ACCEPTING CASES · MON–FRI 9:00AM–5:00PM
Certified Data Recovery Professional · Phoenix, AZ ☎ (602) 686-2622

USB Flash Drive Powers On but Isn't Detected (Phison U17 Controller)

A USB flash drive that powers up but never appears on a computer — no drive letter, no device, a dark activity light — is a common walk-in. On drives built around the Phison PS2251-17 (the "U17"), what we see on the bench is often a controller that starts and then stops partway through boot. Here's the honest part: professional tooling can recover some drives on this controller family, but not every variant yet — and you can't tell which from the outside. Here is what is happening, what not to do, and why a bench evaluation is the only way to know whether yours is one of the recoverable ones.

Free evaluation · No data, no fee · Talk directly with a technician.

Status: partial tooling support — some variants, not all, pending firmware development.
Recovery tooling supports some drives built on the Phison PS2251-17 (U17) controller, but some variants are not supported yet. Which one a given drive is can only be confirmed on the bench. For a variant that isn't supported, we hold the drive so no repair tool overwrites it while that support is developed — nothing we do writes to it. It is an honest status, not a quote.

Most "my USB stick isn't showing up" cases are simple — a bad cable, a dead bridge chip, a snapped connector. This is the harder version: the drive is electrically fine and the controller is alive, but it never finishes starting up, so the computer never sees a disk to mount.

What this looks like

These drives use a Phison controller that talks USB directly, with no separate bridge chip, next to a single NAND memory package. When one comes in not detected, the pattern we typically see is:

Put together, the board is healthy. The controller powers up, begins running its firmware, and stops at the stage where it reads its own startup data and translation tables back from the NAND. That points to a firmware or service-area problem on the flash, rather than a dead chip — though on any given drive that stays a working theory until it is confirmed on the bench.

A flash drive that powers on but isn't detected still has its data on the NAND — until a repair or format utility rewrites it. If the files matter, stop running tools and get it evaluated.

Why you can't just read the NAND chip (LDPC)

The obvious question is: if the data is on the memory chip, why not lift the chip off and read it directly? On an older, simple controller, sometimes you can. On a controller like this one, generally you can't — and it is worth understanding why, because it is the same reason a lot of modern flash recovery is controller-level work.

Modern Phison controllers don't store your files on the NAND in a form you could read off the chip. They scramble the data before writing it, and they protect it with LDPC error correction — a heavy, math-based coding scheme that lets the controller reconstruct bits the aging flash no longer returns cleanly. Both of those are undone inside the controller, using parameters specific to that controller and firmware. Pull the NAND off the board and dump it raw, and you get exactly that: scrambled, LDPC-coded data that is noise until you can reverse the controller's own scrambling and ECC.

That is why chip-off recovery is not a shortcut here. Reading the flash is the easy part; interpreting it the way the controller would is the hard part, and it needs support for that specific controller. Until that support exists, the chip's contents can't be turned back into files.

What not to do

Where this stands

We're being direct because it's the honest thing to do: recovery tooling supports some drives on this Phison U17 controller family, but not every variant yet, and the ones that stop at this stage of boot may or may not be among the supported ones. Which variant a drive is, and whether it's currently supported, can only be confirmed on the bench — which is exactly why it's worth having evaluated instead of guessed at. The data is still on the NAND, and nothing we do writes to it — but reaching it needs either the drive's own controller to finish starting up, or tooling that can take over this controller and read the NAND through it. For a variant that isn't supported yet, we hold the case so nothing overwrites it, and if a drive identifies at any point, we image it immediately.

Update: the drive that prompted this post, a 512 GB PNY on a U17, was recovered in full. Its NAND was failing with bad pages, and a firmware repair followed by file-system reconstruction got the data back. The details are in this case in our case log. That is one drive, not a promise about yours, but nobody had run a format tool on it, and that is why the data was still there.

This is also a good illustration of why no honest lab can promise a turnaround on sight: when a case depends on tooling or firmware that still has to be developed, the clock isn't ours to set. If your drive is a different controller, or a mechanical/electrical problem rather than this firmware stop, it may be perfectly routine — which is exactly what the free evaluation sorts out.

When to still bring it in

Bring it in, or mail it in, if the data matters and you haven't run anything that writes to it. We'll evaluate it for free and tell you honestly whether it's a drive we can work now, a drive we'd hold pending controller support, or a different problem entirely. What we won't do is charge you to run a utility that erases it, or promise a recovery we can't yet deliver. The work is flash drive data recovery; browse real outcomes in the case log.

Desert Data Recovery is a Phoenix-area lab, founded in 2013 and now a division of ioLogik LLC. Arizona drop-off is on the locations page. Everywhere else is mail-in. Call (602) 686-2622, Mon–Fri 9:00am–5:00pm.

Phison U17 flash drive not detected FAQ

Is my data gone?
Usually not erased. On the drives we have seen, the controller stops before it ever writes anything, so the data is still sitting on the NAND. The problem is reaching it. Getting it back needs the drive's own controller to finish starting up, or lab tooling that can take over that specific controller and read the NAND through it. Tooling supports some variants of this controller and not others yet, and a bench check is the only way to tell which one a drive is.
Can't you just read the memory chip directly?
Not usefully on a controller like this. Modern Phison controllers scramble what they store and wrap it in LDPC error correction, so a raw dump of the NAND comes back as noise unless you can reverse the exact scrambling and ECC the controller used. That is why chip-off is not a shortcut here — without controller or firmware support, the bytes off the chip are not readable data.
Should I run a Phison format or "repair" tool (MPTool, MPALL, a mass-production utility)?
No. Those utilities rebuild the drive from scratch — they rewrite the controller's firmware and translation tables and re-initialize the NAND. On a drive you are trying to recover, that is the step that destroys the data. If the data matters, stop before running anything that writes.
So should I even send it in?
The evaluation is free, and we will be straight with you. If a drive is one of the ones we cannot yet support, we say so rather than charging to hold it indefinitely. Nothing we do writes to the drive, and if support for the controller is developed, a drive that hasn't been overwritten by a repair tool is still a candidate. We don't advertise a success rate, and we won't promise a result before we have looked.

A USB flash drive that powers on but won't show up still has its data — until a repair tool rewrites it. Don't format it, don't run a mass-production utility, and start with a free evaluation.

Request free evaluation →

Free evaluation · No data, no fee · Talk directly with a technician.