No, this is the strange part. The CPU tokins have a great deal of influence on the Noise that bleeds over to the RSX side. So both sides are important, even though they are electrically independent.
If you're still getting 3004, then I the capacitors you chose might be sus. What are you using? Model number so I can lookup ESR would be helpful. Also, does you eerlog show any 1002's?
Those are great caps and should be fine. I really like the middle post for them as well. Did you confirm that middle strip is indeed GND? I'm sure it is, I just haven't used them before. I would definitely recommend them based on their rated specs.
Could We see a picture of your installation?
A few more things to check:
What's the resistance from +/GND? CPU side and GPU side? You should be getting 1-4 ohms of resistance! Otherwise, thoroughly clean off the board with IPA. Hint, flood with 99% IPA and use a toothbrush to agitate. Then lay a paper towel over the area and rake the toothbrush over the top to soak up the IPA+Flux mixture. Repeat aggressively! I have noticed that flux residues are conductive and can decrease the resistance across the +/GND. Since your caps have a small gap between the 2 rails flux will get trapped between and there is a very small distance between the +/GND rail. That flux could be acting like a short. Once thoroughly cleaned with IPA, use some acetone! Or goof off if you have it. This dissolves flux even better! Try to angle the board so the acetone runs underneath the caps between the rails, hopefully dissolving the residues there. It wont hurt the caps or motherboard components. They are made of chemical resistant resins, not plastic. Ideally you would use an ultrasonic bath which uses cavitation to dissolve the flux residues, but they are expensive. I also recommend this stuff. It's much more harsh blend of solvents, but wont hurt the motherboard or it's components. This is the what I use to clean the BGA after reflow.
What gauge bridge wires are you using and how many?
Those are great caps and should be fine. I really like the middle post for them as well. Did you confirm that middle strip is indeed GND? I'm sure it is, I just haven't used them before. I would definitely recommend them based on their rated specs.
Could We see a picture of your installation?
A few more things to check:
What's the resistance from +/GND? CPU side and GPU side? You should be getting 1-4 ohms of resistance! Otherwise, thoroughly clean off the board with IPA. Hint, flood with 99% IPA and use a toothbrush to agitate. Then lay a paper towel over the area and rake the toothbrush over the top to soak up the IPA+Flux mixture. Repeat aggressively! I have noticed that flux residues are conductive and can decrease the resistance across the +/GND. Since your caps have a small gap between the 2 rails flux will get trapped between and there is a very small distance between the +/GND rail. That flux could be acting like a short. Once thoroughly cleaned with IPA, use some acetone! Or goof off if you have it. This dissolves flux even better! Try to angle the board so the acetone runs underneath the caps between the rails, hopefully dissolving the residues there. It wont hurt the caps or motherboard components. They are made of chemical resistant resins, not plastic. Ideally you would use an ultrasonic bath which uses cavitation to dissolve the flux residues, but they are expensive. I also recommend this stuff. It's much more harsh blend of solvents, but wont hurt the motherboard or it's components. This is the what I use to clean the BGA after reflow.
What gauge bridge wires are you using and how many?
Not enough amps? Could you add photos or tell what wires are did you use to power both ic not sure if one or both, seems power fail.
About the bridge made to give power for ic.
Thank you both for the copious advice. Its good to hear that my thinking so far sounds relatively similar to the both of you.
I have to admit its going to look pretty bad. All the flux I had is some crappy fake "Kester" super messy sticky rosin flux in a squirt bottle. I ran out of my alcohol and so all I had is 70%, but it hasnt been cutting it.
I have an authentic syringe of Kester flux direct from Kester on order in addition to 99% alcohol from MG Chemicals.
OK on the RSX chip side, I'm getting about 0.1 Ohms or nothing. On the CELL chip side I'm getting about 3 Ohms. The same on the reverse. So that tells me something is shorting the RSX side then correct?
I'll run out to Autozone and get some of that electronics cleaner tomorrow and clean this puppy up.
Thank you both for the copious advice. Its good to hear that my thinking so far sounds relatively similar to the both of you.
I have to admit its going to look pretty bad. All the flux I had is some crappy fake "Kester" super messy sticky rosin flux in a squirt bottle. I ran out of my alcohol and so all I had is 70%, but it hasnt been cutting it.
I have an authentic syringe of Kester flux direct from Kester on order in addition to 99% alcohol from MG Chemicals.
OK on the RSX chip side, I'm getting about 0.1 Ohms or nothing. On the CELL chip side I'm getting about 3 Ohms. The same on the reverse. So that tells me something is shorting the RSX side then correct?
I'll run out to Autozone and get some of that electronics cleaner tomorrow and clean this puppy up.
