for people who are interested this guy is also flying one looks beasty. He is having trouble though http://www.mikrokopter.us/the-gener...her-things)-on-redundant-system/msg45898/#new
What steve said. The problem he has is the U8's with the 28" props. SO far I have had great luck with my MK 3.5 bl and redundant system. Gonna do the Holger test with a servo to pull the FC when I get home. On the bench it works fine and the simulated I2c works too.
Great thread btw dave=) Some things I have learned about the dual MK FC setup. (more to come) The UART board I had came with some pins out the back. Made me a bit nervous so I hit them with the clippers. As you can see the pins COULD cross if they get bumped or damaged. I did not like the molex to molex two pin they included and it was far to short. So I clipped it and went straight to the board at one of the caps (this is safe and tested). I put my main FC bottom, secondary top. Working on modifying a hood from QC to do a stack and will then swap the FC's, but for now since I fly in the snow/rain/dust on a regular basis the cover is a must. The GPS gets a shower cap.
Dave - how did you decide to mount the batteries on the booms like that instead of above the central stack? We were just debating that and have decided to go above the stack on our redundant build. Our pilot was worried about too much weight down low which may cause a pendulum effect he's experienced with other crafts that were bottom-heavy.
Derek: I'd really be surprised if the Cinestar + Gimbal + Camera wasn't already waaay bottom heavy. The combined C of G, is probably down near the top of the camera....just guessing, but... Andy.
Thus our reason for thinking we are better off mounting two very heavy LiPos on the top of the centre stack? Two 16,000 LiPos outweigh the camera.
I would be concerned that if you nick or roll the copter, those batteries might then cause the copter to roll on its back....but you can suspend the full-laden copter from one motor (so the booms are vertical) and get a sense of where the balance point is. Andy.
HI Derek I always run 2 batteries and when I switched over to the QC10,000 batteries it made the copter extremely top heavy. It put the CG too high. Even though it was still below the hub it was about 1/2 inch below the hub with the red camera and gimbal on it. I made some initial flights last year with the batteries mounted on the traditional stock one plate that mounts above the flight control board and I did not like the way it flew. It was not crisp or responsive, and not very wind resistant. Then I remembered a friend of mine on this board Steve Maller made runner plates for his X8 configuration so I mimicked his setup and it flew much better. I love this setup because its so much easier to dial in the Horizontal CG with slight adjustments of the battery plates.
Hey Guys, Ok, you've convinced me - I will add two "battery," booms to our X8 and test with the batteries out there instead of above the stack. Just finishing up the build on the redundant BL with two FCs now so should be in the air soon. Will report back on findings. Question - for the redundant build, does anyone know where on a Graupner MZ-24 you'll see that the MK is in redundant mode? I know on the other Graupners you see a little "R" on the main screens, but the MZ-24 is a wee bit different.
Excellent - I'll dig down to that screen. I think it's about 5 deep in the telemetry section of the MZ-24. Hope to be done the redundant build this aft.
Dave / Ryan, Did you have I2C errors initially when you were setting up the boards? Can't seem to get it resolved.
Been working on the issue most of the week. Busy with Royce at Quad trying to isolate the problem. Firmware are all installed on the boards (2 FCs, navi, BL). MK Tool reports error 17 and wildly, the motors spin up and shut down randomly. It's truly bizarre. Neither Royce nor I have seen that behaviour before. I have packed all our electronics in a Pelican case so Royce can have a look on his end to see if we can't isolate the issue. We are thinking we have a short I2C on the BL at this point. Once Royce has a look I'll report back.
Derek: There is the possibility of I2C errors as the boards start up, so an initial non-zero count is not necessarily a cause for alarm. The time to worry is if the count of errors increases from the initial value. This has been a known phenomenon for quite a while on MK boards. It might be worth posting on http://forum.mikrokopter.de/ just to confirm though... Andy.
It's interesting - I should upload the video we shot as we were attempting to diagnose the problem. I don't think I can upload video to this forum, but I'll pop it up on my YouTube channel and link to it from here. Will also post some Screen Flow videos we shot of MK Tool while connected to the drone. It's a thrill! Give me a bit...
In these videos, the copter has two FCs connected to the redundant BL. It's the behaviour we've been seeing and can't get it resolved. And some video of the drone after start-up sitting on our stand (props removed)... that is the sound of the drone spinning up and shutting off various motors. Applying no throttle. Nothing. Just boot-up and this behaviour surfaces.
As a general diagnostic strategy, I'd start the process of removing the redundant stuff and see if, when you've regressed it back to the non-redundant configuration it still does that weird start/stop thing. It does certainly sound like a problem on the I2C bus between the various boards. The Screenflow isn't at quite high enough resolution to be certain, but it looks like it keeps looking for the Flight Control board(s)? Andy