This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

CC8520: PPW Commander RF Test issues with my hardware designs

Part Number: CC8520
Other Parts Discussed in Thread: PCM1808, PCM5102A, CC2590, CC8531,

Hello,

I am having a frustrating time trying to figure out why PPW Commander is not working, or at least not working consistently, with my hardware designs. I am new to the world of PurePath wireless and have worked with the CC85XX design kit for a number of months. Recently we have produced prototype PCBs, one as a dedicated transmitter with on-board PCM1808 ADC and one as a dedicated receiver with PCM5102A DAC. The performance of these prototypes has mirrored that of the DK very closely with no issues with programming, paring, or signal strength in real world use. As we get closer to a production design, I have started investigating RF and audio performance using PPW Commander and the included tools. I have two CC Debuggers and have run tests using one of the DK boards as Tester and the other DK board as Master or Slave. I can perform prod_test_simplerf_auto and prod_test_simplerf_tx on the DK boards with no issue. I repeated testing with both the supplied antennas and with an RF cable connecting both antenna jacks. Almost without fail everything with the DK boards works correctly.

When it comes to my hardware, using one of the DK boards as Tester or one of my boards as Test, I am running into all kinds of issues with functionality, test results, etc. I am concerned that some of the built-in tests are geared towards the DK hardware and might not be fully compatible with other hardware designs. I noticed one instance where having pull-up resistors on the I2C lines would produce strange results with PPWC:

I should note that my RF sections are perfect copies of the eval board with CC2590 attached and I am using right-angle SMA surface mount RF jacks in the identical positions to the eval boards. All critical RF components are the identical type and manufacturer to the eval boards. I don't use the I2C bus on any of my hardware. My programming connector has all the connections required as outlined in the CC Debugger manual for the CC85xx family. I have no issues with programming my hardware using PPW Configurator.

Here is what I am seeing:

When I run prod_test_simplerf_auto and press Run All I get inconsistent errors including Communication Errors, SPI Errors, etc. These don't happen consistently and sometimes the specific channel Rx or Tx test will complete properly after a series of SPI Errors. This same test with the same CC debuggers on the same computer runs fine and consistently using the DK hardware.

If I manually go through each sequence in this test the test program loads perfectly every time without error. Each Rx or Tx test can usually be completed without an error going one by one, but sometimes I still get a Communication Error. Doing this manually I can pass all Rx channels and Tx on Channel 18, but both Tx CH1 and CH9 fail consistently regardless if I am using antennas or and RF cable. This behaviour is exactly the same using both my transmitter and receiver PCBs. 

Running each step of prod_test_simplerf_tx on the DK hardware using a second instance of PPWC as an RSSI Plot monitor, and running the Tester as receiver, I can see the Tx signal running continuously until I say Passed in the prompt window when the Tx signal disappears. Using the same setup but substituting my transmitter for one of the DK boards I can still see the Tx signal for all three transmission channels, but the signal only lasts for 1 second and then disappears. It's almost as if the software thinks I have pressed the Passed button but I have done nothing. What;s interesting is that Tx Channel 18 behaves the same way but passes during prod_test_simplerf_auto when performing that step individually!

This behavior is the same if I use my receiver as Master Transmitter. Something is not adding up and I'm really confused as to what might be causing all these issues. Again, the only effective difference between the DK and my hardware are differences in CODEC/ADC/DAC. Even the voltage regulators used are the same as the DK. I would really appreciate any help as I am running close to a deadline to release these designs for production.

Thanks!

AW

  • Hi Andrew,

    I have notified our PPW expert about your struggles with PPW Commander. As this is Easter week please excuse a delay before coming back to you on this.

    In the meantime, have you looked at the SPI communication from the Debuggers with a logic analyzer to see if there are any differences between your HW and the DK?

    Cheers,
    Fredrik
  • Hi Fredrik,

    Thank you for responding so quickly and for passing my issue on to your PPW expert. So far I have only had a chance to compare the SPI lines on the MASTER side while performing Tx Test Channel 1 of prod_test_simplerf_tx. All CC Debugger lines (MISO, MOSI, SCLK, CSN, and RSTN) behave identically during this test on both the DK and my hardware. Tomorrow I will check the lines on the TESTER side to see if there are any differences on the receiver end during this test. I will also compare the lines during prod_test_simplerf_auto. Somewhere something is different!

    AW
  • I have not been able to find any functional difference in the CC Debugger communication between the DK and my hardware. Something very strange is going on and I stumbled on this:

    See the note from a Ti employee regarding programming a production test image for your setup? What constitutes a production test image? I programmed my normal Master or Slave code that I have confirmed to work with the attached audio devices on my boards. Is there something different that needs to be done for this testing? There is a woeful lack of any proper documentation for this potentially very useful software, other than a VERY minimal set of help pages.

    One thing that dawned on me is how the software knows about the audio configuration when performing the audio loop test in PPW Commander? Is it pulling this information from the loaded code and using that as part of the test code that is being loaded? How can I look at what the hex test code is actually doing? Maybe this is why things are working for the DK and not my hardware?

    One test I also performed is to take the DK out the equation and use my transmitter and receiver hardware with one programmed to be the Tester in PPW Commander. The result? ALL RF tests fail, both Rx and Tx, with antennas or a hard wired connection. Something is obviously very wrong with this software and non DK hardware as I am able to program with no issues and the system basically WORKS. Now I need a way to confirm the operation of production boards before sending them off to customers. Hopefully your PPW person will be able to look into this soon.

    AW

  • Oh great...now I have discovered other users who had the SAME problem with the Tx test signal running for only one second when using their hardware and not the DK, but it's a locked thread with no resolution!?
    Could this be some sort of bug related to the fact that the DK uses CC8531 and my hardware uses CC8520? It seems the others with the same problem were also using CC8520.
  • Hi Andrew,

    Have you programmed your board with a customized production test image?

    For your project, go to "Flash Programming" . In the "Select Device Configuration" section change "Image Type" to "Production Test". This file must be programmed on your DUT.

    You also need to make a custom ppwptc file, this is well documented in the Commander documentation.

    Regards,
    Fredrik