Reissen konnen sie kaum da es nur sehr wenig kraft auf die lenkungen gibt wenn die MK in luft is. Un bei die landung werden die lenkungen zusammen gedrukt, nicht oder kaum aus einander. Dabei sind die locher mit eine scharfe schere geschnitten, also auch keine einschit um ein riss zu starten.

Bis jezt nur "labo-testen", noch kein wirkliche flugtesten.

Erste motortesten heute waren kein success. Glune leuchten auf motor 3 und 4 blinkten schon aber kein motor umdrehungen. Motor 1 und 2 hatten erst gut geklapt aber nach ein paar testen gab es ein bischen rauch aus ESC 1. Konte nichts schlecths sehen aber werde morgen weither untersuchen. Sehe das mal als einer dere ubliche problemen.

Habe heute weitergearbeiten an Mikrokopter 2 und 3 (die miet YGE30i und mit modificierte TowerPro 25A ESCs arbeiten werden), und auch ein weitere set Holger ESCs und FC platine nr 4 zusammen gebasteld. Und habe mich eine losung gefunden om die armen in mienes centrale hub ein zu klicken. Eines tags werd es bestimt hier eine ganze luftflotte von verschiedene MK's geben.

Die rauch aus einer der Holger regler war --wie anders-- ein zerzossene shunt. Weiter keine schaden. Habe jetz alle A20-20L motoren getestet. Allen mit vollgas auf 3S LiPo 18-19.5A mit EPP1045 props. Also das sollte gehen.

Habe gerade auch die HYX C3542 (Axi 2820 ahnliche) motoren mit APC 12x3.8 props getestet. Das war interessant. 3 motoren gehen bei vollgas uber 40A (bis ungefahr 45A), aber ein motor geht nur bis 33A...... :=(( Also da gibt es vielleicht zwei problemen: zuviel Amps fur die YGE30i ESCs, und ein motor mit warscheinlich ein problem. Glucklich habe ich auch nog APC 9x6 props im hause. Werde mal versuchen wie es damit geht. Und ich hatte 5 motoren angefordert. Also ich habe eine reserve.

Arthur P meinte

Die rauch aus einer der Holger regler war --wie anders-- ein zerzossene shunt.

Mein Beileid: diese Geruch kenne Ich auch
Es war auch meine eigene Schuld: Ich habe etwas geändert in eine BL-regler ohne zu testen mit Strombegrenzung (1A) hinter ein Netzteil und KopterTool...Nächstes mal teste Ich immer hinter ein Netzteil. Mit 1A dreht der Motor auch (aber leistung < 20% oder so).
Aber was Ich noch nicht verstehe: wie kann man 10A (standardFETmax) oder sogar 20A (pimpFETmax) durch diese Shunt bekommen

Well, I-m in luck with the spare HYX C3542. Although it arrived with the axle inserted wrong way around, in other respects it is OK, pulls about 44A at full throttle with a 12x3.8 APC SF prop. So now to cut down three more APC-E 12x6 props to 26.5 or 27 cm diameter to get the max amps down to 30-35A which the YGE30i ESCs should be able to handle for short bursts. Surprising how much difference cutting 1.5cm off the tips of a prop can make....

Repaired the shunt on the Holger ESC by soldering a bit of wire across it. Still not sure why motors 3 and 4 were not turning. Just in case I mixed up with the HEX files, I think I-ll try reprogramming Holger ESCs 3 and 4.

At last some progress again. The Holger ESCs and the unitedhobbies.com A20-22L Hacker like motors clearly were not happy with each other. 1 ESC worked OK. On 4 others the shunt burned out on connecting power with the motors connected. Motors had tested out OK with Amp draw on normal ESCs. On one of the Holgers one of the FETs also burned out.

So now I-ve changed to converting TowerPro 25A ESCs to I2C using Quax's manual and Hex files. And that seems to work better for me. I now have 2 good conversions with the motors giving high thrust at full throttle. However, two others did program correctly, but the motors only kindoff kick back and forth and the ESCs quickly get warm. Most likely the little wire bridge between the two resistor arrays or the connections for SDA and SCL have little solder bridges. These are the real difficult areas of the conversion due to limited space to work in for the connections. For the latter I-ll probably redo any suspicus wiring and see whether that solves the problem.

One thing that I have done differently from Quax is that I actually taped a small piece of prototype board 8 pads wide and 2 pads deep to the motor wirde end of the ESC over the restisor arrays. All wire connections from the processor, 5V, GND and RST are connected to that. I then connect a 3 wire I2C connection (SDA, SCL, GND) to that, and temporarilly connect a programming connector (a mutliplex battery connector with 6 pins) to that as well. Once the ESC works correctly the programming connector is removed and the ESC heatshrink covered. This allows me to still have the normal servo connector so that I can bring out the 5V from the BEC and use that as power for servos, etc.

