Hacking Hardware Picofly - a HWFLY switch modchip

[Dee87]

Yeah props to that guy! Does anyone know if the firmware Rehus is graciously developing for us is in any way related to the original craigslist discovery? Or was the Picofly firmware we use a result of starting from scratch? Either way it's very impressive and I can't say thank you enough to everyone involved!

Picofly is a firmware created by @rehius to proof that is was possible to create a modchip with a RP2040, since a lot people said it wasn't possible to do it with a Rp2040.

I don't he had any relations with the original Craigslist offer

so this is the result of a find on Craigslist where someone offered a Rp2040 modchip for the Nintendo Switch.

 

[Switxh]

Picofly is a firmware created by @rehius to proof that is was possible to create a modchip with a RP2040, since a lot people said it wasn't possible to do it with a Rp2040.

I don't he had any relations with the original Craigslist offer

so this is the result of a find on Craigslist where someone offered a Rp2040 modchip for the Nintendo Switch.

Cool thanks! That's very impressive!

 

[RatchetRussian]

What does it mean if sometimes (maybe 1 out of 10 boots) the switch will just boot into OFW and the picofly will not even light up at all? As though it has no power.

weak mosfet? I thought even with a weak mosfet, the pico will at least give an error light before booting OFW.

 

[deeps]

oled (From @raksmey1231)
View attachment 381314

I did some measurements and it is worth noting that in this specific location on the back of the OLED motherboard, there is only 1.5mm of height available between motherboard pcb and chassis. The mosfet (irfhs8342) is about 0.75mm high, so the solder joints on top of the mosfet must not be as high or higher than the mosfet itself, or they will press against the chassis which is always bad for reliability. It's a good idea to look from the side after soldering and make sure that the solder joints aren't sticking up too much.

 

[abal1000x]

What does it mean if sometimes (maybe 1 out of 10 boots) the switch will just boot into OFW and the picofly will not even light up at all? As though it has no power.

weak mosfet? I thought even with a weak mosfet, the pico will at least give an error light before booting OFW.

Cheap guess, theres problem on 3.3v/GND line on the picofly.

 

[RatchetRussian]

Cheap guess, theres problem on 3.3v/GND line on the picofly.

You are probably right. This is the same batch of rp2040 as the one that was previously confirmed defective.

Serves me right for getting the cheapest possible clone on AliExpress.

So it's true that an unstable glitch would show an error on the pico?

 

[abal1000x]

You are probably right. This is the same batch of rp2040 as the one that was previously confirmed defective.

Serves me right for getting the cheapest possible clone on AliExpress.

So it's true that an unstable glitch would show an error on the pico?

I have never meet with a problematic rp2040-zero from waveshare.
I always choose the cheapest one.

If the glitch failed, theres always error code throws.

Since you said, the picofly doesn't even shows light, it might the power line troubled, could be cold joint in the 3.3V/GND, or the wire shorted if using enameled, or the 3.3v / GND point choose from the board is not suitable, e.g. the microsd 3.3v.

 

[RatchetRussian]

I have never meet with a problematic rp2040-zero from waveshare.
I always choose the cheapest one.

If the glitch failed, theres always error code throws.

Since you said, the picofly doesn't even shows light, it might the power line troubled, could be cold joint in the 3.3V/GND, or the wire shorted if using enameled, or the 3.3v / GND point choose from the board is not suitable, e.g. the microsd 3.3v.

Its not waveshare. Even cheaper ($3 per board)

Maybe I am killing it with too much hot air though. I am not very skilled with a rework station.

I'm using the suggested 3v3 and gnd points from the guide.

I will try with a known good rp2040 to test.

All day I was researching the i2c glitch thinking it will fix my problem. Thank you for informing me otherwise!

edit: known good pico works every time

I feel like I have had enough failures to start suspecting these rp2040 clones as troublesome: https://www.aliexpress.us/item/3256804781611696.html

 

[abal1000x]

Its not waveshare. Even cheaper ($3 per board)

Maybe I am killing it with too much hot air though. I am not very skilled with a rework station.

I'm using the suggested 3v3 and gnd points from the guide.

I will try with a known good rp2040 to test.

All day I was researching the i2c glitch thinking it will fix my problem. Thank you for informing me otherwise!

Using the back mosfet better than using i2c.

Can you give the link, i might try to buy it

 

[RatchetRussian]

Using the back mosfet better than using i2c.

Can you give the link, i might try to buy it

Edited post above with the link.

It has over 2000 sales and good reviews. Maybe I am the problem. Too much heat on my rework station maybe?

Also the chips I got do not have the "waveshare" silk screen like in the photos.

Also yes I am very excited to try your back mosfet method. It looks amazing but I still have a few flex cables left that I need to use.

