DSI failure in room meed temperatures


We have noticed that while the module TX2i temperatures read from the module are in the range +5⁰C up to +31⁰C there is no DSI video coming from the module. Currently we do not know the nature of the problem but the FPGA down the road which converts the MIPI signal from the module to HD-SDI receives no video, while when the FPGA is requested to display HD-SDI (in the same temperqature range) there is video output from the FPGA – so the problem is between the module and the FPGA.

We did not notice this issue until now because during development, the modules raise to ~50⁰C while in room temperature. During extreme testing the tests were applied in the extreme temperatures -40⁰C and +71⁰C (environment) only. Though, while in the chamber test we have tested by mistake the DSI output while the module was in the range mentioned above (chamber temperature was lower than +25⁰C – room temperature). We’ve tried to replace modules thinking that the manufacture batch is to blame, we have replaced our boards, but with no success the failure is constant and reproducible.

Please advise.

Hi, did you do cross check to verify if module can display with dev kit carrier board? Basically TX2i module should have no problem in this temperature range. And i think HD-SDI signals can pass FPGA successfully doesn’t mean FPGA itself has no problem to run DSI signals successfully.

Also is it possible that the module was damaged partly in your “wrong extreme test”? Do you have another module to do cross check again?

Finally you can probe the DSI output signals with oscilloscope to confirm if the DSI part has normal output.

Hi, we did not check it with the dev kit board.

We have tested a few combinations of modules and boards and on all we got the same results.
The tested modules worked in room temperature only and were not exposed to extreme temperatures, anyway modules that were exposed show the same behaviour.

The signals of the DSI are flowing from the moduel to the FPGA, though the problem (after analyzing with a chipscope on the FPGA) that once in a while the type of the packages (per frame) is provided incorrectly which causes the FPGA to drop the whole frame. Again and we move the temperature of the module to below 5 degress or above 31 degrees all data flows correctly adn HD-SDI is displayed.

BTW, we are using 16bpp @ 297MHz, and the lanes between the module and the FPGA are short so we removed any signal integrity issue as being the cause.

Any suggestions?

Still suggest to do crosscheck with dev kit and DSI displays to confirm if DSI output has problem. Or you can use oscilloscope to get the DSI output signals waveform, then analyze it and compare to DSI spec.


At the momnet we do not have the option to check it with the dev kit.

The wave form is there, as I explained, there are loses of a few bits in the communication, that it is, a few times within a frame there is an incorrect bit coming into the FPGA causing it to not identify the H-Sync packet. so the waveform of the signal is the correct most of the time.

Does NVidia have results of testing DSI in the range of temperatures we have? how the TX2i is tested in these temperatures? how is it approved for being an industrial grade?


TX2i module has been validated under industrial spec. Per current situation, if you think it is module issue, you can run RMA process to find out root cause.

Warranty & RMA Process
Go to NVIDIA Support
Select “Live Chat” or “Ask a Question” from the options near the top of the page.
Enter your personal information.
Select the appropriate product from the drop-down list (“Tegra” in this case).
Submit the request.
If the Service agent is unable to resolve the technical issue, an RMA (Return Material Authorization) will be initiated.

Our internal check suggests to back to comment #2, the module should have no problem in such temperature range, for further investigation, please focus on FPGA side, you can check if any clock shift on the FPGA side first.

you are correct we have been tyrin to cool parts of the board during DSI output and we’ve found that there were some issues with the clock distributor’s control circuit. So I am going to accept your reply, until we have other news. Thanks.