Hacking Hardware Picofly - a HWFLY switch modchip

[lightautodark]

This is a bad communication error between the chip and the console. I would first recheck the wire lengths, solderind integrity and the routing of the wires and then change the chip.Also use the 2.75 firmware as i find it the most stable i have used after many consoles.

I

This is a bad communication error between the chip and the console. I would first recheck the wire lengths, solderind integrity and the routing of the wires and then change the chip.Also use the 2.75 firmware as i find it the most stable i have used after many consoles.

It wouldn't even boot properly with ansem's 2.73. had to use rehius's 2.75. that's why I thought it could be the chip. It's a generic clone chip using rp2040 from AliExpress. But I've used the same chip on a different device and it worked properly.

 

[Nephiel]

Latest attempt at kamikaze. The grinding I can handle, my trouble is with the soldermask. After applying it on the surroundings, it tends to swallow the DAT0 trace too, before I can get the UV lamp in place.

For now, I apply the mask in 2-3 thin coats. When some gets on DAT0, I remove the uppermost (uncured) coat with IPA and try again.

Should I grínd a wider area from now on? I kind of like the smaller footprint, though.

 

[linuxares]

Latest attempt at kamikaze. The grinding I can handle, my trouble is with the soldermask. After applying it on the surroundings, it tends to swallow the DAT0 trace too, before I can get the UV lamp in place.

For now, I apply the mask in 2-3 thin coats. When some gets on DAT0, I remove the uppermost (uncured) coat with IPA and try again.

Should I grínd a wider area from now on? I kind of like the smaller footprint, though.

Can you use hotglue instead of soldermask?

 

[SN0WMAN42]

Tinyfly v2.0 - the layout is based on the well known rp2040-tiny from waveshare but made for switch.

1. All necessary pads are exposed and the LED DIN as well.
2. Resistors and capacitors are 0402 size for easier soldering
3. Default D0,CMD,CLK resistors are 100,100,47ohm respectively

I've attached here the gerber file and bom/pick and place files.

Updates:
Tinyfly v2.0.zip - Initial Release
Tinyfly v2.0b.zip - Added extra pads for GND and 3V3 line on the right side. Swapped CPU and RST at the bottom part. Some components are changed to basic parts if planning to fabricate using JLCPCB.
Tinyfly v2.0c.zip - Adjusted few and changed components to basic parts for cheaper fabrication.

Images:
View attachment 425415View attachment 425418View attachment 425417View attachment 425416


Component Mappings:
View attachment 426981

When I try to order, even in different sites, it says failed cause lack of bottom soldermask after uploading gerbur Do you know what I am doing wrong ?

 

[abal1000x]

Can you use hotglue instead of soldermask?

For me hotglue is a big no.
Its dIfficult to cleanup, and make messy all over the place.

 

[Nephiel]

For me hotglue is a big no.
Its dIfficult to cleanup, and make messy all over the place.

And it will melt when soldering near it.

 

[linkref]

Hi,

Please, someone can help me identify the CPU point on that chip please ?

 

[deeps]

Hi,

Please, someone can help me identify the CPU point on that chip please ?
View attachment 452778

 

[Foxy223344]

Hello guys,

Took me some time to come post, dont know why i didnt earlier. I installed picofly aliexpress on my switch lite with a very professional installer shop, it booted to hekate the first time but couldnt read the emmc, then it purple screen everytime to ofw. After troubleshooting it and resoldering, it just boots to purple screen even with or without chip. We got the emmc out and read the dump, it was locked and we couldnt reflash it, so we got a new emmc, installed it and filled it with dump, but it would stuck on nintendo logo. I saw the video about guide 1 unbrick on youtube and thought lets do it, so we reinstalled the chip and now its back to purple screen again. What can i do in this case? We checked the soldering points and there isnt a short anywhere. Any advice on what to do? We are super lost and i really dont know what else to try. Pics included are the install:

 

[abal1000x]

Hello guys,

Took me some time to come post, dont know why i didnt earlier. I installed picofly aliexpress on my switch lite with a very professional installer shop, it booted to hekate the first time but couldnt read the emmc, then it purple screen everytime to ofw. After troubleshooting it and resoldering, it just boots to purple screen even with or without chip. We got the emmc out and read the dump, it was locked and we couldnt reflash it, so we got a new emmc, installed it and filled it with dump, but it would stuck on nintendo logo. I saw the video about guide 1 unbrick on youtube and thought lets do it, so we reinstalled the chip and now its back to purple screen again. What can i do in this case? We checked the soldering points and there isnt a short anywhere. Any advice on what to do? We are super lost and i really dont know what else to try. Pics included are the install:

Try to check the cmd resistor (pull up) on the board.
Search in this thread if you need more info.

