cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Highlighted
Visitor
Visitor
636 Views
Registered: ‎10-13-2020

IPI self-test not work (Mouser S/O 568713_Lihua)

Jump to solution

part#EK-U1-KCU105-G

We have a customer feedback the KCU105 evaluation board suddenly didn’t work as usual!

Internal IPI self-test didn’t work now, the standard serial terminal also no  response!

But the enhanced terminal of system controller can work as usual. The temperature will rise up to a very high temperature as attached picture  

Could you help to locate and fix this problem? Thanks.

1.png
0 Kudos
1 Solution

Accepted Solutions
Highlighted
Moderator
Moderator
545 Views
Registered: ‎06-14-2010

Hello @JEGranizo ,

1. Is KCU105 powering on appropriately? You can validate this by checking the power LEDS in the Quickstart of this board: https://www.xilinx.com/support/documentation/boards_and_kits/kcu105/xtp391-kcu105-quickstart.pdf

 

Do you/customer sees DS3 as RED:

 

e.g.

 

image.png

 

Power good LED, DS3, illuminated green (normally) indicates power system is good. If in your case it is now red, this would indicated that there is some power issue encountered on this board and that there is some problem in power system.

2. If #1 above is not the case, and the power system is good, then

i) please perform the Built In Self Test (BIST). A tutorial on this can be found at the Quickstart Link Above.

ii) try running the KCU105 Board Interface Test (BIT) as per XTP345

BIT files can be downloaded using this link: https://www.xilinx.com/member/forms/download/design-license.html?cid=edfc4aa6-ce81-45bc-8143-1fc37eca0009&filename=rdf0311-kcu105-bit-c-2017-3.zip

Please let us know if any test failed.

Kind Regards,
Anatoli Curran,
Xilinx Technical Support
-------------------------------------------------------------------------
Don’t forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------

View solution in original post

13 Replies
Highlighted
Moderator
Moderator
546 Views
Registered: ‎06-14-2010

Hello @JEGranizo ,

1. Is KCU105 powering on appropriately? You can validate this by checking the power LEDS in the Quickstart of this board: https://www.xilinx.com/support/documentation/boards_and_kits/kcu105/xtp391-kcu105-quickstart.pdf

 

Do you/customer sees DS3 as RED:

 

e.g.

 

image.png

 

Power good LED, DS3, illuminated green (normally) indicates power system is good. If in your case it is now red, this would indicated that there is some power issue encountered on this board and that there is some problem in power system.

2. If #1 above is not the case, and the power system is good, then

i) please perform the Built In Self Test (BIST). A tutorial on this can be found at the Quickstart Link Above.

ii) try running the KCU105 Board Interface Test (BIT) as per XTP345

BIT files can be downloaded using this link: https://www.xilinx.com/member/forms/download/design-license.html?cid=edfc4aa6-ce81-45bc-8143-1fc37eca0009&filename=rdf0311-kcu105-bit-c-2017-3.zip

Please let us know if any test failed.

Kind Regards,
Anatoli Curran,
Xilinx Technical Support
-------------------------------------------------------------------------
Don’t forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------

View solution in original post

Highlighted
Visitor
Visitor
504 Views
Registered: ‎10-13-2020

Hi Anatoli, 

The customer replied as below,

1. I have done Board Self Test according XTP391 with both STEP2 and STEP3 are correct as expected:

  1. Power good LED, DS3 illuminated green;
  2. DONELED, DS34, illuminated green;
  3. The GPIO LEDs will turn on, then off, when SW3(PROG) pressed.

But the Board Self Test didn’t start as expected!

LEDs are shown as attached picture 1,

2. When try to run kcu105_bit, Standard COM port was not detected. But from windows Device manager, I can see these USB to UART bridge were installed.

Attached pic 2

USB drivers installed as instructed. Attached pic 3

2.jpg
3.jpg
1.jpg
0 Kudos
Highlighted
Moderator
Moderator
490 Views
Registered: ‎06-14-2010

Hello @JEGranizo ,

Thanks for the info. Yes, looks like the power system is good on the board.

Based on the serial ports section that you've shown, this behavior can be due to the way in which the CP210x driver is identifying the virtual COM port.

It sees it as "Silicon Labs Dual Silicon Labs Dual CP2105 USB to UART Bridge: Enhanced COM Port"

As you can see, the "CP210x" is replaced by a fully numerated name.

