Hacking Hardware Picofly - a HWFLY switch modchip

[EchoStorm]

If i were to mod my v1 unpatched just for the heck of it, would this be the right board for nowadays?
really want to try a relatively easy one and this seems like the way to go. Its a bit of a desert out there with availability.

 

[deeps]

Guys is it just a case that every picture I see of single capacitor mosfet installs are on the leftmost capacitor?
Can you perform it on the rightmost one too?

Yes, it works the same. For some reason the left one became the default, but there's no difference.

Post automatically merged: Oct 23, 2023

If i were to mod my v1 unpatched just for the heck of it, would this be the right board for nowadays?
really want to try a relatively easy one and this seems like the way to go. Its a bit of a desert out there with availability.
View attachment 400779

We don't support the chinese clone boards in this thread, but yes, this one works for v1

 

[EchoStorm]

Yes, it works the same. For some reason the left one became the default, but there's no difference.

Post automatically merged: Oct 23, 2023

We don't support the chinese clone boards in this thread, but yes, this one works for v1

noted. thank you for the info.

 

[Alanamaro]

100 ohm
and you need to fix your cmd pad

The pad is ok now, it no longer gives a black screen, now the blue screen lol, can I replace the 0201 100ohm with a higher value, will the lite work perfectly without it?

 

[deeps]

The pad is ok now, it no longer gives a black screen, now the blue screen lol, can I replace the 0201 100ohm with a higher value, will the lite work perfectly without it?

that is unlikely to solve your problem, I'm afraid. Recovery from blue screen is rare.

 

[Eds05]

Hey guys, Im working on an Oled with a picofly RP2040 refusing to boot in OFW and CFW (nintendo logo then black screen), works only with an emuMMC. Sometimes the Picofly give the error **= (about bank1).

- I restored the emmc with hekate, still NOK
- I reballed 2 times the emmc, still NOK
- I can delete my emummc and create a new one and it will work but still only with the emummc
- if I benchmark the emmc, everything is OK,
- no slow mode in hekate

Do you have an idea? If my memory was corrupted I couldn’t do an emummc right ?

Im using a picofly Tiny. This console was working fine before I switch from a bad installation with an HWfly to a Picofly because the glitch was really random.
- I fixed the CLK which was cuted really close to the tegra and the via almost dead,
- I fixed the resistor of CMD which was broken,
- I replaced 1 capacitor on the tegra,

At this moment everything was working with the Picofly disconnected, till I fix the dat0 wire to the emmc and I bring power to the picofly.

Now even after 2 reball I still have the issue.

But I can create an emunand from scratch and play with the emunand. But sometimes the chip gives the error **=.

SOS

 

[jkyoho]

Hey guys, Im working on an Oled with a picofly RP2040 refusing to boot in OFW and CFW (nintendo logo then black screen), works only with an emuMMC. Sometimes the Picofly give the error **= (about bank1).

- I restored the emmc with hekate, still NOK
- I reballed 2 times the emmc, still NOK
- I can delete my emummc and create a new one and it will work but still only with the emummc
- if I benchmark the emmc, everything is OK,
- no slow mode in hekate

Do you have an idea? If my memory was corrupted I couldn’t do an emummc right ?

Im using a picofly Tiny. This console was working fine before I switch from a bad installation with an HWfly to a Picofly because the glitch was really random.
- I fixed the CLK which was cuted really close to the tegra and the via almost dead,
- I fixed the resistor of CMD which was broken,
- I replaced 1 capacitor on the tegra,

At this moment everything was working with the Picofly disconnected, till I fix the dat0 wire to the emmc and I bring power to the picofly.

Now even after 2 reball I still have the issue.

But I can create an emunand from scratch and play with the emunand. But sometimes the chip gives the error **=.

SOS

what resistor combination you use on rp2040?

 

[deeps]

Hey guys, Im working on an Oled with a picofly RP2040 refusing to boot in OFW and CFW (nintendo logo then black screen), works only with an emuMMC. Sometimes the Picofly give the error **= (about bank1).

- I restored the emmc with hekate, still NOK
- I reballed 2 times the emmc, still NOK
- I can delete my emummc and create a new one and it will work but still only with the emummc
- if I benchmark the emmc, everything is OK,
- no slow mode in hekate

Do you have an idea? If my memory was corrupted I couldn’t do an emummc right ?

Im using a picofly Tiny. This console was working fine before I switch from a bad installation with an HWfly to a Picofly because the glitch was really random.
- I fixed the CLK which was cuted really close to the tegra and the via almost dead,
- I fixed the resistor of CMD which was broken,
- I replaced 1 capacitor on the tegra,

At this moment everything was working with the Picofly disconnected, till I fix the dat0 wire to the emmc and I bring power to the picofly.

Now even after 2 reball I still have the issue.

But I can create an emunand from scratch and play with the emunand. But sometimes the chip gives the error **=.

SOS

you need additional resistors on dat0+cmd.

 

[Eds05]

I have the resistors of the tiny + 47ohms for dat0.

I will try to add 47ohms pour le cmd.

