JTAG STM1 stopped working

[eskull]

Hello,

We have a Xilinx VC709 FPGA board (Virtex 7), with a JTAG-STM1 interface. The JTAG programmer no longer shows up in either Windows 10 Device Manager, nor does it show up in Vivado (2018.2) Hardware Manager.  

A second VC709 works correctly so it doesn't appear to be driver/SW issue.

Probed the TCK, TMS, TDI pins - all are pulled high (3.3V) showing no activity. 

We had this VC709 installed in to a PCIe slot on a PC, with the JTAG-USB cable connected between the VC709 and a laptop. Is there a clue in this setup, as to how the STM1 might have failed (e.g., ground loop - what would be most likely failure point on STM1)?

The VC709 otherwise appears to work fine - there is a PCIe design in the BPI Flash that loads on power-up and links to host PCIe, as well as flashing status LED's as expected. We just can no longer reprogram FPGA/FLASH. 

Are there any probe points on the STM1 we can check (with DSO) to isolate problem?

 

Thank you.

[JColvin]

Hi @eskull,

I have sent you a PM with some more details on this issue since I think the EEPROM on the JTAG SMT1 needs reprogrammed, presuming you have set SW11 to 101.

Thanks,
JColvin

[eskull]

14 minutes ago, JColvin said:

Hi @eskull,

I have sent you a PM with some more details on this issue since I think the EEPROM on the JTAG SMT1 needs reprogrammed, presuming you have set SW11 to 101.

Thanks,
JColvin

Will give it a try.

 

However, I do not understand what exactly you meant in the above response:  “presuming you have set SW11 to101”

Do you mean, I need to set SW11 to 101 in order to execute the linked prog app you sent ?

Note that the VC709 UG states to set to 010 (position 3, 4, 5) for BPI access, which is how we want to operate.

 

Thank you for the help.

[eskull]

Re, ground loops, what I meant by this, the SMT1  UBS port is connected (via USB cable) to an external computer potentially on a separate ground plane than what the VC709/SMT1 is connected to. This could cause a current to flow over the USB cable ground via the dissimilar ground planes. Something we hadn't considered before making this connection. 

Thank you again for  your time and help.

[JColvin]

Hi @eskull,

Digilent does not have this board to test for ourselves so I am not certain how the switches on the VC709 adjust and change configuration options, so I suggested the JTAG configuration as a safe option. I do not know if leaving it to be powered on via the BPI flash would change anything, though switching it to JTAG will not hurt.

I suppose with your setup there is a chance that a ground loop could have occurred though Xilinx will better know what sort of potential damage could occur with this board in terms of any grounding issues.

Thanks,
JColvin