Hacking Hardware Picofly - a HWFLY switch modchip

[RatchetRussian]

So it sounds like the back mosfet has a similar problem to the processor mosfet install which is some switches might require more current (2 mosfets)?

Would there be a benefit then to using a stronger mosfet on the back or even dual mosfets?

 

[chronoss]

Someone test a Tiny chip, it's ok now and have a good resistance ?

 

[abal1000x]

Does anyone know what causes the Picofly to reset training data and swap firmware? I did an OLED today and initially I flashed the Pico Zero over USB with 2.73 but after a few reboots it was randomly resetting the training data. I flashed 2.74 using the picofly toolbox because it was the version I had on my testing SD card but now what happens is it will randomly reset training data and then ALSO swap firmwares between 2.73 and 2.74. It happens both ways too. It will be on 2.73 and I reboot it a few times then it will reset training and swap to 2.74. Then I do the same checking glitch stability and it will reset training data AGAIN and swap back to 2.73.

Any idea what this could be? Is it a bad Pico zero maybe? When it's glitching it glitches fast on both firmwares but I just cant get more than around 10 reboots in to test stability before it resets it's training and swaps firmware back and fourth.

Resetting the statistics happened when all the glitch trial failed. It will followed by === led light.

For me this error occured when the dat0 adapter are not stably connected.

 

[jkyoho]

Someone test a Tiny chip, it's ok now and have a good resistance ?

https://gbatemp.net/threads/picofly-a-hwfly-switch-modchip.622701/post-10189577

 

[abal1000x]

Excited to be trying your back-mosfet method @abal1000x

This is my first attempt and it is on a lite. I am getting the mosfet error (==*) and booting into OFW. Maybe there is not enough solder on the mosfet?

Yeah thats not enough especially the D part.
The better method is you connect the D part to the center pad.

It could also because of your solder material you use to join the component.
The conductivity (/cm Ω)
Sn = 91.7k
Pb = 48.1k
Higher better. Ag have the best conductivity, 630k.

No-lead (low pb) solder have a better conductivity lower resistance. Also good for your health, but difficult to work on. Since in back mosfet method you need not worry on size and heat, then using no-lead solder is recommended.

In previous i use leaded solder in the apu's cap, since its small and suck the heat. So to reduce the difficulty i use the leaded solder. I used to hold my breath, and turn on the fan to suck the leaded gas, to avoid it into going my lung.

 

[chronoss]

https://gbatemp.net/threads/picofly-a-hwfly-switch-modchip.622701/post-10189577

Thanks but the guide is the pdf or the one on op post ?

 

[abal1000x]

So it sounds like the back mosfet has a similar problem to the processor mosfet install which is some switches might require more current (2 mosfets)?

Would there be a benefit then to using a stronger mosfet on the back or even dual mosfets?

View attachment 382252

In my logical brain, using two mosfet for the same point is useless.
Its equal to adding the maximum current could flows which means its equal to widening the power rail of 1 mosfet.

 

[Dee87]

Someone test a Tiny chip, it's ok now and have a good resistance ?

The 3 I have from amazon have correct resistors, I added 100ohm to dat0 though

 

[chronoss]

The 3 I have from amazon have correct resistors, I added 100ohm to dat0 though

12 a 14€

 

[Dee87]

12 a 14€

Nooo I pay 8.99€ each without flash label.and 11.99 with flash kabel.

Next day delivery.

I did order 10 from aliexpress with my last order but it went back after there was apparently no name on the doorbell .

So that's not even 10€ for everything I need for the pico install

 

[roxzii]

If you have manual air pump
View attachment 382141

Or usb vaccum cleaner you might use it to clean debris i.e below bga.


You could also check when booting to manuallly press the dram, or the cpu, or the emmc carefully. If the issue gone then you found your problem.

Okay, turned pico off again (3.3v and ground) to boot to OFW. Tried pressing all those parts to no avail, always blue screen. Also tried to clean the traces between the APU and the DRAM modules, same results, I have high res pictures attached. Not the best since the pico is in the way but I think all the traces are visible. (Btw pay no attention to the botched capacitor on the apu :c)

 

[jkyoho]

Okay, turned pico off again (3.3v and ground) to boot to OFW. Tried pressing all those parts to no avail, always blue screen. Also tried to clean the traces between the APU and the DRAM modules, same results, I have high res pictures attached. Not the best since the pico is in the way but I think all the traces are visible. (Btw pay no attention to the botched capacitor on the apu :c)

what happened to your CMD resistor?

 

[Dee87]

what happened to your CMD resistor?

really good eye u have there i didnt see that when i looked at his picture Xd

 

[jkyoho]

really good eye u have there i didnt see that when i looked at his picture Xd

lol
I doubt the clk cable(exposed part) is too close to ground pad from shield base as well

 

[roxzii]

what happened to your CMD resistor?

Well, the CMD point was giving me problems sticking (30awg I know) and I ended up knocking it off and losing it. Since I could only find that giant replacement I made that work, and it solved my purple screen. Now "just" the blue

 

[cgtchy0412]

Yeah add some more solder , that's what worked for me

This "add solder" thing is mostly mosfet dependent, as i use aon i only connect mosfet and caps using awg30 core wire(so no solder blob) and it doesnt have any issue to glitch.

 

[roxzii]

lol
I doubt the clk cable(exposed part) is too close to ground pad from shield base as well

Yeah, since I don't have solder mask yet I keep checking the value each time I turn on the pi to make sure it's not contacting ground

 

[abal1000x]

This "add solder" thing is mostly mosfet dependent, as i use aon i only connect mosfet and caps using awg30 core wire(so no solder blob) and it doesnt have any issue to glitch.

price wise the aon is pricier than 8342

 

[jkyoho]

Well, the CMD point was giving me problems sticking (30awg I know) and I ended up knocking it off and losing it. Since I could only find that giant replacement I made that work, and it solved my purple screen. Now "just" the blue

Are you able to go into Hekate though?

 

[abal1000x]

Okay, turned pico off again (3.3v and ground) to boot to OFW. Tried pressing all those parts to no avail, always blue screen. Also tried to clean the traces between the APU and the DRAM modules, same results, I have high res pictures attached. Not the best since the pico is in the way but I think all the traces are visible. (Btw pay no attention to the botched capacitor on the apu :c)

i've no other idea.

if using ums-loader work, then the cpu highly probable okay. if the backup raw emmc works via hekate, than the emmc an okay. Left with dram as the minimum systom to boot, which might be problematic.