Cap value around MAX77621A south to APU

That’s good, atleast partially rules out the SoC (though not conclusively)

Hi there,
so, i replaced again the BQ, the M92, Reflow all Max ICs and all the one at the edge on the back of the board, i reflowed EN IC above the Realtek IC, and still issue. From SherrifBuck voltage chart, i have the following :

  • 3.3V = 3V
  • SYS = 3.8V
  • 1V35 = 0
  • 1V1 =
  • 5V =
  • 1V8 alt. = 0
  • 0V8 GPU = 0
  • 1V15 =
  • 0V8 CPU =
  • 1V8 =

I compared with another similar board, which show bluescreen (that’s all i have from the same revision), only difference is :

  • 3.3V = 3.3V
  • SYS = 3.9V
  • 1V35 = 1.35V
  • 1V1 = 1.1V
  • 5V = 0
  • 1V8 alt. = 0
  • 0V8 GPU = 0
  • 1V15 = 1.15V
  • 0V8 CPU = 0.98V
  • 1V8 = 1.8V

From what i can tell about heat, most come from BQ and fuel gauge region (without being specific to the IC, but the heat is in that region), the SOC and weird, from the BGA IC at the very top of the Switch, i don’t even know the purpose of that IC, it is looking like a smaller RAM IC, labeled “GCBRG HAC STD”.

I think i have something, i just can’t tell what exactly, but if i look to this diagram :

First we agree that VBUS can either be 5 or 15V depending on the charger type right ?
But more interesting, is that my TS = 3.5V. This looks really high.
The other voltages are
VSYS = 3.8V
VBUS = 15V (or 5 if i use a 5V smartphone charger)
VBAT = 3.9V
TS = 3.5V

if that helps.

I would avoid randomly reflowing ICs without proper evidence to do so, all you’ll achieve is stressing the IC’s or the board out which may lead to unintentional secondary issues which can make diagnostics later on more difficult :slight_smile:

None of the rials from the main PMIC are being generated, so either it’s a fault with the PMIC itself or the console is not being prompted to boot which implies USB or M92 isses (or the surrounding support components) I would frst attach the power/vol flex cable (or short the pins with a tweezer for pwr button) and press the power button and then check to see if anyof the rails are then present, if not I’d check the two ESD diodes for the USB and check they aren’t shorted to ground, check the fuse near the USB isn’t open, check M92 is soldered properly and check there is no issues on the surrounding components (open resistor for example)

icbw but think it’s related to game carts … maybe authentication of some sort… I forget, this rarely ever fails so unless you can see signs of prior rework or corrosion or unless you measure a short on the nearby bypass caps (though I doubt you will based on your previous measurments elswhere) then don’t think this is your issue

I’ll look into this, can’t remember tbh - what does it measure on your bsod board?

Ok got it, i won’t do it anymore, i promise :slight_smile:

ESD showing wrong value (no 0L in diode mode for the both edge corner opposite to ground), i removed them.
USB already tested with a test board, looks mint, no short between the leg and continuity is perfect.
Fuse is ok, M92 is perfectly soldered, no bad joints, no shorts, all resistor surrounding showing a value (in between 10 and 100K i think), no shorts on the caps.
Replaced PMIC, got additional voltages, now the list is as following :

3.3V = 3.3V
SYS = 3.9V
1V35 = 1.35V
1V1 = 1.1V
5V = 0
1V8 alt. = 0
0V8 GPU = 0
1V15 = 1.15V
0V8 CPU = 0.95V
1V8 = 1.8V

So pretty much in line with my BSOD board, but nothing on the screen even if being prompted to boot.
Around BQ i have the same as before, and the BSOD board is also having a TS of 3.5V.

And suddenly it booted, without anything additional made to the previous post :slight_smile:
i will monitor this, make a good ultrasonic bath because flux is everywhere with all the ICs reworked, and that is one fixed :slight_smile: thanks for the help again

Cool :slight_smile:

Based on your previous voltage measurments when it wasn’t booting, you were getting CPU voltage present, so usually you would see the first Nintendo logo, so perhaps you have an issue with the backlight connector or ribbon (somtimes the backlight ribbon is prone to corrosion or pad damage on the flex) or LCD ribbon or connector issues

That all being said, you were missing GPU voltage which is a bit disconcerting, you would expect that being present after a few seconds after prompting to boot (though if I remember right this rail drops when the console is in sleep/standby so maybe that’s why it was missing in your previous measurments)

Glad to hear you’ve got it up and running and hopefully after a good scrub it will continue to do so :+1:

1 Like

Thanks, i will give a little feedback after 1h Zelda play with it, usually that game is my stability benchmark :slight_smile:

and it went ko, crash with 21001-001, weird behavior with artifacts in the picture, then bluescreen, then back to life, but 0 stability, sometimes i have the backlight on but nothing on the screen, sometimes i don’t have anything on the screen and no backlight, but the console is on (i can hear the menu clicks), sometimes i have the screen but no touch, it is different each and every start :slight_smile:

What’s charging current looking like throughout this?

