Hacking Hardware Picofly - a HWFLY switch modchip

[Jagovi]

Yes, if those 2 caps are damaged/missing/broken then Wifi will not be working.
https://gbatemp.net/threads/no-wifi-after-hwfly.625830/

A long time ago someone mentioned that he messed up Wifi because he damaged some traces when removing the shielding.
In his case the caps were still present/working.

It would be curious because it was working for several hours until it froze, and upon restarting, the error appeared. I would rule out physical damage from scratched traces, but I will cut the processor shield frame to see if that is the issue. Thank you very much!!!

 

[thesjaakspoiler]

It would be curious because it was working for several hours until it froze, and upon restarting, the error appeared. I would rule out physical damage from scratched traces, but I will cut the processor shield frame to see if that is the issue. Thank you very much!!!

You don't need to cut the shield. Just look for damage around it.

 

[Jagovi]

You don't need to cut the shield. Just look for damage around it.

It is clean no traces damaged . It must be something related to the touchscreen malfunction. Some component that links the touchscreen and the Wi-Fi must be shorted or something similar, since both the Wi-Fi and the touchscreen stopped working at the same time. The touchscreen is already fixed, but I can't locate the fault with the Wi-Fi.

 

[steverebo]

just installed my picofly on my switch oled and Im getting blue light short yellow long yellow, this shows as CMD nor recognised but I have gone back in and checked my connections with a multi meter and they all read 0.65v so the connection is present. What could cause this?

Switch boots absolutely fine OFW with absolutely no issues

 

[0DanDan0]

You know you're responding to a post from almost two years ago right?

Dang didn't realize lol im a bozo

 

[Nontre]

Put 47 ohms on dat0 and cmd point. If you still have the issue try with 100ohms instead.

I fixed one like that.

It keeps working randomly for me, at least I need to completely restart the console more than 20-30 times until it finally goes into Hekate.

At first when I log in for the first time, I couldn't access the emmc memory, it gave me the typical "emmc in slow mode" error. But after a couple of hours, suddenly it read the content correctly and I was able to make backups.

I am using a picofly v2 from Aliexpress. Does what you're telling me have anything to do with this?

In the installation, could using AWG30 or AWG36 cable have anything to do with it? of many or a single strand?

What worries me is that when I uninstalled the entire modkit it did not boot from the emmc. The console just wouldn't turn on. I know that the eMMC memory is not corrupt because I have even been able to boot from hekate in OFW mode. And I have been able to create an emunand from the sysnand.

 

[Danook28]

Any One know the (sda) and (scl) size resistor on oled motherboard and value coz it damge from the point that soldring and the console not tearning on black secreen.

 

[jkyoho]

Any One know the (sda) and (scl) size resistor on oled motherboard and value coz it damge from the point that soldring and the console not tearning on black secreen.

1k ohm from that measurement picture

 

[Danook28]

1k ohm from that measurement picture

You know the size is it 0201???

 

[jkyoho]

You know the size is it 0201???

yes 0201

 

[Jagovi]

It is clean no traces damaged . It must be something related to the touchscreen malfunction. Some component that links the touchscreen and the Wi-Fi must be shorted or something similar, since both the Wi-Fi and the touchscreen stopped working at the same time. The touchscreen is already fixed, but I can't locate the fault with the Wi-Fi.

Any suggestions ? Thanks

 

[Danook28]

yes 0201

God bless you my console working thank you so much for help me.

I have oled sp1 sp2 low voltage when the chip glitch white blue then blue then 4 time yellow light boot ofw.i use 2.74 and 2.73 and 2.75 all same one chip on lite console working fine i put it in oled it is failed.

 

[Takezo-San]

Hey y'all, whats good in the hood. Had an issue I'm needing help with. I ran into this problem with 2 SKHynix OLEDS. Not sure if its exclusively the issue with these SK EMMC type OLEDS for sure though.

