Testing for Counterfit Flash Drives

[billko]

---

@rolfk wrote:

---

@Hooovahh wrote:

---

@Bob_Schor wrote:

The Amazon comments on the $26 1TB drive largely say it is a phony, a counterfeit, etc.

---

Right but regardless of the vendor (and the reviews) a 1TB drive costing less than say $800 is likely a counterfeit.

---

But it's X-mas time and one can always hope for a magical wonder! 

The thumb of rule for all these things really is:

If it sounds to nice to be true, it really always is!

---

My wife sounds too nice to be true.  Should I worry?

[rolfk]

---

@billko wrote:

---

---

My wife sounds too nice to be true.  Should I worry?

---

Seems you are still madly in love. 

Keep it that way!

[emw129]

Here are some free Thumb Drive testers that work, but most of them take days to test a 1TB thumb drive or many hours for Large GB Thumb drives:

https://www.raymond.cc/blog/test-and-detect-fake-or-counterfeit-usb-flash-drives-bought-from-ebay-wi...

Some of them however destroy the thumb drives firmware to do it, others don't.  So what is the trick, how is this code able to access internal specific memory and do it quickly?   Each have their on quirks about them.

There must be an internal BIST inside these Thumb drives.  But how to access it?

[rolfk]

There most likely is. But the magic is to detect the actual controller used and do specific things depending on that detection. And to reverse engineer this for every type of controller.

Read about the attempts to use regular USB flash drives to create USB Rubber Ducky clones. Basically someone figured out how to do this for a specific sort of USB flash drive that used a certain 8051 based controller type, modifying the firmware for that controller to act as a keyboard with embedded keyboard sequences being replayed. It only works with one specific type of controller which nowadays is hard to find drives that contain it. There are several new controller versions since and which one you get can vary even for the same type of flash drive from the same manufacturer. And besides various controller versions from this specific chip manufacturer there are of course numerous other completely different controllers from many other chip manufacturers.

Getting the technical data sheet for these controllers is in most cases an exercise in vain unless you can show the Chinese manufacturer that you can and will purchase a few millions of them.

[emw129]

I would not only have to write 128GB, but I would have to read 128GB, the test time would become astronomically long.  The short cut would be to write at address 0, then 128GB, then 64GB, then 32GB & 96GB, 16GB & 112GB, etc., ... until I wrote and read at those specific address i could more quickly do a sanity check on those drives without taking days and weeks to write and read 128GB of data.