Mikrokopter I'm sorry, Are these things supposed to fly?

[MombasaFlash]

The first flight test attempt was rubbish with the MK Octo wanting to Pitch, Roll and Yaw all at the same time.

The initial level test (Throttle high/Rudder right) and the Compass Calibration were both redone - several times.

Yesterday the Roll and Yaw axes looked as if they were going to behave but the Pitch axis (sorry, NICK - pah, BS) was ridiculous and patently uncontrollable. It appears to be far too slow to react to being off-level and when it does finally correct it way-over-reacts ... and then takes far too long to realise it has gone too far. If the Octo were allowed to attain any altitude the resulting flight would be a jolly event with an ever increasingly violent and jerky forwards & backwards seesaw.

As it would inevitably end up bent, so I have not allowed it get above a few inches yet - although even then it has ended up on its back several times.

Of course EVERYBODY is stupid some of the time but I cannot find any physical assembly mistakes i.e. clockwise/anticlockwise assignment for the motors and wiring from motors to BL-Ctrl's (verified in KopterTool with the individual motor test facility). The Configuration Setting is the straightforward default 'Beginners' configuration or the same with GPS and Compass disabled (Altitude is already disabled by default).

I have no idea which numbers should be changed within KopterTool to smarten up the Pitch axis performance - remember, the Roll and Yaw axes appear to be relatively stable, it is only the Pitch axis which is so obviously wrong - but then, when I am told here on this forum that the default 'Beginners' configuration is sufficient to permit the MK to get off the ground and keep itself under control, why should anything have to be changed at all?

 

[BorisS]

Hi MombasaFlash,

are you sure everything is still in beginner setup ?

Please post your Setup:

FC ?
NAVI ?

Whats happening in your scope to your gyros ? If your using Navi and Gps disconnect everything and disable them in MK tools. Try a solo flight with the FC. Can you show the actual behaviour during flight take off ?

Boris

 

[MombasaFlash]

Hi Boris,

Noting those values before Take-Off is a bit cumbersome since a computer has to be attached but, given that this particular tabletop is not exactly as level as when the Throttle Up/Yaw Right calibration was performed, the AngleNick/AccNick and AngleRoll/AccRoll values are relatively close at -4/-6 and 8/10 respectively at Switch On.

They do tend to drift apart after a few minutes of doing nothing.

The configuration is defaulted using the emergency triangle and then updated with:
a) correct 'poti' settings
b) motor switch on idle speed is reduced from 8 to 1 - because 8 is just plain too fast
c) GPS and Compass both disabled.

Stick & gyro settings are not touched i.e. Nick/Roll P&D, Gyro D or Dynamic Stability etc. because I have no idea as yet what effect changes to those values will have.

I do not even try to fly it now. Instead, I hold it above my head whilst my beautiful lady increases the throttle. I can then feel whatever it is trying to do. As before, Roll and Yaw don't do anything unless I deliberately Roll it in which case it applies opposite pressure.

However, it takes very little time before it starts to fall forwards (or backwards). Then, instead of immediately applying some backwards pressure it continues to fall forwards until it 'wakes up' at which point it shoves itself backwards - but too much, because then it throws it forwards again - and again, by too much.

I am waiting for it to be a little bit more alert and make gentle compensations earlier before I permit it to try and fly on its own but don't yet have a clear idea of what to look for, beyond the physical installation, or change within KopterTool.

The 'setup' is what was supplied in the Octo Basic Kit (FC 2.1, SW 0.84), plus NC 2.0 (SW 0.24e) and GPS (SW 6.02). It all responds correctly to Tx commands in respect of the three axes and throttle.

I will be emailing MK to ask WTF?

 

[Bartman]

M-B-F
It sounds to me like something isn't set right in your copter's brain matter. MK's simply don't behave the way you're describing so let's go back to square one. You sound like an intelligent person so I trust you when you say it's put together correctly and the motors are all turning the same way.
Hook it up to MKtool and let it get itself connected. Hold the Alt button and click on Settings at the same time. If you go to the Mixer-SETUP Tab, there's a small white box in the upper right-hand corner. For your style of Okto frame it should say Octo1 in the box. If it doesn't say that then you'll have to click Load, find the Octo1 file and load it so the FC knows what is being asked of it. Also, the little diagram in that page should show something representative of your motor layout, does it?
In that same tab, near the bottom is should say Parameterset and you should have a 3 there.
If you've made those changes, click "Write" and then OK.
You've got an older style Okto frame, do you know what firmware version is running on the flight control board?
Another issue that I hope you don't have is with the mounting of the gyros. Can you please post a pic of the bottom of your FC in the area where the gyros are? The gyros are the two little squarish boards that are mounted on-end on the bottom of the flight control board.
Lastly, before you try to fly again, remove your Navi board and GPS board. They really shouldn't be installed until after you've got your MK flying more like an MK.

