When my CS8 is sitting on the bench the it's measured altitude jumps around (up to 3 meters.) The effect of that when it's flying is that if I tell it to hover in place it will move up and down quite a bit in about a 10 foot range. It's always done this. When I first built the CS8 I contacted Quadrocopter and they said this was normal. I would just like to double check. When on the bench, does your MK flight controller report it's altitude as going up and down a lot or is it steady?
William I'll take a look today. Do you have something like a piece of foam or small piece of velcro covering the small hole in the center of the sensor? This will make a big difference how well AH works. Also consider getting the new altitude ACC board added. Makes a huge difference. AH is now rock steady.
Yes I do. That's to control the effect of wind gusts, right? I had an upgraded flight control board in my hand. Quadrocopter shipped us that wrong board when we ordered a navigation navigation control board, but we decided to do the honest thing and ship it back to them.
Ok just looked at mine. Caveat that I have the new ACC installed. If I put the copter on the bench, power up and do a normal calibration (left stick up to the left) then the Altitude bale in the Analog list is varying from 0 to 1. in the data screen 1 the height and setpoint move from -1 to +3. I think all of these values are in 1/10m. To see a 3 meter jump the numbers would need to be 30.
I see the same thing as Gary. With the new ACC, that's typical ranges. Also check that ACCZ value in MKTOOL is < 700 to make sure that the new ACC is being used. Andy.
I think I have ACC-Z set to 0. I think that makes it so that the accelerometer is not used in the determination of the altitude. I thought about trying different values, but I had to ready the craft for it's next expedition. It's out of my hands at the moment. (Hopefully, this time they will bring it back in one piece.) Would you mind setting the ACC-Z value to 0 and telling me what values you get for the altitude?
I think we might be talking about two different values, William. The one I was was referring to is in MK TOOL, using the Scope function, where you can display a value called ACC Z -- this is the value coming from the Accelerometer and should show (for those flight control boards that have been upgraded with the new ACC chip) a value that is less than 700. The value to which you are referring might be the parameter Z-ACC (note the different name) on the MK Tool "Altitude" window. The Z-ACC value is shown in this document http://www.mikrokopter.de/ucwiki/en/FlightCtrl_ME_2_2_ACC with a default setting of 30, although the folks at Quadrocopter seem to prefer a setting of 50 (which is also, curiously, the value shown in the screen capture on that Wiki page. Hope that helps. Andy.
Oh, AccZ not Z-ACC it's all so clear now. OK I wonder when my Z-ACC got sent to zero, assuming it was 30 to begin with. I'll make the adjustment and see if that clears up the problem.
My mistake, Gary -- it was your posting where you said you have it installed and I attributed that to William. William:I think the default setting for Z-ACC in MK Tool is 0, so I'm not quite sure what you are planning to set it to? Andy.
You said that the default setting for Z-ACC was 30 and that the recommended setting was 50, but maybe that's only if you have the new accelerometer?
William the default without the new ACC was 30. I ran mine at 0 and altitude hold was pretty good. From the MK Altitude definition page Z-ACC Slows down the oscillation behavior by using the z-axis of the MK. But as with everything MK there are other variables that also come into play. Barometric D Slows down the oscillation behavior of the height regulator. A high value will make the AH more precise, but also a lot more sensitive to wind-gusts. A low value will protect the algorithm from gusts, but could drift up and down. So give it a try. Try adjusting the Z-ACC and see if it makes any difference. In the end you should probably do the upgrade.
Hmmm... sounds like I have a problem. When it's high up in the air I don't really notice the problem, but if I hover it at 7 feet or so I really notice how it can't stay level. It sometimes almost touches the ground. The engines sound like someone is gunning them. It never settles down. I can buy a new sensor for 16 bucks and it looks easy to unsolder. Will do. I have to wait for it to come back from the field. I would love to do the upgrade. I am mostly interested in the fully automated flight that is possible with the upgrade. But the boss has to authorize the expenditure of funds. We have spent so much on repairs recently and I don't think he wants to spend any more money on the thing at the moment. It depends I suppose. If they don't crash the CS8 this time (We did a lot more practicing and we understand what went wrong last time. So I'm hopeful) and they get good data then maybe I can talk him into getting it upgraded.
The kids brought the CS8 back from the field. This time undamaged. Hurray! Anyway, I went to test out what setting Z-ACC to 30 would do, but when I went to try it the system was not behaving as badly as it was last Tuesday. Now the sensor was only jumping up and down a meter. So the aircraft was flying a lot better. Go figure! Well I tried it anyway. I can't really tell whether it helped or not. Maybe it did a little. It's hard to say.
I got a new flight controller board in and it's much better. When sitting on the bench, the altitude on the OSD screen stays at 0 meters instead of jumping around up to 3 meters. And when if flies it runs much smoother. It seems I have always had a faulty board.
Thanks, for posting William. That explains the mystery, should anyone else have a similar problem. Andy.