News:

Attention: For security reasons,please choose a user name *different* from your login name.
Also make sure to choose a secure password and change it regularly.

Main Menu

Recent posts

#1
Trenz Electronic FPGA Modules / Re: Problem with TEM0007 DDR s...
Last post by John - November 04, 2026, 11:38:47 AM
Hello cone83,

did you wrote a email to support@trenz-electronic.de?

br
John
#2
UltraScale / Thermal solution for TE0820
Last post by AdiE - November 04, 2026, 06:43:10 AM
Hi,
I am a thermal engineer working in a team which seeks to integrate TE0820 in our system.

I have seen the proposed cooling heatsink for the PCB (and specifically the UltraScale+ MPSoC).

Are you able to provide some thermal parameters for this cooling heatsink? For example a thermal resistance in units K/W or any thermal experiments results that you conducted to prove the cooling heatsink design.

My system works in 55 Celsius reaching a steady state so cooling the MPSoC is really an important part of the integration.

Thank you!
#3
Trenz Electronic FPGA Modules / Re: Secondary UART from TE0720...
Last post by John - November 03, 2026, 07:19:25 AM
Hi,
did you checked that TX/RX is maybe swapped. X17 should be TX and X16 RX (view from FPGA)?
Did you you try a simple design with only uart1, maybe it's some driver issue for multiple uart?
Which vivado version did you use?
Can you tell me the serial number of the carrier and the module. It's the number of the smaal sticker with QR code.

br
John
#4
Trenz Electronic FPGA Modules / Secondary UART from TE0720 sem...
Last post by DrFluteDos - November 02, 2026, 11:44:50 AM
I have a bare metal application running on a SoM TE0720 that pushes messages out to both PS uarts and also to PL axilite uart. I can receive these messages on my development PC from PS uart0, and PL axilite uart, but not PS uart1. According to the TE0703 carrier TRM (see attached image), the SoM uart1 Tx and Rx lines (nets MIO12, MIO13) pass through the carrier CPLD and onto the VG96 connector (nets X16 and X17). However, there is a suggestion that this is CPLD firmware dependent. I've download the CPLD source code and it seems fine. However, I can't tell what version of CPLD code is actually currently running.

Can you help please.

Kind regards
   
   
 2026-11-02 10_36_07-TRM-TE0703 TRM-v.45 -03_21_2024 - DuckDuckGo.png
#5
Trenz Electronic FPGA Modules / Re: Boot image compatibility b...
Last post by blee - October 29, 2026, 03:58:02 AM
Hi,
I successfully booted the board using the prebuilt images, as well as the custom images generated from the .xsa file.

Thanks a lot.
#6
Trenz Electronic FPGA Modules / Re: Boot image compatibility b...
Last post by John - October 26, 2026, 02:37:13 PM
Hi,
labtool fmeter is not critical it's a simple frequency counter to calculate frequency of the connected clocks. so the timing ignore constrains is missing in the design for these interclock relationships.
When you download the reference design again, this should include prebuild binaries. Can you use this one please and not the generated one from your regenerated project (only to narrow down the issue).

From the boot log it looks like you has a problem with emmc (if I haven't confused it with the SD (I'm not entirely sure which of the two is the controller).

Can you tell me also the variant name and  serial number of the module which you has bought. I want to make sure that you has select the correct assembly variant.

br
John
#7
Trenz Electronic FPGA Modules / Re: Boot image compatibility b...
Last post by blee - October 23, 2026, 06:51:19 AM
Hi,
I downloaded the official reference design package (TE0955-te0955-skrd-vivado_2025.2-build_1_20260724094052.zip) and executed the command "TE::hw_build_design -export_prebuilt" for Hardware Build and Export.

However, I encountered the following issues during implementation and board execution:
 -> According to the timing report, setup violations occur on asynchronous Inter-Clock Paths (aclk6 to clk_pl_0)     
 -> attached files are "Design Timing Summary& & aclk6 to clk_pl_0" & clk_pl_0 to aclk6"

And, Even though the .xsa file was generated despite the timing issue, building the image with PetaLinux and booting the board results in the system hanging and freezing during the Linux kernel boot process.
Specifically, the kernel repeatedly triggers timeouts at the MMC controller initialization (mmc0: Timeout waiting for hardware cmd interrupt.), followed by an RCU CPU stall (rcu: INFO: rcu_sched detected stalls on CPUs/tasks), causing the boot sequence to completely stall.
 -> attached file is boot log data

Thanks
#8
Trenz Electronic FPGA Modules / Re: Boot image compatibility b...
Last post by blee - October 22, 2026, 02:32:20 AM
Hi,
I tried to follow your support guide recommends (a):
"a) (Recommended) Recompile your design with different configuration/mapping: UART0 instead of UART1 for use with TEB0955."

However, when following this recommendation inside the provided Vivado project, the versal_cips_0 IP is locked by AMB0010-01 board presets and design constraints, making it impossible to select or configure UART0 directly through the GUI or standard property changes. (used Official reference design:TE0955-te0955-rd-vivado_2024.2)

Could you please confirm if downloading and using the latest 2025.2 TE0955 Reference Design with Linux example resolves this issue, or provide guidance/a Tcl script patch to properly compile the project for the TEB0955-02 carrier board using UART0?

Thanks.
#9
UltraScale / Re: ps_aux on TE0817
Last post by Martin R @Trenz - October 20, 2026, 06:13:12 PM
Hi,
sorry for not answering for such a long time, this should not happen.
Without making more measurements and test, it is hard to guess what is going on.
I would expect some back sourcing or loading of the rails.
I checked our oscilloscope bring up measurements for TE0817 REV02 on a TEBF0818 base board:
At the state you describe (LP_DCDC is up) there is no other voltage visible  until they are powered up in their sequence. To be fair, we measured only the normal startup sequence without stopping. So we see that zero of PS_AUX and PS_1V8 only a few hundred milliseconds. Meaning if there would be a very slow loading we would have missed it.
Did you already find out what is going on?
Did you measure the time dependency when this voltage appears? Who does the rise look like? Does it coincide with some other rail? Adding a small load (e.g. 1K) to the Testpoint of the rail could also give you a hint, e.g. does it respond like a voltage divider?

Best regards
 
#10
Hi,
Quote from: martinius on October 15, 2026, 03:03:34 PMOh! thanks. I did not notice that so far. Then it is clear to me.
Is populating myself already an modification regarding warranty?

Note: these POF have to be soldered. Any case where you have to use a solder iron may affect warranty because it is often no more possible to figure out if a failure was already there before.