deviation-devo F12E

The DevoF12E is not supported as yet. I can no longer justify buying expensive transmitters to make them work in Deviation. If someone (in the US) is interested in loaning me one, I would consider trying to develop support, but unless another devloper takes up the task, it is unlikely any progress will be made.

I'd probably need it for about a month. I expect I could get Deviation running as is in a week, but I want to understand the video capabilities, and get the hardware working in Deviation (even if I don't have full GUI support). My guess is that would take about a month to get everything flushed out. That doesn't mean you'll have a video-enabled version of Deviation by then, but I'd hope everything else would be working.

FYI, I've made this offer in the past, and it stands:
If you send me a radio:
I will guarantee it works when I ship it back. If I break it, I'll buy you a new one
I WILL open the radio. That may void your warrantee.
I MAY solder pads to the radio if needed to access the debug port (this will not affect performance). Generally this shouldn't be needed, but I can't promise anything
I am in the US. you will be responsible for getting it to me. I'll get it back to you.

If you are interested PM me.

I assume the Devo F12E will work with a Ladybird V2 FPV? I think the 5.8GHz video stream is standard, and not tied to a specific protocol from what I've read. I have a Ladybird FPV, but if it won't work, I need to find a camera that will work before anyone lets me borrow an F12E

I decided to buy an F12E for myself. So I'll work on it once it comes.

I started to document the Devo F12E hardware based on gektor's images here:
walkera-fans.de/devo-f12e-von-innen/

They's used the same CPU they used on the Devo 6/8/10 (256K Flash, 48k RAM). The SPI flash, like the Devo F7 is only 64kB and still uses 4kB sectors. That will be really challenging to deal with, but at least there is enough RAM to do a read/copy/write, so we should be able to fit the models (probably with a 512k sector size). Multi-language support may not be possible though.
It looks like the Deviation port will be based upon the Devo10 GUI. I'm not sure if the overlay supports color, but with limited Flash, graphics are pretty much out.
I haven't fond any documentation on the screen as yet, but I've seen images of it doing simple graphics, so I assume it is better than the F7. I have not yet found any documentation on the display/overlay chip on the F12E, so if anyone has any info on it, I'd be interested in that.

Documentation project is here:
www.deviationtx.com/articles/17-hardware...hardware-connections

We'll see. I have several of the 32Mbit chips, because I had planned to replace them on by DevoF7. I would much rather find a way to make it work unmodified though. Most people would not be able or willing to replace the flash chip, so it would need to provide a significant gain to support it. I may do it on mine since the extra space would make development easier.

LCD driver
www.deviationtx.com/forum/builds/3334-is...-hack?start=20#26997

Images
www.deviationtx.com/forum/builds/3334-is...k?limitstart=0#25588

Thanks vlad, I can't count how many times over the years you've come through with pointers to stuff like this.

SadSack, as I mentioned, If I require modifying the transmitter to use Deviation, the number of people who are interested will be much lower. If it offers some significant advantage, I'm sure some will do it, but I will try to squeeze everything into what we have. Unlike the F7, at least we have a reasonable amount of on-die Flash and RAM to play with which makes this much more likely.

Looks like the TW8816 uses I2C which I haven't used on the STM32 before. A bit of googling indicates it is pretty finicky. Also, I've not found any useful TW8816 example code for developing a driver. I found this:
www.codeforge.com/article/100785

Which I can't download, but the individual files are accessible. Not really very helpful. I can hookup my Saleae to the Tx once I get it and capture the initialization routines, but actually writing using the screen effectively will likely take some effort.

I found some more tw8816 (or very similar)code:
github.com/Si4463Project/RadioCode/blob/...er/Source/drv_tw88.c

I got my f12e today. screen looks nice and there's lots of capabilities. I think it will be easy to share the code being developed for the f7 with the f12e. I need to snoop the i2c bus, and write a display driver, but getting initial functionality after that shouldn't be too hard I think.

I finally started work on my devo-f12e.
Getting i2c working has been a pain. I tried to follow the ST reference design using DMA, but I just can't get it to work. I gave up and switched to a polling solution which is working fine.
I snooped the init sequence from the i2c bus (The pins are open-drain which is annoying, but it is possible to attach my Saleae and still connect the LCD header at least well enough that it gets a signal. I am now able to power on the backlight and init the screen.
I haven't written any of the screen-writing utilities yet, but those are next on the list.

Screen is working:


It is too bad that there isn't much flash. If they had just spent $0.30 more, this thing could have 4MB of flash, and we could have a full graphics (264x162x7color) screen. As it is, we'll need to use the text-display mode instead I think. I need to work out the buttons, sticks and switches next, then it should be possible to get the work we've been doing on the F7 ported over.

I mapped the analog, switches, and buttons today too, so most the hardware is done. Still need to figure out the 5.8GHz stuff, but I've gotten far enough that the GUI work can start. I probably need to setup a emulator to speed up porting the interface.