DeadEnd
Senior Member
Alright. So the error normally referrs to HDMI chip. I found a ruined thermistor next to HDMI output, however replacing it changed nothing. However I have now discovered another fatal problem. The resistance at one of the caps around RSX, in the corner, is 0. When reference values stated it should be around 111 ohms. So basically something went wrong again... Ugh. I just can never get it right. Ironically , in the board number 1 there is around 100 ohms so that chip got seated perfectly, yet syscon on it is still dead. It's getting all the voltages btw, but still no dice. At this point I just feel like I can't keep this up. My apartment is too small for these experiments. Perhaps I'll just send it over to Victor until I ruin the last board.
vyktormvmpay25
Senior Member
As I asked before soldering on board, right after added balls must check resistance to power line of ic and ram power line of ic out of board. This will tell you if you can go further. If you stop right at 205 is not damaged on reball process, could be damaged in process of soldering balls to ic.
It is hard to hold a workshop everywhere in world. I feel lucky home with 3 houses (one only as workshop). Here I don't pay rent so I can afford not to charge for fixing.
It is hard to hold a workshop everywhere in world. I feel lucky home with 3 houses (one only as workshop). Here I don't pay rent so I can afford not to charge for fixing.
DeadEnd
Senior Member
Which pads are those? I forgot to test it, yeah. Another issue is that I lifted this RSX from a dead slim. I had assumed the gpu would be fine since slims usually have Tokin faults. But completely forgot to check resistances... Soo my own flop again. For all we know, I might have been soldering a dead gpu all along. Can't think of everything I guess.
vyktormvmpay25
Senior Member
Ok I will add to this reply photos with example of measurement.
On blue corner is the common ram power line. Between any blue points and gnd should be 170 ohms like for 65nm rsx and 400 ohms like for 40 nm rsx.
Now 65nm power line of ic in center should be 4 ohms. On 40 nm should be 1.8~2.7 ohms on center power line of ic. All measurements are on ic out of board, could be same for power of ic on board but not less than 80 on power of ram when on board. At least none near to short of ram power line.
Think it's working fine view now.
On blue corner is the common ram power line. Between any blue points and gnd should be 170 ohms like for 65nm rsx and 400 ohms like for 40 nm rsx.
Now 65nm power line of ic in center should be 4 ohms. On 40 nm should be 1.8~2.7 ohms on center power line of ic. All measurements are on ic out of board, could be same for power of ic on board but not less than 80 on power of ram when on board. At least none near to short of ram power line.
Think it's working fine view now.
Last edited:
DeadEnd
Senior Member
There is some problem with your attachment. I can't view it.
vyktormvmpay25
Senior Member
Can you compare those corners of untouched slim?
DeadEnd
Senior Member
Ok, so the video is such low quality that I thought he added the voltage wire to the two pins, but actually it's only at the third pin? Can someone clarify what is going on in there?
Now I soldered it to the third only, turned it on and and it's a GLOD... Controller wont sync. More error codes now. But bear in mind, I switched it on without modchip once, so it might have registered those earlier codes. Then with the mod it went back to the dreaded 2120... Since I don't have output at HDMI, maybe the HDMI IC is fried...
Now I soldered it to the third only, turned it on and and it's a GLOD... Controller wont sync. More error codes now. But bear in mind, I switched it on without modchip once, so it might have registered those earlier codes. Then with the mod it went back to the dreaded 2120... Since I don't have output at HDMI, maybe the HDMI IC is fried...
Last edited:
vyktormvmpay25
Senior Member
Even with hdmi dead, av ic is individual side of rsx and should work. Controller won't sync to any unit without hdd if I remember correctly?
Remember long time ago running this type of exchange rsx 90 with 65 got 5 seconds ylod and halting system was on always , so probably is a loop somehow.
Did not know about resistors that had to be moved.
I was looking a way to test motherboard that time, so assuming was starting fine halted it was rsx 90 defects.
While glod you can look after voltages around rsx /cpu. Most likely you have a good board, don't know what to say about this modchip. I was sceptic but didn't want to argue with anyone.
Would it be possible to work with 40? Don't start now to think about changing it for the moment, wait for more opinions.
Remember long time ago running this type of exchange rsx 90 with 65 got 5 seconds ylod and halting system was on always , so probably is a loop somehow.
Did not know about resistors that had to be moved.
I was looking a way to test motherboard that time, so assuming was starting fine halted it was rsx 90 defects.
While glod you can look after voltages around rsx /cpu. Most likely you have a good board, don't know what to say about this modchip. I was sceptic but didn't want to argue with anyone.
Would it be possible to work with 40? Don't start now to think about changing it for the moment, wait for more opinions.
Last edited:
DeadEnd
Senior Member
If you're talking about regular AV output with yellow cable, I am embarrassed to admit I don't have that one. So I can't even test it. Not sure what else can be done at this point... Too much time went into it now. I am probably going to take a break from these experiments.
Here are pics of how I soldered the mod board to get from YLOD to GLOD. (In case it still looks confusing, on the bottom side two pads are bridged, third is voltage wire, 4th pad is empty.) From my understanding voltage wire shouldn't touch any of the nearby pads (This is not 100%)
(The soldering here is insanely sloppy, I know. This was done in a rush.)
