I’m working on some factory testing for our WP85 based boards.
So far I’ve implemented some testing around sensor readings and digital inputs/outputs, but I’m looking to expand this testing to the hardware within the WP85 itself (in particular, cell modem and GPS). An example of this is a WCDMA failure we’ve seen on ~3 of our units, so it would be useful to have tests in place to catch this failure.
Has anyone defined a list of AT commands that could potentially catch something like this in a factory? I’ve had a look through the AT command reference for the WP85 but couldn’t really find what I was looking for.
Any guidance is greatly appreciated.
Have you taken a look at section 14.4 in the Product Technical Specification? (https://source.sierrawireless.com/resources/airprime/hardware_specs_user_guides/airprime_wp75xx_wp8548_product_technical_specification_rev_13/). Note the link might change if we publish an update. It’s Sierra document 4116440 (searchable on Source).
This section describes a test to verify the WCDMA transmit path, and there are similar tests for other access technologies and receive paths. Let me know if that doesn’t help.
Thanks for providing the link. Is there anyway to check WCDMA antenna power without a signal generator/power meter?
Hey @rkirk, I remember you had mentioned there were some AT commands that you had run that indicated the transmit power on the WCDMA module was stuck at 0 (or something like that). Basically we just want to test for that dead module on the WP8548 through an AT command, if possible. At this point we don’t necessarily need to test the transmit power of every unit as it comes off the line, we’ll only test every X number of units when they’re at our Canadian warehouse for those types of discrepancies.
Hi @coastalbrandon, @nvd,
In the case you’re referring to, the modem reported the transmit power to be normal, but the symptom was that the attach procedure was timing out in an environment where we would expect a quick network attach, based on testing with the same SIM and other samples of the same hardware, in the same location. We were able to see that the RX power was as expected (comparing to other working units), and the symptom of lengthy but failing attach procedures was a sign that the Network may not be recieving the signals that the device was attempting to transmit. We guessed at the root cause and proved it with essentially the recommended test in the PTS. In that case, the issue was in hardware, and the modem was not aware that the signals were not being effectively transmitted.
An end-to-end test on the live network, which exercises the transmit and receive paths in every band, might be a valid compromise in place of the recommended testing. Any reduced set of coverage leaves some degree of uncertainty.
Hope that helps,