vyktormvmpay25
Senior Member
Please don't do blind reflow/reball ,it becomes harder to fix.take uart log,resistance measurements ,voltages , etc.. First of all is a must read all around forum then you get more knowledge where to start.
PS3 Fault finding YLOD with the SYSCON - First steps and Error reporting
Timbo9876
Member
Many thanks for your advice. I do have a method in my madness. If I get a 3 second plus YLOD I heat up the tokins. If I get a boot I replace 2 RSX tokins and 1 CPU tokin.
I now if that doesn't work a lot of the time the failure goes from YLOD to GLOD. I do a reflow of the main chips. If that don't work then I do the CPU pressure mod and increase the pressure until I get a stable boot. So it's not completely blind but I take your point, a syscon read would be the way forward
I now if that doesn't work a lot of the time the failure goes from YLOD to GLOD. I do a reflow of the main chips. If that don't work then I do the CPU pressure mod and increase the pressure until I get a stable boot. So it's not completely blind but I take your point, a syscon read would be the way forward
RIP-Felix
Senior Member
Yes.
Interested in both. But particularly interested to see if the GLOD issue was with the RSX BGA, CPU BGA, or was not resolved at all. In the end it could be bad CPU die, but seems unlikely. @SkaziChris looks to have good enough equipment and skill to find out. His tests are the most methodical we've seen at eliminating the variables. No offense to you vic, your tests on these codes and the special GLOD have been super useful too. It's just harder to follow the train of evidence in your posts because of the language barrier (google translate perhaps?). Anyway, I'm watching the errrlogs and tests carefully to see if we can finally pin these codes down to the causes. Like my post above...
I'm just trying to build an understanding of these HDMI/DVE errors. It's obvious they can be caused by bad HDMI/DVE chip. It would be simple and clean if that were the only cause of 2024/2124. Be we 're seeing them in consoles that have bad Tokins. And we see them in consoles that have BGA/Bumps defects. Not just the RSX, but CPU too. They have been a confusing error for while now.
RIP-Felix
Senior Member
Well, @botakompong's been saying all along that RSXRAM can cause GLOD. Maybe that is what this is. Didn't you try reballing RSX ram (hynix) at one point? I vaguely remember a post, but not sure if I noted it in my spreadsheet. We'll see. I'm up to page 90 ATM. Hoping to finish my review of the thread this weekend.
vyktormvmpay25
Senior Member
Yes sure rsx can give that special glod no output very confident as I've resoldered it back to his work jsd that was starting fine with image, I didn't touch cell not even delid, board on jig straight, same glod type. So by now all those cases were cell is fine we should have that SB debugging connection on uart. Sometimes it happens, those 40nm very hard to delid even outside of board, even like that bending it with a space of rasor blade is still a bit tricky.
RIP-Felix
Senior Member
Yeah, I noticed that the 40nm RSX's thermal adhesive is stronger than that used on COK-001. I worried about breaking a solder ball when I delidded the 40nm RSX, after I had already reballed it, I should have done it when it was off the board - like you say.
I wonder if the RSX RAM balls are being damaged by the process of delidding?
Honestly, I haven't seen very many RSX's that actually need to be delidded! It's always the CPU that overheats (1200). Not a single overheating RSX (1201)! We have never seen a 1201 (upto page 91 in this thread at least).
SkaziChris
Member
I will continue the work after the weekend (occupied by my daily work at the moment).
Will post an update once reballing is done.
vyktormvmpay25
Senior Member
@SkaziChris the reason I have said previously it may not work is because I see you desolder rsx without ihs. Anyway sorry for that, I don't know what everyone is doing as there are different rework stations/profiles to do that without ihs.
RIP-Felix
Senior Member
Do the owners of these console mind if you share which user's they belong to?Here received today two packages from users in forum. We will see what we can get sorted and and later will see more Frankenstein units and hopefully more gaming tests.
Edit
@RIP-Felix got something nice to test.
