PS3 Frankenstein PHAT PS3: CECHA with 40nm RSX

[stefanotk]

Do you have tantalizers?

Yes I've already put the tantalizers, however I fixed the problem by doing a GPU reflow with a new profile with a slower ramp and reaching higher temp peak (220°C), now I am testing it since 15 hours ago with gran turismo 6 demo and it hasn't frozen.

 

[shnyaps]

Yes I've already put the tantalizers, however I fixed the problem by doing a GPU reflow with a new profile with a slower ramp and reaching higher temp peak (220°C), now I am testing it since 15 hours ago with gran turismo 6 demo and it hasn't frozen.

Did you fixed both frankies with new profile?

 

[stefanotk]

Did you fixed both frankies with new profile?

On the first one I haven't tried yet

 

[shnyaps]

On the first one I haven't tried yet

Which bga do you use?

 

[stefanotk]

Which bga do you use?

I used CXD5302DGB, I bought from aliexpress but now they are out of stock

 

[iKon]

Yes I've already put the tantalizers, however I fixed the problem by doing a GPU reflow with a new profile with a slower ramp and reaching higher temp peak (220°C), now I am testing it since 15 hours ago with gran turismo 6 demo and it hasn't frozen.

Interesting. Are you actually nudging the BGA during reflow, or just letting it sit and fully reflow on its own?

My two cents: I only pull RSXs from known-working Slims. I try to avoid random loose chips where you have no idea what profile was used, how many times it was heated, or what damage may have already been done.

With leaded balls, around 200°C at the chip should normally be plenty, assuming your thermocouple is accurate. I've seen cheap Amazon/China thermocouples read way off, so I only use lab-grade Omega TCs now. They cost a lot more, but they are consistent and trustworthy.

The only reason I could see needing to push closer to 220°C is if the pads were not wetting properly. I would make sure the pads are extremely clean. I use a three-step cleaning process: MEK, acetone, then final clean with 99.9% IPA using a nano cleaning pad.

Not saying your fix didn't work, but if 220°C solved it, I'd be looking closely at pad prep, TC accuracy, or possibly marginal wetting before assuming the RSX itself needed that much heat.

 

[stefanotk]

Interesting. Are you actually nudging the BGA during reflow, or just letting it sit and fully reflow on its own?

My two cents: I only pull RSXs from known-working Slims. I try to avoid random loose chips where you have no idea what profile was used, how many times it was heated, or what damage may have already been done.

With leaded balls, around 200°C at the chip should normally be plenty, assuming your thermocouple is accurate. I've seen cheap Amazon/China thermocouples read way off, so I only use lab-grade Omega TCs now. They cost a lot more, but they are consistent and trustworthy.

The only reason I could see needing to push closer to 220°C is if the pads were not wetting properly. I would make sure the pads are extremely clean. I use a three-step cleaning process: MEK, acetone, then final clean with 99.9% IPA using a nano cleaning pad.

Not saying your fix didn't work, but if 220°C solved it, I'd be looking closely at pad prep, TC accuracy, or possibly marginal wetting before assuming the RSX itself needed that much heat.

Hello, I didn't nudge the BGA, however it worked without freezes for two days and now after testing again it freezes again after few hours with gran turismo 6, I had cleaned the pads with goot wick for removing the old solder and then I used only IPA to remove the residues of flux. I can try another time eventually with acetone and a flux remover solvent too. I don't think it's caused by a not working gpu, also because it's the second ps3 that has the same problem, I had bought the GPUs from two different suppliers so it's more unlikely that both the GPUs are not good.