Hacking Hardware Picofly - a HWFLY switch modchip

[lightautodark]

not its definetly *==, I saw the video he sent to my fb messenger

I also suggested to him to change the resistor values to 100/100/47 as posted here, but he already tried that ages ago as he says and it doesnt work.

this issue is definetly not a soldering issue, resistor values or something, @rehius definetly need the hardware that have the issue to check it out.

else, all I can say is, instict-nx is the only fix or you can just deal with the issue, at least it works.

Just seen your post history. Seems like you had this issue back in 2023 and people were saying it was the chip used, the dat0 adaptor, poor soldering job, resistors etc.

I'm backing you up now! Albeit a little bit late

 

[killeriq]

Installed - Picofly from AE.
First it didnt boot , just black screen. So it tried to double check all solder points.
But the same result.
Tried different FW 2.73, 2.79, 2.80 chip flashed few times BLUE, then White/Green and OFF.

1. When I unplugged the Flex Cable from Chip and Started Switch I sound hear Touch sounds , but LCD Screen was Black all the time

2. Not sure where was the problem, but unfortunately as i was moving IN/OUT the B cable on Board (backside RST i assume) broke the connection and now is not holding anymore Any clue how else should i try to fix it?



Thanks

 

[lightautodark]

Installed - Picofly from AE.
First it didnt boot , just black screen. So it tried to double check all solder points.
But the same result.
Tried different FW 2.73, 2.79, 2.80 chip flashed few times BLUE, then White/Green and OFF.

1. When I unplugged the Flex Cable from Chip and Started Switch I sound hear Touch sounds , but LCD Screen was Black all the time
View attachment 457928
2. Not sure where was the problem, but unfortunately as i was moving IN/OUT the B cable on Board (backside RST i assume) broke the connection and now is not holding anymore Any clue how else should i try to fix it?



Thanks

View attachment 457920

Erm. possibly damaged the display's fpc connector or the flex of your oled screen not inserted properly or damaged...?

 

[remlei]

Just seen your post history. Seems like you had this issue back in 2023 and people were saying it was the chip used, the dat0 adaptor, poor soldering job, resistors etc.

I'm backing you up now! Albeit a little bit late

yep, I first consulted here what exactly is the issue but... well, the usual replies of check soldering and other stuff was the usual recommendation.

But when I personally consulted the installer with the issue, he was actually not surprised about it and says "this is expected with picofly based modchips" but he redid the wiring changing resistors and swapping it with different rp2040 just to be sure and clean the heck out of it but nope, same, but he knows a solution that can grantee fixed the issue which actually cost me extra 20 bucks for that chip, which me was like "well fuck it, use that instead". He even swapped the modchip right in front of me, test it 100 power cycles and all of it are success glitch.

 

[Jkop]

Installed - Picofly from AE.
First it didnt boot , just black screen. So it tried to double check all solder points.
But the same result.
Tried different FW 2.73, 2.79, 2.80 chip flashed few times BLUE, then White/Green and OFF.

1. When I unplugged the Flex Cable from Chip and Started Switch I sound hear Touch sounds , but LCD Screen was Black all the time
View attachment 457928
2. Not sure where was the problem, but unfortunately as i was moving IN/OUT the B cable on Board (backside RST i assume) broke the connection and now is not holding anymore Any clue how else should i try to fix it?



Thanks

View attachment 457920

Wire is way too big. Just use jumper wire to repair. I just had to repair one this morningView attachment rst.JPG

 

[abal1000x]

View attachment 457877
Tried rp2040-tiny with 47/100/100 and it still boots ofw 1 in 10 or 15... This is depressing.. firmware 2.75

Edit: cmd dat0 100 ohm

If you have time to experiment. Try to reboot 500times, in lot of power condition.
Everytime new glitch point founded, it will be saved.

Maybe the glitch point is different on battery condition or usb powered.
So the more you rebooted on lot of scenario, the more 'better' it is to find the glitch point.

 

[Gobracket]

remove dat0 adapter and boot again

Thanks. Removed dat0 but still no display.. flex screen looks ok..

 

[snaker]

The symptoms are exactly the same as the generic pre-assembled rp2040 that is being sold on AliExpress.

I don't come across this very often so I wanted to test it out before resorting to instinct. This is just the second unit actually. Just sharing my finding. So please be kind.

In order to be sure about the results i would consult you to completely desolder the flex cable, clean the board well and make shorter wires. Route the dat0 cable accroding my pics (this is very important). Insulate the cpu flex before placing the cpu shield. I am pretty sure that the cause of that is not chips fault. Or you are so unlucky that your model is so wierd that instinct nx timings are used in that way that make your cpu glitch never fail. This may be extremely rare case and you may never encounter it again.

 

[killeriq]

Wire is way too big. Just use jumper wire to repair. I just had to repair one this morningView attachment 457941

Thanks for tip, so you scratched off the line and also connect it to one side of that chip/resistor?
Thats too tiny, I will bring it to someone more experienced, as it might be also other issue there and I don't have 2 devices to troubleshoot

Erm. possibly damaged the display's fpc connector or the flex of your oled screen not inserted properly or damaged...?

Flex is this one on the Left and FPC is the tiny on Right?

In case there is damage to one of those, is there a way to replace or what are the steps?
Need to find someone skilled who has extra parts on the table as this way we are only guessing , unfortunately

Got An Idea - if the LCD/Display or cable is messed up - would HDMI screen still works?!