70% IPA has too much water in it. I wouldn't trust it without drying the board first.
Looks like a reflow? Start with a thorough cleaning. Afterwards, if you're sure there are no shorts on the RSX side of the board then that reading means there are shorts underneath or on the RSX itself! So be doubly sure there are no shorts!
Did you read the resistance before installing any AlPol caps? If not it may be worth it to remove everything and re-measure. If you still get the shorts then I'm afraid your issue is either in the BGA or the chip is dead. It will be necessary to reball or replace it.
Also, you can get away with fewer caps. 3x 470uF per tokin is ideal. 4x is potentially counter productive.
Yes like in most cases 90nm always fails to reflow/reball. A simple test remove only those bridges on rsx and taking resistance measurement on point to rsx power.
And if you still think you have issues with caps can remove only the side where you get final power, it seems as short-circuit but from what I see looks good soldering, unless they aren't well soldered to gnd. Just reversing process will tell more.
OK Thanks to RIP-Felix reply in the NEC Tokin Topic with all his important links I got my error log out of my COK-002(60GB CECHC).
And yes they are all the same for all 32 bar the last 1 since I did power on more than 32 times since the YLOD(less than 3s) and the battery was pulled so no times.
Anyway every log bar the last is this:
00000000 A0022110 FFFFFFFF
The last log ERRLOG GET 1F is this
00000000 FFFFFFFF FFFFFFFF
PDF says this
2 1 1 0 Clock Generator Error (IC5001)
So I guess I need to start probing with a meter.
I have 0.2ohms (short) between + and ground of the tokins as well. That is also a short from any Tokin Positive to any ground on the board.
I did remove one RSX Tokin as per the Tokin thread before knowing anything about SYSLOG. No Tant Caps have replaced it yet so I am sure that is not the cause of the short.
The small power connector does not have a short between either 5V/3V and ground. Neither does the 12V
I have 1 fuse on B side(top side) of the board that is no good.
F6001.
I am assuming if that fuse is blown there is a cause. So what next after that?
Hello, I'm hoping that I'm posting in right thread and wont flood too much with some info from my broken PS3's console. Maybe it will give some useful information for future researches. Did this just out of curiosity.
Fix intentions with those consoles:
1st console: I will not attempt to fix it, posting just for the info.
2nd console: Maybe some day, posting just for the info.
3rd console: Some day when I will get tools to rewrite NOR memory.
1st console. Jailbroken, CECHJ03, DIA-002.
Failed delid attempt, maybe it will provide some useful information about error logs. Console was overheating. Being stupid and impatient with improper tools, damaged tracks.
With this PS3 I forgot to plug in BD drive, bluetooth board (I think that how its called). Maybe it is still good for reading current errors? Or should test with everything fully connected? Doing this just out of curiosity, after failed delid attempt, what kind of errors it will throw and maybe it will be helpful for future YLOD investigations.
Before delid attempt, console would give overheating warning and would turn off. After jailbreaking it showed 90C+. At startup fan spins a little and stops. After some time tries to spin full power and instantly shut downs.
So it seems, that those tracks were for Thermal sensing (1103) and for communicating (2203) with IC3001?
Or this error is the main one? "[ERROR]: 0xb0002001 (FATAL) XDR Link not initilized." Cant communicate with BE?
2nd console. Jailbroken, CECHG03, SEM-001.
Same story as the first one, I know, I didint learned from the first time.
Thinking to do micro soldering practice on this board. If I will get microscope some day. Damage is only on top right corner, other corners looks good. (looked through 20x magnifying lenses/glass)
It seems like tracks were damaged for Thermal sensing (1103) and comminication (2203) with IC3001?
Or this one is the main reason? [ERROR]: 0xb0002002 (FATAL) XDR Link not initilized.
3rd console. Failed jailbreak, CECHL03, VER-001.
This one story is a bit different. Would power on, but randomly power off, sometimes pretty fast, sometimes after 10mins~, but never went over 30mins for sure and it didint gave overheating error while it was on.
And as stupid I am, I decided to risk it with jailbreaking and it went off while it was flashing NOR.
And when I went to solder wires for syscon reading I noticed that there were liquid(?) damage? I think.
And when it worked before, when pluged in hdmi cable, would get 1 second of static image (like that one from crt monitors) and would happen when changing resolutions. And when on lowest resolution, monitor would go to sleep and wont wake up until I turn off and turn on monitor again.
And every try, everything is the same. And here command "lasterrlog" doesnt work, it says, command unknown.
This one was more interesting because it wont give error log or it is unknown error for syscon because of corrupted NOR memory? When I will get NOR flasher (probably Teensy 2.0++), will continue investigation on this console. But its not goin to happen very soon.
