Hacking Hardware Picofly - a HWFLY switch modchip

[abal1000x]

.

 

[Adran_Marit]

Never get your firmware from Github. Always get it from here.

https://gbatemp.net/threads/picofly-a-hwfly-switch-modchip.622701/post-10090767

The only issue with the github firmware is that it might be a tad slow to update. Other than that they are identical to what is posted here.

If people aren't downloading the uf2 properly or can't read basic instructions they have no place modding in my opinion, unless they want to risk bricking /s

 

[FreeLander]

@FreeLander do you have the usb voltage tester? Did you check for voltage change/reset when connecting the charger?

Hello. Yes. It's drawing 0.48amps on 15v. Nothing changes when I power it on. Will try to change the CMD resistor and see

Post automatically merged: May 31, 2023

Just read the updated code from rehius.


So heres the summary of the flowchart:

First, the code will detect the voltage
Check the voltage of RST (1.6-2.0v) failed throws **
Check the voltage of CMD (1.6-2.0v) failed throws *=
Check the voltage of D0 (1.6-2.0v) failed throws =*

All in the supposed voltage. This detection doesn't remove the possibility that the D0 short with the D1, only to make sure if the D0 line is not disconnected. In your case, this test is passed (OK).

Then the code check on the activity of the emmc.
It should do somekind of cmd1 request, then cmd1 response, then read block 0 then read block 1.

Here is the boot process in the datasheet as reference
View attachment 374998

In your case, nothing is detected (no cmd1 request, cmd1 response, block 0 and block1).



So this is my one cent guess:
If your device could boot normaly to ofw after that, then the apu or emmc, everything is okay. Your wire is definitely at fault, because the pico couldn't spoof the emmc communication.

Heres the probable scenario, if your device (apu + emmc) is okay:
CLK line is not connected. Even though in the section "Led Indication" stated that the CLK also detected, but i dont see it in anywhere in the code, which detecting the CLK line. I am not expert at signal, maybe you could test using multimeter on the pico, whether there is a voltage. If there is activity then the CLK line is connected. Would be better if you have oscilloscope. I don't have it, so i couldn't suggest further.

Since CMD is connected (voltage detection), and i dont see any possibility that it goes wrong (some interference), then the pico could read the line if CLK is connected. And at this point, the Dat0 not yet to be used. Dat0 will be used in the next stage, which is writing the loader to the emmc. So Dat0 is less probable to be the cause of the problem. Buuuuttt stilll, if the Dat0 short circuit the Dat1, then the emmc might not function properly. And again if its working normally goes to ofw, then the emmc function properly means the Dat0 is not short circuit with Dat1.

So in conclustion, your CLK is the most suspect. But since its difficult and risky to do anything on that point. You might just do some simple measurement using multimeter on the CLK pin in the pico side right before the resistor. Just to makesure if it is really connected. There should be some voltage measured. I am not yet confirmed this. Maybe others who still have oled could confirm this.

Thanks for the time you spent writing this. Actually, it sounds very reasonable to suspect CLK. I don't usually take much time on it, but because my soldering station is new I took more than I should to throw solder in there. When this code appeared, I resoldered it specifically as I thought it was the problem, but still. Regarding the diode mode, It's giving 0.7v.

 

[POPOLO]

Could you explain this to me a little more? I am using fw2.74

PI2040-Zero PIN13 -> SCL PIN12 ->SDA

 

[QuiTim]

Hello. Yes. It's drawing 0.48amps on 15v. Nothing changes when I power it on. Will try to change the CMD resistor and see

Post automatically merged: May 31, 2023

Thanks for the time you spent writing this. Actually, it sounds very reasonable to suspect CLK. I don't usually take much time on it, but because my soldering station is new I took more than I should to throw solder in there. When this code appeared, I resoldered it specifically as I thought it was the problem, but still. Regarding the diode mode, It's giving 0.7v.

After second thought and after reading Abal analysis, I think that CMD is Ok. Leave that be for now.