As per AR#68521 (https://www.xilinx.com/support/answers/68521.html)since the CP210x driver is showing up as CP2105 / CP2108 or anything other than CP210x, then it is necessary to change the SCUI configuration file (config.json).

Please ask the user to follow the steps as per AR above and then, SCUI should detect the board. 

Please ask the user to follow this and let us know the output then.

Hope this helps.

Kind Regards,
Anatoli Curran,
Xilinx Technical Support
-------------------------------------------------------------------------
Don’t forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------
0 Kudos
Highlighted
Visitor
Visitor
452 Views
Registered: ‎10-13-2020

The customer replied as below,

Sorry I can not find SCUI from kcu105_bit (rdf0311-kcu105-bit-c-2017-3.zip) bit package.
The serial port can be detected!

But kcu105_bit on my board still didn’t work as expected, vivado.log file attached.

Anyway, System controller (SYSMON) can connect with Tera Term

But FPGA (Standard COM port) Tera Term no response!
Could you help advise? Thanks.

Capture 1.JPG

2.jpg
0 Kudos
Highlighted
Moderator
Moderator
429 Views
Registered: ‎06-14-2010

Hello @JEGranizo ,

As an advice, if you'd like to get someone's attention in a post, please use @ symbol followed by their user name (e.g. @anatoli ) and that way that person would be contacted directly and get a notification of your response. I actually didn't see your response until i manually scanned the posts that have replies.

Thanks for the info.

Indeed, with older boards, we had variables.xml and not config.json (which is the new format for the latest Eval boards).

Anyhow, please ask the user to edit this variables.xml and update the below 2 lines from CP210x to CP2105 and save the file.

image.png

Once done, ensure that no other Terminal program is connected to the KCU105’s Standard or Enhanced COM ports while the Board Interface Test is running
– See XTP352, page 17 for details on COM Ports

As you can read from XTP352, 

The Standard COM Port is the FPGA UART COM Port
The Enhanced COM Port is the System Controller COM Port

Then please try again and see if this time the Standard COM Port (i.e. the UART COM Port) can now be detected by the Board User Interface.

Hope this helps.

Kind Regards,
Anatoli Curran,
Xilinx Technical Support
-------------------------------------------------------------------------
Don’t forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------
0 Kudos
Highlighted
Visitor
Visitor
378 Views
Registered: ‎10-13-2020

Hello @anatoli 

The customer replied as below,

After change variables.xml file as instructed, both COM port can be detected while the board interface test as the picture shown below: (attached pic 'COM port' )

But, it still can not run the test as expected: When  press button Read Board Info (Bottom Left of the window) , but no Serial Number or MAC ID updated in the display as in the picture below: (attached pic 'CAPTURE')

The vivado.log file attached below:

#-----------------------------------------------------------

# Vivado v2020.1 (64-bit)

# SW Build 2902540 on Wed May 27 19:54:49 MDT 2020

# IP Build 2902112 on Wed May 27 22:43:36 MDT 2020

# Start of session at: Fri Nov  6 11:37:14 2020

# Process ID: 5792

# Current directory: C:/kcu105_bit

# Command line: vivado.exe -mode batch -source C:\kcu105_bit\tcl\kcu105_eeprom.tcl

# Log file: C:/kcu105_bit/vivado.log

# Journal file: C:/kcu105_bit\vivado.jou

#-----------------------------------------------------------

source {C:\kcu105_bit\tcl\kcu105_eeprom.tcl}

# open_hw_manager

# connect_hw_server -url localhost:3121

INFO: [Labtools 27-2285] Connecting to hw_server url TCP:localhost:3121

INFO: [Labtools 27-2222] Launching hw_server...

INFO: [Labtools 27-2221] Launch Output:

 

****** Xilinx hw_server v2020.1

  **** Build date : May 27 2020 at 20:24:38

    ** Copyright 1986-2020 Xilinx, Inc. All Rights Reserved.

 

INFO: [Labtools 27-3415] Connecting to cs_server url TCP:localhost:3042

INFO: [Labtools 27-3417] Launching cs_server...

