RE: Walkera/Dragonfly/Esky 4ch transmitter/receiver technical review.  
View related threads: (in this forum | in all forums)

Tower Hobbies
Enter up to 4 keywords or Tower stock numbers
Logged in as Guest



Users viewing this topic: none
  Printable Version
       


Miniature Aircraft USA Ion-X - RTF
Seller:  1surfer
Details:   $5,000.00   |  10/5/2008   |  Classified Ad
We will rotate YOUR AD in this spot if you select "Forum Featured" when placing or editing your ad!

All Forums >> RC Helicopters >> Electric RC Helis >> Walkera Helis - Dragonfly, etc.. >> RE: Walkera/Dragonfly/Esky 4ch transmitter/receiver technical review.
Page: <<   < prev  1 2 [3] 4   next >   >>  

Login
Message << Older Topic   Newer Topic >>
RE: Walkera/Dragonfly/Esky 4ch transmitter/receiver t... - 11/21/2007 2:15:06 AM   
e1000


 

Posts: 24
Joined: 1/23/2007
From: san juan, ARGENTINA
Status: offline 		Yes, the "metal casing" chip is the gyro, youll find hardly any spec or datasheet. Its a piezo gyro, most of them are manufactured by Murata... i cant tell you much further than that. If youre interested, ill try to explain how it works, from what ive gathered after reading several "almost" datasheets and briefs... if you are not, its enough to know that they give a signal, an output voltage proportional to the angular speed on one axis. They are active, it means they need proper feeding, and output 2 signals, a reference Vref and the output itself, Vout, referenced to Vref.

The uP is a common, 8 bit, 10bit ADC, 16bit-2 channel PWM, RISC arquitecture microcontroller. Quite much like any PIC... but better . And yes, it runs on 4MHz.

The diodes are protection and/or freewheeling diodes.

Drags my attention both pots, apart from the gain pot.... the Zero pot, and the other, what do they do?...

As i can see, there is a recepcion PCB and a processing PCB...thats good, perhaps we can try some home-made controller.

(in reply to huguesdb)
       Post #: 51

RE: Walkera/Dragonfly/Esky 4ch transmitter/receiver t... - 11/21/2007 3:20:03 AM   
e1000


 

Posts: 24
Joined: 1/23/2007
From: san juan, ARGENTINA
Status: offline 		here i post two images, one showing everything before any change and the other with the nasty artwork done. Basically its adding more power, thats all. Ive bypassed the onboard mosfets and added a plug, mosfet and mosfet driver of my own. Ive not tested any of that yet.
quote:

rx601 is made for the CCPM helis - i have one on my DF52 - which is 120° swashplate - and the DF#4 is 90°
yes, ive a DF52 too! in fact, i got this receiver from walkera after purchasing my DF52, i thought it was a good idea to have a #4 controlled by a #52 radio. The rx601 is just a 6 channel PCM receiver, the hard work is done at the controller, and it has on the back some DIPswitchs for configuring operative mode. So, you put CCPM/NORMAL in NORMAL mode, and is just a transmitter. Plug the servos on the AIL and ELEV plugs, and.... well, everytime i go on after that, something gets broken. But it will fly....eventually. What i really like is that you can control exponencial curves on the sticks and can modify gain/type of gyro...

quote:

I have some stuff for PICs, including the dsPIC 28 pin
Ok. So, we can program a PIC, wich accepts THROTTLE, AILERON and GYRO signals and translate them into two signals PPM for MOTOR1 and MOTOR2.... or at least, that would be a starting point.
Which model is the V-Mixer and gyro you have? I want specs.... im still convinced something can be done with those.

Attachments
Click to see fullsize image.
Click for fullsize 		Click to see fullsize image.
Click for fullsize

(in reply to huguesdb)
       Post #: 52

RE: Walkera/Dragonfly/Esky 4ch transmitter/receiver t... - 11/21/2007 7:52:17 PM   
e1000


 

Posts: 24
Joined: 1/23/2007
From: san juan, ARGENTINA
Status: offline 		ok, here is everything glued together. #52 receiver and #4 frame. The servos are e-sky, they must be modified too to work properly, because the center PPM widht pulse walkera uses is about 1250 uS...minimum near 700uS and maximum near 1800uS.

