Hacking Hardware Picofly - a HWFLY switch modchip

[FreeLander]

I use around 0.3mm

@rehius says 0.2mm minimum
https://gbatemp.net/threads/picofly-a-hwfly-switch-modchip.622701/post-10103574

Great. I've got a spool 0f 0.3mm. Don't you struggle with how little the cap is compared to the wire thickness? I'd love to hear your technique + what to do to make sure it doesn't snap off. Thanks

 

[AntonIX]

@rehius I tried picofly but I am getting pink after blue and OFW. Double checked CLK and CMD. The only point that I am not sure is cpu mosfet, but in this case light should be cyan.

 

[detilmalala]

Hi guys, I had an accident while I was removing the metal cover from the cpu, a capacitor was torn off and I don't have it, can the switch work without that capacitor?

 

[abal1000x]

Great. I've got a spool 0f 0.3mm. Don't you struggle with how little the cap is compared to the wire thickness? I'd love to hear your technique + what to do to make sure it doesn't snap off. Thanks

With 0.3mm i ripped 3 pad of the caps. 1 caps on the left is totally ripped.
1 pad on the right caps also ripped, force me to use another pad on the right bottom.

Next month i might work on my friend's oled, i will try to use 0.2mm.

The ripped always happened after i close the apu shield, and put everything tight, the force from the shield forwarded to the wire and forwarded again to the pad.

I might considered using a flat copper tape, or something similar.

At least for now the 0.3mm wire is successfull and working smoothly.
Goes to hekate almost in instant, the longest is like 3-5 seconds.

 

[rehius]

With 0.3mm i ripped 3 pad of the caps

0.3 is large. I suggest to use 0.2, I know a guy who does 0.1 but that could work worse.

UV glue/mask may help keeping the wires safe

 

[ZERO2028]

Hi guys, I had an accident while I was removing the metal cover from the cpu, a capacitor was torn off and I don't have it, can the switch work without that capacitor?

View attachment 364780

Yes, it will work.

 

[detilmalala]

Yes, it will work.

another question, can the 3.3 of the oled be used with the 2 capacitors that appear on the white line or does it have to be only the one on the right?

 

[MegaDeKay]

I will use silicone so that there is no movement in the solder points

View attachment 364747

You should be a bit careful here as some silicones are corrosive. Use something like Loctite Superflex - it is non-corrosive and often recommended for electronics.

 

[nqtal]

Hepl, please! How to determine what is killed in the switch, memory or processor? Are there any signs or method of verification? We changed the power controller, the short circuit is gone, at startup there is consumption. But the console still won't start.

 

[juanvlc]

another question, can the 3.3 of the oled be used with the 2 capacitors that appear on the white line or does it have to be only the one on the right?

View attachment 364791

Both, but why touch 2 components if you can touch only one?

 

[cowboy619]

Hepl, please! How to determine what is killed in the switch, memory or processor? Are there any signs or method of verification? We changed the power controller, the short circuit is gone, at startup there is consumption. But the console still won't start.

Post pics of the areas you worked on


Sent from my iPhone using Tapatalk

 

[detilmalala]

Both, but why touch 2 components if you can touch only one?

because the tin ran, so can those 2 capacitors be connected?

 

[Phantomas77]

Well, had some fun time today while installing double MOSFET for the first time on the Lite with Sammy eMMC.

Followed @Dee87 's diagram and went full double transistor right away I used AWG30 wire for Caps to MOSFET connections and UV solder mask for fixing stuff.

Here're my 2 cents from what I learned during this install (pics below):

• MOSFET prep

1. place the transistors side by side (source oriented) on an adhesive tape in deadbug style as originally coined by our dear @binkinator (man, where have you been?) and solder the source cable with AWG30 by being careful to not to short the Gate pin of the bottom MOSFET.
2. solder the gate cables (enamel wire should be used here).

• SoC prep

1. solder AWG30 wire by slightly bending the end to each capacitor and gently fix it with some UV solder mask.
2. place two MOSFETs by the SoC side by manipulating it by the source wire.
3. solder the Source and the two Drains by following Dee's scheme.
4. cover everything with solder mask.

Side note: I placed the PicoFly on top of the RAM, though when I assembled the metal and plastic shell it wasn't too smooth, but everything seem to fit OK. The console is glitching, though I had a feeling that it was glitching faster before the metal + plastic shell that probably put some extra pressure on top of Pi. Does this make sense, or should I just chill even if I'm not looking into cutting the metal shell or moving the Pi?

 

[Dee87]

Well, had some fun time today while installing double MOSFET for the first time on the Lite with Sammy eMMC.