Last edited:
vyktormvmpay25
Senior Member
Yes better wait. If I don't get it working, nobody will be able to make it. I am not so advanced but very confident so I should not fail 99 percent. 1 percent would be modchip.
Yes AV output should have image if there is an defective hdmi ic, done in many different cases.
Now if you are sure you soldered good those tiny points on spi motherboard then is not working as it should with 65? Not sure what to suggest. Let's wait more opinions.
Edit
right now got mine, it will take some time to find untouched cok002 boards.
Yes AV output should have image if there is an defective hdmi ic, done in many different cases.
Now if you are sure you soldered good those tiny points on spi motherboard then is not working as it should with 65? Not sure what to suggest. Let's wait more opinions.
Edit
right now got mine, it will take some time to find untouched cok002 boards.
Last edited:
DeadEnd
Senior Member
Yea and one more thing, I actually quickly reballed it today one last time before getting those new errors (tried to get 113 ohms back in the corner). And I also forgot to put the heat spreader when I turned it on and the fans slowly went wild. Must have been tired. Went full OCD on this. So my results are not very reliable. I guess I killed one of the cell's tokins in the process (error 1001) and may have ruined some other joints (Even error 3034 appeared). It's toast now most likely...
About spi, I am not fully sure. Instructions are all over the place, but the only time he used the white booard is in the video and it's horrendous quality. Can't be sure if I've done it right or not.
I won't claim it doesn't work well with 65 nm. Somebody else should confirm it. My BGA station and skills are rather amateur . I have never actually reballed these things before. Not even once. This is all new and plenty of mistakes could happen. Nevertheless, maybe someone could use this data for more research... All this is a lot harder than it looks, if you're new to this. Not for the faint of heart.
About spi, I am not fully sure. Instructions are all over the place, but the only time he used the white booard is in the video and it's horrendous quality. Can't be sure if I've done it right or not.
I won't claim it doesn't work well with 65 nm. Somebody else should confirm it. My BGA station and skills are rather amateur . I have never actually reballed these things before. Not even once. This is all new and plenty of mistakes could happen. Nevertheless, maybe someone could use this data for more research... All this is a lot harder than it looks, if you're new to this. Not for the faint of heart.
Last edited:
DeadEnd
Senior Member
Just realized that squeept had a similar error, 2120 and 3034. So there are two of us now...
Pacorretaco
Member
Yeah... don't beat yourself up too much. This is experimental stuff and the failures are valuable too. Also quite expected. Reballers don't grow from trees.
Besides... You probably didn't cause as much destruction as you think either.
Error 1001 doesn't necessarily mean anything is broken. Don't be misled. The tokins are fine. If you want to see me barking the same thing in more detail you can check the SYSCON thread. I say It's something misleading that should be made more clear.
RIP-Felix
Senior Member
@Pacorretaco beat me to the punch! I was going to say...1001 doesn't necessarily mean CPU tokins. That error can happen when you flip the PWR switch off, so I wouldn't pay any attention to it.
The fact you got a GLOD was progress! I wish you would have gone for internal access mode in the SYSCON, because then you could use the powerstate command and get a more detailed errlog with startup sequence information.
Don't despair! You're just tired. You'll feel better in a few days. I've been where you are now and know the disappointment. The excitement to try again will come back once you've had a chance to rest and process what might have gone wrong.
For example, what's the purpose of bridging those 2 pads? We don't know! No one has tried leaving them un-jumped. Maybe that selects between the 65nm and 40nm. The video was for using a 40nm. Also, no one has tried it on a 40nm RSX yet, which is what it was made for. Maybe the mod chip doesn't even work for the 65nm. So really, there plenty we haven't tried to still get hyped about.
The fact you got a GLOD was progress! I wish you would have gone for internal access mode in the SYSCON, because then you could use the powerstate command and get a more detailed errlog with startup sequence information.
Don't despair! You're just tired. You'll feel better in a few days. I've been where you are now and know the disappointment. The excitement to try again will come back once you've had a chance to rest and process what might have gone wrong.
For example, what's the purpose of bridging those 2 pads? We don't know! No one has tried leaving them un-jumped. Maybe that selects between the 65nm and 40nm. The video was for using a 40nm. Also, no one has tried it on a 40nm RSX yet, which is what it was made for. Maybe the mod chip doesn't even work for the 65nm. So really, there plenty we haven't tried to still get hyped about.
vyktormvmpay25
Senior Member
Not to much time but wait my test results. If it will work I will send you that board to test your modchip. Now working on different things. Probably nearly Easter or after will be less crowded.
DeadEnd
Senior Member
Thanks for the support guys. Took another look at the video. Well, what do you know, turns out he soldered the wires in a different way. Will give it another go and might as well take a deeper look into syscon powerstate.
Edit: Now that I soldered the IC exactly like he did in the video, we are back to YLOD...
Edit: Now that I soldered the IC exactly like he did in the video, we are back to YLOD...
Last edited:
patricksouza472
Member
Did you fix the checksum before grounding diag?
Similar threads
-
-
PS3 PS3 CECHA RSX Swap (40nm) – JP Limited – SYSCON Patch, VDDC Mod – Need Detailed Help- Started by jw0710
- Replies: 4
-
-