I-ll post pictures later.

Go Arthur!

Hab jetzt ein der problem-Towerpros wieder von alle drahten gelost. Da gab es eine kleine lotbrucke hinten die SDA und SCL pins, und vielleicht auch zwischen zwei pins der wiederstandsarray. Also jetz gute hofnung das es mit ein neuversuch gut klappen sol. Also hoffentlich in wenige tagen endlich erster flugversuch mit ein oder zwei MKs.

Weil es mit zwei anderen gut geklapt hat, habe ich mich schon einige weitere TowerPro 25A ESCs bestellt. Und zum testen auch diese neue 860kV 22A motorchen: http://www.unitedhobbies.com/UNITEDHOBBIES/store/uh_viewItem.asp?idProduct=5691

Hier die bilder von das Konvertieren:




Centimer scale, TowerPro 25A ESC with the connection between the two resister arrays at the bottom left already grinded away, and the original jacket of the ESC. Of note this is one of the ESCs which caused some problems on first conversion attempt, so I removed everything and am here starting over. That is why there are already traces of solder on some of the processor pins.



Centimeter scale, wirewrap wire (blue wire), TowerPro 25A ESC, and tip of soldering iron. Note the limited space between the top side of the processor and the doubled stack fo 7805s. Not much space to position the wires, get the soldering iron in cleanly over just one wire, and see what you're doing. The nice thing about the wirewrap wire is that it is as thin as the processor pins. The bad thing is that it does not solder very easilly. A real thin copper wire with reasonably heat resistant sheath which could be well seen for removal would be better. With electromotor winding wire you can't see the sheathing very well.



Bridging wire between two resistor arrays at the bottom right of the ESC in place. That, together with connecting SDA and SCL cleanly, is the most frustrating part of the conversion.



SCL and SDA wires connected to the processor.



Wires for programming connected to the pins of the processor. See Quax's manual for explanation.



8 pad prototype board held in place by double sided tape and with pads already with solder on them. From left to right:
1= SCK, 2 = RST, 3 = MOSI, 4 = 5V, 5 = MISO, 6 = GND, 7 = SDA, 8 = SCL.



All control wires connected to the prototype PCB. Still have to connecto 5V and GND coming from the 7805.



I2C interface soldered to the PCB sticking towards the top. Brown = GND, Red = SDA, Orange = SCL.
The program connector soldered to the prototype PCB sticking towards the bottom. Blue = SCK, Green = RST, Orange = MOSI, Red = 5V, Yellow = MISO, Brown = GND.



The program connector, a MPX battery connector. Blue = SCK, Green = RST, Orange = MOSI, Red = 5V, Yellow = MISO, Brown = GND.

Once the ESC has been repogrammed and has tested correctly I remove the programming connector and use cyanoacrylate to glue the wirewrap wires to the top of the ATMega8 to reduce risk of tracking causing the wires to come loose.

On future conversions I-m going to use only a 3 pad prototype board to connect GND, SDA, and SCL and am going to just temporarilly connect the programming connector to the CPU. The programming connections are rather easy to do, haven't caused any problems so far. The rerouted connection, SDA, SCL and resistor array bridge are much trickier and are in part covered by the wider PCB. With only 3 pads I-ll have less mess and easier access to those difficult connections if they cause problems.

Hadn't posted in this thread for quite a while. But had been quite busy over the last months working on different parts of my Aerial Photography Multikopter's (APMk's) Had quite a number of setbacks with shunts on Holger ESCs burning out, motors being destroyed by an error on my side converting TowerPro 25A ESCs, having to develop an interface for the Spektrum AR7000 together with Mr. Cam Man and others, running into the small hiccup of TowerPro 25A/2 ESCs overheating at low load, and a few other little distractions, and eventually deciding that I was just going to take my time and progress everything in a more or less stepwise fashion towards the concept I had in mind: a multikopter frame designed specifically for aerial photography with the following requirements:
* modular,
* space enough to mount different controlers and sensor in the hub,
* ability to quickly remove the click-locked motor arms as integrated units with lighting and ESCs,
* ability to mount up to 8 motor arms (4, 8, 12, or 16 motors),
* landing gear quickly removable and versatile enough to allow mounting of skids or floats as option,
* space for batteries of 3-4S from one or two 2100 up to two 5000 mAh,
* quickly removable integrated self-leveling camera mount (with optional servo roll and nick stabilisation) with tilt servo,
ability to mount different compact cameras, or exchange for a larger mount for SLR's or video cameras,
* small package size disassembled into motor-arms + camera mount + hub so that the machine could be easilly taken along on business trips.
* low cost of frame.