BTW I always do a reball on the EMMC and use a 40awg wire on the dat0 trace. 40awg on the CLK, 32awg on the CMD and RST, and 30awg on the ground and 3.3v. Usually use a mosfet for CPU line. So thats my setup on things.

But its like after 3months, the switchs start to freeze on booting into atmosphere and or randomly won't wake from sleep. No ofw either. Just pure dead and symptoms of a dead EMMC. I tried switching over to the V6s chip but its the same story (and on a different console too). I figured out the CMD line shows a low resistance and all other lines show normal voltage drops and funnily enough, a reflow of the EMMC fixes everything. But its good for another 3 months and then it happens again!? Anyone ran into this issue or had this issue? Any permanent fix? I did try reading the backlog of this mammoth thread to hunt a repeat of this issue by anyone but its been an hour of reading already XD and no cigar yet. Am I doing things in an antiquated way? Has the switch mod scene moved on that much?

Appreciate the help.

 

[thesjaakspoiler]

I figured out the CMD line shows a low resistance and all other lines show normal voltage drops and funnily enough, a reflow of the EMMC fixes everything

A while back someone mentioned that the flux he was using was transforming into this nasty sticky goo within several months, causing some major chemical conductive circus under his emmc.
There have been some discussions on what flux to use and how to clean up afterwards.

 

[Phantomas77]

...I figured out the CMD line shows a low resistance and all other lines show normal voltage drops and funnily enough, a reflow of the EMMC fixes everything.

What type of solder paste are you using for the reball and what's your cleaning procedure for the old solder?

If the cleaning of the old solder is not done correctly (e.g. not removed completely) and let's presume that you use Sn/Pb 63/37 solder paste for the reballing, there could happen an alloy mix that after some time due to heat exposure could create cracks within the solder balls and the contact between PCB and eMMC would brake.

Try this procedure instead and clean nicely with IPA in the middle steps:

• remove eMMC with hot air
• PCB side:
  • add a small amount of Sn/Pb 63/37 on top of the existing solder and mix everything with iron
  • add flux and wick the solder mix
  • use a very small amount of the same Sn/Pb 63/37 for fixing the DAT0 wire
• eMMC side:
  • add a small amount of Sn/Pb 63/37 on top of the existing solder and mix everything with iron
  • add flux and wick the solder mix
  • use Sn/Pb 63/37 paste for reballing
• Apply flux to the PCB eMMC area (small amount)
• Put everything together with hot air
• Clean eMMC and sides with abundant amount of IPA
• (Optional) use compressed air bottle to remove rest of IPA and flux between PCB and eMMC

 

[Takezo-San]

What type of solder paste are you using for the reball and what's your cleaning procedure for the old solder?

If the cleaning of the old solder is not done correctly (e.g. not removed completely) and let's presume that you use Sn/Pb 63/37 solder paste for the reballing, there could happen an alloy mix that after some time due to heat exposure could create cracks within the solder balls and the contact between PCB and eMMC would brake.

Try this procedure instead and clean nicely with IPA in the middle steps:

• remove eMMC with hot air
• PCB side:
  • add a small amount of Sn/Pb 63/37 on top of the existing solder and mix everything with iron
  • add flux and wick the solder mix
  • use a very small amount of the same Sn/Pb 63/37 for fixing the DAT0 wire
• eMMC side:
  • add a small amount of Sn/Pb 63/37 on top of the existing solder and mix everything with iron
  • add flux and wick the solder mix
  • use Sn/Pb 63/37 paste for reballing
• Apply flux to the PCB eMMC area (small amount)
• Put everything together with hot air
• Clean eMMC and sides with abundant amount of IPA
• (Optional) use compressed air bottle to remove rest of IPA and flux between PCB and eMMC

Hey @Phantomas77, good to see you again.

I don't use solder paste. My results were to inconsistent with paste and I found it too finicky. Instead I use 0.3mm solder balls (mechanic brand 63/37), dump them on a jig, do the heat cycles. Works the best for me and no residual paste between the grid arrays.

