Radiolink AT9 as a platform for Deviation?

Here we go. Deviation running on an AT9:


I used PBM image format (same as for the F12E) so each image is only 2 colors, but I added a small extension that allows setting the foreground and background colors for each image. I then converted the Devo8 bmp files into pbm with imagmagick.

The filesystem is devofs running on the upper 256kB of the MCU. It currently takes up 61kB.
I have the rotary encoder working properly and mapped to left/right. there is no way to do a long-press on left/right, and there is no up/down, so I've temporarily mapped the trim up/down to the Up/Down buttons

The filesystem is currently read-only because If a write goes bad it could overwrite the bootloader. Once I back that up, I'll work on enabling write.

So basic functionality is now there. No protocol support yet. Figuring out how to control the interface with only 5 buttons will be the big deal. One trick would be to convert the rotary encoder from left/right to up/down when Mode is also down. IT would work, but I don't really like it.

I think you will need to use the rotary for both left/right and up/down, and the function will depend on weather you are editing a field or not. That you need to switch with a button, like ENT.
So you scroll by the rotary switch up and down between the fields, you press ENT to edit that particular field, then use the rotary switch again to change the value. After that you have to use the ENT (or EXT) switch to finish editing the field, so you can scroll between the fields with the rotary again.
Quite different working...

FDR, what you describe is likely the only reasonable solution, but changing the button controls is likely to be a huge task. I need to look at it and see if I can come up with something reasonable. As Victzh said, the rotary switch has a push button embedded which will be 'ENT' the buttons on the left are 'EXT' and 'Mode' (which I don't know what to do with).

Anyhow, I need a break before I tackle it. I'll get the AT9 checked in tomorrow.

Also, I have no idea what to do about all the icons and backgrounds and whatnot. They all need to be redrawn to look ok in the 2-color mode.

Probably will be better upgrade memory chip first (hardware upgrade) and develop Deviation for Radiolink AT9 with upgraded memory chip only. I think it will be not too hard to implement memory chip upgrade.

also, while we do have a plan for adding more memory to the AT9 though an easyish mod, we don't actually know it will work yet. I'm not sure there is a real need for the extra space. We should have plenty of space for a nice clean interface using what I have now. coming up with a nice set of bi-level icons shouldn't be that hard.

If anyone wants to play with it, here is an emu build for the AT9:
www.deviationtx.com/downloads-new/catego...s-at9-emulator-test1

It has roughly the same functionality as the button based build I posted in the 'Post 5.0' thread (i.e. there are lots of bugs, and it is quite easy to get stuck in a menu that you can't get out of)

But except for the fact that we have no protocol support, this is nearly ready to roll

I'll continue working on fixing the bugs in the new button code. It will take some time to work out all the issues, after which it'll be time to tackle an SPI mod so we can actually use this Tx with Deviation.

I do not plan to use PPM-out to control the module. We will have our own board that works with SPI to drive the 4:1.
Of course, Deviation supports PPM, so of course it could be used, I just don't plan to do it that way myself.

victzh wrote: So I can use the branch AT9 to build a version for real one? I ordered discrete logic version of add-on board, it should come around 22nd.

You can, but I wouldn't quite yet. I am still working on fixing up the button handling further. It is mostly functional, but it is still possible to get stuck on a few pages.

Also, I haven't setup a way to install the filesystem yet in the builds. it needs to be compiled into a dfu and then installed, so it REALLY won't help you as it is today until I add that part.

I'll try to get things mostly tidied up over the weekend.

I have checked in working (mostly) code now.
The devofs module is currently hardcoded to only support 64k filesystems. I will need to fix it, but for now, I just reduced the size of the fs (do a make distclean 1st)
I also had the dial set for left/right instead of up/down. It will work properly now (more or less, there is more work needed on the button handler, it doesn't detect scrolling very reliably)
I create 2 dfu file at9.dfu and at9-fs.dfu that way you can replace the firmware without rewriting the FS.

And writing to the FS is still disabled.

the emu_at9 does work, and is a good way to figure out what works and what doesn't (for instance, there is no way to access the list-boxes yet or the main-page config).

To use a text-select box, you must scroll to it, then press ENT, then you can scroll within it. It works pretty well.

But it is getting close to functional now.

I found a rather large issue that will take some time to resolve. The devofs filesystem I wrote only supports having a single file open at a time, but the font code requires that we have at least 2 (one is dedicated to the font) one for general-purpose I/O. If we enabled datalog, that would require 3 (with 2 of them being in write mode).

I have 2 options for the fonts: either enhance the devofs code to keep track of multiple pointers, or separate the fs like on the Devo10 such that the media dir is kept in a separate address space. I need to give it more thought. On the positive side, I found several bugs in the devofs code today, and having fixed those, it should help the f12e/f7/f4 stability.

I'm surprised the usb drive works. it really shouldn't yet. I don't think it can break anything, but I don't recommend using it until write support has been added.