I have been compiling the errors/console/user in a spreadsheet. This allows me to track each individual consol's repair history on the thread. I would just like to know (for example) that the motherboard is the same one @userdude posted initial results on page 83. This way I don't have duplicates of the same motherboard.
vyktormvmpay25
Senior Member
One user from USA only has those to me. I did not asked for spears pcb for tantal on right time. Once is done I will ask for sharing there. I may fix one cok001 (not sure yet), one dia001. Both will have those pcb.
DeadEnd
Senior Member
They probably won't overheat, but heavy weight of IHS would cause the balls to merge during installation. In fact, it happens quite easily even without IHS on cok001 boards. Additionally, some will have dried up paste so the temps could still improve a bit.
Last edited:
RIP-Felix
Senior Member
I'm just wondering if delidding is worth the risk. I'm always afraid the BGA will pop. Now I'm worried about the VRAM too.
Man, I hate the RSX package! Terrible design.
SkaziChris
Member
Yes, it is totally worth it, but only if you have the right tools and right technique for it.
Frankly, cpu delid is waaaay more risky than the rsx.
Rsx takes maybe 1-2 minutes to delid, without any excessive force.
Cpu... well... without the razor sharp thin blade, it is very risky - and I have had plenty of ps3's with cpus scratched or permanently damaged by previous owners...
SkaziChris
Member
@RIP-Felix
Reballing is done.
Console starts and works for 20s and then shuts down.
(I think I need to resolder tantals after bga process, as they were held only by capton tape).
Current errlog:
C:\Users\chris\Desktop\PS3_syscon>ps3_syscon_uart_script.py COM3 SW
>$ AUTH
Auth successful
>$ bringup
00000000
# [SSM] Bringup Start.
# [SSM] PS0 ok.
# [SSM] PS1 ok.
# [SSM] PS2 ok.
# [SSM] PS3 ok.
# [SSM] PS4 ok.
# (PowerOn State)
OK 00000000
#!
#!Boot Loader SE Version 2.5.0
#!(Build ID: 3318,35708,
#!Build Date: 2008-10-11_00:31:58)
#!
#!Copyright(C) 2008 Sony Computer Entertainment Inc.All Rights Reserved.
#!
#![INFO]: Connecting to Debug Device (SB UART)
# [UCMD] Unknown command.
>$ powerstate
00000000
# ATA :ON
# PCI :OFF
# PCIex:OFF
# RSX :ON
# GDDR :ON
# XDR :ON
# EURUS:ON
# SB :ON
# [SSM] Cond/Fatal received, msg=24AE.
# [SSM] Fataldown Start.
# [SSM] Fataldown ok.
# (PowerOff State) (Fatal)
# [SSM] Clearfatal Start.
# [SSM] Clearfatal ok.
# (PowerOff State)
>$ errlog
00000000
# CODE CLOCK
# A0801601 FFFFFFFF
# A0801701 FFFFFFFF
# A0801802 FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
>$
>$ AUTH
Auth successful
>$ bringup
00000000
# [SSM] Bringup Start.
# [SSM] PS0 ok.
# [SSM] PS1 ok.
# [SSM] PS2 ok.
# [SSM] PS3 ok.
# [SSM] PS4 ok.
# (PowerOn State)
OK 00000000
#!
#!Boot Loader SE Version 2.5.0
#!(Build ID: 3318,35708,
#!Build Date: 2008-10-11_00:31:58)
#!
#!Copyright(C) 2008 Sony Computer Entertainment Inc.All Rights Reserved.
#!
#![INFO]: Connecting to Debug Device (SB UART)
# [UCMD] Unknown command.
>$ powerstate
00000000
# ATA :ON
# PCI :OFF
# PCIex:OFF
# RSX :ON
# GDDR :ON
# XDR :ON
# EURUS:ON
# SB :ON
# [SSM] Cond/Fatal received, msg=24AE.
# [SSM] Fataldown Start.
# [SSM] Fataldown ok.
# (PowerOff State) (Fatal)
# [SSM] Clearfatal Start.
# [SSM] Clearfatal ok.
# (PowerOff State)
>$ errlog
00000000
# CODE CLOCK
# A0801601 FFFFFFFF
# A0801701 FFFFFFFF
# A0801802 FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
>$
Did this thermal pad trick on the rsx last year. The SUR-001 slim died after 2 months with graphics glitches. After delidding the rsx, found one corner of epoxy case of the VRAM broken possibly caused by the pressure.