Otherwise, I have black screen if I remove de 3.3v on the pico

 

[Vigintiduo]

After browsing the thread, I've been getting quite curious about the Kamikaze method for the dat0 point in oled units.
I know that reballing is safer and probably better but... Kamikaze just looks more fun as a hobbyist... I guess?
Is there any documentation or video about it? I'm VERY close to buying a oled Switch just to try this lol

 

[jkyoho]

After browsing the thread, I've been getting quite curious about the Kamikaze method for the dat0 point in oled units.
I know that reballing is safer and probably better but... Kamikaze just looks more fun as a hobbyist... I guess?
Is there any documentation or video about it? I'm VERY close to buying a oled Switch just to try this lol

this should be a practical one

 

[sw3]

Hi,

until now I thought with the RP2040-Tiny (original one/ Waveshare), theres no need for additional resistors. But as i read now, sometimes, if it doesn't work properly, there are additional resistors needed (47 Ohm?).
Is this summarised correctly? Thanks a lot!

 

[Vigintiduo]

this should be a practical one

That's pretty clear! Looks like two ground layers and then you get to dat0.
But people in this thread have been using a different drilling spot or am I wrong?

 

[jkyoho]

That's pretty clear! Looks like two ground layers and then you get to dat0.
But people in this thread have been using a different drilling spot or am I wrong?

back side also has a spot under 2layer ground. People use front side one because they don't need to take the board out and do everything at front-side.

 

[Eds05]

you need additional resistors on dat0+cmd.

I tried with additional 47ohms for dat0 and cmd and I still have the same issue.

I will try a layer 2 restore.

 

[CarlosCruz]

That's pretty clear! Looks like two ground layers and then you get to dat0.
But people in this thread have been using a different drilling spot or am I wrong?

For me it is safer to drill than to lift the nand. Not everyone has a good soldering station, the nand can break due to heat, break when handling it, break a pad, a lot of dirt spread by the flux, and thermal stress in general. It is very likely that the board will end up broken due to not knowing how to recover a pad or the chip itself. A jumper under the chip by joining dots, you connect the battery and goodbye. It is a job that is not paid, the client does not care if you install an adapter or a fixed point, if the adapter comes loose, the problem lies with the installer. Drilling is cheaper, cleaner, safer and you get the same result. A fixed point.

 

[abal1000x]

For me it is safer to drill than to lift the nand. Not everyone has a good soldering station, the nand can break due to heat, break when handling it, break a pad, a lot of dirt spread by the flux, and thermal stress in general. It is very likely that the board will end up broken due to not knowing how to recover a pad or the chip itself. A jumper under the chip by joining dots, you connect the battery and goodbye. It is a job that is not paid, the client does not care if you install an adapter or a fixed point, if the adapter comes loose, the problem lies with the installer. Drilling is cheaper, cleaner, safer and you get the same result. A fixed point.

Why not the client choose it?
If i were the client, i want the least destructive undo-able mod, yet solid.

 

[Hassal]

For me it is safer to drill than to lift the nand. Not everyone has a good soldering station, the nand can break due to heat, break when handling it, break a pad, a lot of dirt spread by the flux, and thermal stress in general. It is very likely that the board will end up broken due to not knowing how to recover a pad or the chip itself. A jumper under the chip by joining dots, you connect the battery and goodbye. It is a job that is not paid, the client does not care if you install an adapter or a fixed point, if the adapter comes loose, the problem lies with the installer. Drilling is cheaper, cleaner, safer and you get the same result. A fixed point.

Now I wouldn't go as far as saying its cleaner and safer, perhaps for you but for everybody else its not. In fact you're far better reballing and killing your eMMC than doing this since by the end of the day the eMMC can be replaced with another but in your case if something goes wrong its hasta la vista.

Engraving by hand using a drill requires a skillset. Its quite apparent you're good at doing engravings.

 

[cicci084]

Well, I also always got DAT0 by reballing.
Seeing the latest video (the one of the drilling) I wanted to test it and risk it just yesterday.
I say risk because I won't deny that it was very risky (I did it on a customer's console).
In my experience, both, if done well, are infallible.
It must be said that both methods are very risky if you don't have the right equipment.
As regards the reball, since I carry out repairs on smartphone mainboards (including IC replacement), I already have excellent equipment and therefore I have not had any kind of difficulty, but it must be recognized that a novice hobbyist could cause quite a bit of damage , so my advice is to practice on faulty mainboards.

As for the drilling, as mentioned, it was the first one I carried out yesterday, it is the method I would prefer if I were a beginner.
I went there very CAREFULLY, I used a smartphone grinder (Maant D1) for the first and second layer of GROUND, but then, to find the DAT0 trace I used a sharp blade and scraped off the PCB little by little.
In practice it took me the same time as it takes to reball.
It must be said that the tips of my grinder are terrible, worn and not precise (they are crooked, so I have to get new ones).
If I had better tips, it would definitely have been a quicker and safer job.

I attach some photos, if you have any doubts I am ready to answer.

 

[deeps]

Hi,

until now I thought with the RP2040-Tiny (original one/ Waveshare), theres no need for additional resistors. But as i read now, sometimes, if it doesn't work properly, there are additional resistors needed (47 Ohm?).
Is this summarised correctly? Thanks a lot!

yes, you are correct. extra 47 ohm each for dat0+cmd. or replace the on-board dat0+cmd resistors with 100 ohms.