HDD GURU FORUMS

This controller gets extremely hot. Too hot to touch. Under the infrared the only part that gets hot is the controller so I am presuming the controller is shorting. There are also shorts on caps all over the board. Can someone confirm this is likely to be a controller short, or offer advice on where to test other parts of the SSD?

Thanks in advance.

I would start by measuring the resistances between ground and each of the inductors surrounding the PMIC. I suspect that one will measure as a short, and will probably correspond to your shorted capacitors.

I'm betting that your diagnosis is correct, ie the controller is probably shorted. BTW, the Vcore resistance is typically in the low ohms range (~ 100 ohms).

I've just realised that 3 of the supply voltages (V1, V2 and V3) are isolated from their respective loads by load switches.

This makes it easier to identify the shorts. If the short is on the Vin side, then this would point to the PMIC or load switch as the culprit. Otherwise, if the short is on the Vout side, then this would point to the respective load as being faulty. If there is a short on both sides, then the load switch may also be faulty.

I managed to do some testing. Results seem confusing and maybe punts to a more widespread issues (controller).

Top Right Load Switch - Vin and Vout shorted
Cap @ Vi - shorted
Cap @V3 - shorted
Cap at V6 - shorted
Bottom Left Load Switch - shorted
STEF4S - 3.3v in / 2.6v out
VBoost - 0v
When powered on none of the caps V1 thorugh V6 show voltage.

The e-fuse appears to be in current-limit mode.

When you say that "Vin and Vout [are] shorted", do you mean that Vin is shorted to ground, and Vout is also shorted to ground, or do you mean that Vin is shorted to Vout?

It's quite possible that a current overload has killed the load switch. A typical failure mode is an input-to-output short.

The PMIC may not be enabling certain supplies if it is perceiving that its input voltage (2.6V) is below the Under Voltage LockOut (UVLO) threshold.

It could be that a switchmode regulator in the PMIC failed with an input-to-output short. This would then apply the full 3.3V supply voltage to the corresponding load via its load switch. The load would then fail in short-circuit mode, causing a failure in the load switch due to overcurrent.

fzabkar wrote:When you say that "Vin and Vout [are] shorted", do you mean that Vin is shorted to ground, and Vout is also shorted to ground, or do you mean that Vin is shorted to Vout?

Just to clarify
Vin - shorted to ground
Vout - shorted to ground

ISTM that there was a cascading failure, ie one device failed and took others with it.

In retrospect my question was unanswerable. That is, if the load and load switch were both shorted, then Vin would also test as a short, even if it were OK.

Yikes! The operating temperature can reach 81C during high workloads.

Samsung 970 EVO Plus NVMe M.2. (1TB) SSD review - SSD Operating Temperatures:
https://www.guru3d.com/articles-pages/samsung-970-evo-plus-nvme-m-2-(1tb)-ssd-review,6.html

fzabkar wrote:Yikes! The operating temperature can reach 81C during high workloads.

Samsung 970 EVO Plus NVMe M.2. (1TB) SSD review - SSD Operating Temperatures:
https://www.guru3d.com/articles-pages/samsung-970-evo-plus-nvme-m-2-(1tb)-ssd-review,6.html

Put it in a laptop that is sitting on someones bed with the air ducts blocked and a bad fan.....