1
\$\begingroup\$

I plan on reading and probably writing NAND flash on some embedded devices or USB flash drives.

I did not come across many examples for in circuit reading though I'd give that a try at first if possible. Otherwise I'm going to desolder the NAND flash. For reading I want to use a FT2232H board. I was wondering which of the 48 pins are really required to be connected. Sure there are the obvious ones which are labeled DNU or NC on the spec sheet.

But could I omit connecting some of the ground or voltage pins or some of the IO pins? There are multiple of the latter ones and maybe connecting just one or two of those would be sufficient though maybe reducing the read and write speed.

\$\endgroup\$
  • 1
    \$\begingroup\$ The datasheet of the FT2232H and your flash memory are your friends. Why do you want to omit a few pins? That sounds like a very unusual thing to do – especially things like ground pins. I smell XY Problem. What is your reason to not connect the 48 pins, which aren't neither all that many, nor high density, and to top it all off, relatively conveniently laid out on the FT2232H side? \$\endgroup\$ – Marcus Müller Mar 9 at 23:02
  • \$\begingroup\$ Connecting the pins on the FT2232H is not the problem. But on the NAND flash directly ... it depends. When reading in circuit it maybe gets a bit more problematic depending on how you interface the flash. TSOP 48 adapters are not readily available around here. My other idea was either using ffc cables with the same pitch or directly soldering to the pins. The latter is going to be not that straightforward at least for me. So the less soldering the better. I'd prefer doing everything in circuit since I intend to use the device afterwards and don't know how much (de)soldering it will survive. \$\endgroup\$ – curiouskid Mar 9 at 23:10
  • \$\begingroup\$ Ah, interesting! You'd want to add that to your question – it makes a lot of difference to a potential answerer. \$\endgroup\$ – Marcus Müller Mar 9 at 23:16
  • \$\begingroup\$ Beware that the NAND chip probably won't match the USB mass storage device 1:1 - there will be a whole translation layer with data tables and remapping intelligible to the original controller, but not to you until you do a lot of reverse engineering. You might find block-sized chunks of data scattered around, but I wouldn't even rule out some sort of treatment of the data itself. \$\endgroup\$ – Chris Stratton Mar 9 at 23:53
  • \$\begingroup\$ If you're not fairly confident about transplanting wide format TSSOPs with hot air, it's probably not worth diving into this. If you really can't find a breakout, that would be one of the more worthwhile uses of iron transferring some laser printer toner to copper clad and etching with whatever is handy. \$\endgroup\$ – Chris Stratton Mar 9 at 23:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.