MicroSD Eletrical Re-Flashing ??

[igounfazed]

Hello all,

First, this is my first post, as I just joined to ask this question. I figure this user base is better suited for the question, than any other Operating System User base.

I have a Galaxy Tab 2 - 7", for which I purchased a new 32gB MicroSD card (SanDisk), and have had no issues with either since purchase in December 2012, until now.

Recently, when attempting to take a picture, I was greeted with some error stating I could not save to the MicroSD. It progressed into the MicroSD card not being recognized by the tablet. I tried reformatting it, as the tablet suggested, but nothing would happen. The card would not mount either. I figured I could have it read on my PC via USB dongle. But the MicroSD would not register on the computer as ever being inserted.

I searched forums looking to see if the error I was receiving is a popular occurrence, and found that it was. What I could not find, was a way to fix the card back into a usable state of any sort. Meaning: with or without losing the data on the chip.

The consensus I found was: dump the chip in the trash and buy a new one; or have the manufacturer send you a new chip. At which time you 'dump the chip in the trash'. Unless, of course, you have money to spare and send it to some factory to have the data recovered.

I came across the idea that, as MicroSD's are just a form of EEPROM, non-volatile memory, there could be a way to electrically return the chip to a zero state, kind of like I did many years ago in my electronics school; thus allowing the chip to be readable once again.

I looked up MicroSD's pin-out, and placed my DMM on impedance across pins 4 (+), and 6 (-). I figured using the impedance mode, the current going through the chip would be low enough not to electrically damage the chip.

My reading was about 105K ohms. Not sure if that matters, but mentioning it just in case.

My result was a partial recognition -within my PC- that allowed me to "Quick Format" the chip. At that point, I tried the MicroSD within my tablet, and still had the same problems.

After that failure, I placed the chip back into the PC, and it was no longer recognizable.

My thinking on the initial PC recognition is now that: having placed the USB dongle within a different USB port, rendered the chip momentarily recognizable. I think ohming pins 4 and 6 was just a 'fluke' of luck... no pun intended.

I have now tried ohming other pins, starting from pin 1, with the positive DMM lead, and going across the pins until an impedance is read.

Pin 1 to pin 4, provided a small impedance reading. I placed the chip back within my PC, and it immediately popped up, and opened to the drive.

There is no data, as I did perform a 'Quick Format' earlier. But this time around, I am doing a 'Full Length Format'.

The format settings are:
FAT32 (Default),
Allocation Unit Size: 16 kilobytes,
No Volume Label

It's currently still going through the formatting. About a third complete.

I'm sure some of you may have heard of non-mounting MicroSD's, that are NOT recognizable in any device (mobile or home). Some advice you've given in the past could be helpful here as well.

If you have heard of someone electrically 'flashing' a corrupted non-mounting / non-readable MicroSD card. Could you point me to their successes or failures?

Once I get through my testing, working or not, I will post my findings. I'm hoping it works, but not going into this with high hopes. Just taking the challenge.

One last thing: I'm currently under the impression this type of problem ONLY occurs with Android OS devices, as I did not find anything stating it occurred with other non-Android devices. If you know of it happening outside of the Android OS, please point me in that direction as well. Google hasn't helped me in that regard.

 

[igounfazed]

Sorry it's a lengthy post. I just want to be as informative as possible.

 

[Joelgp83]

A fascinating read, sir. About 8 months ago I too lost an SDcard like that, an 8gb microSD. Apparently it was so far gone that even Win7/8's Disk Management tool could not lock onto the card for partitioning/reformatting.

For whatever reason I kept the card, and now it seems your technique could possibly be used to revive mine. I just have one question:

What is this DMM you speak of using?

 

[igounfazed]

Hello, I'm providing an update.

*DMM = Digital Multi-Meter*

It seems that the process I go through above works, but solely temporarily. I've been able to 'reset' the card, but not for long term usage. After a remove and reinstall into any equipment able to read cards, the card goes back into its 'defective' / 'non-responsive' state. I've stopped short of putting straight voltage into the card. As I figure anything from a battery would make it go 'pop'.

It is interesting to me, that a small amount of voltage from my DMM can cause it to be recognized, if even for a short term.

In it's short term fix. I was able to add and delete data onto the chip. I was able to perform formats. And any other activity associated with the file placed within the chip (ie. opening, closing, editing, etc. etc. etc.). But once the chip was not recognized, I began anew.

I've not found any information regarding electrically flashing memory chips as a fix. And I don't have equipment to further investigate the actual circuit paths on the chip for 'actual' damage. So I cannot go any further.

