T&L Publications

Hi, folks
I am using the Zynq Ultrascale+ MPSoC on a custom board and have an 8bit NAND Flash memory to boot from. When I try to program the Flash using the Vitis Program Flash Memory tool I am able to measure the CE pin and it does go low and high so I believe the tool is trying to program the NAND flash but in the Vitis Console window it shows 8 bad blocks and fails to successfully write data.
The part number of my device: MT29F8G08ADBDAH4-AAT Souce
(sale item)

How do determine if the memory I have is compatible with the Vitis Program Flash Utility? Does anyone have ideas? Many thanks!

Hi

Searched zynq and it is a Xilinx product.

Seems like their forum would have people that would be off and running with a question like that.

Hope this helps a little

Just saw your question at bottom. That should be in device datasheet.

Electrical specifications section maybe?

In programmer manual might be sort of a stretch.Unless they mention in passing.

Thank you for reply.
Yes, btw, can I modify the PCB and get the LOCK pin to control it. In the case of using the Flash Programming Tool, does that tool control this pin when in NANDx8 mode? Would you consider it ?

Hi Markkim

Will take a look at manual for the cpu and the programming software and help you figure this out.

Have a nice day

Bob

If you can attach to it with your programmer then you are probably addressing the device properly. I suppose if the programmer expects a certain size and if its not, it might report that as bad blocks but I'd think modern programmers are smarter than that.

I'm not sure what the AAT suffix means but I do know that this segment of the market is rife with counterfeit devices and resold defective parts. I suggest you obtain a golden device from a trusted supplier like Mouser or Digikey and compare that to the device you got. If you obtained it from a smaller distributor or on the resale market for low cost or especially ebay, I'd be suspicious at least a little. The P/N implies an 8Gb device but a fake 4GB device would drive you nuts.

Also an open on one or more address or IO lines could also cause read write errors that may appear to map to memory space, so double check the assembly, an open BGA pin can be impossible to see and hard to fix. A solder fault is actually the most likely possibility, try reflowing the part on the board if you have the gear.

While Micon makes this one, there should be equivalents from WDC, and Sandisk and they may also offer development tools on the website you can try. I'm pretty sure these devices are programmed and read like a SRAM so no need for complex timing of RAS and CAS signals like with DRAM. so it would function partially with open or shorted IO and address pins especially if they were the MSB, that would only knock out the higher range of addresses.

Don't know if you have looked but there are now some super low cost digital debugging tools like logic analyzers that are in an accessible price range, 8 channels with USB interface are going in the $15 range on ebay, add some microhooks and you'll be able to look at the device signals with much better resolution than a 2 channel o-scope. Used benchtop equipment is a good value too.

Hi Bob, I saw your email, not sure how to reply back that way. I don't write firmware, My perspective is more from a general device testing perspective but that gets me looking at a wide range of datasheets along the way and experience with DC characterization on these parts.

Are you using the TSSOP package or the BGA package in your project?

Thanks you all again.
I got the ideas, and I try to program it after apply the patch and report the log back, and I also tried the Flash Program Memory Tool to write a Boot image to the memory and have attached the console output.