 

[lightautodark]

In order to be sure about the results i would consult you to completely desolder the flex cable, clean the board well and make shorter wires. Route the dat0 cable accroding my pics (this is very important). Insulate the cpu flex before placing the cpu shield. I am pretty sure that the cause of that is not chips fault. Or you are so unlucky that your model is so wierd that instinct nx timings are used in that way that make your cpu glitch never fail. This may be extremely rare case and you may never encounter it again.

A slightly clearer picture. But why are you so sure it's not chip's issue when they are a few people reporting the same problem?

 

[snaker]

View attachment 458061
A slightly clearer picture. But why are you so sure it's not chip's issue when they are a few people reporting the same problem?

Because i never saw something like that after over 300 switch units. Maybe you have a newer model? I dont know.

Also remove the resistor next to led in order to see what error code you get when you fail to glitch.

 

[lightautodark]

Because i never saw something like that after over 300 switch units. Maybe you have a newer model? I dont know.

Also remove the resistor next to led in order to see what error code you get when you fail to glitch.

**= Is the error. Why do I need to remove it? Mine is a genuine waveshare rp2040, from pihut, an authorised distributor. I thought you only remove it if you use AliExpress clone?

 

[snaker]

**= Is the error. Why do I need to remove it? Mine is a genuine waveshare rp2040, from pihut, an authorised distributor. I thought you only remove it if you use AliExpress clone?

If you have the original you are ok there is no need to remove anything.
**= No eMMC block 1 read error has happened to me only from poor quality flex cables from aliexpress clones. Remove the flex cable completery and use good quality wires with solid connections. It is caused from poor wiring that doesnt pass the signals correctly.

Dont get me wrong the flex cables work almost every time but i had 3 cases with the error code you mention and it was the flex cable every time.

 

[lightautodark]

If you have the original you are ok there is no need to remove anything.
**= No eMMC block 1 read error has happened to me only from poor quality flex cables from aliexpress clones. Remove the flex cable completery and use good quality wires with solid connections. It is caused from poor wiring that doesnt pass the signals correctly.

Dont get me wrong the flex cables work almost every time but i had 3 cases with the error code you mention and it was the flex cable every time.

I think I'll not dwell on this anymore. You are clearly set on your opinion, which is fine. Thanks for the help and suggestions.

 

[snaker]

I think I'll not dwell on this anymore. You are clearly set on your opinion, which is fine. Thanks for the help and suggestions.

I never dwelled my friend! Use instinct-nx if it works for you. I have used a lot of instinct-nx too! But time showed me that instict-nx is more expensive and slower at glitching. I must have used around 100 instinct-nx before switching to picofly and i never had an issue. Who cares if it glitches at 4 vs 6 seconds. Both work well. Choose whatever works well for you!

 

[MixMasterMizzy]

I can't find this issue elsewhere.
1. I have no problem getting the "no SD detected" screen
2. I keep getting this screen with an SD card that has hekate as payload.bin.

Any suggestions?

 

[thesjaakspoiler]

Got An Idea - if the LCD/Display or cable is messed up - would HDMI screen still works?!

Yes and no. Hekate only works in handheld mode afaik.
Once everything is set up with Hekate and you boot to CFW automatically, then the HDMI output will work as normal.
So if you have a working setup at hand, then you're fine.

 

[RexSonic]

This post was crazy in hindsight

 

[psycho-neon]

I think I'll not dwell on this anymore. You are clearly set on your opinion, which is fine. Thanks for the help and suggestions.

Why not try the suggestion? Some fpcb causes interference with the signal if I remember correctly. I've modded a lot of switches with the og RP2040-tiny from waveshare, pure wires only, but never had an issue the same as you are experiencing.

it's worth the try though.

 

[lightautodark]

Why not try the suggestion? Some fpcb causes interference with the signal if I remember correctly. I've modded a lot of switches with the og RP2040-tiny from waveshare, pure wires only, but never had an issue the same as you are experiencing.

it's worth the try though.

Everything is wire except for the mosfet to sp1 sp2 is fpcb. I've removed everything and redid everything from scratch. Since I'm installing instinct later, I'm using the cpu flex from instinct. And it still had the same issue. I even removed the solder mask on dat0, cleaned, and reapplied solder mask to make sure there's no grounding to other layer of the pcb, still it boots ofw 1 in 15 tries. I'm moving on. It's not worth wasting more time on it. The salvaged chip and flex works on another console. So it's a mystery. Switch belongs to my work colleague and I've already had it for a while.

Post automatically merged: Sep 16, 2024

Update: as a last ditch effort, I made my kamikaze point a lot bigger, change the dat0 wire from 0.1mm enamel to UL10064 wire and that solved my glitching issue. Now the console glitches all the time (2.75 firmware). But I suspect it wasn't the wire that was causing the problem. It must be the soldering. I wanted the kamikaze point to look nice so I kept it tiny (around 0.5mm). That made soldering difficult and even though I did get the right value at C point on the chip (0.5-0.7), it mustn't have been a solid connection. The correct reading gave me a false sense that my soldering was good. The reason I think it wasn't the wire because in the past I have used both 0.1mm enamel wire and insulated 36awg wire interchangeably when soldering the dat0 point with no issues.

To anyone having this 1 in 15 boots to ofw please check your dat0 connection, even if it gives you the right reading. Thanks to everyone who have given helpful suggestions, really appreciate it. Resistor value does not matter in my case, I tested with 100/100/47, 240/47/47 (AliExpress clone), 47/47/47. They all glitch 100% of the time. With 47/47/47, the switch may fail to boot and stuck on black screen after Nintendo logo once in a while. If you're using rp2040 zero or tiny, go 100/100/47 as advised or just stick with the AliExpress pre-assembled clone.

As for why instinct works with a dodgy dat0 connection, I have no clue