Re-Short at MAX77621 left from the soc SYS line

Nintendo Questions Nintendo Switch
#1

I have a Switch with short on the SYS-line. After pulling all related ics (bq24193, max77621 south,max 77621 west, 8316 (@lcd), ENxx, max77620, +AJZ, U25) and the connected caps I found that one of the last three caps was cause of the short. So I resoldered every part back and replace the shorted cap.

First readings in diode mode were fine. Than I connected a battery and my usb-meter. Readings 0.45 amps and after a few seconds 1.37. Than I realizied that the soc is getting realy hot. After plug the mainboard in its shell a try a second time the short on the SYS-line is back again… :face_vomiting:

The max ics I replaced with new ones.

MAXshort
MAXshort998Ă—754 31.9 KB

Any ideas or similar experience?

#2

Had a BSOD board and it been through reheat cpu from bottom side by someone and left-side max77621 short similar to yours

#3

Whats your resistance to ground on this shorted rail Calvin?

It must be someting intermittent for it to come and go like that, suspect would be another cap or impact damage causing a ball under an IC to come loose.

Might be worth connecting your meter up to the shorted rail in resistance and with your hot air at 100C go round the board - main PMIC, SoC BQ area etc and seeing which causes the reading to change the fastest.

Alternatively you could hook up a bench PSU to try and find the short, though there is a high chance it will clear once you apply power making it troublesome to find.

#4

The resistance to gnd is 0.40 Ohm.
I desoldered the three caps and the max ic again. But this time the short is still there.
Before pulling all ics befor I found the defect cap the first time, I tried with injecting voltage to the line and thermal cam to find heating spot. But the defect cap never warm up.

#5

Man you’d think you’d be able to find it quite easily whilst providing power ext, if at 4V you have a potential maximum of 10A even at 1V would net 2.5A. which would be scorching. What was the current readout on your PSU? do you think it was clearing on power application?

You may have to resort to the local heat and resistance monitoring method mentioned above, while finicky and not very accurate it’s useful in these situationa, especially when they’re intermittent

#6

I limted the psu at 3 amps and injected 4V in the SYS line.
But I found the cause of the short. This time with the thermal cam and battery and power supply. The max77620A shorted the ciruit… now I m wondering if my soldering was bad or the ic was bad…

#7

Nice one!

I find the max77620A is typically pretty solid and doesn’t fail often, though from time to time you do get one that is bad but more often than not the cause is liquid or impact.

Try measuring the corresponding ball/pad on the IC itself relative to (corresponding) ground and see if you still measure a short, try also heating the IC up to 100C and taking that reading again on the IC and see if it approaches anywhere near a dead short, in case it’s an internal failure temp resolved by heat. If all tests fine then likely the IC is fine and can be reballed.

Sometimes you just get unlucky and a ball decides to wander off, though your soldering might not be to blame depending on the results above.

#8

I think it is my soldering skill. Now after reballing the max77620A the Switch boots up fine but have no backlight. So I will also resolder the backlight ic the next days. :stuck_out_tongue:
I thought the little 3x3 bga would be easier, but for me it s hard to hold the hot air nossel with my working hand and with the other hand the tweezers without moving the ic out of its position.

#9

Oh no, one IC after the other, I have days like this :slight_smile: Glad you were able to resolve the PMIC issue though :+1:

Check your getting your enable on pad near backlight IC before removing incase that’s the issue.

After wicking the pads on the board, try adding a super thin layer of flux and setting the airflow on your station to the lowest it can go and see if that helps, sometimes too much flux and airflow will cause the IC’s to take off.

On my station i place the IC into position and then just hot air, don’t have to use tweezers to keep it in place or otherwise, but maybe your stations min air is higher than mine by default.

#10

After reballing the backlight ic and reballing the lcd power ic again the Switch is working fine. :grin:

In the end it was one defect cap at the max77621 (right from the soc) which cause the short on the SYS line.

And it was not getting hot enough to find it during thermal inspection. Bad luck that it was one of the last three components I removed from the SYS line. I removed eight ics and sixteen caps. :stuck_out_tongue_closed_eyes:

#11

The re-short appeared due to a bad soldering at the max77620. On of the balls disappeared during the soldering process. After powering the Switch it shorted the SYS line again. And after reballing the max77620 again and rework the before mentioned backlight and lcd power ics the Switch is working.

1 Like
#12

Hola cuando hay corto en esas lineas, se remplazan los IC, pero, si no se va el corto, el que está mal es el procesador, saludos, me dedico a reparar nintenido Switch en Mexico y ese problema lo he visto pocas veces.