 

[Phantomas77]

What's the best way to go about fitting a wire underneath the emmc? I tried with 34 awg magnet wire but it was just a little too thick and raised the corner of the emmc.

Solder balls or paste?

I've just been practicing with donor boards and can consistently reball the emmc but not with a wire attached.

You should use thinner wire, AWG40 or AWG42 would be more suited. Otherwise, you can use these.

One trick that I learned from an experienced repair person: do you have a pair of those old cheap headphones that are broken or you don't use them anymore? The enamel wire inside the cable is very thin and it's perfect for this type of mod

 

[MAI87]

Good again. I have a friend with a pretty weird or complicated problem and we can't get this nintendo switch back so far. I explain the situation.

The switch worked fine with picofly until the day I took it out of its base after playing it fell to the ground and since then when I turn on the switch it shows a purple screen but with an error on the screen.

missing lpo sleep lib missing minerva lib/
update bootloader folder/
press any key...

if we remove the memory card we get the no memory card icon, but if we put it back we get the purple error screen again.

if we remove the installation we get a blue screen and it is not possible to start the system at all.

tests we have tested:

changed several memory cards = no success
tried different updated files latest version = no success
tried different card reader = no success
has done a reflow to gpu drive = no success
you have checked the female connector and made some measurements on the lines it seems that everything is fine here = no success

We don't want to run dangerous tests until the last minute to avoid possible further damage or to get to a worse point than we are now. We think that the problem may possibly be in the Nand or Ram memories.
what do you think? where do you think the problem is here?
Where would you go to carry out new tests or new measures to rule out different areas?
I hope you can help or give some advice and we can bring this nintendo switch back to life, thank you all in advance.

 

[QuiTim]

Good again. I have a friend with a pretty weird or complicated problem and we can't get this nintendo switch back so far. I explain the situation.

The switch worked fine with picofly until the day I took it out of its base after playing it fell to the ground and since then when I turn on the switch it shows a purple screen but with an error on the screen.

missing lpo sleep lib missing minerva lib/
update bootloader folder/
press any key...

if we remove the memory card we get the no memory card icon, but if we put it back we get the purple error screen again.

if we remove the installation we get a blue screen and it is not possible to start the system at all.

tests we have tested:

changed several memory cards = no success
tried different updated files latest version = no success
tried different card reader = no success
has done a reflow to gpu drive = no success
you have checked the female connector and made some measurements on the lines it seems that everything is fine here = no success

We don't want to run dangerous tests until the last minute to avoid possible further damage or to get to a worse point than we are now. We think that the problem may possibly be in the Nand or Ram memories.
what do you think? where do you think the problem is here?
Where would you go to carry out new tests or new measures to rule out different areas?
I hope you can help or give some advice and we can bring this nintendo switch back to life, thank you all in advance.

Easiest test to do is this:
- Remove cover and shield and remove the pico and all the wires
- Press hard and I mean quite hard on RAM module while starting the Switch
- If it starts OK then you need to refllow RAM.
- If it does not start repeate the procedure but this time press hard on APU.
- If it starts you need to refllow APU (good luck with that)
- If it does not, we need to investigate further

 

[CarlosGo]

Hello everyone. This works? I saw him in a group. I have a flex v3 and I want to install it on a lite. Could this work so that the cable does not get in the way so much?

 

[MAI87]

Easiest test to do is this:
- Remove cover and shield and remove the pico and all the wires
- Press hard and I mean quite hard on RAM module while starting the Switch
- If it starts OK then you need to refllow RAM.
- If it does not start repeate the procedure but this time press hard on APU.
- If it starts you need to refllow APU (good luck with that)
- If it does not, we need to investigate further

Sorry. The pressure tests have already been carried out previously without any success. My friend is from another country, I can't be with him I don't know if I'm applying enough pressure here either. I'll tell him to try all this again just in case before continuing to try other options. Thank you

Post automatically merged: May 31, 2023

Hello everyone. This works? I saw him in a group. I have a flex v3 and I want to install it on a lite. Could this work so that the cable does not get in the way so much?

