« PreviousPages: 1 [2]
| Loading_M_ | Posted: 12 May 2026, 10:47 PM |
|---|---|
 Member Posts: 12 Joined: 7-May 26 |
From what I can tell, it's looking like there are some bad modules, but it may not be the only issue. What's the expected variance between modules? Some of the problematic modules are showing voltages around 2.1V on the diode tester, there is at least one that's failed open, and at least one that's showing ~0.01V. Specifically, the 3 modules furthest from the connection point on the date side are showing problems. All the other lights are showing 1.78-1.79V and appear to light up as expected. (Curiously, the colon LEDs measure ~1.75V, but I assume that's normal / within spec - and they seem to work totally fine). The ones causing issues on the time side might be something else - they measure and light up normally, but after a little while start to dim / flicker. They appear to act as a unit - which would make sense if they are a matrix group. ------------- |
| [top] | |
| Loading_M_ | Posted: 13 May 2026, 04:04 AM |
|
Member Posts: 12 Joined: 7-May 26 |
The issues on the time side look like it might be something to do with the led driver - I'm also getting inconsistent brightness levels. When covering the light sensor, most of the digits adjust as normal, but a couple digits don't dim nearly as much as the rest. ------------- |
| [top] | |
| mit | Posted: 13 May 2026, 02:29 PM |
 yeah whatever Admin Posts: 684 Joined: 4-May 16 |
The expected results with the diode tester is that all of the segments should read about 1.8V, but anything 1.7 to 1.9 is fine. The colons will read slightly lower because it's got two diodes in parallel. It does sound like you'll need to replace some of the digits. I am just a bit cautious because we still don't know what the original problem was. The display has four separate matrices of digits (5 and 5 on the date side, 5 and 4 on the time side). From the picture, the date side seems to have random segments failing, so that could be that they were overheated and damaged during soldering. On the time side, it looks like all of the same segment in one matrix has failed, which is a bit more worrying. There is a display test mode on the clock ( MODE_DISPLAYTEST=1 in the config file) which might help you spot which segments have failed. ------------- |
| [top] | |
| Loading_M_ | Posted: 13 May 2026, 10:38 PM |
|
Member Posts: 12 Joined: 7-May 26 |
Thanks for the tip on DISPLAY_TEST. I took a video of a full cycle: https://immich.loadingm.xyz/share/P9xfm5QIFpL1594com4ObcTr7wXYRqLwFKKj5K61bwU7KvhPbRk0IlN3_ajwr53vuE8 At 6 seconds in the video, you can see the flickering of the top segments on the first 5 digits on the time side. It also appears (from the 13th second) that the decimal point on the first 5 digits is always off - is that expected? It looks like it might just not be wired up. I've also seen several cases where, after a minute or so of leaving it running in the long configuration, it just starts rapidly cycling the modes, even though the buttons aren't being pressed. It also seems like the date side is lagging a bit behind, or just not displaying any digits at all. Sometimes it picks back up almost as normal. https://immich.loadingm.xyz/share/wKfB8tW71aev1qW4rpFLabbXL4l4kdfVvPtulojIF1gMC5UgZDSFUI2PIOyCtnsAAw8 A quick still of the display test with all 8s showing.  ------------- |
| [top] | |
| Loading_M_ | Posted: 13 May 2026, 10:52 PM |
|
Member Posts: 12 Joined: 7-May 26 |
As a side note - I haven't had a chance to measure the voltage on the date side while it's bugging out, but I'm guessing it's sagging - causing the processor to brown out. ------------- |
| [top] | |
| mit | Posted: 14 May 2026, 12:27 PM |
|
yeah whatever Admin Posts: 684 Joined: 4-May 16 |
Thanks for the videos. Yes, the decimal points on the hours/minutes aren't expected to light up in the display test. There's definitely more than one problem going on. It's very strange (and unlucky) that it affected both sides of the clock. How comfortable would you be replacing the surface mount display drivers? At this point I think the options are: - you could send it back to me, and I will repair it, replacing whatever is needed - we can keep trying to figure things out, and I can send you some replacement digits and other parts as needed ------------- |
| [top] | |
« PreviousPages: 1 [2]
Sign in to post a reply.