Proxmark 3 LCD version [Null Space Labs] build log

So we decided to build Proxmark’s at [NSL] and not just the well tested Proxmark3’s that most people use, the LCD version which there appears to be only possibly one other of, maybe not even  that, that had  a lot of missing parts and only one guy who knew anything about it. Luckily he was a cool guy and helped us out


So we order boards from BatchPCB they came in, not bad quality, only one board of the 18 so far had a problem in where a trace was shorted out. Then fixed up the BOM and ordered enough parts for one board, built it and didn’t work. So we did what any sane space would do and ordered another 10 sets of parts. Parts that were going out of stock as we were creating the order!


Some parts were hard to get the Atmel’s seem to be almost legendary these days, but they were found in China, the 8050 had a newer replacement part.. We used the 512K part.


The first board was built by someone who’d never soldered before, and a hotplate, it worked pretty well only one chip had to be removed and a few pads lifted, but some Kynar wire took care of that.

proxmark3 lcd, the blew up the fpga version yes the FPGA blew up, don’t know why, as it was replaced and the board gone over but didn’t blow up on the second power up, another $13 to the cause.

Debugging SPI for the LCD

Compiling the software and getting OpenOCD were a fairly massive pain (took longer than 10 minutes), so I ended up just throwing it in IAR and using a SAM-ICE which meant i could debug and flash the board quickly. All that had to be done was put the bootloader, fpga  and code into one program and change some no longer needed section attributes, add a few pack pragmas and off it went , the later proxmark3 code has done some of this but we are using a branched build.

Once the software was up and running, mostly, we started building antennas and what not. This mostly meant wrapping bits of magnet wires at random and see what happened. M showed us his superior skills at picking up a partially used spool of magnet wire and connecting it directly to the PM4, and immediately getting a perfect LF antenna.

I attempted a milled antenna, the first one didn’t go so well as I’d forgotten that sometimes when debugging FTDI hardware or designing a board with it, it causes a bluescreen in windows during the power down mode, but only much later, like for instance when you’re halfway through milling an antenna that you didn’t save the project file for and had to restart it from scratch, but only after buying a new V60 bit because it wore out during this board and I’d run out.



Making a patch PCB for the broken LCD connectors that some people had, these would be affixed to the board with glue and kynar wire, awwwhh yeah!