Discussion Mikrokopter / UAVP / quadrocopter / quadrotor build

[Arthur P.]

Well another little step for mikrokopter, large step for me, so to speak. In the RC Cam forum there was a bit of a discussion whether my simple diode/resistor/single transistor circuit would result in a usable PPM signal. Connected it to the MK controler this evening and the MK Tool and controler both happily indicated a good Rx signal. I could even asign the right channels for throttle, jaw and nick, but for some reason I couldn't get any input off of the roll stick on any channel. When I later went on testing with the USB scope I first had the same problem but then found that one single pin connector on the Rx probably hadn't been plugged in correctly. So further testing with the MK controler tomorrow, but I-m 99% certain that I will be able to use the Spektrum DX7 / AR7000 with the MK with this simple circuit in between

I also received my compass from Robotiknet.de today. And that came with a CD full of stuff, including manuals and free software, seemingly also including a copy of Eagle. Who knows, my try and make a little PCB for the diode circuit for the fun of it (and for practice; that's almost 30 years go that I did that last).

Tomorrow is probably going to be a very critical day as I-ll be trying to convert the ESCs to understand I2C. Hope that goes off well. Then I-d have the most dreaded hurdles licked.

One thing I-m still not fully happy with are the offset values I-m seeing for different sensor channels, but not much reference for that. Nick and Roll gyro and accel offsets and changes with movement are certainly OK, as is the air pressure sensor. But still not sure about Z-axis accel which has a high offset, and about the jaw gyro which has a normal offset but seems to have very limited response to movement. Hope those with mikrokopters (and UAVPs) will start a nice thread on normal values on testing.

@Roland,
Certainly willing to help out, as will all others present, but then isn't that what RC Groups is all about

[Arthur P.]

Build two more AR7000 PWM 2 PPM conversion circuits, now paying a bit more attention to looks. They're sized in such a way that they can be attached to e.g. the top of the RX with double sided tape.

Also did some more checking of the jaw gyro. And I-m convinced it's dead. The only thing that changes when taken out of the circuit are the offset values. In other respects little change. Luckily I ordered a spare last week which hopefully will be in house tomorrow or day after.

And now the ESCs....

[Arthur P.]

Although superficially the PWM2PPM multiplexer seems to work I kept having problems getting the MK to see all 7 channels. Ik kept either not seeing the roll (1) or jaw (7) channel. I even used an additional resistor and transistor to reverse the signal, assuming it might be due to the signal being inverted. However, that didn't help. Then I did some further research.

Here is the scope image of my signal. 7 nice channels, don't you think ???



But here is wat a real PPM signal should look like. Notice the difference ??



My signal misses the opening 0.3msec spike. And seemingly the MK wants to see a full 0-5-0V transition to trigger the timer. So with the signal straight out of the converter it sees channels 2 through 7 but thinks these are channels 1 through 6. And with the inverted signal it sees channels 1 through 6 as 1 through 6, but doesn't see 7.

Now here's the xxx USD question: how to get that extra little spike into the signal (or get the MK folks to solve this in software). It would be ideal to be able to use the 2.4GHz system as it is highly probable --if not certain-- that MK's will often be flying in urban environments.

[thanhTran]

Originally Posted by Arthur P.

...
My signal misses the opening 0.3msec spike.
Now here's the xxx USD question: how to get that extra little spike into the signal (or get the MK folks to solve this in software). It would be ideal to be able to use the 2.4GHz system as it is highly probable --if not certain-- that MK's will often be flying in urban environments.

What's the circuit you use for converting the channels? Maybe the component you use is NOT fast enough, so the combined signal is missing some narrow gaps

[Arthur P.]

Very, very simple. It isn't a matter of not fast enough. It is just putting all the PWM pulses back onto a single line and inverting the signal (which you can correct with a second transistor).

The circuit was taken from the UAVP diagram. I think it works there because they look at uneven channels and probably are just clocking on every transition from 0V to 5V or 5V to 0V instead of only positive or only negative transitions. That is in fact a simple software solution, but it may be a bit more prone to glitching causing shifting of the channels if the glitch falls within the acceptable 1-2mSec from last transition range.

[w.a.s.p]

Hey Arthur ,I found this thread by accident,glad to see you're making some progress on the MK.Off course there are always some minor bumps in the road with projects like these but I must say I admire your electro-technical skills.
Did you cancel the uavp order from L$##$&Y.DE or is it still "out there"?
@Djago (and everyone else who might have a clue)
Are you using carbon rods as a frame? I cant seem to find a shop/supplier in holland which sells +/- 20mm rods (gewikkeld).
Greetings,
Tom.