I went through quite a number of prototype frames which were never completely right. Finally, over the last weeks, after putting drawings into QCad I made a series of three frames, 2 lightweight and one with 8mm carbon rods glued into the 10x10 alu arms to mount much heavier motors. Frames were made primarilly of 2mm modelbuilding plywood (about Euro 12.50 for a piece large enough for the parts of 1 frame, including the camera mount and motor mounts). In the same period I developed a programming cap from Fimo clay to be able to quickly reprogram ATMega8 processors without having to solder too many wires to them, and reprogrammed and converted some more TowerPro 18A and 25A ESCs, and integrated the LED lighting with them, using their BECs. And I developed a motor thrust stand which allows those integrated power-unit arms to be pulled out of a frame and put into the test stand.

Today APMk nr.1 (I've collected electronics for four over time ) had all electronics mounted, including Turnigy 2213 1050kv motors and EPP1045 props, except for the receiver and PWM2PPM converter. And then the final test, connecting everything, putting in a battery, hooking up the MySmartUSB, connecting the battery, hearing the musical Quax's , turning on the FC with v.0.68d firmware, hearing that come to life, and then turning on the motor test feature and getting bombarded with that horrible beeping (is there some way to turn that off ??). With one hand on the top of the canopy I opened up the throttle. And the thing just lifted off, most of the time just pressing up a little bit against my hand but otherwise more or less sitting still in the air. Space in my study is too limited and my cables too short to even try letting go of it. But this was the first little flight, finally

Over the next days I-ll be finallizing APMk nr 2 and nr3 (with Turnigy 2217 and with HYX 3542 motors), doing the finally installation of the AR7000 receivers with PWM2PPM converters, finalize the camera mounts with installation of the tilt servos, and doing some indoor testing with the MK tool to ensure settings are all OK for the PWM2PPM converter. But the first real RC testflight should occur coming weekend. And then I-ll post pictures

gratulation arthur P
i´m waiting for the weekend pics

Hi Arthur, welcome back! We hebben je gemist

Are you ready before march 1st ?? (Köln!)

I expect so, but not sure I can make the meeting due to other commitments.

Leider bin ich durch das slechte Wetter und Arbeit nicht zum Hallenfliegen gegangen. Aber heute gute vortgang: drei APMk's haben heute geflogen unter RC control met Spektrum DX7/AR7000/PWM2PPM converter. Nur indoor und mit die hand mitcontroliert. Aber doch, geflogen haben sie. Und alle drei waren sehr stabiel. Habe jetzt auch die kamera tilt eingestelt. Das ist mit die DX7 ein bischen schwierig weil die nicht 127 oder 255 bit auflosing hat, statt 64 oder so. Also offset auf ungefahr 30 und dan limits auf ungefahr -35 und +40 oder so etwas. Muss jetzt nur noch herasfinden wie ich die Kamera aufloser vom FC aus kontrolieren kann....

Nachtste tagen gibt is noch immer schlechtes wetter. Aber am nachtsten tag mit gutes wetter solte es endlich gelingen ausen zu fliegen und fotos von APMk nr 1, 2 und 3 zu machen.

Just figured out how to control the shutter servo through the FC. With v.0.68d pads J3, J4 and J5 are channels 5 though 7. Now the AR7000 has a rather strange channel sequence: 1 = roll, 2 = flap, 3 = (landing)gear, 4 = nick, 5 = aux2, 6 = throttle, 7 = yaw (gier). For the normal FC controls this is not a problem as you can assign each real channel to a virtual channel. However, J3, J4, and J5 are bound to real channels. Would have been nicer if they had followed the same potX approach as with the tilt servo on J7 as that would have allowed free allocation of the channels to the outputs. But this isn't the case. So J3 = ch 5 = AUX2, J4 = ch 6 = THROTTLE, J5 = ch 7 = YAW. That does mean I'll have to change my normal routine and use AUX2 to control the shutter, FLAP to control the tilt and GEAR to control the altitude sensor, but I should be able to get used to that quick enough, particularly if I change the setup on my EasyStar and EasyGlider to follow a similar switch layout.

Finished binding all wiring late this afternoon. And decided that it was time for the outdoor test-flights. Conditions: overcast, dark, 8C, wind gusting up to 45km/h. So excellent testflight conditions. Wouldn't ever have considered a first testflight with an airplane and certainly not with an helicopter. But in view of the indoor stability, I considered the APMk's should be able to handle. Of course in all the excitement not all went well.



