Hacking Hardware Picofly - a HWFLY switch modchip

  • Thread starter Thread starter mathew77
  • Start date Start date
  • Views Views 3,668,579
  • Replies Replies 17,052
  • Likes Likes 15
A friend sent these to me. I only tested the adapter on an old HWFLY modchip and its giving me conflicting results. I'm still waiting for my batch of rp2040 tiny to arrive tomorrow or after tomorrow to fully test on a fresh console.
I like the idea! You could use rp2040 tiny to eliminate those giant 0805 package resistors and you could use back mosfet locations to unnecessarily opening apu shield. This way everything can be on back side of the board and as a bonus you can decrease flex cables to one piece.
 
I like the idea! You could use rp2040 tiny to eliminate those giant 0805 package resistors and you could use back mosfet locations to unnecessarily opening apu shield. This way everything can be on back side of the board and as a bonus you can decrease flex cables to one piece.
My friend everything was accounted for, it isn't just for show. Resistor pads are crucial part of the design to make it easy to solder resistors when required. The design is meant specifically for RP2040 Tiny as I teased it last month, but could also be used with Zero. In the case of Zero you can neatly add them here while the Tiny you can add the extra ones when required.

I thought of doing one flex holding everything but that would be a bad design as different thicknesses are required and people might want to use different type of modchips so having the adapter separate makes it useful for more people.

Feel free to do back mosfet. I will only finalize the current design I've made.
 
  • Like
Reactions: achm3t
Most likely put some kapton ape on the chip for isolation. With the backplate it will be probably shorted.
I redid the kapton tape and now it functions even with back plate on (I still have like 5 screws of the metal shield removed because I got too lazy). So yeah, thanks for making me re-check the tape.

Now only problem with this is that SysMMC and OFW goes to a black screen and does not work, but that's... fine. EmuMMC works perfectly well.

Edit: I can now access SysMMC and OFW. My eMMC was SK Hynix and was known to not play well with the Flex cable. 30 minutes of Direct Soldering via Enameled Copper Wire later and Its working.
 
Last edited by Ravenext,
  • Like
Reactions: QuiTim
An error occurs when using rp2040 zero "*== No eMMC CMD1 request (poor wiring, or dead CPU)", OFW is loading. But if I solder the old hwfly, then everything works fine. I've tried different rp2040 zero, all works in other Nintendo Switch
 
Last edited by heinrich_frei,
So, I installed the flex on a brand new Switch and it glitches fine, not sure why its giving me messed up RGB.
Though I have several remarks,
1) Adapter needs to be below 0.1mm for comfortable installation for that I made a single layer adapter but will not be be compatible with the master flex. I will try to remake it at a later time.
2) There's too much stress on RST point, I feel the tension could tear it. Solution: bend it before you install, this way you relieve some stress, or better soldermask like I did.
3) CPU attachment needs to be longer to relieve stress on flex.

There will be a follow up revision that address these.
 

Attachments

  • 1c8ee9ae-d475-43b3-b3e2-1515b2aec9f0.JPG
    1c8ee9ae-d475-43b3-b3e2-1515b2aec9f0.JPG
    203 KB · Views: 66
  • 126277e5-7f9a-4816-8bf2-b7de0d5a89fd.JPG
    126277e5-7f9a-4816-8bf2-b7de0d5a89fd.JPG
    268.4 KB · Views: 78
  • 466947a9-1a98.MP4
    4.1 MB
I had an oled that glitched really slowly. anything between 10 seconds to a minute. mosfet on the back. Added another mosfet on the back - no change. Added a mosfet on the apu (so 3 mosfets in total) - still no change. disconnected mosfets on the back and only used the one on apu - boom, 3 second glitch every time.
most other oleds I've done have one on the back and glitch fast every time.
The point of the anecdote is that changing things around can make a difference in unexpected ways.
Has anyone figured out what factors affect the glitch time? From my experience I know the thickness of wire on the APU caps does but now with OLEDS and installing a single mosfet on the back I’m also getting different results.

Yesterday, had one that took forever to glitch. Moved the mosfet to the MAX ic location and still no improvement. Gave up and went with the cpu flex and almost instant glitch times.
Today, with another OLED went back and used the SAME mosfet on the back and this one glitches super quick.
I know it’s not the brand of eMMC as both were Samsung.
 
I'm doing adapter first, then emmc. Seems more logical, and it's a good idea to not expose the emmc to more heat than necessary. So the workflow is as follows:

- detach emmc from mainboard
- clean mainboard with solder wick
- clean emmc with solder wick
- reball emmc
- reball adapter (on the bottom side)
- attach adapter to mainboard
- attach emmc to adapter
Did you still add resistors to the picofly?just about to finish up with the install and wanted to double check.
 
I normally use rp2040-zero, so yes, 100/100/47 on it. When i used rp2040-tiny i added 47ohm extra on dat0+cmd
I figured they were still used but wasn’t sure. Just finished the install and got the correct sequence of lights! 🙌🏼 thank god lol. I thought this one wouldn’t be able to be modded because I tore the pad off CLK. So happy to finally have a modded OLED.
 
  • Like
Reactions: deeps
Hello, newbie here. YES I did reaearch before posting, but I got the HYfly on my OLED, and when I power chip, it flashes yellow/red mix 13 times then it resets and does this in a loop. I've left for 2 hours and nothing. Anyone come acress this?
I tested the solder points as shown on youtube and everything was reading as it should, including the emmc chip solder point
 
Lot of time but a easy work that second time done. @abal1000x
 

Attachments

  • A11.jpg
    A11.jpg
    1.8 MB · Views: 63
  • A10.jpg
    A10.jpg
    1.6 MB · Views: 71
  • A9.jpg
    A9.jpg
    1.3 MB · Views: 67
  • A8.jpg
    A8.jpg
    826.5 KB · Views: 60
  • A7.jpg
    A7.jpg
    614.7 KB · Views: 52
  • A6.jpg
    A6.jpg
    1.3 MB · Views: 60
  • A5.jpg
    A5.jpg
    1.4 MB · Views: 69
  • A4.jpg
    A4.jpg
    1.3 MB · Views: 56
  • A3.jpg
    A3.jpg
    1.6 MB · Views: 63
  • A2.jpg
    A2.jpg
    1.6 MB · Views: 57
  • A1.jpg
    A1.jpg
    3.2 MB · Views: 56
  • A12.jpg
    A12.jpg
    1.7 MB · Views: 71
  • Love
Reactions: abal1000x
i flashed my picofly-one with 2.75 but after that i reconnect and is not showing again in file manager even pressing boot button. how to check if is ok?

edit: nvm solved
 
Last edited by nitrozz,
Hi everyone, what does this mean?


LED goes from blue to blinking yellow, and I'm stuck in sysNAND even though it should boot to Hekate.
I had someone else install it for me, so I'm not even sure what kind of chip it is.
 
Hi everyone, what does this mean?


LED goes from blue to blinking yellow, and I'm stuck in sysNAND even though it should boot to Hekate.
I had someone else install it for me, so I'm not even sure what kind of chip it is.
Is it an Oled model? Looks like =* dat0 not connected
 
Sorry, I forgot to say the most important thing. Yes, it is an OLED.
Most likely the dat0 adapter lost contact.
Since you don't do this kind of installations yourself you will need send it back to the installer and tell him to fix the dat0 point. If possible he should do it properly by setting a permanent dat0 point (reball the emmc).
 

Site & Scene News

Popular threads in this forum