Attachments
Click to see fullsize image.
Click for fullsize

(in reply to e1000)
       Post #: 53

RE: Walkera/Dragonfly/Esky 4ch transmitter/receiver ... - 11/22/2007 12:32:23 AM   
huguesdb


 

Posts: 41
Joined: 4/29/2007
From: , SWITZERLAND
Status: offline 		Now the burning question: does she fly more than 5 seconds in the air?

quote:

Ok. So, we can program a PIC, wich accepts THROTTLE, AILERON and GYRO signals and translate them into two signals PPM for MOTOR1 and MOTOR2.... or at least, that would be a starting point.

OK, looks good but i think we need gyro, rudder and throttle. I'm assuming we will work this with a simple simon 4 channel dumb transmitter.

Here are some rules :
0) Basic rule is :
Motor 1 = throttle + M1 * (Rudder-midpoint +C1) + M2*(Gyro-midpoint+C2) where Mx is a gain and Cx an offset.
Motor 2 = throttle - M1 * (Rudder-midpoint +C1) - M2*(Gyro-midpoint+C2)
It may be necessary to increase the rudder effect more than decrease it. Same for Gyro.

1) if there is no throttle, all motor outputs are inhibited, even if the gyro sends signals. Maybe there is a throttle threshhold below which the gyro is ignored. Otherwise, with no throttle, u turn the heli and one balde starts turning. Can be annoying.

2) motors are totally inhibited untill throttle comes down to zero. This may complicate ESC programming but is a safety feature.

3) In the absence of inputs, some neutral value is assumed, gyro action continues operating

4) If we include channel 5 or 6 input, say, this could be used to tune the gyro gain - low end gyros cant be adjusted remotely.

5) In a second stage, we can think of some noise reduction in case of short transmission cuts. The Spektrum DX7 seems to do a pretty good job at this.

6) If we include the aileron and elevator inputs, its possible to program some nice turning and banking sequences.

quote:

Which model is the V-Mixer and gyro you have? I want specs.... im still convinced something can be done with those.



The V-mixer is some gereric 5$ thingy with no name and no documentation. I did test it with servos and it functioned as expected. I will redo some tests when i have a bit of time!


Have fun!


Hugues

(in reply to e1000)
       Post #: 54

RE: Walkera/Dragonfly/Esky 4ch transmitter/receiver ... - 11/22/2007 1:58:06 PM   
huguesdb


 

Posts: 41
Joined: 4/29/2007
From: , SWITZERLAND
Status: offline 		I tried the Gyro/V-tail mixer setup again - and it does seem to work.

Here is the configuration:
The Lama receiver part of the 4-in-1, with servos connected to aileron and elevator channels.
Throttle receiver channel directly to the 'elevator' input of the V-Tail mixer, rudder receiver channel to the gyro input. Gyro output goes to 'rudder' input of the v-tail mixer.
This behaves just as one would expect - throttle is distributed equally to both motors, rudder + to one, - to other, gyro correction + to one, - to other.

I tried a walkera G011 gyro but there was oscilation of the power on both motors - none of the possible combinations of gyro settings worked (ds/nor, delay,extent, gain, AVCS or normal).
With an e-sky 0704 gyro, it does work - as long as the gain is turned down to about 15%! Otherwise it oscillates like snake!

As long as the battery is well charged, it does fly very well.. really nice hovers, very fast horizontal spins, in both directions, without a significant change in altitude, and generally quite responsive.

However, as soon as the battery gets somewhat discharged (about 2/3 of its normal flying time), there is some pretty unpredicatble behaiour - like the rotors turning very fast and not responding to the throttle and more.
My guess is that this is due to the gyro needing 5V - I will do tests with a 11.1V LiPo.

So your hunch was right e1000, it does work.

The disadvantages of this solution are that one cannot set the rudder mixing as the v-tail mixer has a fixed gain, nor compensate for the notoriously unreliable electronics in these small machines.

The advantage is reusing the original electronics (radio and receiver) and low cost.. 10$ for a mixer (I said 5$ before.. that was a mistake - forgot shipping) and simplicity.

Have fun!
Hugues

(in reply to huguesdb)
       Post #: 55

RE: Walkera/Dragonfly/Esky 4ch transmitter/receiver ... - 11/24/2007 7:22:32 PM   
e1000


 

Posts: 24
Joined: 1/23/2007
From: san juan, ARGENTINA
Status: offline 		Ok, so hands to work. Perhaps the set up youre using cant perform any better, i mean, that is the best youll get off it. Maybe, with another gyro, a V mixer giving you more freedom with variables, etc., something better could be done... but, well, that is another field.

quote:

Here are some rules :
0) Basic rule is :
not many people would understand why you start with 0, and not with 1... im surprised you havent ended the sentence with a semicolon!

