LCD garbled on new uConsole with A-06

Ordered my uConsole in June 2025, and it was just delivered yesterday (26 March 2026). I ordered one without a core since that’s all that was in stock, but I also ordered the A-06 core separately.

Assembly went very well – no troubles at all (I’ve been a tinkerer and an electronics guy since I was a kid, and I’ve spent the last 10 years building all sorts of things with microcontrollers and SBCs, so I know my way around this kind of stuff). I was very slow and meticulous, mainly because I wanted to enjoy the assembly experience, but also to be careful about what I was doing.

Although, I have to admit I was a bit concerned about connecting the LCD ribbon cable – the flap on the slot was difficult to close and would try to pop open on its own. I know those connectors are very, very easy to break if you’re not careful, so I was being extra-cautious with how I handled it. When it felt like I was using too much force, I’d back off, pull the ribbon out, re-seat, and re-try. After a few attempts, I got the ribbon cable seated correctly and the connector to close firmly without too much force, and without it trying to pop itself open.

I flashed the latest OS image for the A-06 (which is three years old) onto a fresh, new high-quality PNY SD card. I popped it in, hit the power button, and… nothing (except the power button glowing green).

Eventually, it did finally come to life – I’m guessing it was busy with its first-boot filesystem resize. But… the LCD isn’t behaving right at all. It’s difficult to describe, so I’ve attached a photo:

If I work the trackball, this odd pattern of snow does change ever-so-slightly, so that tells me the OS was running.

Of course, my first thought was that I’d somehow fudged the LCD ribbon cable. So, I disassembled it and checked – it looked perfect to me. It was straight, inserted all the way in, and the flap on the connector was closed tight. And, of course, no damage.

I decided to try an external display just to make sure. And yes, an external display confirmed that the OS was running.

As I type this, I’m waiting for apt upgrade to finish. I’m hoping this problem is due to a bad display driver, and not faulty hardware (I saw the Updated Images for new uConsole screens topic, where it mentioned “new screen,” but I haven’t found any further information on that).

Evidently, choosing the A-06 might have been a bad idea, since every OS image and every forum topic about the uConsole is within the CM4/CM5 context, so I’m kinda out in the ocean on a rowboat, heh). But I haven’t found any discussion anywhere about a uConsole having this behavior with its LCD, so it makes me worry that I do, in fact, have a hardware fault.

Before I send an email to the Clockwork folks, I thought I’d post my issue here in hopes anyone might have any insight.

because you are the first one who bought A-06 with this screen :slight_smile:

2 Likes

you’re going to have to create your own disk image for the A-06, none of the existing images support the new screens
Or at least provide your own kernel. But a new image is better since the 3 year old OS is likely also out of date for many other things

1 Like

it definitely looks like the new panel issue

2 Likes

That old image doesn’t have the necessary drivers to run the new panel correctly. I’m not sure if you’d be able to port the current panel driver in my kernel to the old kernel for the a06 and compile it.

1 Like

it definitely looks like the new panel issue

Aha! I did my my due diligence in searching the forums (and I read many, many discussions) before posting, but somehow I missed that one. Thanks for pointing it out.

Yes, that looks exactly like the problem I have. So, that’s a bit of a relief that it’s not faulty hardware. Whew! :sweat_smile:

Okay, then I can stop troubleshooting hardware and get on to figuring out how to build an up-to-date image for the A-06…

Wait, so clockwork didn’t update their image to support their current hardware??

1 Like

no only the cm4 images got updated for the new screen drivers, but also only cm4 uconsoles got shipped with the new screen by that point all the other variants were already “out of stock”

I saw that you’ve been maintaining the images (thanks for taking that on!), but I see it’s only for the CM4/CM5 variants.

I poked around your ClockworkPi-linux repo a little, but haven’t taken the time to fully understand how you figured out what driver you needed, and what my next steps are for following in your footsteps with an A-06 kernel.

I guess the easy thing to do would be to swap out the A-06 for a CM5. Heck, I think I have a CM4 sitting somewhere. But as this is uncharted territory for uConsole users, I may as well give it a shot…

I studied the original patch for the display panel, which added the CWD686, then looked at Rex’s changes, which added a CWU50 display panel on top of that. Okay, now I know what to do.

EDIT:

Looks like someone else already did this (two years ago).

1 Like

But not tho only one ^^’

@guu A06 kernel indeed needs an upgrade…

yeah , I agree
However, I haven’t been able to locate the A06 chip in my messy room for the time being.

and A06 is out of stock right now, So I have to say, need a bit longer to wait for me to upgrade A06

When I find my a06 or get a New one , I will upgrade the A06 os image right away

3 Likes

If you provide me some instructions, maybe can I do it :slightly_smiling_face:

Heads up for other **A-06** people — I’ve got an A-06 with the new CWU50 panel (same garbled

snow as the first post) and I finally got it fully working. Clean now, even at full brightness,

banding and all gone.

**Before (stock A-06 image):** **After (my own kernel):**

Rough version of what was going on:

- The old A-06 image’s panel driver just doesn’t match this newer panel, so → garbage.

- I pulled Rex’s new CWU50 init sequence (from his rpi/linux commit) into the ClockworkPi

A-06 5.15 panel driver. That got it displaying, but with faint horizontal banding.

- The banding was a clock/timing thing. The RK3399 VOP can really only do ~61.538 MHz

(800/13), but the mode was asking for 62.5/61.02. Once I matched the mode to the real clock

(`clock 61538`, `htotal 787` ≈ 12.78µs line) the banding just disappeared. Funnily enough the

same banding seems to show up in the CM5 “horizontal lines” thread too — probably the same

root cause.

Built on Armbian (rockchip64, 5.15.94). Honestly took a bunch of fighting with the build to

get there, but it’s rock solid now.

Credit where it’s due: Rex for the new CWU50 init

saeziae for the earlier cwu50 driver patch, and ClockworkPi for the base panel + audio.

@guu — might be handy whenever you get round to the official A-06 image :slightly_smiling_face:

5 Likes

@maffooclock
@capsule

Just made a minimal archlinux for A06 with new screen driver,hopefully can resolve the garbled problem.

welcome to test

Here is the whole process How I made this image,included all patches

kernel version 6.2

root password is root
also has a normal account: alarm,password is alarm ,too

Notes

After flashing,to expand the filesystem and use the entire SD card capacity, run the following commands on host Archlinux:

sudo growpart /dev/sdX 3
sudo resize.f2fs /dev/sdX3 

replace sdX with you real node of your SD card

4 Likes