Walkera MTC-01 'Magic Cube'

[j0bro]

Did anybody spot this Walkera MTC-01 'magic cube' already?

Walkera product website: http://www.walkera.com/en1/particular.jsp?pn=MTC-01
RCGroups forum thread: http://www.rcgroups.com/forums/showthread.php?t=1716207

It is supposed to be an adapter for the older WK-series TX-es and Android phones to be able to control Devention based models. It would be great to hack this thingy onto a 9X and be able to fly Devention helis, such as the wonderful New V120D02S. Of course, it will not support telemetry, but still a nice option. Will it be possible?

[Kilrah]

Should work, simply by plugging it into the trainer port, maybe with a little cable hack.

[cre8tiveleo]

Looks like 'Anylink' , should work without issue.




---
I am here: http://tapatalk.com/map.php?grlmg2

[kaos]

Looks like my wK-2603 just been revived. it even gives more programability than original Tx (if I read it correctly).
does that thing actually pugs into the earphone jack?

[j0bro]

Should work, simply by plugging it into the trainer port, maybe with a little cable hack.

Well, they claim that it only works on WK-2x0x models (correct me if i am wrong) and Android phones. I would like to know whether plugging it into a Turnigy 9x would work That would mean that the Devo 10 I just ordered along with a New V120D02S was not necessary.

[Kilrah]

It depends if those radios (and the android software) output standard PPM on the trainer jack or the proprietary Walkera PCM. If it's PPM it should work with no problem.

I guess the best way to know is to scope the output from the Android software, but AFAIK it hasn't been released yet. The adapter isn't shipping yet either anyway.

The thing is it's low power, so kind of parkflyer only.

[algo]

The thing is it's low power, so kind of parkflyer only.

Yeah but radios like the Devo 10 only output 10 mW anyway so in theory it should be at least as good as that if you put a normal antenna on it. Plus it's possible that it is using the same TX chipset as the other Devo's and those can be hacked to run at 100mW with firmware changes (that's a huge unknown though; grain of salt etc).

It may very well be PPM based because apparently it's designed to convert non-DEVO WK PPM radios as well as the Android app.
http://www.oomodel.com/walkera-mtc01-mu ... 63879.html

But who knows for sure, have to get hands on one. Would be great to get the 9x working with a Mini CP or similar.

[j0bro]

Since 2 days I am the lucky owner of a very nice Walkera New V120D02S with a Devo 10 TX. In the setup menu there is an option to change the output power, it goes up to 20 dBm which is equal to 100 mW. The default is 10 dBm (10 mW). See here: http://en.wikipedia.org/wiki/DBm

Too bad the TX's menu is not as feature full as the custom 9x firwmares, let's see what the DeviationTX firmware (http://deviationtx.com) will bring as there is currently a Devo 10 version in the works

[Rob Thomson]

Interesting... Makes me think a 9x port could be possible to this tx.


Sent from my iPhone using Tapatalk

[etien96ne]

Hello,
I received my magic cube today with a mini cp ! I tried to use it with my Turnigy 9X in open9x and....

It doesn't work !

I think this is normal because walkera don't use a standard ppm signal...
I want to know what is the specification of the ppm of walkera (positive ? 22,5ms ? 300u ?...).

Please help me

[algo]

Finally got mine today. I'm really interested in getting this to work with the 9x. Taking it apart I see two ATMEGA168PA processors. It is using the CYRF6936 radio which is the same as the big Devo radios which hopefully means we can pump up the power output (with a real antenna) but that depends on if the big radios use an amp or not. I guess I need to ask in on one of the Deviation TX threads. It has some sort of chip connected to the wire antenna, not sure what that is (balun or something?).

The first thing I need to figure out is how to upgrade the firmware without Walkera's $pecial cable. I believe the headers on the side are just using a standard serial connection to the CPU's. Somewhere I have the Walkera document that says which pins are which. I do wonder what bootloader they're using. In any case I think I can build a cable that will work with their firmware upgrade software. Even better would be if it's running a STK500 type bootloader then I can upgrade the firmware from avrdude/Linux.