Sorry for long message and if this kind of information was already known or it is pointless. Just did it out of curiosity. Some other day Im going to continue CECHA01 YLOD investigation with syscon. Currently already did some (created thread), but only used external mode. Maybe I will find out something more with internal mode.
Yes like in most cases 90nm always fails to reflow/reball. A simple test remove only those bridges on rsx and taking resistance measurement on point to rsx power.
And if you still think you have issues with caps can remove only the side where you get final power, it seems as short-circuit but from what I see looks good soldering, unless they aren't well soldered to gnd. Just reversing process will tell more.
70% IPA has too much water in it. I wouldn't trust it without drying the board first.
Looks like a reflow? Start with a thorough cleaning. Afterwards, if you're sure there are no shorts on the RSX side of the board then that reading means there are shorts underneath or on the RSX itself! So be doubly sure there are no shorts!
Did you read the resistance before installing any AlPol caps? If not it may be worth it to remove everything and re-measure. If you still get the shorts then I'm afraid your issue is either in the BGA or the chip is dead. It will be necessary to reball or replace it.
Also, you can get away with fewer caps. 3x 470uF per tokin is ideal. 4x is potentially counter productive.
Resistance of ic power line is on orange point and gnd, if same issue after removing those bridges spotted with red x and still 0.02 ohms you can remove that red line of caps and test again. Be sure you set multimeter on smallest gain of resistance like 200 ohms or if any automatic would help you to understand.
Resistance of ic power line is on orange point and gnd, if same issue after removing those bridges spotted with red x and still 0.02 ohms you can remove that red line of caps and test again. Be sure you set multimeter on smallest gain of resistance like 200 ohms or if any automatic would help you to understand.
Ok so you're saying just test resistance between positive and negative. I am confused how is that different from what I was doing? I will try with the caps removed.
Ok so you're saying just test resistance between positive and negative. I am confused how is that different from what I was doing? I will try with the caps removed.
Yeay, +/GND needs to have a small resistance of greater than 1 ohm. 2 is good, 3 is like new. If there is no resistance when you remove the caps/wires, then it's proves the short is elsewhere. So probe everything around that filter island! If they are all good, then you have narrowed it to the chip itself, which means it's dead. Be sure your multimeter can distinguish small resistance, though! I recently bought a POS that didn't. Bad equipment will have you tearing up chasing gremlins.
Ok another strange story with RSX INT error 1802 unsorted what can be motherboard or cpu itself or interface between them.
Board kte001.
Received Sunday one costumer slim dropped floor unit while it was working. All errors on the motherboard 1802, 14ff and 1701.
Board was sometimes started glod without beep for reset image , sometimes 3 seconds one beep.
OK looks like Rsx damage. Desoldering been looking for oxidation for pins. It seems that rsx was missing pads. On motherboard nothing strange. Took another rsx, clean errors, same 1802 error only after adding new rsx. OK Now with all knowledge I was confident my second rsx is fine, cpu from faulty motherboard was looking fine for measurements but still couldn't boot, 3 seconds with one beep.
Moved rsx back to know working board kte001 starting from first try.
Now it's like 4 times in different units I can not understand why slims will give this error while cpu is the real thing that causes this problem and not rsx itself.
Probably the problem occurs with cpu when is working and shock damaged cpu internally or costumer pushed mobo in order to fix it itself. Don't know exactly but if I ever came across to this error, board will be left for scrap. Not necessarily to follow my advice this is as reference. I've done reball both so I am confident for myself not losing time.
At least costumer will get another motherboard.
Full error was A0611802
After that I inspected CELL/BE die and noticed that it was chipped:
Its not a thermal paste. And after seeing this I decided that it is dead. Assembled enough to put psu on top and tried to boot few times more and it booted? Kinda. Green LED would stay on and fan would ramp up slowly, due to heat, since it had no thermal paste between cooler. And here what I saw in syscon:
Code:
>$ bringup
bringup
[SSM] state: 0000 -> 0101
Bringup Mode #0 (0xFF)
[SSM] ssmCb_OnStartingBePowOn() called.
[SSM] Bringup mode : syspm_stat=00000000/00000000
[POWSEQ] PowerSeq_Setup called.
[SSM] state: 0101 -> 0201
[POWSEQ] AV Backend Setup
[SSM] state: 0201 -> 0102
[SSM] state: 0102 -> 0202
[SSM] state: 0202 -> 0103
[SSM] state: 0103 -> 0203
[SSM] ssmCb_BeforeBeOn() called.
