontal-run-in-alt-stabilized-modes
I have a hood on our Cinestar X8 that may shield the sensor for fast sideways flight. I'll try and get time to stick bang it out the back door in AH (FC 2.5). However since it is somewhat a wing there might be a low pressure created underneath. Cheers, Jeff
I've had a dialog with Holger and the summary of it is as follows: There are two factors involved when the copter translates: 1. The pressure difference caused by airflow over the pressure transducer. The airflow is caused in part by airspeed, but also because the copter is now flying into the turbulence that its own props are producing (the prop wash). 2. The center of lift tilts over and thus the vertical component of lift is reduced (according to the cosine of the tilt angle from the vertical). The pressure difference, although small can be compensated for by gust filters (both mechanical, i.e. foam on transducer) and by software averaging. The reduction of the vertical component of lift can be compensated for by increasing the thrust produced by the motors to restore the vertical component back to what it was when the copter was on hover. There is a new beta version of the flight control firmware that has added an additional Tilt compensation factor for the thrust compensation. You can find more details at http://forum.mikrokopter.de/topic-post525395.html#post525395 The firmware is 2.07f. MK has also fixed some additional aspects that are very important: - the default time for RC-Lost failsafe is set to 60 sec- the MK decends automatically at the home-position, no matter if the "use vario for failsafe" was selected in the settings- bugfix: when a seperate channel is used for the failsafe, the nick/roll channels were not set to zero- the correct failsafe behavior is activated as default Hope this helps. Andy.
I'm curious at what angle the kopter is at in sideways flight where the AH is not maintained. Also with the diminished vertical thrust component how much translational left will replenish? Cheers, Jeff
It starts happening even at a very slow forward flight. 5mph even. As you go faster it drops more. I haven't tried the new beta firmware... I'm not fond of using anything in "Beta" on a rig that I use for work... I am also confused since I thought the Z-Axis ACC was supposed to help with this? Shouldn't that be measuring the copter dropping and thus it compensates to bring it back to correct alt? Also I don't use any type of dome here in AZ. It's just way to hot most of the time for me to think that is safe... Josh
I have a Skyjib x8 with 15" props, and mk 2.08, same issue, AH isn't very good. Ive spent hours tuning it.
Josh, does the landing gear come close to the props when up? How long are the booms on that frame? Thanks,