Anyone know what the 4 pins on the "audio" connector do? Obviously if it works with a phone it must have a mode that only uses 2 or 3 of the pins. Does the WK-2801 have a 4-pin connector on its trainer port?

I have an Android phone to test with to help figure out the protocol in that mode.

My main problem is going to be that I don't have a Devo receiver or model to bind with so that's going to slow me down.

[algo]

Before I start tearing into this thing I need to know what steps people have taken to try to get it work and why you think it doesn't work out of the box.

First, did you use the supplied cable that converts from the 4-pin connector to the mono plug?

Second, did you do the binding procedure? If it won't bind the normal way did you try binding with the Android app then put the module on the 9x?

I'm going to try to get hold of a Mini CP for a while to do some testing. I need to upgrade my firmware first though because there was a critical flaw that was fixed.

[Kilrah]

I don't have the adapter, however I now have an Android phone, so I've downoaded the app and hooked a scope to the audio output and it's definitely not PPM that comes out. So yeah, I'd guess it's the proprietary Walkera PCM that's needed.

So the options are simple... add that protocol to the 9x, make a PPM-> Walkera PCM converter for example with an arduino, or find a way to write an alternate firmware for the magic cube.

[algo]

I know the audio output from the Android app is proprietary. It's sort of "encrypted". It might require a lot of code added to er9x or an adapter to sit between the MTC and the 9x.

However, there may be no point in doing that. The MTC-01 can be hooked up to the WK-2801's trainer port which is just the standard mono plug and I don't think it uses any special protocols, it may be standard PPM. If that's the case then the MTC-01 should work in the 9x trainer port if you're using the 4-pin-to-mono cable that comes with it. But that's the part I'm unsure of and is what I was asking about in my last post.

[Kilrah]

Hmm, possibly:
http://www.helifreak.com/showthread.php?t=183689

[MikeB]

What does the protocol look like? Maybe it could be added to er9x without too much flash being used up.

Mike.

[Kilrah]

I found this:
http://www.smartpropoplus.com/Docs/Walk ... 01_PCM.pdf

Not sure it's the same though.

[algo]

What does the protocol look like? Maybe it could be added to er9x without too much flash being used up.

Unfortunately this is kind of a gray area. I said "encrypted" but it actually is using encryption which I assume they did for some ridiculous reason. I think I'm going to leave this alone for someone braver than I or in a country without restrictions. I'm paranoid about that stuff and don't want trouble.

Plus there is a reason they used two processors in the MTC, a single atmega probably doesn't have enough horsepower to do the protocol decoding and run the radio (although I can't say for certain).

But lets not get ahead of ourselves. If the WK-2801 uses PPM or even if it's that custom PCM format Kilrah posted above then I don't think we will have any problem using that interface.

[algo]

Alright, I finally acquired one of the WK radios and a Devo RX for testing. It looks like the radio does in fact output just plain PPM on its DSC port. It has a 22ms frame time and seems to use 0 to 1 volt signaling level.

The MTC-01 will become "active" when it detects movement on channel 2 (aileron on my WK) and will start a binding procedure when channel 2 is moved again.

However, even after adjusting the frame time to 22ms in er9x and playing around with different resistors inline on the signal wire trying to get it down to 1 volt I still can't get it to work. I move channel 2 on the 9x and the MTC does not respond. I can't figure out what is going on. I may need to come up with a proper circuit to make sure it's using 1 volt signaling but I don't know.

Unfortunately all I have is a soundcard for an oscilloscope so I can't get 100% accurate readouts due to the capacitors. See the following links for traces of each. These large images may need to be zoomed to 100% to see the trace correctly.

[Trace from Walkera WK]

[Trace from 9x]

I don't get it, they look similar but I must be missing something. Maybe the WK is not using 300us for the pulse time. That does look a little different, hard to tell.

[algo]

It works!! After acquiring an old WK radio I was able to analyze its output on the DSC port.

Here's the scope, complete with fully working er9x models for DEVO Genius CP:
http://www.rcgroups.com/forums/showthread.php?t=1781132

[algo]

It seems that a 2.2k resistor on the PPM signal to the MTC-01 works. 4.7k was too big. I think the actual required value to get it around 1v would be 1.3k to 1.5k. 1.2k might be too small for 1v but would probably be safe for 3.3v if that's what the MTC-01 can handle. I wasn't able to trace the signal on the MTC-01 to see where it goes so I'm not sure what it's connected to.

A simple resistor like this is not really proper level conversion so maybe someone has a better idea that would be more reliable.

[algo]

It looks like the MTC-01 will accept a frame rate of no more than 24ms and a pulse separator of no less than 350µs and no more than 400µs.

Apparently there is a "channel mixup" issue on these MTC-01's where the channels get mixed up and things go bad. There is a firmware patch and possibly a new hardware revision that may address the issue. I wonder exactly what the issue is. My MTC is not patched with the new firmware because I don't have one of those DP02 upgrade cables that cost more than the MTC did (!). Could the issue be related to the low signaling voltage from the WK radio? Maybe the WK's short 22ms frame time ("normal" is 22.5ms)? Or maybe when enough channels are maxed out the frame sync pulse is not long enough. Or something else completely? What I'm wondering is if the issue can be fixed on the 9x side without having to upgrade the MTC's firmware. The reports of problems seem to come from WK2801 owners that are probably running 8 or more channels and if it's a frame sync pulse length issue that could be fixed by running less channels or a larger frame size. I'm currently running only 6 channels with a 22.5ms frame size 350µs separator and I will see how that goes. I have a kill switch configured that puts all channels to -100% just in case it goes haywire and I'm unable to shut off the throttle.

Still looking for a good solution to the signaling voltage problem. I think a ~3V zener would probably be the simplest solution but unfortunately I don't have any zeners below 5.1V. I tried diodes series with a pulldown and a voltage divider but both pull the line down too much or add too much impedance and the MTC stops working. Another option would be to wire a CMOS logic device into the MTC and use the MTC's power supply to do level conversion or I suppose the same could be done inside the 9x with an added 3.3V regulator, converting the whole trainer PPM output to 3.3V or less.

[algo]

I did a more detailed analysis of the PPM signal from my WK2403 today. The pulse separator is actually around 422µs to 423µs so I recommend setting it to 400µs on the 9x (I edited my previous post to reflect this). I was having glitch issues where heli randomly freaks out and goes into failsafe at 350µs and it seems to be better (so far) at 400µs. Is there any way in er9x to get more fine grained control of the timing here? Just in case.

I got lucky and found a 3.3V zener in an old mouse. I'm currently using that with a 1K resistor to the MTC for signal conversion. 470 ohm might be better to give a stronger signal but time will tell.

Some information on the Genius CP channels:
Channel 1 = Elevator
Channel 2 = Aileron (needs to be reversed)
Channel 3 = Throttle
Channel 4 = Rudder (needs to be reversed)
Channel 5 = Unknown, probably gyro rate/sensitivity or something but it doesn't seem to change anything no matter what I set it to
Channel 6 = Collective

Do not use "heli" CPPM swash mixing on the 9x, the Genius does that stuff itself. If I ever get this thing working reliably I'll post a er9x model because there is a lot of stuff that needs to be set up.

The binding procedure that has to be done every time is a little involved. First you need to set all the trims on the radio before binding. For me this usually means full left rudder trim, some good amount of left aileron, and a bit of back pitch but I haven't set up my heli yet; the rudder/yaw has to be done no matter what though. Then after binding is 100% complete move the trims back to center. If you don't do this the heli will drift and yaw all over the place. This is a known issue with the GCP and other Walkera and is made worse because it seems you have to re-bind every single time to the MTC-01. To me it seems the heli RX does some sort of initialization to set the center of the PPM signal (or something) at bind time and it just doesn't work right.

To figure out how much trim you need, you first have to set the trims to center, bind, then try to fly (watch out! it can get crazy). From then on you will need to dial in that trim before binding, bind, then move the trim back. My plan that I think will work (thank you er9x!) is to set up a switch on the radio for "bind mode" that sets all the controls to the various trimmed values they need to be at, then once the bind is done I can flip the switch back to fly normal. That way I don't need to constantly set and re-set the trim every single time.

[algo]

I went ahead and linked this back to the rcg topic I spent way too long writing (see post above)

It seems I have it mostly sorted at this point. Good stuff.