Hacking Hardware Picofly - a HWFLY switch modchip

[aguskaryana]

Anda dapat mencoba memeriksa resistansi tanpa mengangkat resistor, seharusnya sekitar 4,7k.

how do you check it?

 

[abal1000x]

how do you check it?

If you couldn't even know how to measure resistance. Then i am reluctant to give further advise. It could be dangerous for a noob to do this. You might broke your device more then what it should be. At least you should know the basics about electronic.

 

[aguskaryana]

Jika Anda bahkan tidak tahu cara mengukur resistansi. Kemudian saya enggan memberikan saran lebih lanjut. Bisa berbahaya bagi seorang noob untuk melakukan ini. Anda mungkin merusak perangkat Anda lebih dari yang seharusnya. Setidaknya Anda harus mengetahui dasar-dasar tentang elektronik.

i know how to measure resistance but which resistance should i measure?

 

[cgtchy0412]

i know how to measure resistance but which resistance should i measure?

Ukur 2 kali, yg hijau harus 0 ohm.. yang merah harus 4.7Kohm.

 

[Hassal]

Agreed. I will follow the three-level tutorial Stheix did and see how it goes. You've always warned us on those adaptors. I'm glad I'm progressing in my reballing skills. Thanks again.

I'm quite skeptical about this being the fault of the adapters alone, there's just too much of a gap if you're using the right adapters. I think some consoles come faulty from factory where the balls are close enough to make a bridge between them. There's no way to conclusively tell until you x-ray these chips to see what is happening underneath them.

 

[Dee87]

I mean I heard a lot bullshit about the picofly
But this dude tops it

I maybe think even it is a bug of the 2.73 fw. I had the fw 2.73 fw installed on one china board and boom i had the purple screen and my cmd was damaged. Replaced it, no purple screen anymore. Then tried again 2.74 on another chip worked. I dont want to test if install the 2.74 on the old chip if it works because if not it could screw my cmd again

 

[Switxh]

I'm quite skeptical about this being the fault of the adapters alone, there's just too much of a gap if you're using the right adapters. I think some consoles come faulty from factory where the balls are close enough to make a bridge between them. There's no way to conclusively tell until you x-ray these chips to see what is happening underneath them.

It's very easy to insert the contact between dat0 and dat1 even with the "right adapter". The contact is more than big enough. There's nothing wrong with the EMMC from the factory you guys need to stop looking for outside sources to blame and just accept when you've made a mistake.

 

[abal1000x]

I mean I heard a lot bullshit about the picofly
But this dude tops it

People in past saying that lightning comes from Zeus.
Lack of understanding, will make people claims something 'weird'.

I encourage people to understanding the 'how-tos' of this glitch worked. When rehius open source the firmware, its like an euphoria for me. Because from the code i could understand the cloud of the mystery of how this works.

 

[Dee87]

People in past saying that lightning comes from Zeus.
Lack of understanding, will make people claims something 'weird'.

I encourage people to understanding the 'how-tos' of this glitch worked. When rehius open source the firmware, its like an euphoria for me. Because from the code i could understand the cloud of the mystery of how this works.

sadly im not really good with coding but i did understand some.

what i just dont get is people saying the picofly is no good it only gives u issues. when i have no installed 100+ without any issues well at the beginning i did also have one issue with the cmd but that was my fault.

why is it so hard for people to just admit making mistake a instead saying its somethings elses fault , i just dont get it we all are humans and we make mistakes thats normal.

 

[QuiTim]

Nice looking good :-)

What was ur issue with the stencil?

U have to put magnet or something around the emmc so the rest of the stencil is laying on something so it doesn't have that much chance of bending.

Then add ur solder paste , if u have enough set ur tweezers in the middle of the stencil and push it on the emmc.

Then set ur hot air at 180 a bit more then ur pastes melting point and slowly best ur stencil up don't go straight to the stencil so its not a heat schock otherwise it Gonna bend really fast due to the heat.
Wait till everything is melted but always keep pressure on the middle of the stencil.