Followed @Dee87 's diagram and went full double transistor right away I used AWG30 wire for Caps to MOSFET connections and UV solder mask for fixing stuff.

Here're my 2 cents from what I learned during this install (pics below):

• MOSFET prep

1. place the transistors side by side (source oriented) on an adhesive tape in deadbug style as originally coined by our dear @binkinator (man, where have you been?) and solder the source cable with AWG30 by being careful to not to short the Gate pin of the bottom MOSFET.
2. solder the gate cables (enamel wire should be used here).

• SoC prep

1. solder AWG30 wire by slightly bending the end to each capacitor and gently fix it with some UV solder mask.
2. place two MOSFETs by the SoC side by manipulating it by the source wire.
3. solder the Source and the two Drains by following Dee's scheme.
4. cover everything with solder mask.

Side note: I placed the PicoFly on top of the RAM, though when I assembled the metal and plastic shell it wasn't too smooth, but everything seem to fit OK. The console is glitching, though I had a feeling that it was glitching faster before the metal + plastic shell that probably put some extra pressure on top of Pi. Does this make sense, or should I just chill even if I'm not looking into cutting the metal shell or moving the Pi?

nice looking awesome

that "style "isnt actually from me i copied it frome someone else
:-)
and installed it myself that way and made the diagramm after that

but great job u did there

 

[DonZott]

Well, had some fun time today while installing double MOSFET for the first time on the Lite with Sammy eMMC.

Followed @Dee87 's diagram and went full double transistor right away I used AWG30 wire for Caps to MOSFET connections and UV solder mask for fixing stuff.

Here're my 2 cents from what I learned during this install (pics below):

• MOSFET prep

1. place the transistors side by side (source oriented) on an adhesive tape in deadbug style as originally coined by our dear @binkinator (man, where have you been?) and solder the source cable with AWG30 by being careful to not to short the Gate pin of the bottom MOSFET.
2. solder the gate cables (enamel wire should be used here).

• SoC prep

1. solder AWG30 wire by slightly bending the end to each capacitor and gently fix it with some UV solder mask.
2. place two MOSFETs by the SoC side by manipulating it by the source wire.
3. solder the Source and the two Drains by following Dee's scheme.
4. cover everything with solder mask.

Side note: I placed the PicoFly on top of the RAM, though when I assembled the metal and plastic shell it wasn't too smooth, but everything seem to fit OK. The console is glitching, though I had a feeling that it was glitching faster before the metal + plastic shell that probably put some extra pressure on top of Pi. Does this make sense, or should I just chill even if I'm not looking into cutting the metal shell or moving the Pi?

Beautifull job you did there, the solder joints look great!

I don't think the metal shield or the shell should have any impact on glitching times. You could retrain the timings, maybe it will improve.

 

[FreeLander]

With 0.3mm i ripped 3 pad of the caps. 1 caps on the left is totally ripped.
1 pad on the right caps also ripped, force me to use another pad on the right bottom.

Next month i might work on my friend's oled, i will try to use 0.2mm.

The ripped always happened after i close the apu shield, and put everything tight, the force from the shield forwarded to the wire and forwarded again to the pad.

I might considered using a flat copper tape, or something similar.

At least for now the 0.3mm wire is successfull and working smoothly.
Goes to hekate almost in instant, the longest is like 3-5 seconds.

Figured as much. I've ripped many tiny components when using 0.3 on them. I might as well go for 0.2mm. I was surprised to know you could use 0.1mm on 3v3. I tried it once and smoke came out of the console!

 

[Dee87]

Figured as much. I've ripped many tiny components when using 0.3 on them. I might as well go for 0.2mm. I was surprised to know you could use 0.1mm on 3v3. I tried it once and smoke came out of the console!

the rp 2040 doesnt use that much power thats why u can use 0.1mm

 

[rehius]

v2.65 - added RP Pico support. might be cut after flashing to fit:

you can attach a ws2812 RGB LED to pin 16 for debug.
a default RP Pico LED also works (before you cut the board), but only in the ON / OFF mode

P.S. sure this soldering sucks, I don't care. you get the firmware and the diagram, that's it

 

[Dee87]

v2.65 - added RP Pico support. might be cut after flashing to fit:

View attachment 364810

you can attach a ws2812 RGB LED to pin 16 for debug.
a default RP Pico LED also works (before you cut the board), but only in the ON / OFF mode

P.S. sure this soldering sucks, I don't care. you get the firmware and the diagram, that's it

is the pinout same as before?
or should i do changes in diagramm
?

never mind a few changes will need to be done , il get them done tommorow so u can add pinout if u want to

 

[rehius]

is the pinout same as before?

I decided to go with reliable detection by shorting pins, so the pinout is not the same as planned.