I'd have immediately said that this was an LVDS cable or panel setting wrong, but here's a panel which is showing a solarised picture where the transitions aren't particularly smooth, and the gamma IC is the cause. So looks like I was wrong on that one... it is possible.
Interestingly, a certain band of colours does appear OK - it's only the darkest or lightest which go all crazy. For example on Jake Sully's face, the mid-blue details are accurate; but the darker and lighter colours aren't.
Has an AUO 37 inch 720p panel in it.
VGAMA voltages, only some are available for test but we can still see it is bad:
1. 14.91V (should be near AVDD)
2. 14.55V
6. 10.36V
7. 9.56V
8. 9.42V
10. 7.63V
11. 7.42V (midpoint - should be 1/2AVDD so this is OK)
12. 10.77V (notably bad)
13. 10.35V ^^
15. 7.53V ^^
16. 7.07V ^^
17. 6.22V
21. 6.02V ^^ (should be nearer to 0V by now)
22. 0.37V (seems OK, should be about AVDD-VGAMA1)
I compared REF voltages and they match, so it seems like that's not a reliable test: what appears to be happening is the input bias of the device is failing, and this is causing it to malfunction, but it sure is an odd failure. The references are derived with a long resistive divider, with 1% resistors.
A quick check: device was running 55C which is hot, in a 21C room. Calculated rise of temperature at 67C/W T(j-a) as specified in datasheet puts it at 480mW... according to the datasheet the quiescent power should be 130mW so we can see that the IC is dissipating 3x more power than it should, probably due to an internal short.
It's an AS15-U gamma IC, but this is identical to AS15-F, AS15-G, etc.
My first gamma IC failure! woot.
Interestingly, a certain band of colours does appear OK - it's only the darkest or lightest which go all crazy. For example on Jake Sully's face, the mid-blue details are accurate; but the darker and lighter colours aren't.
Has an AUO 37 inch 720p panel in it.
VGAMA voltages, only some are available for test but we can still see it is bad:
1. 14.91V (should be near AVDD)
2. 14.55V
6. 10.36V
7. 9.56V
8. 9.42V
10. 7.63V
11. 7.42V (midpoint - should be 1/2AVDD so this is OK)
12. 10.77V (notably bad)
13. 10.35V ^^
15. 7.53V ^^
16. 7.07V ^^
17. 6.22V
21. 6.02V ^^ (should be nearer to 0V by now)
22. 0.37V (seems OK, should be about AVDD-VGAMA1)
I compared REF voltages and they match, so it seems like that's not a reliable test: what appears to be happening is the input bias of the device is failing, and this is causing it to malfunction, but it sure is an odd failure. The references are derived with a long resistive divider, with 1% resistors.
A quick check: device was running 55C which is hot, in a 21C room. Calculated rise of temperature at 67C/W T(j-a) as specified in datasheet puts it at 480mW... according to the datasheet the quiescent power should be 130mW so we can see that the IC is dissipating 3x more power than it should, probably due to an internal short.
It's an AS15-U gamma IC, but this is identical to AS15-F, AS15-G, etc.
My first gamma IC failure! woot.
Comment