APMk nr 1 (the lightest of the three with Turnigy 2813 motors) lifted off very nicely, immediately hanging into the wind. Of course with me not yet used to the controls I was bouncing about a bit. And with its light weight it clearly was being pushed around a bit by the wind. Sadly I was getting low voltage alarms straight off. I-d put in a 3A 11.1V battery, but a 10C and hadn't recharged it. Still I lifted off 4 or 5 times for short hops, going up to about 3 meters max.



APMk nr 2 (mid-weight with the Turnigy 2817 motors) also lifted off nicely. But I was still too twitchy on the controls. So I bounced it back on the ground a bit hard. One of the hardwood landing gear block split and the leg came off. I did find I could keep it to stay upright on a little slope, and took off twice more, going up to about 10 meters max. Handled the wind excellently. And my twitchiness was decreasing. Still, decided not too push my luck too much as it of course tipped over on landings.



APMk nr 3, the heavyweight had a problem. Although the ESC armed correctly, one motor would not start. Unclear why not. Will have to look at e.g. the wiring and the TP25A ESC this evening. Shame though. With its heavy weight it should be the "wind-beater" of the three. Maybe tomorrow. Expectations are even worse than today re wind and rain

Hi Arthur,

Impressive photo's, owning 3 mikrokopters.

Vraagje: waar heb je die props vandaan?

NICE Arthur !!
They LOOK very nice. Hope they will all fly without problems soon.
looking forward to see a video.

Or live...... I live around the corner.

ps... ik stuur je nog een email over de Benelux-dag.

GOOD LUCK FURTHER!!

Wow Arthur,

1 MK, no... let's start with 3!
Maybe time for a "bakkie koffie"?

Groeten Chris

Actually, I have the parts already for nr 4. Only my remaining AR7000 is defective, and I want to use it for an updated frame design and other mods, i.e. for an "experimental" ;=))

Looked at that TowerPro 25A ESC which isn't responding to TWI/I2C commands from every angle. Couldn't find anything wrong with it. Even redid most of the wiring, and reflashed the firmware. But no effect. So I think the cause is a defective SDA/SCL port on the processor. It was an early conversion, so maybe I had to redo some soldering at the time and overstressed something leading to early failure.

Anyway, I still had 8 TowerPro 25A's around, so I just converted those. Conversion time is down to about 45min per ESC and an 8/8 success rate. Practice helps ;=)) I also modified the landing gear, retaining the 6mm carbon rods in small hardwood blocks, but adding some carbon tube cross-braces and two 50cm carbon skids. Much stiffer at not too much additional weight and stil a small package for travel. Hope that will be better able to handle the medium and heave APMk. With these changes APMk nr 3 is now again ready for testflight. Weather coming days is not going to be very good with lots of rain and stormy winds. But later this week should be better.

It's indeed weather to stay inside. Congrats Arthur with the 3 babies

After the incident with the landing gear leg on the medium weight APMk I decided to modify the landing gear on all three. I-ve added 6mm carbon rod horizontal braces between front and rear legs, and between the rear legs, about 5cm up. This already distributes forces better and significantly stiffens the LG. struts are help in place with simple 8mm plastic tubing 6cm long with a hole cut in the side halfway along the length. Easy to disasemble, and weighs little. Also added two 6mm carbon rod "skids" with "pool-noodle" pipe isolation around them. Doesn't look very flashy, but with the rear 1/3 of the skids coloured red, I should be able to better orient even at a distance. And it does add some additional stiffness and "camera-protection".



Sadly the weather is way beyond anything resembling test-flight conditions. But the forecasts for the weekend are a bit better.

Well APMk nr 3 flew. But just for about 15 seconds. Then there was suddenly very serious vibration of one of the motors necessitating a quick landing. Due to an uneven landing the craft tipped over, shortly blocking 2 props. Result: smoke, then fire from the two TowerPro 25A/2's.











After the battery had been disconnected the fire was quickly out. Other ESCs and FC seem OK on inspection. Motormount of motor 2 is cracked. And motor 3 seemed to be stuck, but after opening it was probably just some dried clay it had dug up. So all motors seem to have survived. But some serious testing to do to figure out which motor suddenly vibrated and why.

On the bright side APMk nr 1 flew fine on Turnigy 2213/20 motors with EPP1045 props on 2150mAh 3S lipos with RTF weight of 1345g. Probably need to loose some weight on this one.

And APMk nr 2 flew very nicely on Turnigy 2217/20 motors with EPP1045 props on 4000mAh 3S lipos with RTF weight of 1570g. Could also do with some weight reduction, but I think these motors are a better match.