[thanhTran]

Thanks for the info Arthur. Please let us know how that works for you.

I'm thinking about building the UAVP system. But so far, not sure how to order things and what to order yet . Wondering why they said opensource, but I haven't seen source code yet. Do you think the MicroKopter is better? I want to use the UAVP since they seem to use the good gyro and accelerator there (maybe better performance? )

[Arthur P.]

Hi Tom, I cancelled the UAVP and will not do any further bussiness with named company.

With respect to the frame I-ve decided on a combination of 10mm aluminum square profiles with 8mm carbon tubes glued in with polyurethane for the arms, 6mm alu tubes with carbon tubes glued in, and different densities of multiplex (3mm 3 layer light ply from the DIY, and 1-3mm model building ply which is stronger but heavier) for the camera mount, battery tray and housing for the controler.

I haven't seen many shops with cross-woven carbon rods, other than fishing rods (but they taper). At the prices and very light weight of some of those fishing rods they may still be interesting though, even if you need to buy 4, and throw most of the things away because you can only use certain parts. There is quite big fishing utensil shop in Lelystad with a huge selection including quite cheap but very lightweigh carbon rods.

[Arthur P.]

The UAVP source code is available here http://dev.uavp.ch/trac/browser/ as is the design for the Q4 frame.

The two designs are gradually growing together. Both use 300degrees per second gyros. Both are very stable when properly trimmed. So far the MK was lower powered but used I2C for higher control speed. UAVP used fast commercial ESCs of higher power but was restricted to PWM. They are now also switching to I2C. At the same time the MK will be out with upgraded Holger ESCs which will narrow the power gap significantly. Processorwise the UAVP used a PIC controler, MK a ATMega644. One of the MK adepts has adapted the MK PCB to take a ARM7 processor. Sizewise the MK PCB is about 30% of the size of the UAVP PCB and still offers the same types of sensors standard with only the compass (semi-standard) being external. Some of the MK users are further along with GPS integration. The MK group are clearly more flexible and imho are progressing at a more rapid rate at the moment even if some of the progress is in offshoot branches.

Main reason for me to cancel the UAVP and got to MK: I was just fed up with waiting for the UAVP kits after some 2 months while they should have been available according to the site within 1 week or less when I ordered. The UAVP would have been a bit easier for me to implement hardware wise (Spektrum DX system) and would have immediately resulted in higher power for more complex camera mount and heavier camera. But I expect to see ongoing progress with the MK and can wait for what I knew up front to require a wait.

[w.a.s.p]

Hi Arthur,
There are some kite shops which sell the cross woven(thanks for the translation) carbon rods,tapered and non tapered.Problem is that the largest diameter is around 10mm which is not big/stiff enough for my/our purpose.Maybe a solution like the Draganflyer's cross braced frame is an option?
I would like to loose as much Aluminium as possible and yes even fibreglass rods ,not uncommon in the kite-scene as well are an option I considered.
About "that" shop which sells some uavp stuff,I emailed them repeatedly about the fact that they sent me the wrong motors but no answer at all..Just a big waste of money,for all those who know what a single 2217 style axi motor costs.
Well we will see how your MK turns out,and maybe you will have to build me a mainboard in the near future?.My new roxxy motors should arrive soon so I hope that improves a lot (no hot esc's and empty/swollen batteries after 3 minutes flight time).

[Arthur P.]

I would certainly have a look at those fishing rods (karp poles). They are a lot thicker. They do taper, but that to some degree probably helps stiffen them.

I am not showing much progress but have been quite busy this evening. I-ve finally mustered the courage to give the first TowerPro 25A ESC a second try to get the wires attached to the little legs of the processor and several other locations. In the weekend I was probably tired and frustrated and was using the larger soldering iron which has a reasonably small point. But I actually have a 12Watt soldering iron with an significantly sharper point and thinner heating end. And that did make a lot of a difference. In addition, I had found myself some of those stereo magnifying glasses. And you really can't do this without. In fact a stereo microscope would be even better. If you think those little mikrokopter parts are small, .... you're talking minuscule and crowded on this little ESC PCB. And you have to solder the wire to adjacent legs of the processor. In the weekend I-d get one wire on, and it would come off when I tried the next leg. With the much thinner soldering iron, some old wire-wrap wire (nice and thin), and some patience it actually went quite quickly today. So I now have an ESC with two bridging wires, and 8 wires connected to verious parts and just sticking out. Tomorrow I-ll solder then to a small scrap of experiment board with just sufficient copper leads on one side to then allow me to solder thick wires on to that as well. That little scrap will fit nicely over the lower part of the ESC under the new heatshrink. A programming connector will come out just outside of the capacitator. And the I2C interface will come out on some longer wires. I-ve actually left the normal servo wire intact for now. That way I still have the option to use the BEC 5V power supply to e.g. drive servos and thus not overload the MK power supply.