Ill try to get a small PIC, for example, 12F675 or alike, so theres very little space and weight in the project. See if the hardware you have (programmer, compiler, etc) support this family.
To begin with, the ecuations youve stated are excellent and self-explaining. But i dont know if the would work well with different gyros (Normal and Head-lock, for example). However they are good point to start tests.

quote:

OK, looks good but i think we need gyro, rudder and throttle. I'm assuming we will work this with a simple simon 4 channel dumb transmitter.
Yes, i meant Rudder.
Ive never learnt the names in english, its like larboard and starboard, or prow and poop...never could tell them apart on a single shot.

The rules are OK, ill add some little bits as an extra precaution. For example:

quote:

3) In the absence of inputs, some neutral value is assumed, gyro action continues operating
-> The program "creates" a THROTTLE signal wich goes from the value it had untill it lost signal, gently decreasing to zero, and then both ESC and gyro stop working.... and stuff like that. Ill keep you informed, it might take me a week to get some PICs and maybe 2 more to get some decent working-code.

I never replied some things,
those probes you showed me are good. They are active probes, it means you need to feed them with a 9 volt batt or something, but they are bound to work. The signal goes in through the LINE IN jack. I tell you again, be careful, try not to roast anything in the motherboard. Ive tried the soft, Audiotester, it seems awesome. Its a bit buggy, but considering its shareware, its great.

You can try to get a capture of the signal that comes out of the receiver an goes into the 4 in 1 uP. You might want to start with the signal present in pin 1, thats the NMI of the uP, and most walkera receivers use this as an input for the incoming PPM/PCM signal.

OK, ill go for now. Mail me if you need any help.

(in reply to huguesdb)
       Post #: 56

RE: Walkera/Dragonfly/Esky 4ch transmitter/receiver ... - 11/25/2007 5:23:58 PM   
huguesdb


 

Posts: 41
Joined: 4/29/2007
From: , SWITZERLAND
Status: offline 		I have received a E-sky 0704A gyro so I will try with tis as well. Its half the weight for twice the price!

OK, so we have a project starting!!! I sent u an email with some questions. Let me know if they dont arrive.

quote:

But i dont know if the would work well with different gyros (Normal and Head-lock, for example). However they are good point to start tests.

I will try to have my PC oscilloscope soon so that I can see what comes out of the gyros, etc. Hopefully I wont burn my PCs. I'll use an old one that I have been wanting to replace! Any excuse will be good!

quote:

-> The program "creates" a THROTTLE signal wich goes from the value it had untill it lost signal, gently decreasing to zero, and then both ESC and gyro stop working.... and stuff like that.


Yes, something like that. The only problem I can see is when to decide that its no longer a temporary transmission problem but a total break in communication. And then, once we have stopped the motors in mid air because there is a transmission failure (!) what do we do if we have a short revival in transmission?

Some pilot assistants / transmitters / receivers I have played with have a programmable servo position to assume in case of transmission failure. They suggest something like all controls neutral and power 40% - this should provide a non fatal descent.

However, this should not be applied if the heli is on the ground - so the failsafe function is something like MIN(last good position, 40%).

OK, off to dinner..

Hugues

ps. looks like we are the only crazy ones in this forum! Maybe we should start a thread just for this subject.. what do u think?

< Message edited by huguesdb -- 11/26/2007 2:51:41 AM >

(in reply to e1000)
       Post #: 57

RE: Walkera/Dragonfly/Esky 4ch transmitter/receiver t... - 11/26/2007 2:39:19 AM   
huguesdb


 

Posts: 41
Joined: 4/29/2007
From: , SWITZERLAND
Status: offline 		OK, so I made the PC oscilloscope - kept me working for a while till I saw the microphone was muted! I was getting a small but real signal at my test frequency but it was barely above noise. I was wondering if my approximations had some unexpected effects (I used a LM324 and took some liberty on resistor and capacitor values).

Right.. so here is a pulse from a receiver at 1.5 ms. 3:30 am so sleep time now! Pity the shareware software stops runnig after some minutes.. annoying to have to restart it! But then, nice of the guyy to make it available for tests.


Hugues

Attachments
Click to see fullsize image.
Click for fullsize

(in reply to e1000)
       Post #: 58

RE: Walkera/Dragonfly/Esky 4ch transmitter/receiver ... - 11/26/2007 9:19:57 PM   
huguesdb


 

Posts: 41
Joined: 4/29/2007
From: , SWITZERLAND
Status: offline 		Dammed. I started this post and waited so long to finish it that I lost all the text.