If I'm not mistaken, I think the cable should be soldered here. let the companions confirm if this is correct.

 

[abal1000x]

.

 

[FreeLander]

After second thought and after reading Abal analysis, I think that CMD is Ok. Leave that be for now.

Alright. I removed everything and prepared a fresh rp2040 and wiring. Will do it all over again. CLK was a tough one on this OLED, but when I got a diode reading I thought it was all good.

Will see how it goes, reporting back later tonight.

 

[MAI87]

Already tried reinstall the hekate?

yes, it has changed all the new files and sd card and always the same result.

 

[abal1000x]

.

 

[POPOLO]

Sorry. The pressure tests have already been carried out previously without any success. My friend is from another country, I can't be with him I don't know if I'm applying enough pressure here either. I'll tell him to try all this again just in case before continuing to try other options. Thank you

Post automatically merged: May 31, 2023

correct

 

[MAI87]

https://gbatemp.net/threads/picofly-aio-thread.628951/post-10170289

Seems like the same issue with that.
Have you tried that ums-loader?

It seems that this problem is not the same, although it looks a bit similar. The name has no error nor does it have a purple screen when executing hekate. Even so, I keep it in mind and I will tell my friend to review this in case it helps anything. thank you

 

[QuiTim]

Good again. I have a friend with a pretty weird or complicated problem and we can't get this nintendo switch back so far. I explain the situation.

The switch worked fine with picofly until the day I took it out of its base after playing it fell to the ground and since then when I turn on the switch it shows a purple screen but with an error on the screen.

missing lpo sleep lib missing minerva lib/
update bootloader folder/
press any key...

if we remove the memory card we get the no memory card icon, but if we put it back we get the purple error screen again.

if we remove the installation we get a blue screen and it is not possible to start the system at all.

tests we have tested:

changed several memory cards = no success
tried different updated files latest version = no success
tried different card reader = no success
has done a reflow to gpu drive = no success
you have checked the female connector and made some measurements on the lines it seems that everything is fine here = no success

We don't want to run dangerous tests until the last minute to avoid possible further damage or to get to a worse point than we are now. We think that the problem may possibly be in the Nand or Ram memories.
what do you think? where do you think the problem is here?
Where would you go to carry out new tests or new measures to rule out different areas?
I hope you can help or give some advice and we can bring this nintendo switch back to life, thank you all in advance.

Try this: https://gbatemp.net/threads/hekate-v4-6-bootloader-library-sleep-mode-is-missing.528305/post-8935862

 

[themine1500]

PI2040-Zero PIN13 -> SCL PIN12 ->SDA

Ahh thank you!

Post automatically merged: May 31, 2023

Hello everyone. This works? I saw him in a group. I have a flex v3 and I want to install it on a lite. Could this work so that the cable does not get in the way so much?

Yes this works.. I’ve done it 2x with lite and works fine. It was v2 flex I used though. Modzville posted that to Twitter.

 

[AceCuba]

Ey guys what is the BEST place fpr 3.3v on v1 and V2 . I dont like That little cap on de power ic

Enviado desde mi Redmi Note 8 mediante Tapatalk

 

[themine1500]

That blue flashing means glitching.

The first time pico run, it will try the best parameter for the glitching. It will do that normally around 10seconds.
It will stopped when the parameter found. If theres error than there will be yellow blinking the error code.

In your video, its glitching. Its normal, just wait until its done. The parameter will be saved, and the next boot it will be faster.

If it doesn't found it will timeout and throws the yellow code of timeout.
=== Glitch attempt limit reached, cannot glitch

I've meet this timeout glitch error in oled, shifting the dat0 adapter little bit and its works.

I let it run into it turned off and never got any yellow lights. I will try again today. I think it may be the dat0.. it is one with 3 tabs. I think the tab that goes over to the cap is blocking me from getting it in the right place. I have it over against the cap.. I’ve seen older video where sthetix cuts the small extra piece off that tab.. I will try that.