Good luck!
Bart

 

[BorisS]

ah okay i thought you have the possibility to read out the data via xbee or similar, during flight. Hook the bird up to your computer just calibrate it and check out the nick gyro action when you tilt it backward and forwards compare it to the roll gyro. Should be doable with a longer serial cable or usb if you have mkusb. Should give you some insight what going on with the nickgyro !

Look thats the testing i did after my crash to check the gyros and compass http://sprealinvest.at/still.mov

and than on a setup were the kopter can swing back and forward http://sprealinvest.at/swing.mov

I woud suggest you try something similar moving it by hand. I cant say since there are to many variables but it seems like the gyro has an issue.

 

[BorisS]

ah and mobasa did you get rid of this peeping ? what was the origin. If you didnt fix that first !!

 

[MombasaFlash]

Dunno what the peeping was. It comes and goes. Now I am in the office with it powered up from the 12v charger supply and it hasn't uttered that alarm-sounding chreep once all day.

I have been in KopterTool looking for which parameter to change to bring the Nick axis under control. I discovered that reducing ACC/Gyro-Faktor from the default 27 to 01 allowed the ACCNick to follow AngleNick almost perfectly. Unfortunately it completely screwed the previously well behaving Roll axis until the value was returned to 27.

So that's not the right one.

Gyro P, Gyro I & Gyro D do nothing whatsoever apparently because changing their values to 01 or 200 makes no difference to anything - at least, not on the scope.

In the end I have not randomly changed everything in an attempt to find the needle. Instead I have sent an email to MK asking them to provide some information because I am not going to wade through miles of incomprehensible german wiki or equally incomprehensible Google english translation for a problem that is conceivably a manufacturing issue.

We will see whether or not anybody bothers to respond and whether or not they respond with anything useful.

To answer some of Bart's points: the FC is V2.1 with SW 0.84 and there is only one way of mounting the FC board - if you want all the connectors to line up properly later.

The KopterTool scope illustrates graphically precisely the symptoms I outlined previously.
1. The Roll axis behaves reasonably well with acceleration following attitude fairly well - but not perfectly ...

View attachment 753

2. ACCNick lags way behind AngleNick ....

View attachment 754

3. ACC/Gyro Faktor reduced to 01 ...

View attachment 755

improves Nick axis performance immeasurably (on the scope)...

View attachment 756

... but destroys the Roll axis (continued)

 

[MombasaFlash]

... but destroys the Roll axis ...

View attachment 758

4. Gyro P,I & D do nothing whatever their setting ...

View attachment 759

5. And the rudder gyro does not respond at all ...

View attachment 760


And to complete the response to Bart, the Mixer-Setup is for 'Octo'. Other Octo options are for 2, 3 and U.
Individual motor tests within KopterTool show that all motors are correctly assigned and operate in the correct direction. The physical installation is neat, tidy and correct and the Tx channel assignment is correct and the MK responds correctly to Tx commands.

I am not surprised that I have bought the 'lemon'. I have had four years of the same sort of thing with the AP business and the bigger helis.

 

[Bartman]

M-B-F
can you shoot a picture of the gyros on your FC board? You may have an infamous zero-ohm resistor board with loose gyros.
If the mixer set-up is correct and everything else is stock, it should fly dead on right out of the box.
Bart

 

[MombasaFlash]

Hi Bartman,

Now you have me concerned! I took it off to take a photo and I now see there is this great big lump of solder crud smothering one of the (presumably) resistors top left on the Pitch axis gyro board. Ooo-er.

View attachment 761

 

[Bartman]

the gyro on the left doesn't look like it's mounted very well. it maybe even looks like the mounting pad is pulling off the board? and yes that resistor (presumably?? ) looks like it's got some parasitic solder lifeform residing on it.
the good news is you don't have a zero-ohm resistor board.