After that I usually add a bit flux heat it up one more time and then take the stencil off

Well I did pretty much everything like you described but I think my problem was that the temperature of the air was too high and it experienced the heat shock that you describe. I guess this was just me being impatient and trying to do this in the middle of the night after a bussy day at work but I'm sticking with it and eventually I will find the sweet spot and develop some consistency. Thank you for everything.

 

[Hassal]

It's very easy to insert the contact between dat0 and dat1 even with the "right adapter". The contact is more than big enough. There's nothing wrong with the EMMC from the factory you guys need to stop looking for outside sources to blame and just accept when you've made a mistake.

I would like to be proven wrong but testing this on a naked chip makes it near impossible to bridge these two point especially if this adapter is pushed to the far right. I'm not making a claim without proof. I just think there's more at play here than just the adapter.

 

[FreeLander]

I'm quite skeptical about this being the fault of the adapters alone, there's just too much of a gap if you're using the right adapters. I think some consoles come faulty from factory where the balls are close enough to make a bridge between them. There's no way to conclusively tell until you x-ray these chips to see what is happening underneath them.

Got it. I have never had such issue. I've used this adaptor on 10+ OLEDs. I'll open it up once arrives.

Post automatically merged: Aug 15, 2023

I finally got my hands on the Zelda Oled that I ordered 3 months ago and I decided to try and set a proper dat0 by removing the emmc.

A couple of things to point out:
- It was easier than I thought but I've been preparing/practicing for this for quite some time so for anyone reading this I would suggest you do your homework before diving in.
- The amount of beating these chips can handle is quite impressive.
- Same as Takezo-San wrote a couple of post above, I had no luck with stencil and paste (maybe I was stressed and tired because I started doing this at 11pm) so after trying a couple of times and failed I just did it manually one by one with 0.3 balls.
Everything went smoothly afterwards and since the board was already off I also decided to try the abal's mosfet placement on the back of the apu.
I just want to say that this is an amazing thread with some amazing people and the list of people to say thank you is way too long but I have to point out @Dee87 and @Takezo-San without whom I probably would have not had the courage/will to go through with this process and a special thanks to @abal1000x who’s dedication and contribution to this whole mod is priceless.
Below you can see some pictures, I should have done more but I got carried away in the heat of the moment
Edit: It goes without saying that without Rehius none of this would even be possible so...

Looks amazing. I'm currently practicing reballing on a lite. I recommend this AMAOE reballing kit. It really does the work for you. No holding the stencil with forceps or anything like that. I'm too lazy that I also got a heat gun holder (powerful combination)

With my setup, these balls get created while I'm preparing the wicks and tinning my iron.

 

[twins333]

Then tell me why my china board rp2040 just destroyed my cmd resistor, which was working completely fine with the hwfly chip. Im not talking about your blue chips, im talking in another thread about a china board.

It seems to me that you already have the answers in your question. You used an unsupported clone.

 

[Switxh]

I would like to be proven wrong but testing this on a naked chip makes it near impossible to bridge these two point especially if this adapter is pushed to the far right. I'm not making a claim without proof. I just think there's more at play here than just the adapter.

Solder balls are larger in diameter than the pads on their own. This isn't something that needs proving. Everyone knows you can bridge dat1 and dat0 with the adapters. It's the cause of 90% of these purple screens you see in this thread. You should do some reading of the older pages and catch up on knowledge.

 

[abal1000x]

Then tell me why my china board rp2040 just destroyed my cmd resistor, which was working completely fine with the hwfly chip. Im not talking about your blue chips, im talking in another thread about a china board.

First i don't know the diagram of your china board rp2040
Secondly, i don't know how you installed it, how you handle your equipment, and i don't know the quality of your equipment.

If you follow this thread, theres people saying cmd resistor broken. And i made a theory of why this could happened. Even i push the hypothesis that the cmd is destroyed the picofly send too high current to the 0201 4.7k cmd pullup resistor on the motherboard. And someone refute is, and he correct. If we calculate with 1.8V its only 0.38mA which is okay for 0201 max rated current.

I mean you need to explain it in logical sense when you made a theory. How could the picofly destroy the cmd line, when the current it flows only 0.38mA maxed?

 

[Motordriver122]

First i don't know the diagram of your china board rp2040
Secondly, i don't know how you installed it, how you handle your equipment, and i don't know the quality of your equipment.