All in all I have a tremendous respect for what Quax accomplished with these ESCs. This is really like trying to solder threads of hair together.

With respect to the PPM output problem I-ve sent Spektrum RC an email asking whether they might be able to help find a way to get a good PPM signal out of the AR7000. I would really like to be able to use the 2.4GHz as I tend to fly in urban areas with a lot of RF.

I-ve also started working on two frames. I felt 10mm square aluminium profiles were a bit weak (think Slow-Stick fuselage). But thicker profiles immediately go up in weight. Carbon of course is lighter but the tubes tend to split on impacts. So I'm going to try combining the two. For the arms I glued 8mm carbon tubes into 10mm aluminum square profiles, and for the landing 4mm carbon tubes into 6mm alu tubes. Did this using polyurethane glue (gorilla glue) as this tends to fill up open corners and is quite lightweight. For the motormounts I-m still looking for aluminum mounts, but may fall back on homemade ply mouts. And the battery carrier, MK PCB housing and cameracarrier will be a mix of differents densities of 1-3mm plywood

I should have some new pics this weekend.

[eekflyer]

Originally Posted by Arthur P.

Main reason for me to cancel the UAVP and got to MK: I was just fed up with waiting for the UAVP kits after some 2 months while they should have been available according to the site within 1 week or less when I ordered.

Hi Arthur, I liked the UAVP, too. Works with Spectrum, open source, modular, etc.

If I might ask, did you have trouble getting populated boards, full kits, or what? I was hoping to get me a populated board and go from there. All the other parts can be had in the States.

EEK

[Arthur P.]

I had ordered the complete kits from the associated shop. According to the site at the time of ordering I should have had those within 1-2 weeks. After 8 weeks they were still coming up with half-truth excuses for further delays. That is if they responded to questions at all. And they tended to only do that when you started posting about the delays and poor responses in forums. After 8-9 weeks and some final lies I cancelled the orders and will not do any bussiness with that shop in the future.

You may have more luck if you can order the PCBs partially populated directly fromm the UAVP group. But I also was not very impressed with their customer focus. In particular one person in that group doesn't want customers at all, just wants to see this as a real high level research project.

You would certainly be better off ordering the sensors directly from Sparkfun. They've provided excellent service on all my orders even if they got slapped badly by import taxes due to their packaging seemingly attracting more attention. But that wouldn't be an issue if you're in the US already.

The only problem you will have is how to get I2C capable ESCs as even the UAVP is moving to those. Not all PWM driven ESCs are fast enough. YGE30i's and the mikrokopter Holger ESCs are currently the only ESCs which come with I2C interface standard. The only shop claiming to be able to supply the YGE30i's is said shop. The standard Holger ESCs are rather low power. I expect them to be out with a higher powered version with better FETs any day now though. And as a final alternative it may be possible to modify existing ESCs. Some people have succeeded in modifying some dirt-cheap TowerPro 25A ESCs and I-m trying. If I also succeed, it should be doable by most ;=D

[jesolins]

Hello Arthur,
I like your reviews/guidance and am impressed with your study and knowedge of the different Qcopters. I am interested in the MK too. Sounds like you have had a good experience ordering the PCB's from their site? If I just ordered the main board and ESC pcb's from them, what other parts/prices would I be looking at to get one in the air. What parts would I not be able to get from a USA online source? You mentioned Sparkfun as a source. I have had good luck with Mouser.com too. Any others? I guess I would need the programmer for the microprocessor too. Do I have to get the programmer and microprocessor from the MK site too? I would be building my own CF/alum frame too. I am currently flying a highly modded Silverlit Qcopter with piezo gyros, better radio/rx...yada-yada...
Cheers,
Jim

[eekflyer]

Thanks, Arthor for the information on 'said' company.

-EEK