SkaziChris
Member
Adding CPU tantals did not help.
So the console starts, goes from PS0 to PS5 ok, then works for about 20s,
and I hear disk starting and spinning, and then it shuts down with these two error codes:
# CODE CLOCK
# A0801601 0B49D885
# A0801701 0B49D884
# A0801601 0B49D84C
# A0801701 0B49D84B
# A0801601 0B49D818
# A0801701 0B49D817
# FFFFFFFF FFFFFFFF
# FFFFFFFF FFFFFFFF
SkaziChris
Member
I desoldered the RSX chip properly.
Mobo is still in good condition, all pads on mobo socket and on the RSX are perfectly fine.
Previous reballing was properly carried out...
Right now, the RSX VDDIO has 4 Ohms...
Previously it had 2.5 Ohm...
Anybody has an idea about these 601... 701... error codes?
RIP-Felix
Senior Member
Unfortunately yes, I do have an explanation for this.
If a YLOD turns into a GLOD after reball/reflow then we think 1601 (with or without 1701) means the RSX RAM was damaged and the chip needs replaced. This is a loose association based on user reports. But your results seem to suggest the same progression. Also the 2124 errors you previously had with the "Special GLOD" are another indicator of RSX RAM issues. That lone 1802 is a telling sign. That's an RSX initialization error. SYSCON Initializes the RSX at step 20. And the CPU/SB do some AV backend initialization/corrdination with the RSX (not sure what), which is where the 2120 errors come in if the RSX or VRAM is unhealthy. We know that a dead or missing RSX returns 20 1802.
You got an 1802, but it occurred at step no. 80, not 20. That could mean it's not completely dead. It did respond during initialization, but failed in the power on state with the 1601/1701. VRAM is still highly suspect, but I also suspect the Redwood FlexIO interface at the CPU, where the CPU/SB communicate/corrdinate with the RSX's VDDIO.
Did you reflow/reball the CPU? That would rule out a dodgy CPU/SB FlexIO connection. After that they RSX VRAM is probably dead. Try replacing the RSX. If that still doesn't work, the CPU is probably dead. You can try exchanging the entire married chipset if you feel like getting to the bottom of it and confirming the hypothesis. But most techs call it dead if the CPU dies.
That's what I'm currently thinking.
SkaziChris
Member
Mine is CXD2991CGB
I just ordered two CXD2982GB, as they should be compatible with
- CXD2991GB
- CXD2991BGB
- CXD2991GGB
- CXD2991CGB
Shipping... 2 days...
I will get back to you once I have it and measure it first.
RIP-Felix
Senior Member
Converting your colors to voltages:
AQUAMARINE = +1.8V_RSX_FBVDDQ = 25 Ohm
BLUE = +1.8V_RSX_PLL_VDD = 0.355 MOhm
RED = +1.2V_RSX_VDDC = 2.5 Ohm
HARLEQUIN = +1.8V_RSX_PLL_VDD = 0.63 MOhm
GREEN +1.2V_RSX_VDDR = 0.53 kOhm
YELLOW = +1.5V_YC_RC_VDDA = 0.264 MOhm
Striped YELLOW (\\\) = Not Connected on RSX = OL (normal)
BROWN = +1.5V_RSX_VDDIO = 0.815 kOhm
PURPLE = +1.2V_YC_RC_VDDIO = 1.44 kOhm
You can compare these voltages to the new RSX to see if any stand out of the ordinary.
Similar threads
-
-
PS3 CECHA00 with several SYSCON errors (3004, 1001,1002, 2120,3011)
- Started by LSL
- Replies: 7
-
PS3 Interesting PS3 Errors (1802, 1701, 1601) and Capacitor Replacement- Started by Cheshire UK
- Replies: 2
-