[SSM] state: 0203 -> 0104
Psbd_SbTransMode_Half:0x20e2
>$
[SSM] state: 0104 -> 0204
[SSM] state: 0204 -> 0105
[SSM] state: 0105 -> 0400
(PowerOn State)
[mullion]$
>$ powerstate
powerstate
ATA Power : ON
PCI Power : OFF
RSX Power : ON
XDR Power : ON
Eurus Power : ON
SB Power : ON
RSX Thermal Sensor : AVAILABLE
BE Thermal Sensor : AVAILABLE
[mullion]$
>$ shutdown
shutdown
[SSM] state: 0400 -> 0500
[POWSEQ] AV Backend Letup
[SSM] ssmCb_AfterBeOn() called.
[SSM] Shutdown mode ... req_wake_src = 000000F4, ctxt=00/00
[SSM] Shutdown mode : syspm_stat=00000000/00000000
[POWSEQ] PowerSeq_Letup called.
[SSM] state: 0500 -> 0000
(PowerOff State)
But I couldnt receive signal on monitor. And in powerstate ther was no PCI Power. Nothing reported in error logs.
After many many tries I managed to startup console few more times, with everything connected (HDD stolen from other PS3, didint had its own). Green light would stay on and fan would ramp up to max speed, like first time. No display on monitor, controller wouldnt light up when connecting to USB ports.
Is it still fixable with those errors? Or it acts weirdly like that because of damaged CELL/BE? Console was already delided and by taking many times apart this console I damaged its die or it was already like that.
Have you read my earlier post before your?
My advice leave it for a while. Just try clearerrlog and boot about 3 times. If 1802 comes back then you must wait until someone will test another cpu with his set to a board with same error. I did not get this test but have to. This is about 5 times with different motherboards. On slims will be easy to test, but this is going same error to phats. Do you know if any pressure was applied to any ic, or floor dropped unit?
I still want to create a way for quick test rsx for the moment with a socket which will sit on probably a dia002 motherboard with a modchip. For that I must ask here help to create a pcb for rsx footprint double layer only with pads for balls from front to bottom only. Need someone good at software to create those gerber files to be sent to production. I may fail if I do this with normal things in house pcb. Or at least advise to convert rsx footprint to pcb. This socket it has to something cheap in case it fails to hi frequency it won't be a big lose. It is not as a permanent fix for light of dead, just as universal jig test rsx for one minute maximum until it take image to screen.
We have a MF70 to create different supports for this, now done over 3 years back a 4 mm support and holes to a teflon that will match to rsx layer, rsx can be sit on with balls already, now need some cpu socket pins to insert and solder to a pcb that needs to be created.
Have you read my earlier post before your?
My advice leave it for a while. Just try clearerrlog and boot about 3 times. If 1802 comes back then you must wait until someone will test another cpu with his set to a board with same error. I did not get this test but have to. This is about 5 times with different motherboards. On slims will be easy to test, but this is going same error to phats. Do you know if any pressure was applied to any ic, or floor dropped unit?
Yes, I have read. But after seeing a bit chipped CELL/BE die I think that its dead and wanted to get some info about it. And error A0801301 was more common than A0611802. Most likely I damaged it with loose IHS, cant tell for sure because only yesterday cleaned thermal paste and inspected die. (Yes, stupid of me for not doing it at first ant making sure it was good from the beginning)
While I had this console for 1-2 years it wasnt dropped, didint put pressure on any IC. Before that I dont know anything about it. Bought it for cheap, it said that it just gives ylod if I remember correctly. Though top cover was cracked and on right side (opposite of HDD) plastic is broken. Console was already delided, came without HDD and almost had no screws, only CELL/BE, RSX, power button and antenna was screwed in. Top shell had only one screw. Console also was shipped for 200km~
70% IPA has too much water in it. I wouldn't trust it without drying the board first.
Looks like a reflow? Start with a thorough cleaning. Afterwards, if you're sure there are no shorts on the RSX side of the board then that reading means there are shorts underneath or on the RSX itself! So be doubly sure there are no shorts!
Did you read the resistance before installing any AlPol caps? If not it may be worth it to remove everything and re-measure. If you still get the shorts then I'm afraid your issue is either in the BGA or the chip is dead. It will be necessary to reball or replace it.
Also, you can get away with fewer caps. 3x 470uF per tokin is ideal. 4x is potentially counter productive.
Ok I ripped all the caps off and am getting same results. RSX side very little resistance and Cell side I get about 3Ohms. Should I try to reflow the RSX chip?
Ok I ripped all the caps off and am getting same results. RSX side very little resistance and Cell side I get about 3Ohms. Should I try to reflow the RSX chip?
Hmm, if you are getting 0 ohms short on the RSX... It sounds more like it already was "reflowed" or something.
I don't think a reflow will help. Especially if it's reading short in the core.
Might as well remove it completely and check out of board. Then you will clearly see how your short is internal to the chip.
