Diagnose Corsair Flash Voyager 64GB w/ SM3252R BA controller

[SOSdonnees]

Hi,

I have this defective Corsair Flash Voyager 64GB USB 2.0 key for data recovery. (P/N: CMFUSBHC-64GB)

Corsar Flash Voyager 64GB USB 2.0

This flash drive use several kind of circuit boards, with 4 NAND chips, 2 on each face.

Was initially briefly detected but froze, could not be ejected, and had to be unplugged manually.
Then when plugged, had the diode flashing every 5 seconds as shown in this video: https://imgur.com/mvW7bvo

On the back face, I found a defective resistor 2.2Ω at the location R22, which was swapped from a working one, same value.

I also soldered a new inductor at location L1 on the back face, as the original one was absent. I assume there was some broken solder here.
Not sure however what the inductor value should be.

The USB connector appeared in good condition ; was replaced so that I could check the tracks below it.
I could no observe any defect at the tracks.

The controller pins were tinned again and checked as properly soldered to the boards.

The flash diode is no longer emitting any light.

When plugging the flash drive, an "unkown device" is detected, but only after about 12 seconds.
Looks like the flash drive is underpowered in some way.

The controller is a SiliconMotion SM3252R BA.
There is some info about it on Flash Extractor and PC-3000 websites, but I could not find a datasheet with the pinout.

The PCB is labelled 201ZZZA1. I could not find any info about it.

I measured the voltages on both sides of the PCB, which are shown in the pictures attached.

The continuity of all resistors, capacitors, inductor and diode was checked.

Not working PCB 201ZZZA1 with controller SM3252R BA, top face, with measured voltages.

Not working PCB 201ZZZA1 with controller SM3252R BA, back face, with measured voltages.

I still assume likely that the problem comes from the circuit board or the controller.
If possible to avoid a chip off, would be nice.

[fzabkar]

AFAICT, inductor L just connects the metal shell of the connector to circuit ground.

The 2.2 ohm resistor (2R2, R22) is in series with the 5VUSB supply, so an open circuit would suggest that there was a current overload.

Voltages of 3.3V and 0V at the crystal would suggest that it is not oscillating, but this is often the result of loading by your meter probes.

I see 1.8V and 3.3V supplies. I assume that 1.8V is Vcore.

This product sheet (not a datasheet) confirms that 3.3V and 1.8V are the internally generated supplies:

https://www.datasheetarchive.com/datasheet/c94e00089b58ffee?type=P&term=sm3252

The crystal is 12MHz.

[SOSdonnees]

Thank you.

AFAICT, inductor L just connects the metal shell of the connector to circuit ground.

Correct. Here's a picture of with the inductor at L1 removed.

I initially measured no current (0 Amp) at through the 2,2 Ω resistor at R22, and around 850 / 900 mA at the capacitor located at C4.

I replaced the 2,2 Ω (2R2) resistance at R22 by one labeled 2R00, i.e. 2,0 Ω but that I measured as 2,5 Ω.
I also replaced the capacitor at C4 by a 35 μF one.
After that, I measure current of 31 mA through R22 and initially little above 900 mA (~910 mA) at C4.

Here is another detailed picture. I notice a small damage to the resistor at R19 but I don't think the problem coming from here.

The crystal is 12MHz.

Yes. It is labeled 12.0 KAe.

[SOSdonnees]

Seems that the "SM3252R-BA" controller can still be found at some distributors, but strangely I can nowhere find the pinout information.

[SWM]

I would solder another quartz resonator to check. Maybe a different type. There are two pads connected to the ground anyway.
Out of several dozen flash drives, two had a "broken" resonator.

[SOSdonnees]

I would solder another quartz resonator to check.

Thank you for the suggestion and sharing your experience. Sorry for the late reply as I was ill this weekend.
I could find a 12 MHz resonator on another Corsair Voyager 64GB having a different board. The logic a the resonator was same.

I first soldered the resonator to the pads, but because of different width of the crystals, could not be absolutely sure that the contact was properly done at all four connection point.

Hence, I unsoldered the resonator and wired it.

I also swapped the light emitting diode.
Still no result.

solder-pads-on-patient-board.jpg (36.18 KiB) Viewed 15619 times

crystal-oscillator-soldered-on-patient.jpg (43.02 KiB) Viewed 15619 times

crystal-oscillator-wired-to-patient.jpg (73.96 KiB) Viewed 15619 times

[SOSdonnees]

I found that user Davor from Flash Extractor's forum (which is now read-only) worked with the same* Corsair Data Traveler 64GB, also 4 chips, and same controller SM3252R BA : https://flash-extractor.com/forum/viewtopic.php?t=3918

If yes, can you please PM me how to contact him?
Thank you.

(* So far as I know, only the model of the NAND chips differ. Mine are 29F128G08CJAAA)

[SOSdonnees]

The Corsair Flash Voyager 64GB USB 2.0 flash drive (P/N CMFUSBHC-64GB) with

• circuit board 201ZZZA1
• controller Silicon Motion SM3252R BA
• 4 NAND chips 29F128G08CJAAA (or possibly 29F128G08GJAAA)

is a known case for data recovery by chip off.

Likely because it is located an past version of Flash Extractor's forum and undocumented in FE's online library, this case was buried in the bowels of the web and during weeks did not appear for me in search engines despite very specific keyword searches.

Let's bring it to the surface : https://flash-extractor.com/forum_old/viewtopic.php?t=2971