HP Compaq Presario CQ42 powering on, no display. Missing +3VLCD.

[gjgib]

Hi again guys. I have this motherboard that won't show display even tho it was doing it yesterday. I think something could've past away in the circuit.
I'd appreciate if anyone who is used to MOBO LCD voltages and circuits could enlighten me since this is the first time I have to deal with it.

I'm missing +3VLCD.

This is the circuit


With laptop OFF:

Values in LCD motherboard connector:
+VIN_BLIGHT: 19.55V
+BLONCON: 0V
+VADJ1: 0V
+3V: 0V
+3VLCD_CON: 0V

Also some voltages I assume are important here:
+15VALW: 12.8V
LCDONG: 12.4V
LCDON#: 0V
DISP_ON: 0V



With laptop ON:

+VIN_BLIGHT: 19.55V
+BLONCON: 0V
+VADJ1: 0V
+3V: 3.3V
+3VLCD_CON: 0V
--
+15VALW: 12.8V
LCDONG: 0V
LCDON#: 5.06V
DISP_ON: 0V

Is very strange to me how in R36 has 12.8V in both ends when laptop is OFF, but when laptop is turned ON, one end has 12.8V and the other one is 0. Blows my mind.

I don't know the timing of these voltages/signals but this is my conclusion:

-GPU is not giving DISP_ON signal; therefore
-Tansistor Q5 is not pulling +5V to GND
-+5V reaching Q7-Gate and now pulls +15VALW to GND
-+15VALW doesn't turn into LCDONGate, therefore
-+3V don't go through Q8 for +3VLCD


I'm very noob still but I'd like to know if anything I said makes sense.
Thanks in advance.

[piernov]

It is probably a similar effect as with the Nvidia "bumpgate" ICs (https://www.badcaps.net/forum/troubl...vidia-bumpgate), where heating the IC to a high temperature softens and expands the underfill and the bumps between the die and the substrate, which causes them to make contact again. But thermal cycles will deteriorate it again. This is not about the balls between the IC and the motherboard, but between the silicon die and the substrate of the IC itself, hence why reflowing and reballing is useless.

[mcplslg123]

Does your mb has discrete video chip? If yes, then its the video chip gone bad. Thats the norm on discrete models of CQ42.

[gjgib]

mcplslg123 Thanks for answering. So, is my conclusion accurate?

Btw I took this picture. I assume this is discrete since it's not integrated inside the CPU(?

[gjgib]

I think I might be wrong in some way. DISP_ON signal is supposed to shut down Q5 INVERTER transistor. So, since DISP_ON is not there, +5V is going to GND through Collector to Emitter, therefore, not turning into LCDONG.

Very interesting stuff.

Conclusion will update like this:

-GPU is not giving DISP_ON signal; therefore
-Tansistor Q5 is pulling +5V to GND
-+5V not reaching Q7-Gate and not pulls +15VALW to GND
-+15VALW doesn't turn into LCDONGate, therefore
-+3V don't go through Q8 for +3VLCD


Seems weird to me that +5V reaches Q6 GATE but not Q7 GATE aswell. Schematic design couldn't show properly why?

[alexfts20]

It appears that your board has the "popular" AMD 216-0752001 North Bridge chipset with a built-in GPU, which is known to have such issues. One method to verify it is by applying hot air to the chip for a short period. If it temporarily starts working again, it could confirm the issue, although this is not a recommended solution.

It's worth noting that most 2001 "new" chips on the market are remarked, meaning they have been rebranded, so replacing the chip may not resolve the problem.

[gjgib]

alexfts20 True! I already saw them on AliExpress for a future purchase. Saw both 'old' and 'new' ones. Old ones say 2010+ and new ones 2019-2022+. I guess that's the year of manufacture.

Would you suggest to buy the old ones instead of the new ones?

Also, why would that be that they wouldn't work? Different build?

edit1: Would appreciate if you could link where you got that info. Very helpful and money saving..

[alexfts20]

In my opinion, they are the same. Unfortunately, it doesn't make any sense to replace this chip. Finding a genuine new 2001 chipset is almost impossible.
I obtained that information from my past experience with many laptops on this chipset.

[piernov]

These BGA ICs form Aliexpress are for sure pulled from dead boards and reballed, and guess why the boards would be dead… In the past, see-ic had some good ones (and some bad ones too) but I don't think that's the case anymore. Not much you can do on this 15 year old platform. Best would be to replace board+CPU+heatsink with an Intel platform without dGPU, but it's not really worth it.

[gjgib]

I'm kind of a rookie at the moment. Would you consider it a waste of time to buy it or even try to reflow it?

[alexfts20]

gjgib Do you have hot air rework station?

[gjgib]

alexfts20 Yessir! Baku-601D.

Also, how would you interpretate the circuit that I posted first? I mean, what are the parameters to have the missing +3VLCD.

[alexfts20]

gjgib You don't have DISP_ON which comes from 2001 chipset. Please check PM.

[piernov]

No point in reflowing as always. Just heat it to 200°C for a minute and a half (no flux, the point is not to melt the balls) and it'll come back to life, but who knows when it will fail again. The ICs you'll buy from Aliexpress will be in the same state so it's completely pointless to try to replace with one of these.

[gjgib]

alexfts20 Hey! I've been kinda away. Thanks for the video. Okay I'll try that. Thanks!

piernov Thanks. Oh, okay. I thought it was about reflowing but its just heat?. Btw I have a G62 that seems to give video after a bit of push on the GPU chip. That one is indeed about reflowing?

[Jesse James]

I agree it is pointless to fix. I had 5 or 6 AMD CQ42 and CQ62 all with this same issue. If you try to search other faults on the board you will always come back to heating up the chip and it will work temporarily. I've also had it where DVD is not recognized.

[gjgib]

Thanks for insights. I did the reflow on the 2001 GPU chipset since customer wanted to try. You guys were right, it came back to life but went off again. What does the reflow do exactly in that case? 'Repair' the structure within the chip and make it work or is it something with the BGA solder? Now I'm shipping in the stencil because customer wants me to try reballing since she saw it work and plus got nothing to lose and I can practice manual reballing.

What do you guys think?

[piernov]

It is probably a similar effect as with the Nvidia "bumpgate" ICs (https://www.badcaps.net/forum/troubl...vidia-bumpgate), where heating the IC to a high temperature softens and expands the underfill and the bumps between the die and the substrate, which causes them to make contact again. But thermal cycles will deteriorate it again. This is not about the balls between the IC and the motherboard, but between the silicon die and the substrate of the IC itself, hence why reflowing and reballing is useless.

[gjgib]

piernov Ohh, I didn't get before why people was so pesimistic about this. Crystal clear now.

Very interesting stuff! Thanks a lot for the insight, piernov. I'm digging more about it.

Thanks to everyone who contributed to the post.


CASE DISMISSED.