APMk nr 3 had HYX 3542C motors with 12x6 APC-E props cut down to 11 inch to keep amps manageable, and a 5000mAh battery, RTF weight 2077g. It had a bit of a bad rolling tendency on first take-off, then stabilized up to about 1m50 into a hover, then developed the vibration requiring quick landing. Was at or just below half-throttle, so enough power for its weight.

Started on APMk nr 4, using Turnigy 2830 motors with integrated motor mounts and 3mm prop shafts. Mounted the motors directly to the alu arms. Am aiming to get the total frame weight with camera mount ex battery and camera down with at least 200g, and will be moving the battery to a location below the central hub (much closer to prop level in the current frames) for easier access and to see wether a slighty lower CG will help get rid of some oscillation.

The pipe isolation around the landing gear skids may not look very flashy, but the two colours (red = rear) really help with orientation, particularly at a distance. And just a bit more shock absortion. Landing gear on all three held up well with several "bouncy" landings.

Little update on the fiery mishap with APMk nr 3. Tested the FC with the MKTool. Everything seems in order. Haven't reconnected ESCs yet, but as it was still controling the last two ESCs while the other two were burning, the TWI/I2C interface should be OK as well. Also tested each of the motors, two with their original ESCs and 2 with newly converted TP25A/2's. All work fine. So they also survived their little adventure.

The vibration problem seems due to the motormounts. One mount had longitudinal cracks. On two motors the 4mm modelbuilding plywood mount clearly starts lifting up a bit at about 1/3 throttle and at 1/2 to 2/3 thottle suddenly serious vibrations start. With the other two motors this is slightly less. Although all are mounted on 10x10 alu arms strenghtened with 8mm carbon rod glued in, the two worst vibrators have a less tight fit in the motor test stand than the 2 better ones. However that probably is just an additional factor. I had previously tested the same motors on differently shaped mounts on a rather flexible wooden test stand and had no serious vibrations. So most likely resonance based motormount vibrations.

So next steps are:
* Recheck the balance of each prop. I am sure I balanced them very carefully after cutting them down to 11 inch to keep amps under control. But just in case...
* Increase the thickness of the motor mounts to at last 6mm and glue some further stiffeners along the back side of the mouns and along the alu arms.
* Consider a small bolt through the frame and arms to ensure no play of the mount in the frame. Probably wise as well as these high powered motors do want to pull the arms out of the motor-teststand, certainly when there is some vibration.

Had a very good test flight in our small garden this evening with APMk nr 2 with the Turnigy 22172/20 motors. Almost 10 minutes on 4A battery which had lost just 2388mAh. Flight was much more relaxed than previous flights which may explain the somewhat lower battery drain. And calibrating it prior to flight probably helped as there was no adverse yaw anymore and quieter hover and thus much less need for control input. Still will need a few more flights before I-m so comfortable with it that I can start focussing on camera tilt and shutter while flying. Weather forecasts are not very good, but maybe sometime over Easter...

Received 4 Turnigy 3-5A UBECs and 4 Turnigy 2217/20 (860kV) motors today.

Modified the 3V power supply on all my FC's even though on 3 the baseline curves before and after installation of the 3.3 uF tantal capacitator did not differ much. On the fourth there was a rather severe irregulatory of the ACC curves before installing the cap while after installing the cap the curves were comparable with those of the other three FC's. On APMk nr 3 I also installed one of the above UBECs. The other two APMk's have 2A UBECs as additional power and seem fine with that so didn't change for now.

While testing one of the new motors with 10x6 GWS three bladed prop and one with an APC 12x3.8 SF prop it became clear that with high power and the APC-SF prop there is also resonance vibration risk with a 3mm plywood plate motor mount. Propellor was well balanced. The 10x6-3bladed seems a very good match, possibly even underutilizes the motor (about 15A and almost 1kg at full throttle). The APC-SF also seemed very usable with quite low amps at half throttle, but I didn't reach full throttle due to the vibration problem. This really seems due to the flat plate design. So I will have to modify all my motor mounts to be a bit thicker and probably have a bar glued against the back to stiffen them up significantly. Oh well, that's what testflights and thrust-testing of motor-prop combinations is for And in fact for the HYX3542C's I have mounts of an earlier design which are a lot stiffer and have additional bars to avoid vibration. So that's just a matter of moving the motors over.

Weather forecasts are not too good for tomorrow, but I-m hoping to get some testflights with APMk 1, 2 and 3 in on Sunday, hopefully even get some first flights in with a camera mounted.