Start a new topic

Problem with FMC112 on ZC702

Dear Friends,

I'm working on data acquisition using FMC112, on ZC702. I followed chapter 11.3 from 4FM_Get_Started_Guide.pdf precisely, to evaluate the precompiled design files.

After completion of programming in ISE cmd, teraterm output remains at "Waiting for PHY to complete autonegotiation".
In addition, evaluation of the command "Fmc116APP 1 0 0" returns "could not obtain sipcid ..."

The point is that I've already designed an Gig.Ethernet data transfer bw the ZC702 board to PC and actually achieved gig throughput. So the firewall and other problems associated with the link is simply not the case.

Any input is very welcome.

Dear Karen,

I think you must be using a Rev D ZC702 where we have done our works using Rev C ZC702.

Can you please try to modify download_elf.tcl, comment out the rst like so:

fpga -f $bit
connect arm hw
source ps7_init.tcl

That should get the fabric not upset, it looks like there is an issue with the new Zynq silicon or differences in their toolchain causing rst command to crash the PS.

Can you confirm?

Best Regards,
Dear Arnaud

Thank you for your good insight, since it solved the first part of the problem, and I get "StellarIP TCP/IP server started on port port 1000" in teraterm.

Question1: Is this an indication that my zc702 is rev. D? since wheneven I designed using XPS, i set it to the only avalable option: rev. C.

What about the command for data acquisition "Fmc116APP 1 1000 0" am I missing something?

if i enter:  Fmc116APP 1 0 0    i get :could not obtain sipcid ...
if i enter: Fmc116APP 1 1000 0  i get : could not open device 1000 ...
Dear Karen,

1) I don't know about that. We both have a revc and a revd and it wont work on revd without commenting the rst line. I don't know why this is that or even if revD are fully supported by Xilinx tools yet. This is something you should assert with Xilinx , via webcase support as it is not really related to our design.

2) You should use 2 instead one as the design uses TCP/IP, Fmc116APP 2 1000 0

I hope that helps,

Thanks again, I'm so happy with my purchase from 4DSP, since you are here and answering the question from time to time, I know the forum is extremely crowded with questions and I appreciate your help ;)

the command "Fmc116APP 2 1000 0"
returns the following. Could you please imagine what might be the problem?

[i]"Connected FPGA device type: UnIDENTIFIED DEVICE[/i]

[i]start of program[/i]

[i]FMC supported :FMC112 on ZC702[/i]
[i]could not detect FMC116 hardware, exiting"[/i]
Dear Karen,

Can be two reasons:

- The first one would be the SW12, GPIO_DIP_SW0 which should be set properly, set to "0" which actually drives the FMC present signal.
- The second one, PG of the ZC702 ( might not be set properly. You can check whether or not PG is asserted, if not you will need to follow steps in this guide.

Second one should cause most of the voltages reported wrong in the console output.


Dear Arnaud

the SW12 is in the correct state.

Do you have any idea whether there is any way to verify the assertion status of the PG signal??, without having to buy a "TI USB pod to update the TI controller" ?

Dear Karen,

I guess you can guess where to probe looking at the ZC702 schematics.

Best Regards,
Dear Arnaud,

Thanks for all of your advice, I think you are right and the problem is about the PG signal, I'm going to order the cable to reprogram the TI controllers.

I'm pretty sure that my FMC112 card is working as expected, since I got a clean signal using KC705 board.

You may now close the topic, since acquiring the cable is taking quite a while.
Dear Karen,

Thanks for the feedback.

Feel free to open new topic any time you need to do so.

Best Regards,
This topic is being closed because the issue is considered as resolved by 4DSP. Feel free to create a new topic for any further inquiries.