Hacking Hardware Picofly - a HWFLY switch modchip

  • Thread starter Thread starter mathew77
  • Start date Start date
  • Views Views 3,677,512
  • Replies Replies 17,052
  • Likes Likes 15
Guys, I've done 4 switches during the last 4 weeks, 2 lived and 2 died. Both that died were V2 but at this point I'm baffled as to why that happened.

My question is, once they're BSOD is there any way to recover them? I've try reflowing them but I was sure that wasn't going to work as the connection isn't the problem. I was thinking of replacing the caps where the picofly connection goes, but I don't know if I would be wasting my time at that point?

Has anyone successfully reversed a BSOD after damaging their switch with a PicoFly installation?
 
can you show us the error video? For better debug
Here ya go.
Post automatically merged:

Could you give photo of the mosfet installation?
Post automatically merged:

Guys, I've done 4 switches during the last 4 weeks, 2 lived and 2 died. Both that died were V2 but at this point I'm baffled as to why that happened.

My question is, once they're BSOD is there any way to recover them? I've try reflowing them but I was sure that wasn't going to work as the connection isn't the problem. I was thinking of replacing the caps where the picofly connection goes, but I don't know if I would be wasting my time at that point?

Has anyone successfully reversed a BSOD after damaging their switch with a PicoFly installation?
Is yours a blue screen of death? Did you use a flex cable? I used a flex cable on one today and it went bsod on me. Reflow didn’t work either.
 

Attachments

  • IMG_6581.MOV
    36.8 MB
  • ED5459F6-32F3-438D-B03B-B46BA559F77F.jpeg
    ED5459F6-32F3-438D-B03B-B46BA559F77F.jpeg
    1.4 MB · Views: 86
Last edited by BlueBeans,
Here ya go.
Post automatically merged:


Post automatically merged:


Is yours a blue screen of death? Did you use a flex cable? I used a flex cable on one today and it went bsod on me. Reflow didn’t work either.
As debug list in V6 guide, I can say there is something wrong with your mosfet or you mosfet installation.
 
As a Proof of Concept (POC) try to add another cable in D from the cap to mosfet.
View attachment 378003
Lot of trouble happened because the current flows to the mosfet is not enough. The IRF8342 itself could hold 8A, so the problem not lay on the mosfet, but on the delivery of the current, which is the cable.
Ok I’ll add another and report back
 
Here ya go.
Post automatically merged:


Post automatically merged:


Is yours a blue screen of death? Did you use a flex cable? I used a flex cable on one today and it went bsod on me. Reflow didn’t work either.
Yes, Blue screen of death. I used a dual mosfet array as always and it's a coin toss at this point. I made sure I applied all the tips you guys have shared with me here (check for shorts before powering on, making sure just the necessary amount is tinned, etc) and still dies.

At this point I'm done with this mod but I want to know if there's any way I can get those consoles back since I've burned near 600USD in replacing damaged boards.
 
As a Proof of Concept (POC) try to add another cable in D from the cap to mosfet.
View attachment 378003
Lot of trouble happened because the current flows to the mosfet is not enough. The IRF8342 itself could hold 8A, so the problem not lay on the mosfet, but on the delivery of the current, which is the cable.
So it’s still giving the same error code.
Post automatically merged:

As debug list in V6 guide, I can say there is something wrong with your mosfet or you mosfet installation.
I get this error code literally every single time I do an install. I’m not quite sure what I’m doing wrong.
 
So it’s still giving the same error code.
Post automatically merged:


I get this error code literally every single time I do an install. I’m not quite sure what I’m doing wrong.
You can try to Redo D0 with new wire, and/or shorten the wire from mosfet to pico, and/or redo/change the mosfet.


Anyone can confirm/test that if we change the wire lenght then pico needs to learn new timing? so wire length is part of the equation here.
I found that if glitch wire is on 5cm it will glitch in under 3s .. but if make 10cm or > it will take >5s.
 
i just wanna say
thx for your help guys!!!
i ask many questions and you help me:mthr:
you guys really help me alot



View attachment 377996
View attachment 377997
this pic , i shot when about 90% process installation, after that im too lazy takin the shot :rofl2:

glitch work!:grog:
View attachment 377998
View attachment 377999

View attachment 378000

View attachment 378002
Selamat. :grog::toot:.
With that kind of wires all around but still works.. impressive to say the least.
 
  • Love
Reactions: nico77
Selamat. :grog::toot:.
With that kind of wires all around but still works.. impressive to say the least.
ma~acih~
yeah, that shot from 90% in the process of installation(yeah many kapton tape LOL), after that i clean up more. trim some wires.
 
You can try to Redo D0 with new wire, and/or shorten the wire from mosfet to pico, and/or redo/change the mosfet.


Anyone can confirm/test that if we change the wire lenght then pico needs to learn new timing? so wire length is part of the equation here.
I found that if glitch wire is on 5cm it will glitch in under 3s .. but if make 10cm or > it will take >5s.
Re did all the wiring and I’m still getting the same code.
Theres couple of thing you might need to check:
1. Make sure the Gate connected to the pin 15 on pico using multimeter. If its not connected, the same error happened.
2. Measure the D and S/GND resistance in mosfet pin should read around 12 Ohms. Also measure D-G and S-G resistance, should be high resistance (around MOhm).

So I tested from the gate to 15 and it’s connected. And I know this is going to sound terrible but I don’t really know how to test in resistance mode. Do I just put black on ground and test the points in resistance mode? If that’s the case my numbers are totally off.
 

Attachments

  • image.jpg
    image.jpg
    1.8 MB · Views: 54
How about the CMD and the CLK pad. Is it okay?
Post automatically merged:


Check also the cmd resistor might be damaged.
View attachment 377991
I forgot the value in lite, something like 4k7.
I am not in front of lite right now.
in Ω20k scale: ~4.6
in continuity test: 697
 
It seems okay.

The cmd and clk pad doesn't broke right?
The pad connect the emmc and the cpu, if its disconnected the cpu cannot access the emmc.
The dat0 pad still bothers me, because I don't particularly know what the expected default value is for a regular dat0 connection.

My current dat0 connection value, post reconstruction is 655 on the continuity scale. The others (CMD and CLK) are above 700~780. I know that there is a connection between my reconstruction and the alternative DAT0 point, as that is where the picofly wire is soldered. But I'm not sure if the DAT0 pad is connected.
 
Similar when testing resistor.

Put the black to S/GND, and put the red on D, it should read around 12 Ohms.
If its high then its not connected, Could be the problematic cap solder or mosfet D solder. If its low 0-3 ohms, then its shorted, could be problematic solder on cap or mosfet, could be the mosfet damaged, or the cap damaged.

Check the D-G or S-G similar put the black on G, and the red on D or S. It should be high. Its just cheap detection if the mosfet is okay. If its connected, could be problematic solder joint or the mosfet damaged.

Might could also check using diode mode, but i am prefer resistance. No logical reason, just preferential.

Man this seems all messed up. When I put black on S and red on D I get 50ohm. When I black on G and red on D I get 50ohm.
 

Site & Scene News

Popular threads in this forum