Hacking Homebrew Failed (?) Picofly install on Switch Lite. Please help!

summer_star

Member
Newcomer
Joined
Mar 15, 2016
Messages
12
Reaction score
0
Trophies
1
Age
26
XP
128
Country
Greece
Hi all,

I modded a Switch Lite yesterday with a Picofly chip. I've succesfully modded a V1 switch with a Picofly before with ease so I thought I could do this.

Long story short, the install actually went well. I didn't struggle much and my solder joints look mostly good. Not professional, but good. Here are some screenshots.

SP1:
1786646771025.png


SP2 (less good but okay I think):
1786646905651.png


3v3. A bit too much solder on the anchor but its fine:
1786647284223.png


C/D/A:
1786647909470.png


B (Again, a bit too much solder, but that shouldn't cause problems):
1786648206456.png


For the record, I never left my iron at these points for too long (especially the SP caps) to avoid burning anything out. My iron was at 310 and I used 60/40 solder with Chip Quik SMD flux.

The install works and the Picofly succesfully glitches the Switch. But the Lite never stays running for long. After a while, seemingly at random, it freezes. It sometimes freezes silently. Other times, in a more noisy way—the speaker buzzes and the screen is pixelated and green.

The only thing that, I think, I did wrong during the install is I accidentally left blobs of solder on two resistors. I cleaned up the mess but it is possible their values have been thrown off and this is what causes the freezing. One of those resistors is visible in the B picture. It's value is 8.21k.

The other one is this one. it's value is 100 ohms:
1786650071100.png


If anyone knows what the value of these resistors are (or has some kind of board they can probe to test), that would be massively appreciated.

It's possible that the Switch was also this way to begin with. It was sitting in a drawer for months and I never properly tested it before I did the modchip. But I somehow doubt that's the case.

Any help would be appreciated to save this little Switch. Thank you :)
 

Attachments

  • 1786649757128.png
    1786649757128.png
    440.9 KB · Views: 1
Poor solder joints or a badly seated DAT0 adapter can cause those green artifacts on your screen.
Sometimes the DAT0 will warp a little bit due to heat from the APU and thus cause issues after booting successfully.
Sometimes things look fie but still be causing problems.

With a multimeter in diode mode you can check your work.
Expected Diode Readings : (Black on Point, Red on Ground)

Point A / C / D / DAT0 : 0.40V – 0.80V drop
Point B ( Reset): > 1.0V or OL depending on your multimeter

First check those before doing anything else.
My bet is on the DAT0 adapter as other lines would have given you errors during glitching.

You can always share your findings in the PicoFly thread, where all the experts are (although lately it has become a bit quite after the Switch2 release).
 
  • Like
Reactions: summer_star
Poor solder joints or a badly seated DAT0 adapter can cause those green artifacts on your screen.
Hi. Thanks for your input. It's a Switch Lite so no DAT0 adapter necessary. It could be that the Lite FPC solder joints were not the best. It's quite hard to see whether good joints were made because it's not possible to look under the FPC :( I removed it because of that and the joints were pretty good. I had to suck the solder out with my wick in order to lift the cable from every spot, so I don't think it was a bad joint.

Here is what I have discovered so far.

I've experimented with the unit since yesterday and have discovered some interesting things:
  • The crashes are random and happen everywhere—even inside Hekate.
  • In the middle of an eMMC backup, Hekate crashed with an exception, but it actually let me go on. In Horizon, crashes look pretty severe and the system usually freezes beyond control. I've never even seen Hekate freeze. It gives me exceptions.
  • It completed a "slow" RAM test with a Hekate ram test payload someone gave me from the NH discord server.
  • The next day, it failed the RAM test, and a bunch of "TV snow" artifacts came up during the test. But even though those appeared, the console didn't crash. It kept doing the RAM test—showing errors this time.
So, overall, the console works. But, at random, sometimes with an hour between power ons, it entered this broken state. I find it really interesting that it completed a slow RAM test one day (ran fine for a whole hour) and failed the fast one the next day very swiftly.

I suspect that RAM is fine. If it was bad, it would never test good. The APU also appears to be fine because it does appear to execute instructions correctly. I actually had Claude put together a small cpu test program that executes a range of instructions and checking the output a against known data. I have never seen that fail even when the snowy artefacts appear on screen.

What I think it is

To me, it looks like instability because of a bad component. Not anything major like the APU or RAM though. Those resistors I accidentally touched with the iron could have gone bad and be responsible for this oddity. I'm not sure what they are.

The help which I need

Right now, if anyone can tell me what the values of the resistors I posted on the pictures are, that would be massively appreciated. If it isn't them, I at least know where to focus my attention next. I don't think this console is a goner since it does actually work.

Thank you!!!!

Some Screenshots of the Symptoms
(Crash inside Horizon OS)
1786731211060.png

(Failed RAM test and weird artefacts (menu still controllable!)
1786731241979.png


(Succesful RAM test)
1786731323288.png
 

Site & Scene News

Popular threads in this forum