 

[Foxy223344]

Try to check the cmd resistor (pull up) on the board.
Search in this thread if you need more info.

Could you please link me to it? All i am
Finding are the OLED ones and normal
Switches, but not lite.

 

[Nephiel]

I have an OLED here that's giving me trouble modding. It always boots OFW with error =*** eMMC init failure.
Using a Waveshare RP2040 Tiny with 2.75 and 100/100/47 resistors. For the MOSFETs I am using a V6 flex.
EMMC is SK Hynix (it's kamikazed, but I looked behind the eMMC shield to check which brand).

Tried changing resistors to 47/47/47 . No difference.
Already checked the APU flex and the wiring many times, and it's OK. Redid it all, just in case, twice. The error persists.
Also tried with another, brand new, RP2040 Tiny, nuked it before flashing it with 2.75. No difference.

The error has stayed the same =*** on all attempts on this OLED so far. Never reached the No SD screen. What else should I check?

edit: solved, I didn't account for the built-in 47 ohm resistors of the RP2040 Tiny v1.1. See here.

 

[jkyoho]

I have an OLED here that's giving me trouble modding. It always boots OFW with error =*** eMMC init failure.
Using a Waveshare RP2040 Tiny with 2.75 and 100/100/47 resistors. For the MOSFETs I am using a V6 flex.
EMMC is SK Hynix (it's kamikazed, but I looked behind the eMMC shield to check which brand).

Tried changing resistors to 47/47/47 . No difference.
Already checked the APU flex and the wiring many times, and it's OK. Redid it all, just in case, twice. The error persists.
Also tried with another, brand new, RP2040 Tiny, nuked it before flashing it with 2.75. No difference.

The error has stayed the same =*** on all attempts on this OLED so far. Never reached the No SD screen. What else should I check?

picture of your dat0 kamikaze work please

 

[Nephiel]

picture of your dat0 kamikaze work please

It's right up here.

 

[jkyoho]

It's right up here.

and your wiring pictures on the board?

 

[Nephiel]

Ah hell, the RP2040 Tiny v1.1 has 47ohm resistors onboard, right? The cause is probably the extra 47ohm I have on CLK. Will check that and report back.

 

[Takezo-San]

Hello again guys. Just an update and a question (after) too.

So I went ahead an got another OLED (after my frontal-side botched Kamikaze attempt). Though this time I went for the rear side, as I have had better success with this in the past. Long story short, it was a success. I did cut the Dat5 trace....and the Dat2 trace. They were repaired (with Relife kit wires) and the line was connected and went fine. I find this side easier because If I do botch something, I can repair it. The front side was too close to other major ground lines I think and I have poorer success with it. And the depth of the rear is less than the front, so less grinding involved (1 layer less I believe). So In'm sticking with it.

My question was, I have stuck to fw2.73 but I read recently that fw2.75 is better/more stable!? Should I start to use that instead or stick to the current one I've been using!? Whats the major difference. Any one can chime in hear with their results and experiences.

 

[Nephiel]

Ah hell, the RP2040 Tiny v1.1 has 47ohm resistors onboard, right? The cause is probably the extra 47ohm I have on CLK. Will check that and report back.

Yep, that was it. For the usual 100/100/47 install, with the Tiny v1.1, only 2 external 47 ohm resistors are needed: one for CMD and the other for DAT0. Together with the onboard resistors they make a total of ~100 ohm on each line, and CLK already has 47 ohm onboard, so it needs nothing else.

 

[linkref]

Yep, that was it. For the usual 100/100/47 install, with the Tiny v1.1, only 2 external 47 ohm resistors are needed: one for CMD and the other for DAT0. Together with the onboard resistors they make a total of ~100 ohm on each line, and CLK already has 47 ohm onboard, so it needs nothing else.

Hi I just did an install on a v2 with a tiny 1.1, and i did not add any resistor.. They already come with 47ohm resistors from factory. So why did you need to add some ?

 

[Nephiel]

Hi I just did an install on a v2 with a tiny 1.1, and i did not add any resistor.. They already come with 47ohm resistors from factory. So why did you need to add some ?

With 47/47/47, there have been many reports of consoles that can boot CFW just fine but have trouble booting OFW, or have slow EMMC issues. In those cases, the solution is to add a second 47 ohm resistor to CMD and DAT0 lines, so some of us choose to install 100/100/47 by default.