BC118 Temperature Dependent Operation

We need application support for the BC118.

We’ve been having issues with about 5% of our BC118s where they don’t initialize properly or start to broadcast.
We’ve noticed that by heating the units in a thermal chamber or with a hot air gun, the BC118s initialize correctly and communicate as expected. If the units are subsequently cooled, they continue to operate correctly, But a power cycle at low temperature (<0°C) will result in initialization failure and failure to broadcast.

Here is some identifying info. regarding the units that exhibit this behavior:
MAC Address Lot Code
20FABB0B5C3C 1847
20FABB0B66C8 1847
20FABB0B671B 1847
20FABB0B62EC 1847

Someone on my team has also sent in the following 2 units for RMA (W2223572) which exhibit similar behavior:
20FABB046982
20FABB0B5E3E

Is this a known issue? Or is anything pertaining to initialization extremely time/temperature dependent?
(We have ensured that the 3.3V rail reaches 1.8V in less than 1.5ms per the data sheet.)
Perhaps something is marginal in either the BC118 or our supporting circuitry.
I would like a reference design, but have not found one online.

Any help is appreciated.
-Kyle

Here are the relevant supply schematics in our application:

And here is the BC118 portion of the schematic:

@kylehaston,

I would suggest you go back to your channel and push to have a design review performed 0 degrees is not very cold and the units should work quite happily there and/or under it.

Regards

Matt

We test down to -40°C, and our unit meets or exceeds our spec. except for [roughly 5% of] this one peripheral, which does work unless initialized below roughly 0°C. We’ve had several design reviews and tried many topology changes, but we’re always open to new ideas/suggestions.

To that point, can you please direct me to all application info. specific to the BC118?
I currently have the document: “AirPrime BC118 Product Technical Specification” 41110676 Rev 2

I will be examining the initialization process at hot and cold temperatures more thoroughly, and will post here if I find anything significant.

Thank you in advance for directing me to the existing knowledge base.
-Kyle

@kylehaston,

All of the available technical documentation is on the source, including the PTS, which you have which I guess means you have already accessed the below page (you need to create yourself a login)?

https://source.sierrawireless.com/devices/wifi-bluetooth-modules/bc118/

As I say it is unusual because the PTS does state that the unit is operational down to -40 (that does generally mean that you can power it up down to this level, not that it will just stay on.

Regards

Matt

I’ve disconnected the BC118 from our MCU and repeated the same start-up test at -20°C and room temperature.
The BC118 fails to initialize when it is cold (as witnessed by its Tx pin over a 10 second period.)
Results can be seen in this online gallery: https://imgur.com/gallery/xrzyxc3
I’d like to start an RMA process for this sample. Can you help me with that?
-Kyle

@kylehaston,

For RMA’s you need t push it back through the commercial channel from which you bought the unit.

Regards

Matt

Here are the results from the RMA:

Please see the analysis result below,

MAC addr. Product name Investigation result
20fabb046982 BC118 Firmware in EEPROM was changed or corrupted. Module can be recovered after FW re-flash
20fabb0b5e3e BC118 EEPROM hardware fault

In both cases, something “under the hood” of the BC118 was misbehaving.
I have to ask what we can do differently?
And is this is a known issue? (With know resolution?)

If Sierra Wireless is interested, we have several more units that exhibit behavior that is out of spec.