Would ordinarily be M92 related, though I think in your case you potentially have two issues

I’d check both the backlight and LCD connector and ribbon cables, check there is no bent pins in the connector and there is no breaks or corrsion on the ribbons themselves, check the solder joints are good etc, following prompting to boot try wiggling the LCD backlight ribbon and see if the symptoms change

This could be as a result of LCD related issues and/or Ram or SoC related issues (sometimes in rare cases PMIC issues) if you find the above LCD/BL etc doesn’t change anything, remove the SoC/Ram shield (twist latches outwards with blade all the way round) and put slight pressure on the Ram IC’s and try prompting to boot and see if it resolves the BSOD/Artifacting issues.

Hi there, all good ?
Damn, this is definitely the most difficult Switch i ever encountered in my young fixing console life :rofl:
I think this time it is good, beleive or not, the problem was … 2 broken coils. One on top of the Realtek IC and one on the back near the “AZJ 734” component.
I took them from my BSOD board, but they were different model, looks much more robust. By the way, as i would like to fix this BSOD console one day, is there a way to know the value of the coil on the board so that i can replace them ?

I just played Zelda for a hour, looks stable, battery drain is correct, 72% remaining after one hour. Finally, i made it :rofl:


Hey yeah not bad, how you getting on with the house?

Interesting failure mode, if this coil were open you’d lose your primary 3V3 rail which if that was the case the console wouldn’t boot so not so sure this was relating to your previous fault symtoms I’m afraid (no screen but hearing console sounds etc)

This one is specified in the datsheet as a 1 μH inductor and as per the datasheet they’re specifying one of these

Though, pretty much any of the correct value, size and form factor will do

Don’t think I’ve identified this IC but on a Mariko board a similar form factor Max IC part is also using a 1μH incudctor so chances are this one is too

Good stuff, :crossed_fingers: it continues

Not bad, all water pipes finalized, isolation to be projected next week, all looking good so far thanks for asking :slight_smile:

Thanks for the feedback about the inductors. Will check that out.
However the console is still not 100% stable. In game, no issue, i played Zelda for more than 2 hours, it worked. But if i leave the console in standby, battery drain quite fast (around 4% each hour) and may sometimes not come up, i have to reset by pressing long the power button and power the console on again :frowning:
Interestingly, i have another console where i had exactly the same issue (i still have it by the way i kept it in case one day i could fix it). Tried another battery , same symptom, basically, discharge normally while on, no difference compared to a 100% working console, but in standby fast drain.

Good stuff, getting there! :slight_smile:

Hmm this could be a couple of things, either it’s genuine battery drain as a result of an issue on the board such a shorted / dragged down rail (most likely not a primary one if that’s the case based on your previous measurments) to check this I would fully charge the battery then when the OS reports a low percentage turn the unit off, disconnect the battery and check the voltage manually to see if there is any truth to that

Or it’s false battery drain being misreported as a result of fuel gauge issues, there was a thread recently here where I covered this and a few more serious issues which act like fuel gauge fault symptoms but are quite a bit more serious

Another thing you can do is note the battery percentage in the OS then turn off the unit, wait a few seconds and then prompt it to boot and take a look at the battery percentage again, if it’s not within a couple of percent of your previously noted reading then it’s likely fuel gauge (or the more serious issues mentioned above) which is the problem

Sure, after projection, heated floor to install and then concrete floor, tiles and we are good to go for all the smaller stuff.

I haven’t checked manually the battery voltage yet, but i read the thread you are referring to, it is that one right : Fuel Gauge IC issue but fuel gauge not at fault
What i can tell so far, without having taken a part the console yet since (no time for), is that i don’t have the exact same symptom. I launched Hekate to check, and my fuel gauge is reporting good and valid data. But basically i think the battery drain is gone, afaict i haven’t got it anymore since a couple of days. Now i am left with sometimes the error 21001-001. I will replaced once more the M92 and hope for the best.

Hi there, again me :slight_smile:

36h without anymore crash. What i found interesting is, that both joycon rails and both joycons had corrosion on their pins. After cleaning everything, i didn’t got anymore stability problem.
I noticed that the crashed were related to that somehow because one joycon was having a failing lock, and when i was taking the console, i was detaching it a little bit, and sometimes it was triggering the crash 21001-001. I know it looks completely unrelated, but since i replaced the lock, noticed the corrosion and cleaned everything , i ran without any issues since then.

No that makes total sense, if for example the corrosion was shorting out or dragging down the joycon supply rail/s then this is what I mentioned earlier about it not being one of your primary rails but a sub rail etc :slight_smile:

I’d still run through the tests I mentioned above for the fuel gauge though just in case, you never know

Would you still if no more drain ? I have a normal percentage drop now in standby as well.

I’d just verify the percentage remains within a couple of percent over x10 power cycles, if it does each time then it’s likely fine :slight_smile:

If on the other hand for example it reports 95% and you power it off and then back on again and it’s reporting 90% then you can safely say there is a fuel gauge related issue

It went fine with the a bit more than 10x power cycles, i lost 2-3 % i think after all of them.
Looks finally over, customer came back to collect it, finally :rofl: