openrcforums

0ff wrote:Thanks Mike for this awesome port!

Haven't tested it just yet, but after programming the module flashes the LED as expected, so I guess all went well.
While I was at it, I've dumped the stock firmware and eeprom - if you do have a PDI-Prog and you maybe want to go back to stock someday, this is how you can do it with avrdude. Note: It'll dump both intelhex and raw binaries.

What version is this dump from? I have a blue PCB - Label on module only shows "DSM2". I tested recently with my Inductrix and the quad flips instantly to the left. All channels are centered except throttle. I'm thinking this could be the bad firmware. I wouldn't mind going to the stock 'fixed' firmware assuming that's the only thing they fixed.

@Mike...

How frequent will you most likely be updating the firmware of this module?

I'm asking because I have already soldered a micro-JST connector in my Taranis and I'm considering to expose it in the battery bay.
But if it's unlikely there will be frequent updates I can open up my radio, upgrade the Orange-module and close it afterwards.
On a 2nd occasion I can do that extra mod anyway.

But if you already know there will be 2 or 3 updates in this year I will immediately do this....

At present I'm not planning on a need to update the module. As far as I can tell it is working OK. An update will only be necessary if a problem is found.

Mike.

Thanks Mike...
All I need to know...
Again, thanks for all your work you did and are still doing

I chose this morning to flash my recently acquired Orange module (green board).
I (think) I followed the complete procedure, but in the end it doesn't work as I inspect it to.

I connected the Orange module to my Taranis using a 4-pin micro-JST on the Taranis motherboard. I also used the 3.3V pin.
I flashed "x9dp_rom.bin" of the file "ersky9xProvR219q_music.zip" ( download/file.php?id=14053 )

I started the Taranis normal and it erased the eeprom (I saved it before)
Then I started the Taranis with the trims outward.
I came into the maintenance interface where I chose Mega, I chose Orange module and it gave me feedback that it was flashing the Orange module.

Then I turned everything off, re-assembled the Taranis and flashed Plaisthos OpenTX firmware with support for the multiprotocol module.
I used Companion to get my settings back.

Then I copied an existing model, selected Multi-module DSM2, DSM2, 7ch 22ms and tried the bind button.
I didn't notice anything happening (I don't know what to expect), so I used the manual bind button.

I was able to bind it to my 7-channel Orange receiver, but no channel was working.
Just now I selected PPM and that works...

It makes me think it still works as an original Orange module.
I have no experience with this module (or any module) and I'm probably missing something.....

What I would also like to know is if I can check the flashing was successful. No errors were shown during the flashing process, but I don't know if it does if the module is connected wrong.

The multimodule code supports both PPM and serial in, but the module may need a hardware mod (just adding a resistor) to accept serial. See details at the bottom of the first post.

Kilrah wrote:The multimodule code supports both PPM and serial in, but the module may need a hardware mod (just adding a resistor) to accept serial. See details at the bottom of the first post.

It probably does...
But as I understood it this only applied to the blue PCB.
I'm now off on my only day of flying of the week. Will investigate more.

Thanks Kilrah... I didn't do my homework properly as it seems. I will do this tomorrow morning and report.

Heated up the soldering iron, opened up the Orange module......
Turns out my Orange module already has these 2x 10 ohm resistors already in place....

I now again think it might not be properly flashed.
Is there a way to determine this?

As I wrote before, it seemed to flash fine. I just don't know if there's some check on it whilst it's flashing.

When you start the module in serial mode, it flashes the LED around once per second until it sees valid serial data, if you have the radio paused at, say, the throttle check screen, you should see this. If the LED isn't flashing, then the module probably didn't get flashed with the firmware.

When flashing the should be 5 (hex) numbers across the bottom of the display. The first 2 are the XMega ID values, these should be something other than 0000. If they are 0000, then communication to the XMega isn't working. The other 3 values are error counts, and these should stay at 0000.

Mike.

It seems communication isn't working.
All 0000's stay 0000's

I have just checked with a logic analyzer and the signals are delivered to the pins.
There's also a 3.3V delivered on the pin.

There is no reaction of the Orange Module during flashing (no LEDs).

Contrary to your advice I supplied the 3.3V on the 4-pins.
I do think you only gave this advice because it would be quite possible to shortcut the supply line.

Could something be going wrong there?

It seems all the proper signals are give to that 4-pin connector: GND, Vcc, CLK and Data

I think you may have the CLK and DATA signals swapped. Please double check. The JTMS signal is used as the CLOCK, the JTCK signal as the DATA.

Mike.

Ah...
That explains.... I saw the "CK" in JTCK and wrongfully assumed it was the clock-signal.


Edit:

I just swapped the cables on the micro-JST cable and flashed it again. Now serial communication is working.

Awesome work Mike, as usual

Since we're already discussing programming methods, I believe you were using an Atmel PDI Programmer to flash these modules in the past, would you mind sharing your experience with it?
I'm considering to buy an Atmel ICE Basic at a steep student discount in the near future, therefore my interest in your opinion.

Before i shell out the 55€ for it I'll try the method CrazyCoder suggested, tho.

CrazyCoder wrote:Great job! I found a way to make a PDI programmer from ISP programmer and want to give it a try with Taranis: http://szulat.blogspot.ru/2012/08/atxme ... r-050.html [...]?

Even if I get it to work I might want to invest in a proper Atmel programmer for once, and the ICE seems to be the Swiss Army Knife of them.