INFO: [Labtools 27-2221] Launch Output:

****** Xilinx cs_server v2020.1.0

  **** Build date : May 14 2020-09:10:29

    ** Copyright 2017-2020 Xilinx, Inc. All Rights Reserved.

 

connect_hw_server: Time (s): cpu = 00:00:00 ; elapsed = 00:00:08 . Memory (MB): peak = 1037.109 ; gain = 0.000

# current_hw_target [get_hw_targets */xilinx_tcf/Digilent/*]

# set_property PARAM.FREQUENCY 15000000 [get_hw_targets */xilinx_tcf/Digilent/*]

# open_hw_target -jtag_mode 1

INFO: [Labtoolstcl 44-466] Opening hw_target localhost:3121/xilinx_tcf/Digilent/210308AB1099

INFO: [Labtoolstcl 44-467] Setting hw_target localhost:3121/xilinx_tcf/Digilent/210308AB1099 into jtag_mode

# scan_ir_hw_jtag 6 -tdi 0b

# scan_ir_hw_jtag 6 -tdi 3f

# close_hw_target

INFO: [Labtoolstcl 44-464] Closing hw_target localhost:3121/xilinx_tcf/Digilent/210308AB1099

# open_hw_target

INFO: [Labtoolstcl 44-466] Opening hw_target localhost:3121/xilinx_tcf/Digilent/210308AB1099

# current_hw_device [lindex [get_hw_devices] 0]

# set_property PROGRAM.FILE {[pwd]/../bitstream/kcu105_iic_log.bit} [lindex [get_hw_devices] 0]

# program_hw_devices [lindex [get_hw_devices] 0]

INFO: [Labtools 27-3164] End of startup status: HIGH

# close_hw_target [current_hw_target [get_hw_targets */xilinx_tcf/Digilent/*]]

INFO: [Labtoolstcl 44-464] Closing hw_target localhost:3121/xilinx_tcf/Digilent/210308AB1099

# disconnect_hw_server localhost:3121

# close_hw_manager

****** Webtalk v2020.1 (64-bit)

  **** SW Build 2902540 on Wed May 27 19:54:49 MDT 2020

  **** IP Build 2902112 on Wed May 27 22:43:36 MDT 2020

    ** Copyright 1986-2020 Xilinx, Inc. All Rights Reserved.

 

source C:/kcu105_bit/.Xil/Vivado-5792-DESKTOP-NPQJ0I9/webtalk/labtool_webtalk.tcl -notrace

INFO: [Common 17-186] 'C:/kcu105_bit/.Xil/Vivado-5792-DESKTOP-NPQJ0I9/webtalk/usage_statistics_ext_labtool.xml' has been successfully sent to Xilinx on Fri Nov  6 11:37:38 2020. For additional details about this file, please refer to the WebTalk help file at D:/Xilinx/Vivado/2020.1/doc/webtalk_introduction.html.

webtalk_transmit: Time (s): cpu = 00:00:00 ; elapsed = 00:00:07 . Memory (MB): peak = 113.848 ; gain = 17.465

INFO: [Common 17-206] Exiting Webtalk at Fri Nov  6 11:37:38 2020...

close_hw_manager: Time (s): cpu = 00:00:01 ; elapsed = 00:00:08 . Memory (MB): peak = 1037.109 ; gain = 0.000

INFO: [Common 17-206] Exiting Vivado at Fri Nov  6 11:37:38 2020...

 --------------------------------------------------

Is these log help locate the problem?

COM port.jpg
Capture.JPG
0 Kudos
Highlighted
Moderator
Moderator
366 Views
Registered: ‎06-14-2010

Hello @JEGranizo ,

This Board UI was tested with 2017.3 version of Vivado. 

However, the customer is running the tests targeting 2020.1

Please ask the user to install just 2017.3 Vivado Lab Edition (which is much smaller than the entire Vivado 2017.3) and then, point Board UI to 2017.3 version of Vivado. I your case, it is 2020.1, and for the test to work, it requires at least the same version of vivado that it was build and tested with.

This cam be seem from https://www.xilinx.com/support/documentation/boards_and_kits/kcu105/2017_3/xtp345-kcu105-bit-c-2017-3.pdf

i.e.

image.png

Therefore, please ask the customer to navigate to https://www.xilinx.com/support/download/index.html/content/xilinx/en/downloadNav/vivado-design-tools/archive.html

and under 2017.3, there is a section called Vivado Lab Solution.

Here, the customer can download and install Vivado Lab Edition  

image.png

As you can see, this is around 600MB in size, and not 18GB of the full 2017.3 version of 2Vivado.

