Arty S7 - Changes from Rev B to Rev E

[Mike Reynolds]

Gday,

I have been using a set of Rev B Arty S7 boards over the last few months with great success.  Nice board.  I have a UART in the FPGA and access it via the Shared USB JTAG / UART port connected to a Windows 7 machine.

Today I received an additional Arty S7 which is a Rev E.  The UART isn't working.  If I plug a Rev B in it works.

I also notice the Rev E has a different SPI FLASH brand, requiring a different setup in Vivado.

Any advice on how to get the UART port working appreciated.  Is an updated Windows driver required perhaps?

Mike.

[JColvin]

Hi @Mike Reynolds,

Does your set of Rev B's all work or are you just trying a single Rev B?

Could you send a picture of the different SPI flash? I looked at the build list for both the Rev E and the Rev B but they both appear to be the same flash chip (Spansion S25FL128S).

Thanks,
JColvin

[Mike Reynolds]

Hi JC,

Thanks for your reply.  We have 3 Rev B units, all of which are performing faultlessly and have been for several months.  I'll try to arrange a photo of the Rev E unit tomorrow.

My main issue is with the serial comms on the Rev E.  Any thoughts on that?  ie Has _anything_ changed in this part of the board?

Cheers,
Mike.

[JColvin]

Hi @Mike Reynolds,

Are the Arty S7 boards both the -25T / -50T variant? If they are different, they would each need their own bitstream. Another easy thing to check would be to see if Jumper JP1 is loaded (set to load the program from QSPI memory) or unloaded (set to load from volatile memory) to ensure the correct program is loaded.

What I would check to ensure that the Rev E UART works would be to run the General IO demo on the board; I just did this for both the Rev B and Rev E versions of the board and could see the UART portion worked correctly. As far as I know, there are no pin changes that Vivado would be aware of (or that I could find in the schematic) between Rev B and Rev E (aside from the change between the -25T and -50T variant if that is applicable in your situation).

Thanks,
JColvin

[Mike Reynolds]

Hi @JColvin,

All the boards are -50.  JP1 is loaded on all the boards.

We ran the General IO Demo on the Rev E and also on one of the Rev Bs.  The demo project successfully recognizes the FPGA in Hardware Manager.  However, the Rev E FPGA is not recognized in our project. We can program our bitfile from the demo project, and read registers over the UART.  So there’s something weird going on in Vivado, but the JTAG/UART seems OK at a hardware level.

Regarding the FLASH devices here are two photos.  The Rev B photo is clear.  The Rev E photo is less clear since the marking is very faint.

Rev B

Rev E

 

Looking up the Rev E FLASH marking at http://www.bios-chip24.com/epages/63730052.sf/en_GB/?ObjectPath=/Shops/63730052/Products/33320 appears to indicate it is a Micron device.

Does the FLASH device look incorrect to you?  It certainly doesn't look like a Spansion device.

Any further thoughts?
Cheers,
Mike.

 

 

[JColvin]

Hi @Mike Reynolds,

We're looking further into this; I got ahold of a Rev E and can see the exact same Micron chip that you have, so I am verifying if this should be accurate since all the written materials I mentioned indicated otherwise.

Thanks,
JColvin

[JColvin]

Hi @Mike Reynolds,

I learned that the chip you see is accurate and is an alternate build piece, but should not make a difference as the manufacturing test does not change to account for this different part. What version of Vivado are you using? And to confirm I understand this correctly, the General I/O demo worked successfully for you on both boards, but not your project?

Thanks,
JColvin

[Mike Reynolds]

Hi @JColvin,

Thanks for your two replies.  In our Vivado project the Rev E Flash is identified as a Micron device (as its labelling suggests). The Micron and Spansion devices are not identical.  Rev E is therefore unusable for us.  We are using the Arty S7s for prototyping purposes associated with an in-house board using the Spansion FLASH.

We don't have time to further investigate this.  We'll continue using the Rev B boards to get the job done.  Thanks again for your help.

Regards,
MIke.