 

[abal1000x]

Edited post above with the link.

It has over 2000 sales and good reviews. Maybe I am the problem. Too much heat on my rework station maybe?

Also the chips I got do not have the "waveshare" silk screen like in the photos.

Also yes I am very excited to try your back mosfet method. It looks amazing but I still have a few flex cables left that I need to use.

I use that board too, i've not meet any problem.
Its okay even without the waveshare silk. Its only silk, and might increase the cost.
Removing the silk to "cost down" the price, its an okay for me.

I am hoping a lesser price though .
A board without usb and the switch.
We could easily solder to bridge the boot switch, and connect the usb to our own usb adapter.
I assume its only use the old usb d+ and d- not the full feature of usbc.
So outputing only 4 testpad vbus, gnd, d+ and d- enough for us to solder it ourself.
It could reduce the cost further.

 

[postnutclarity]

Alright fellas, thanks for your inputs. I didn’t listen to some of your guys advice because I went back to my modder with the switch and it worked for a good minute. But when I went back home and fiddled around, it kinda crashed and rebooted back to OFW.

I opened it up, and the error codes gave me I think lost connection to D0 and RST again. Additionally, it seems like one (or two) of the resistor is touching that solder joint where the boot button used to be.

What do you guys think? Is this salvageable? Should I go back again? Third time the charm?

 

[abal1000x]

Article from nasa about the standard on soldering for J and gulwing is good to read
https://workmanship.nasa.gov/lib/insp/2 books/links/sections/707 Gull-Wing^L Leaded Packages.html
https://workmanship.nasa.gov/lib/insp/2 books/links/sections/709 J Leaded Packages.html

 

[HackMan37]

Successful installation of mosfet on the back side of my switch lite, i wasn't confident on do it on the APU, but that's safer.

NOTE: Same result as before, can't get any payload than ums loader xD

 

[Arakon]

I still have the issue that after glitching, my V2 switch flickers a grayish screen (black with gray lines) a few times before it boots. I'm on 2.73 now, it did the same on 2.61. However, it does glitch absolutely reliably, takes 3-5 seconds, then does the gray screen thing, then boots to Hekate just fine.
Any ideas?

 

[abal1000x]

I still have the issue that after glitching, my V2 switch flickers a grayish screen (black with gray lines) a few times before it boots. I'm on 2.73 now, it did the same on 2.61. However, it does glitch absolutely reliably, takes 3-5 seconds, then does the gray screen thing, then boots to Hekate just fine.
Any ideas?

Seems like not a picofly issue.
I think its more on lcd issue.

If before install picofly, the gray lines doesn't shows up, then my cheap guess is your soldering quality is low. Need to resolder, and makesure everything is solid, and not short circuiting in anywhere.

You might give a picture of your work, so we could have a gist on your soldering quality.

 

[Dee87]

Seems like not a picofly issue.
I think its more on lcd issue.

If before install picofly, the gray lines doesn't shows up, then my cheap guess is your soldering quality is low. Need to resolder, and makesure everything is solid, and not short circuiting in anywhere.

You might give a picture of your work, so we could have a gist on your soldering quality.

I agree , also I would check the lcd connector is it sits properly.
A pic or video would help

 

[Arakon]

I can take a video later. Don't have it open currently, but if it glitches AND has no LCD issues at all after this, I doubt it's the solder or the LCD connector. It literally only happens between the LED going green and the actual switch logo showing on screen.

 

[deeps]

Alright fellas, thanks for your inputs. I didn’t listen to some of your guys advice because I went back to my modder with the switch and it worked for a good minute. But when I went back home and fiddled around, it kinda crashed and rebooted back to OFW.

I opened it up, and the error codes gave me I think lost connection to D0 and RST again. Additionally, it seems like one (or two) of the resistor is touching that solder joint where the boot button used to be.

What do you guys think? Is this salvageable? Should I go back again? Third time the charm?

it still looks like the rp2040 is directly touching the cpu heatsink clamp (along the left side of the rp2040) which would short your connections and result in the errors you are seeing. Your installer sucks.

 

[Switxh]

Alright fellas, thanks for your inputs. I didn’t listen to some of your guys advice because I went back to my modder with the switch and it worked for a good minute. But when I went back home and fiddled around, it kinda crashed and rebooted back to OFW.

I opened it up, and the error codes gave me I think lost connection to D0 and RST again. Additionally, it seems like one (or two) of the resistor is touching that solder joint where the boot button used to be.

What do you guys think? Is this salvageable? Should I go back again? Third time the charm?

That picture is after you went back again? It looks worse than before almost. Honestly the guy should give you your money back and never touch a Switch again.