Nintendo Switch - No Power 0.4a slow charge, detected device in TegraRcmGUI

Hi guys

I have a Switch that is only taking around 0.45a and there are no shorts around any major IC, it is unresponsive.

However it will connect to a PC (plays USB connect sound), and with TegraRcmGUI it says that a “RCM Device detected”. when it connected to the PC, if i press power button you can hear the USB disconnect sound, and when i press again you can hear the USB reconnect sound, so there is some communication going on.

How can i get this Switch to boot again preferably with ofw

Thanks

I am guessing from your description that you are doing this with the emmc connected? If so, it is entering RCM mode when it shouldn’t. If it can accept payloads, I would inject hetake and see if autorcm is turned on, if it’s not, you can check in console info if it is able to read the emmc.

Assuming it can’t read the emmc (or it wont accept payloads) I would inspect the emmc module and connectors for any damage.

Thanks for the reply,

yes the emmc is connected, (i have no history of the switch, cfw or ofw)

ok, i am unable to inject payloads with TegraRcmGUI, it just goes to “not responding” as soon as i chose a payload, similar thing with TegraRcmSmash, it starts and just hangs, i left it for 30mins and it done nothing.

the emmc looks fine no noticeable damag etc, to the module or connector

If you remove the emmc, does tegrarcm still hang when injecting?

yes it still hangs without the emmc, both gui and cli software

Damn… I have one that is similar to this, with it hanging when injecting. Can’t remember if it was the same one we confirmed was a dead CPU or a different one. I would probably have to go through all my un-solved boards to be sure.

I would probably check the resistances to ground on the coils around the PMIC, follow the post here:
(I cant post links, you’ll have to fix the link up)

www tronicsfixforum com
/t/switch-0-05-a-current-draw-won-t-charge-or-power-on/5331/13

ok im not too bothered if it is a dead cpu, i just want to know there is nothing else i can try to save it.

nice one thanks ill read that thread

here are the resistance readings of the coils around the MAX7762, (some do look a little bit crooked)

short on 1V8PDR (97 ohm)

Remove the EMMC module and see if the short clears, should read in the k ohms,

If not then remove M92 IC and check again, if still nothing then start looking at caps on this rail then the main PMIC itself, Ram, CPU/GPU regs etc.

Yes you are correct Severence, without the emmc module connected i get 1.3k ohms on that coil.

so the emmc is bad? these cant be replaced can they, with them being linked to the soc?

I would still say that is a little low to be honest ( I get around 5k on my metre, others say 10k), there may be an issue with the emmc, but there may be other things wrong also, especially as it hangs with the emmc removed.

I posted a drawing not long ago with some measurements on the emmc module you could compare against.

www tronicsfixforum com
/t/is-this-cpu-failure/5674/21

sorry Insomniac, i read the reading wrong, im getting around 4.5k ohm on that coil.

thanks for the link

Ah! Good! So to confirm it is 97 ohm with the emmc, and 4.5k with it removed?

I would go over the pics I just sent, and see how your reading differ from mine.

Just add it’s worth checking your SYS rail as well as it’s not uncommon for that to short out and put a subsequent short on 1V8PDR.

It’s also worth noting that the temperature of the board can have a serious impact on diode/resistance readings in circuit, so if you’ve attempted to power the board up and it’s 5C (for example) above ambient just prior to taking measurments, your readings can/will skew significantly, this has nothing to do with cap charge etc (which will also skew readings)

yes, on checking again with the emmc it was 78ohm and without it was 4.3k ohm

yes ill compare the the emmc readings and see what is different, thanks

Severance where is the SYS rail?, im still learning the layout of the board and different lines etc

great info thanks, the readings were taken after the boad had been off for sometime

SYS rail can be found at the “2R2” coil next to the BQ IC, to test just put one probe on ground and the other on one side of the coil (doesn’t matter which side)

EMMC causing the low reading is a concern, inspect the few passives on the module and see if anything stands out, or better yet post a pic of the EMMC module either side so we might inspect it.

If it isn’t one of the passives at fault then next thought might be liquid under the IC given that this is quite a high short, and keeps changing, give the IC on this module a flux boil (unless you suspect sugary soft drink was the cause) and try flushing any corrosion out from below, then rinse with alcohol and repeat heat, don’t bring the IC itself up to reflow temps, just enough to activate the flux - so approx 200C should be good.

Don’t waste your time taking readings on the EMMC module itself, it is the cause of your 1V8PDR short. so we already know it’s the cause of the fault on this rail.

Refrain from powering the console on further until the fault is resolved less the EMMC be corrupted

ok thanks, i have 1.2M ohm on the SYS rail

i will inspect it in more detail tomorrow,

here are some pictures of the emmc, it looks pretty clean really, there was a very small amount of liquid on the game cart slot/touchscreen board, so its possible a small amount may have hit it.

1.2M on SYS sounds fine. Agreed, a good IPA bath, flux boil combo is probably the way to start. I only suggested taking reading as it may point to a particular pin / trace on the board, which while not all that useful right now, may help prove the fault if you have to remove the chip for a full re-ball later.

I attempted a flux boil on the emmc and it doesnt seem to have done anything, the resistance has gotten lower when the emmc is connected to the board (30 ohms). how long would you let it soak in the ipa after?

i have an ultra sonic cleaner, if its worth giving it a minute or two in there?

i can take some readings from it to using Insomniacs measurements if that would help