However, I did used to use Kingbo flux but know I'm thinking of getting a no clean flux like 'Chipquik no clean' or something. I think its the flux thats probably forming a conductive goo like @thesjaakspoiler mentioned above.

I'm super diligent in using IPA99 between the steps of the Reballing procedure and my work is super clean. However, I cannot, obliviously, reach the underside of the seated EMMC so I'm guessing theres some unwanted conducting going on after a period of time thats causing a short on the CMD line of the EMMC. And my fix of a reflow might be proof that that gooey stuff when going back to molten, spreads around more evenly again!? Flux seems to be where the roads lead.

You think thats an accurate diagnosis and that I should use no clean from now on!?

Again, appreciate the help on this issue.

Edit - Or in future, is it time to bite the bullet and go with Kamikaze!?
Edit - Do you think after a reball an ultra sonic clean would/should be recommended?

 

[Phantomas77]

Personal thought: I won't assume that in this case flux is the culprit, but yeah, even ChipQuick "no clean" if not properly cleaned could create some conductivity (there was a guy on youtube who done some tests). It could be as well the micro-cracks on the solder joints, since after reball the Switch works.

You can still use Sn/Pb 63/37 paste/wire in order to mix it with "no lead" solder with the iron before wicking everything off, enriching in this case the Pb mixture while applying the balls.

Regarding flux, I've used ChipQuick's SMD291 in the past with good results, but recently I'm more into the viscosity of MG Chemical 8341.

The compressed air with a nozzle pointed between eMMC and PCB could help to remove the excess of IPA and major amount of the flux left --> impregnate well the sides of eMMC with IPA and then spray it off.

 

[Takezo-San]

Personal thought: I won't assume that in this case flux is the culprit, but yeah, even ChipQuick "no clean" if not properly cleaned could create some conductivity (there was a guy on youtube who done some tests). It could be as well the micro-cracks on the solder joints, since after reball the Switch works.

You can still use Sn/Pb 63/37 paste/wire in order to mix it with "no lead" solder with the iron before wicking everything off, enriching in this case the Pb mixture while applying the balls.

Regarding flux, I've used ChipQuick's SMD291 in the past with good results, but recently I'm more into the viscosity of MG Chemical 8341.

The compressed air with a nozzle pointed between eMMC and PCB could help to remove the excess of IPA and major amount of the flux left --> impregnate well the sides of eMMC with IPA and then spray it off.

Oh yeah, forgot to mention. I use low melt solder to get the initial solder off the chip and the board (wick and iron). Maybe I should do the pb mix at the end of the wicking process to enrich the pads with lead!?

Is the MG chemical 8341 less viscous or more than the ChipQuick?

And will try the compressed air with IPA99 on the sides. Is it to effectively force in the IPA underneath the chip and out the side, right!?

Its happened twice now, this issue and Im really contemplating going the Kamikaze route or a sonic cleaner.

 

[LuigiGad]

Oh yeah, forgot to mention. I use low melt solder to get the initial solder off the chip and the board (wick and iron). Maybe I should do the pb mix at the end of the wicking process to enrich the pads with lead!?

Is the MG chemical 8341 less viscous or more than the ChipQuick?

And will try the compressed air with IPA99 on the sides. Is it to effectively force in the IPA underneath the chip and out the side, right!?

Its happened twice now, this issue and Im really contemplating going the Kamikaze route or a sonic cleaner.

you perform the kamikaze, you don't stress any components and you have the same reliability as a mariko.

 

[Phantomas77]

Oh yeah, forgot to mention. I use low melt solder to get the initial solder off the chip and the board (wick and iron).

Well, you just solved your problem, low-melt residue is detaching/cracking the solder balls. Do not use it, especially for applications related with heat. Apply 63/37, wick-off and you're good.

I find the 8341 a bit more flowing and seem to leave less residue in comparison with SMD291, but either of these are quite good. Order from official reseller, there're many fakes on the market.