I looked at the signals on the 4-in-1 and as you expected, the outputs are directly PWM, using the PWM1 and PWM2 oputput of the microprocessor. Logical! And the mixing is done internally. PWM is about 5k Hz, going from 30% to 90% duty cycle (approximatly).

Throttle input goes to the INT(errupt) pin and the rudder to CIN-, the comparator input. I assume a difference between CIN+ and CIN- generates an interrupt?

Nothing else really to note. Maybe it will be easier to take the PWM output and convert it to a 1-2ms PPM for the external ESC. All the logic will be in the 4-in-1 and the device will just convert PWN to PPM.

However, this would lock us into using the 4-in-1 which is not the epitomy of reliability. What we have in mind is much more interesting!

Hugues

(in reply to huguesdb)
       Post #: 59

RE: Walkera/Dragonfly/Esky 4ch transmitter/receiver ... - 11/27/2007 2:40:17 AM   
e1000


 

Posts: 24
Joined: 1/23/2007
From: san juan, ARGENTINA
Status: offline 		ok, lets go step by step.
quote:

PWM is about 5k Hz, going from 30% to 90% duty cycle (approximatly).
Check this value again. If duty is 30%, the 5khz tone would be audible and the motors would have some lor rev. spinning, even with the Throttle at 0....

quote:

Throttle input goes to the INT(errupt) pin and the rudder to CIN-, the comparator input. I assume a difference between CIN+ and CIN- generates an interrupt?
Not likely. how you gather Rudder goes in Cin-? Perhaps, the PPM signal goes INTO INT and you get some signal OUT P55(Cin-). Usually, the comparator is not used (at least, thats what ive saw). If you see a lot of noise when you turn the Remote Controller OFF, then thats the input for the composite PPM signal that comes out the FM demodulator/RF stage. And, some signal for a servo goes out the port pin... at least, thats how it happens in the 4-in-1 DF#4 has. Have you seen a composite PPM? (NOT the servo signal)...
Here youll find some information, some numbers may vary, dont take everything in literally, but on the whole the concept is OK.
For example, walkera uses around 1200uS for the middle position, instead of the "standard" 1500uS.
Here is another real good page, very clear and didactic.

quote:

ps. looks like we are the only crazy ones in this forum! Maybe we should start a thread just for this subject.. what do u think?
Maybe its a good idea! Weve gone a bit astray here. Something with a more descriptive title? I dont know, but the idea sounds great.

OK, you seem to be going really good with the osciloscope! Its pretty good, ive compared with my "real" one and its pretty accurate... at least, in my pc.

Ill try to do some photos of the receiver i have, it uses the same uP, and post where i get wich signal. Thatll help us.

(in reply to huguesdb)
       Post #: 60

RE: Walkera/Dragonfly/Esky 4ch transmitter/receiver t... - 11/27/2007 10:32:21 PM   
huguesdb


 

Posts: 41
Joined: 4/29/2007
From: , SWITZERLAND
Status: offline 		So I checked again all the signals of the uP and there are only 4 signals that look pulse-like : the 2 PWM and the 2 PPM. There is clearly the rudder and throttle pulses which are present only at 2 pins: the INTerrupt and the CIN-. Given there are no other pins with this signal, it has to be an input.

The 2 PWM have a edge to edge of 200 uS, which looks like 5kHz to me! Below roughly 1/3 throttle there seems to be nothing output on the PWM or some funny signals that I cannot interpret very well. Then at roughly 30% throttle, the 5kHz PWM signal starts with a roughly 30% duty cycle. You can hear a strident tone if the motor does not start to rotate.

I show the trace of the PPM signal on the bottom and the PWM on the top for about 20% throttle. It looks as though for low throttle settings, there are bursts of roughly 10% duty cycle pulses at roughly 5kHz. As the throttle increases, these bursts are closer together till at 30%, it becomes continuous.

OK, thats it for now.

Hugues

Attachments
Click to see fullsize image.
Click for fullsize

(in reply to e1000)
       Post #: 61

RE: Walkera/Dragonfly/Esky 4ch transmitter/receiver ... - 11/27/2007 10:42:20 PM   
jtspin


 

Posts: 614
Joined: 5/3/2006
From: Reno, NV, USA
Status: offline 		I've been reading this thread, but I don't understand what you guys are trying to do. I've been playing with my 22E Rx/Tx. I'm thinking about trying to reprogram the microcontroller in the 22E 4-in-1 so I can use a different Tx.

So in a nutshell, what are you guys trying to do?

-JT

(in reply to huguesdb)
       Post #: 62