CMOD A7 - SHA204

[rgc 0]

Hello Folks,

I'm trying to use the ATSHA204A in the CMOD-A7 as a high-quality random number generator.
However, the interfacing is by design limited to the single-wire (SCL is grounded), and I'm having trouble making it work.

Is there an application note or reference anywhere?

Thanks in advance.

/rgc

[Cristian.Fatu 37]

Hello,

This chip is not used by any of our boards.

[rgc 0]

Hello,

Thanks. Not sure what you mean, I see the ATSHA2 0 4A-MAHCZ-T in the CMOD A7 schematics, and xdc. Mind to expand?

[Cristian.Fatu 37]

I apologize, you are right. The chip is part of Cmod A7.

We do not encourage you to use it on specific application therefore we have no reference materials including it.

Still, connecting SCL to GND leads to Single-Wire Interface usage. You can read more on ATSHA204A datasheet.

[rgc 0]

Thank you. Just wondering, if you do not encourage to use it on specific application, what is it exactly there for?

Best

[Cristian.Fatu 37]

If you take a closer look on the Cmod A7 schematics you will find that it has "No load" marking. Therefore the chip was not equipped during the manufacturing process.

The reason for including it in the design was a potential development that was never implemented.

Of course, a user can choose to acquire such a device and solder it on the board.

We are sorry if you foreseen an interesting application and now you lack the on-board device to implement it.

[zygot 153]

4 hours ago, Cristian.Fatu said:

We do not encourage you to use it on specific application

I certainly can't speak for whoever decided to include this part on the CMOD module. A guess as to  what they don't encourage is trying to use the device for secure applications. If you want a 72-bit unique hard-coded serial number for each of your CMODs it can certainly do that. If what you are looking for is a RNG Microchip makes the ATTPM20P with a considerably easier to use and faster SPI interface. This device claims to be FIPS-140-2 module compliant; though there is an effort to replace FIPS-140-2 with something better. I'm no cryptographic expert but what little I know from available literature suggests that such devices aren't robust enough for secure applications.  If all you are looking for is a source for pseudo-random values there are probably easier ways to get that.

Digilent seems to have a penchant for making unfortunate, head-scratching, design decisions that render some good ideas into less than great products.

[rgc 0]

Thank you, Cristian and zygot. I understand. I just wanted to use the 204 to replace an LFSR-based random generator to get a better distribution without using FPGA resources. 

I think it'd be a good idea to remove it from the schematic, and the xdc file to avoid confusion and wasted time... spent a day talking to a non-existant chip 😂

No design is perfect, but I do think the CMOD A7 and Nexys are great products. 

 

Cheers!

 

[Cristian.Fatu 37]

Sorry for this confusion.

We considered that labeling "No load" in the schematic will be enough. In fact we can't remove it from the schematic as the board connections include the connections to this potential device. Its location is there, ready to get a chip on it.

Still, I am forwarding the output of this conversation to the people in charge.

Edited November 5 by Cristian.Fatu