As I can't go any further... I believe I will take a 1.5v "AAA" to the chip, and increase battery voltages until I hear a 'pop' and smell a burn.

Again: if anyone has a direction to point me in (before I pop the chip), please point me towards it.

 

[igounfazed]

For those that were following this, it has been a while. I wanted to provide an 'odd' update over the MicroSD card.

After going through all the previous things I did. I left the M-SD lying around. One day, I picked up the card, slid it into the USB styled adapter, and loaded it with files.

About a week later, as I was going through my M-SD cards, I realized the card I thought I had in my USB adapter was the one I currently had in my video camera recording.

Yes, I didn't realize I had the old M-SD card working.

Well, today the M-SD pooped itself again. Giving the 'format needed' pop-up. I had been slowly filling the card with data (mostly videos) and reached about 16 gigs before the lock out.

I fully believe that a 'sector' is bad, and that upon placing data in that section: the card poops itself.

While it was working, I was able to play video from it on two different T.V.'s, my desktop, and my laptop. Prior to reaching this 'limit', I noticed the stability dropping. Meaning, just like my original posting: It worked like new, then began having issues with readability, then no access (can not format successfully).

Anyway, I figured you guys that were following this deserved to know what I found. -I'm going to zap it again and start all over .

 

[Dngrsone]

This, indeed, is a fascinating read.

I am not an expert in Flash memory. However, I am an electronics mechanic with decades of experience working on computers and computerized systems. So I think I can weigh in on this conversation with a teeny bit of authority. Feel free to research anything I say here and provide links to argue that I am wrong.

I am thinking you are looking at a couple different problems here.



Flash memory, whether it is CF, SD, uSD or USB 'thumb' drive, is basically the same device under different packaging.

They all have the same limitations, which is that flash memory can only be written to a finite number of times. Whether that is 50,000 times or 500,000 times (these are arbitrary numbers I pulled out, I am not looking at any reference at the moment) is dependent upon the newness of the design being used.

Regardless, there is a limit to how many times a piece of flash memory can be written to.

These devices also have memory management processors built in. Their job is to move data to and from the memory itself and to identify and mark off which sections (blocks) that are bad, so that data are no longer written to those portions.

When a flash device is wearing out from too many writes, it starts to become unreliable. For USB drives, it often manifests in a message from your Windows computer telling you that the device is running in a USB 1.1 port and would work better if you used a USB 2.0 port (yet your machine will likely have only 2.0 or a combination of 2.0 and 3.0 ports).

In other devices/machines, the symptoms would be different. At any rate, when your memory begins to act erratically, then you should consider replacing it.


The Samsung effect is an entirely different problem. This is an unfortunate miscommunication between the Samsung device hardware/firmware and the memory controllers hardware/firmware on certain uSD cards. Samsung is not really interested in addressing the problem, and since uSD memory providers are liable to use different controllers and firmware with different batches of memories they sell (and no identifying marks as to which controller is inside), there is no guarantee you will be able to identify which particular card will work well with your Samsung device. I have personally gone through five cards trying to get one that will work well with my tablet.

Typically, an uSD card that is weirding out on your Samsung device but is recognized and operates normally in, say your computer, is fine, just not fully compatible. If it does not work on the computer either, then the card is defective and should be replaced.



Digital Multi-Meters (DMMs) are very common and their quality ranges from the dirt-cheap to very expensive, depending on how accurate you want them to be. There are many designs, but they generally use the same principles to accomplish their tasks.

For resistance measurement, DMMs (as opposed to analog multimeters, which can act differently) typically apply a small constant current through the leads and measure the voltage developed across those leads to calculate the resistance of the load applied between them. The amount of current varies with the resistance range setting-- higher for larger impedances (inherent resistance).

This means that, unless you happen to know the specific design specifications of your DMM, there is no knowing how much voltage is being applied to your memory when you try to read its impedance.

By applying a small current to random sections of your flash memory, you may be erasing the parts of the controllers own memory-- damaging the firmware-- or its knowledge of which blocks are bad.

You might be erasing chunks of the memory, I don't know.



Regardless, this is not going to 'renew', 'reset' nor 'rejuvenate' your uSD card, only make it more unstable and unusable.

 

[Rukbat]

A bit late (for huge values of "bit), but SanDisk Ultra 64 GB microSDXC Class 10 UHS... fix (a long read) will give you contact information. SanDisk is quite a company to work with. No guarantee that any data is recoverable, but I'm certain they'll be interested in your experiences, and might replace the card.