If you follow this thread, theres people saying cmd resistor broken. And i made a theory of why this could happened. Even i push the hypothesis that the cmd is destroyed the picofly send too high current to the 0201 4.7k cmd pullup resistor on the motherboard. And someone refute is, and he correct. If we calculate with 1.8V its only 0.38mA which is okay for 0201 max rated current.

I mean you need to explain it in logical sense when you made a theory. How could the picofly destroy the cmd line, when the current it flows only 0.38mA maxed?

Thanks for giving these values. I didnt measure the output of this clone, so i cannot say that the output is 0.38mA max. I only know that i replaced the cmd resistor on the switch and it is working again. As i said i only swapped the clone one with the previous installed hwfly chip, i didnt even touch the cmd resistor. But the easiest solution is just not using the clone again. I dont want to make a science out of it when it is not necessary

 

[abal1000x]

Thanks for giving these values. I didnt measure the output of this clone, so i cannot say that the output is 0.38mA max. I only know that i replaced the cmd resistor on the switch and it is working again. As i said i only swapped the clone one with the previous installed hwfly chip, i didnt even touch the cmd resistor. But the easiest solution is just not using the clone again. I dont want to make a science out of it when it is not necessary

The best theory right now is the solder the installer used, might leaked electricity. Cheap solder usually like this. And this electricity have high enough power to destroy the cmd resistor on the board, when people soldering the cmd point.

But i don't know the validity of this theory, since i use a 'good' solder station. And no one who experience the broken cmd, share information about their solder station quality.

 

[Hassal]

Got it. I have never had such issue. I've used this adaptor on 10+ OLEDs. I'll open it up once arrives.

To put an end to these claims that these adapters are the main cause of bridges I took a piece of transparent acrylic then reballed and seated the chip on top with a thin layer of adhesive film to hold it in place. And lo and behold, not once did I touch D1 even when doing it blinded. There is a case though when I pushed in aggressively it wrapped.

 

[QuiTim]

indeed, but then tell me why the fw 2.74 worked so far on 3 clones well, and the only one with fw 2.73 did not. Only that is why i suggest that the fw 2.73 does not work well on these clones.

And what is the purpose of quoting me then here at all if he knows that i used a clone? I would be thankful that he could prove to me that there is no possible way that fw 2.73 can malfunction on china boards. But that might be too high for some people, laughing might be easier

The thing here is that there is no reason for any firmware from this thread to work on any AliExpress boards.
This only goes to prove what has been said all along in this thread "they stole the firmware from here, made crappy clones and they dont give any support after sales" because how can they give support for something that they did not make...
On the supported boards (listed on this thread and the guide by lightninjay) the 2.73 works great and it's actually the 2.74 that had issues and that why it was pulled (by the developer).
So the headline here is not 2.73 has isues on aliexpress boards but it's AliExpress boards have random issues because they use a firmware that was not written for them - stay clear

 

[Motordriver122]

The best theory right now is the solder the installer used, might leaked electricity. Cheap solder usually like this. And this electricity have high enough power to destroy the cmd resistor on the board, when people soldering the cmd point.

But i don't know the validity of this theory, since i use a 'good' solder station. And no one who experience the broken cmd, share information about their solder station quality.

Do you maybe know the behaviour of solder flux on that? maybe if there is some left, could it cause also such issues?

Post automatically merged: Aug 15, 2023

Post automatically merged: Aug 15, 2023

The thing here is that there is no reason for any firmware from this thread to work on any AliExpress boards.
This only goes to prove what has been said all along in this thread "they stole the firmware from here, made crappy clones and they dont give any support after sales" because how can they give support for something that they did not make...
On the supported boards (listed on this thread and the guide by lightninjay) the 2.73 works great and it's actually the 2.74 that had issues and that why it was pulled (by the developer).
So the headline here is not 2.73 has isues on aliexpress boards but it's AliExpress boards have random issues because they use a firmware that was not written for them - stay clear

yes indeed, thats why i only talk about the clone ones. Still, my headline was just made by my experience, the second one you said is just more generalized and does not exclude mine