 

[Bartman]

looking at it again i'd desolder that joint, flux it, and resolder it so you know it's done well.

 

[MombasaFlash]

No, it is just the camera flash and poor focus that make the left one look odd. In reality the joint is shiny and secure.

This closer look at the FC board has revealed the Yaw gyro on the top. So that answers a question - a compass module is not required to stabilise the yaw axis.

So why is the Yaw gyro doing absolutely nothing in the KopterTool scope? Do I have TWO bad gyros?

 

[MombasaFlash]

Its sorted. It was that blob of solder that was screwing everything up. Removing the capacitor underneath all that crap, cleaning it and replacing it solved all of the slow gyro issues and it has just proved it by flying in the back garden (which is about as spacious as a postage stamp). All being well tomorrow will be at the field and all of the other functions can be tested.

Freudenberg at MikroKopter insists that the board was checked (twice!) before being shipped and that I am responsible for the blob by overloading my solder iron and flicking solder around.

A-hole. Seems he has already forgotten about the capacitor that fell off the BL-Ctrl board.

So I have read all about the quality control and I have read all about the customer service attitude and I have now experienced both at first hand. I am not impressed.

My company will NOT be investing in any MK machines. This Octo is just a trial of the drone platform before buying bigger payload machines.

 

[Bartman]

so how did it fly?

 

[MombasaFlash]

Just like it says on the box. Steady(ish), level and predictable and slightly less prone to the normal heli 'bar of soap on ice' tendencies. Still has the 'drone bounce' of course but I hope a little familiarity with the software will reveal what to change to remove that.

I will be enabling all the other functions tomorrow at the field (weather permitting) and, providing it all behaves, the next thing will be to try out the flimsy MK camera mount. If it looks promising (it is not going to be good, that is for definite) and the onboard accelerometers keep the camera acceptably level, I'll get a Droidworx 6 or 8 with a decent gimbal and the Octo will probably be looking for a new home!

But in the meantime, if it REALLY behaves I might take it to UK and use it for some aerials at my nephew's wedding (it will be less scary for the dearly beloved than the Copterworks heli).

 

[jes1111]

... with a decent gimbal ...

I was going to make a smart-*** remark like "Good luck with that!" ... but I won't

Have you made a choice yet? This one is interesting - never seen it mentioned in multirotor forums. Construction looks good - the doubled servos are a good idea (Denny uses same setup) to eliminate backlash. I don't think it's got external potentiometers (meant for direct manual control) but it shouldn't be too hard to rig them. You'd need a simple cantilevered mast to suspend it. In any case the chap is very industrious and would probably quote you a price for building one to your spec.

But in the meantime, if it REALLY behaves I might take it to UK and use it for some aerials at my nephew's wedding (it will be less scary for the dearly beloved than the Copterworks heli).

Be careful flying above people

 

[MombasaFlash]

I was going to make a smart-*** remark like "Good luck with that!" ... but I won't

So ... erm ... did you or didn't you?

Even my Copterworks gimbal has had some serious modification and that is probably one of the better ones, but it requires a large heli.

The Roll axis really benefits from a large Roll bearing with the camera sitting inside it.

I do not expect the PhotoHigher/PhotoShipOne gimbals to be usable out of the box but at least a lot of the manufacturing is already done. The belt drive on the Roll axis on both designs is neat and promises to be more stable than those designs that rely on a central shaft to support the camera tray with the Tilt mechanism which can often be plagued by a rolling wobble in the air.

I have used a Raptor as a test bed for a few years and have constructed umpteen camera mounts for it, learning a bit more with each design.

 

[Bartman]

I have used a Raptor as a test bed for a few years and have constructed umpteen camera mounts for it, learning a bit more with each design.

pics?

 

[AeroVision]

Strange behaviours

Hello everybody,
I'm new here, and I was reading some posts when I saw this thread...
Well, I had a similar crash, couple of weeks ago, and I've realized that the problem
(in my case) was mostly due to EMI... Now I weaved all the small wires around, took off
all the motor cables from the center plate (now are connected coming out from the outside holes of the center plate)
and then I've added a "floor" to the electronic tower, between FC and Navi, made by a thin plastic 6x6cm with some
alu foil glued on both surfaces. Today I check with MKTool, and compass/gps crazy behaviour seems to be disappeared.
Hope this will help
Cheers