Missing HDMI on custome Xavier carrierboard

[jzeigler]

Working with custom carrierboard that had previously shown no known issues with displaying HDMI. Currently unable to get either of the 2 HDMI ports to display at all, no HPD, CEC, or 5V measured from the HDMI cables (on the custom carrierboard though we measure 3.3V and 5V for our old carrier’s HDMI voltages and 3.3V for both the HDMI voltages on our new carrierboard). Tried switching to a new carrierboard with no luck with either the new or old. Checked kernel logs and looks like HPD is being asserted high correctly (see attached), but have been unable to identify the cause for this issue. Attached dmesg/kernel logs for assistance looking into them and identifying the HDMI problem and any other red flags that I may be missing within these logs. Please do let me know if you have any advice or questions, thanks!

(I think our JetPack version is 4.4?)

bad XAV kernel msgs from serial.txt (27.4 KB)
new carrierboard with working IO board, devices listed.txt (8.7 KB)
new carrierboard with working IO board, dmesg.txt (65.7 KB)
no eth or hdmi but good XAV kernel msgs from serial.txt (66.5 KB)

[WayneWWW]

Hi,

Is the HDMI pin in use same as the one on devkit or not? Is there only one HDMI port on the board or not?

My question here is could you just share me a log with only HDMI issue? I don’t care about any other kernel issues here and I cannot tell what are the difference of those 4 logs you attached by just the file name.

Please just make sure the hardware would work. I don’t care it is old or new one. This is only for yourself. Just use the board you want to test, connect the HDMI and share log.

[jzeigler]

HDMI pin is the same. Our board has the one HDMI port and then uses the DP as a second HDMI port (have not gotten that to work as of late though so we’ll focus on the one HDMI).

I shared the previous logs to give as complete of a scope I can on the software side (ctrl + F for HDMI), they all have the no HDMI appearing issue. It is the following lines which make me question the HDMI. I do not work in Linux much and could use some help understanding this prod list issue.

[ 1.672285] tegradc 15210000.nvdisplay: hdmi: invalid prod list prod_list_hdmi_board

[ 1.672294] tegradc 15210000.nvdisplay: hdmi: tegra_hdmi_tmds_range_read(bd) failed

I recently switched the hardware out somewhat and it brought HDMI back up (which indicates a hardware issue), but it still shows the above errors multiple times (see attached, ctrl + F for HDMI)

old carrierbaord with working IO board, dmesg, hdmi working.txt (68.4 KB)

Specifically in the previously shared cases which all did not have working HDMI, the following lines appeared. But in the log I just shared now, these lines do not appear?

[ 2.014559] mmc0: SDHCI controller on 3460000.sdhci [3460000.sdhci] using ADMA 64-bit with 64 bit addr

[ 2.020221] input: tegra-hda-galen-t194 HDMI/DP,pcm=3 as /devices/3510000.hda/sound/card0/input0

[ 2.020360] input: tegra-hda-galen-t194 HDMI/DP,pcm=7 as /devices/3510000.hda/sound/card0/input1

[ 2.020537] input: tegra-hda-galen-t194 HDMI/DP,pcm=8 as /devices/3510000.hda/sound/card0/input2

[ 2.020709] input: tegra-hda-galen-t194 HDMI/DP,pcm=9 as /devices/3510000.hda/sound/card0/input3

[ 2.022657] gpio tegra-gpio wake8 for gpio=55(G:7)

[ 2.022766] mmc1: SDHCI controller on 3400000.sdhci [3400000.sdhci] using ADMA 64-bit with 64 bit addr

Any advice I can get on this would be greatly appreciated, I will continue debugging the hardware side to see if there is any damage to the board or our cabling (as this board design was working well for HDMI previously before). Thank you so much!

[WayneWWW]

There are some points need to clarified.

1. Your system is already not the original one from jetpack. From dmesg, I can tell your device tree has been changed. If this is not done by you, you better clarifying with the one who did that can ask what did he change.

2. You should read the log from 15200000, but not 15210000. But again, I don’t have your device tree, so I can only tell from the log. Looks like you are still using nvdisplay 15200000 for HDMI as the devkit.

[ 15.781273] tegradc 15200000.nvdisplay: hdmi: tmds rate:148500K prod-setting:prod_c_hdmi_111m_223m

This log indicates system is able to read the mode from your monitor, 148500k is a common number which stands for 1080p@60 mode. And there is no other error here.

If you still cannot see HDMI monitor, please review your hardware. It is common that lane 2 and lane 0 didn’t connect right. Please refer to the design guide for detail.