Lab Edition is potentially your Vivado Hardware Manager + Vivado Analyzer

If Vivado Lab Edition of 600MB is till too big to obtain and download, i believe the customer can just download and install standalone Vivado Hardware Server

image.png

Once done, the user can then open the Board User Interface and point to this Vivado HW Server or the Vivado Lab Edition (which even they install) and then do the tests. Let us know please how they get on.

Hope this helps.

 

Kind Regards,
Anatoli Curran,
Xilinx Technical Support
-------------------------------------------------------------------------
Don’t forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------
0 Kudos
Highlighted
Visitor
Visitor
316 Views
Registered: ‎10-13-2020

Hello @anatoli 

The customer replied as below,

I have tried 2017.3 Lab Edition, but still cannot get the correct response, the messages from the console are as below:

From board led indicator, it seems board can be programmed and initialized with the #proram_hw_devies tcl command, after this command, soon the board will become very hot, may reach a very high temperature!

1.jpg
0 Kudos
Highlighted
Moderator
Moderator
304 Views
Registered: ‎06-14-2010

Hello @JEGranizo ,

Complex issues like overheating are generally not trivial to solve.

Generally high temps mean the device is sinking a lot of currents.

If the problem only happens with their design then it could be that their design is drawing a lot of current and causing this issue.

Did you/customer try putting your design into XPE tool and see what the estimated thermal and power figures are? Make sure the fields are properly selected to reflect your design set up and board.

However, based on the customer's feedback, it sound that the same issue also occur with the design loaded with BIT test and not just their (customer's) own design? Can you confirm this please?

I’ve reached out to the higher level experts to see if they can advice further on this issue.

I will update you once I hear back from them and hope this is OK with you.

Kind Regards,
Anatoli Curran,
Xilinx Technical Support
-------------------------------------------------------------------------
Don’t forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------
0 Kudos
Highlighted
Visitor
Visitor
268 Views
Registered: ‎10-13-2020

Hello @anatoli 

The customer replied ' Up to now, we only tested with BIT test not our own design!'

0 Kudos
Highlighted
Moderator
Moderator
248 Views
Registered: ‎06-14-2010

Hello @JEGranizo ,

The temperature readings seems overly hot. You can see that for this part in the product guide, where the max temp listed is 85c. Something is definitely not right here. This high of a temp is never expected, and isn’t normal.

1. Can you please check with the customer and validate that the heatsink wasn’t modified or taken on or off on this board?

2. Did they disable the fan? Is fan actually operating OK or is it off altogether? 

3. For the very first screenshot from the customer, are we sure that there are no designs running on the FPGA in that tera term screenshot? That would seem to indicate that the board is sourcing current when it shouldn’t be. This could be due to a shortage on the board.

4. Do you know if that particular board is within 90 days warranty? If so, we may need to create an RMA SR and get this board replaced.

Please confirm on the above and let us know if you have any questions.

Kind Regards,
Anatoli Curran,
Xilinx Technical Support
-------------------------------------------------------------------------
Don’t forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------
0 Kudos
Highlighted
Visitor
Visitor
211 Views
Registered: ‎10-13-2020

Hello @anatoli 

The customer replied as below,

1.   A: We didn’t make changes to heatsink, neither taken heatsink off, nor put it on.

2.  A: We didn’t disable the fan, the fan seems working fine when power on!

3.  A: No customer’s design running on the FPGA. May on board IPI (in flash) running.

The customer ordered the part on 8/10/2020, from PO# 029-UF300, within 90 days warranty. Thanks.

0 Kudos
Highlighted
Moderator
Moderator
189 Views
Registered: ‎06-14-2010

Hello @JEGranizo ,

Thanks for the info.

With that, please go ahead with the RMA SR creations, to get their board replaced. Seems indeed an issue with the board.

To open a new RMA SR, please open this via SRM tool. If you don't have access to SRM and when you'd need to choose the type of this SR, please selected DevSys RMA.

Provide the info of this issue and please indicate this forum link in the notes, to state that WTS (i.e. Anatoli from Technical Support) have fully debugged the issue and we came to conclusion that there is an issue with the board and this needs to be replaced for the customer. That way they won't contact us separately for debugging, and will then continue with the next step of the RMA process.

If you have any questions or difficulties in opening a new RMA SR in the SRM/Compass, please let me know.

Hope this helps.

Kind Regards,
Anatoli Curran,
Xilinx Technical Support
-------------------------------------------------------------------------
Don’t forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------
0 Kudos