[SOLVED] iMac 21.5 backlight flashing/flickering
I've got an iMac 21.5 mid 2011 that have a black screen issue. Right after the boot chime, the backlight goes on for a short period (100 to 200 ms) and then stays off forever. After reading Apple's technician guide and performing some tests, I found that I can see the image using a flashlight. So, the conclusion is that there is some problem with the backlight.
After reading some posts here, I would say that most probably the problem is with the Backlight Board. But, before replacing it, I would like to ensure that this is the real problem. I have knowledge in electronics, so I looked into this board and found that it has 3 connectors:
-> 4-pin Input: (Vin, 0V, PWM and ENABLE in this order)
-> Vsync 2-pin connector
-> 6-pin Output: Output voltage (63V probably...)
I think there are 2 hypothesis:
1) PWM or ENABLE signals are not correct (cable or logic board)
2) Output voltage is not being generated (defective board)
I decided to test the hypotheses 1 by soldering a 4-wire flat cable on the Input connector of Backlight Board and measuring the signals with the system running. Result:
- Vin: +11,5Vdc
- PWM: 3,0 Vdc
- ENABLE: 3,3 Vdc
So, considering that the ENABLE and PWM signals are 3v3 logic, i think everything is correct here and then the Backlight Board may be defective.
Maybe i can measure the output voltage to be 100% sure. However, i was just wondering: could a problem on the Vsync signal turns off the backlight?
I've just replaced the backlight board by a brand new one, but unfortunately the problem remains exactly the same.
Since the backlight turns on for a short period (200 ms I would say) right after the power on, I guess that the problem would be related to some control signal.
Since I already measured the PWM and ENABLE on the 4-pin header and they are ok, I think there might be some problem with VSYNC signal that is located on a dedicated small 4-pin header (only 2 wires).
I read on another forum that the mid 2011 no longer used a V-sync cable as it switched to an internal DisplayPort connection which integrated the V-sync into the main display cable. So, it seems that this signal works as a "backlight sync" signal or something like this.
Anyway, I already tested it with the VSYNC cable unplugged, and the problem symptom is exactly the same.
Does anybody what is the expected voltage/waveform of this VSYNC signal? I could measure it with a multimeter or an oscilloscope.
I've replaced the backlight board by a brand new one, but unfortunately it didnt fix it. The sympthom is exaclty the same, i.e, right after the Mac is turned on, the backlight turns on for a short period (300ms lets say) and then turns off forever.
I suspect about some control signal. Since i already checked the ENABLE and PWM signals, the only one left is the VSync one.
I read on other forum that he mid 2011 no longer used a V-sync cable as it switched to an internal DisplayPort connection which integrated the V-sync into the main display cable.
Could a damaged vsync cable turn off the backlight?
Since the problem persisted even after replacing the backlight board with a brand new one, I decided to take a closer look at the troubleshooting steps on Apple Technician Guide.
I just realized that I missed a basic step, which is to reset the PRAM. I did it, and now the problem symptom has changed completely. Now, the backlight DOES NOT turns on for a short period (200 ms I would say) right after the power on, but It DOES turn on when the Apple logo shows up. However, the whole backlight is flashing intermitently. I also tried to reset the SMC, but it didn't change the behavior.
Apple Technician Guide suggests that if the flashing stops when brightness is set to maximum, then the problem might be the vsync cable. However, it still flashes when it is set to maximum. Anyway, I checked the VSYNC cable continuity with a multimeter (OK) and reseated the connectors, but the problem is still there.
Last thing I tried was to run Apple Hardware Test. It has reported the following error:
It means that there is a problem with the voltage of GPU. I decided to go deep and tried the following software: Bresink - Hardware Monitor
It shows that the MXM graphics module voltage is around 10.78 volts. Probably this is what the error code "VGOR-10.484" wants to tell me :-)
I suspect that this voltage should be around 12V. Can anyone confirm? This tool is paid, but I accidentally bought two licenses...
So, if some good guy wants to help me, I can give one license to them :-)
After the suggestion of our colleague Dan, i tried to measure the 12V output of the PSU.
I already measured it before, but at that time the PRAM was not reset yet and the backlight was off all the time.
So, i measured again, now with the backlight turned on and flashing. I have soldered a flat cable on the 4 pin connector to be able to measure the signals with everything connected (LCD, etc...).
When turned off, the input voltage of backlight board was precisely at 12,0V. It seems that this is the stand-by voltage, but i think its very strange to see it on the input of the backlight board... Anyway.
When turned on, the input voltage of backlight drops to around 10,5V. If i set the brightness to minimum, it rises to around 11,0V.
So, definitely the problem is the PSU. I have checked the 3 output capacitors, but they are perfect. I made a quick test by soldering an additional 3.300uF cap in parallel with the existing ones, but the problem remained. I think this confirms that the PSU is bad (output voltage drops when current is drained)