I tried that CrazyCoder linked method but I didn't get it to work properly.
I have an Atmel AVRISP-II programmer that works fine with the PDI of XMEGA devices, I've never seen the Atmel ICE, but would expect it to work OK.

Mike.

I got an AVR Dragon a (long) while ago in previson but never actually used it...

I was under the impression that the AVRISP-II is not capable of PDI, never bothered checking tho.
I've access to one at university, I'll try to flash my Orange module in some spare time.

I can get a Dragon for the same price as an ATMEL ICE, however I prefer the protective case and the form of the ICE.

Thanks alot!

MikeB wrote:It looks to me as though these are the required mods and connections for that board. The blue line is a wire that is needed if you want to receive telemetry. You need to cut the track where indicated anyway.
The 4 pads are for the PDI programming connections. Remember the Vcc must not exceed 3.3V. For programming, follow the instructions in the pdf file. Take care connecting to the PDI pads, too much heat will lift them off the board.

Omodulex.jpg

Mike.

I think you'll need to cut around that bottom pin on the JR connector as it's bridged to ground on mine.
Grant.

Has anybody had success programming Orange module on the original Taranis.
If lock bits are set then will it erase device?

Have spent a few days trying but always get code 0000 0000 so issue with comms. Can see clock and PDI signals on the correct pins - check multiple times and data signal is clearly different from the clock when viewed on oscilloscope. Tried swapping lines around anyway just to make sure. Tried adding a transistor buffer. May try shortening leads next but wires are currently only about 15 cm long. Tried powering from 3.3v on Taranis and using modules internal 3v3 regulator. The one positive is that module and Taranis still work. Also note that version 219q_music did not have the mega programming menu in diagnostics on my Taranis. Did have on later releases.

So am stuck - signals look good on scope but always get 0000 0000 and modules is not programming. My module is the green circuit with dip switches.

Any help and confirmation that others have succeeded one of the earlier Taranis variants would be great.

Still no luck - tried adding a 10k pull-up on PDI data line. Thought this workes as I got FE00 0000 during programming but module code was not flashed. Tried shortening wires - still no luck Run out of ideas so may try modifying a usbasp and try this way.

I'll admit to only having modified a Taranis Plus when I implemented and tested this. I'll see if I can do the mods to a Taranis, but for this operation I believe they are the same hardware.

Mike.

Thanks Mike - suspect it is not an issue with older tartan is code as signals look good on scope. If I reverse calk and data then flashing progress is slower. If lock bits were set in module then would this be an issue? Suspect module not responding and pull-up on data line just floated line high resulting in fe00 0000 code. Clock looks periodic but does have periods of inactivity. What frequency do we expect the clock to be?

Thanks
Bruce

The clock high and low times should be about 3.5 to 4 uS.
The signature should read regardless of whether the lock bits are set or not.
The lock bits are read, and if any are set, then a chip erase is carried out. But if you are not reading the signature, then you are unlikely to be able to read the lock bits!

The data line of the PDI is unusual. It is bi-directional, but the XMEGA only drives it high or low for a very short time, then relies on a "bus keeper" circuit in the processor to hold the level. This circuit has a high resistance, so there must be NO other circuitry on the signal, other than the connection to the Taranis signal. This is why I had to find the signals inside the Taranis, anything easily accessible has extra circuitry on it that stops it from working.

Mike.

I've modified a standard Taranis and it fails to flash the Orange module. I then checked the two signals carefully, and I found the standard Taranis has 10K resistors (in RP1) on the two signals, one is pulled up, the other pulled down. These will stop the data signal from working correctly. The Plus does not have these resistors.
I can see no reason to keep these on the standard Taranis as the Plus works fine without them.
I'll look to remove them on the standard Taranis I've modified, then test again.

Mike.

Edit: Removed RP1 and the Orange module flashed fine.
It is a very small component, 4 'pins' on each side. I found that if I add a small amount of solder so all the pins are soldered together on each side, then heat both sides and keep the heat on, after a few seconds all 8 pins become unsoldered and the component itself just comes free.

Many many thanks Mike. Removed resistor and module programmed perfectly

Really appreciate your help - fantastic

My only mods to a tx so far have been adding modules to my Devo 10 so I'm no expert at this stuff. I have a question relating to the physical connecting of the Orange module to tx for flashing.

Tx is a 9xr Pro
Orange module arrived a couple of weeks ago.

My understanding is that I connect 3 wires (CLK, GND, DATA) from the tx module port to the Orange circuit board according to the Pdflashing photo. (Upper right points in photo).
Step 5 talks about powering the Orange module which apparently can be done from the tx module port again to ensure correct voltage.

Do I only need to run VCC from the tx to the Orange module tx connection (lower left in the photo) or do I need to run VCC and GND to the module tx connection?
Sent while reciting Vogon poetry.

Don't Panic! There is an infinite probability the two pads labelled Gnd are connected together, so only one ground connection is needed

If powering from the radio, I would recommend connecting Power and Gnd to the module connections (bottom left), then you only need to connect the Clk and Data signals to the pads in the upper area of the module.

Mike.

Thanks Mike! Much appreciated

Will do this tomorrow after lunch and before a Pan Galactic Gargleblaster or 2

Sent while reciting Vogon poetry.

It worked! I can fly....err ummmm bind to my Nano QX and everything works now. Used to bind but no throttle. Now it works!

No idea how you do it Mike but you're worth my weight in gold! Thanks very much.

Sent while reciting Vogon poetry.