Transmitter failure on the nRF51 ?

[Roger Clark]

I've been running a hand soldered prototype board for about a week in an outdoor location where it gets hot and cold (mainly hot), and I noticed it was no longer visible on BLE.

I checked that the nRF51 was responding to the debugger, and uploaded my application with Logging turned on, and as far as I can tell, the MCU is running as it sends strings to the log just like normal.

However I can't detect it at all on BLE, either on an iOS or Android device.

Its possible that the antenna has a dry joint, but I am surprised that if this has happened, I can't detect the device at all, even right next to it.

I presume this is possibly because the external components of the antenna are somehow part of the active circuit and there will be no transmission at all without those components being connected.

Can someone confirm this is the case, i.e a dry joint could completely stop transmission.

BTW. I realise if there was a short on the antenna to ground that could have the same effect, but in this case I think an open circuit is more likely, as it was running for several days before it went wrong

[Michael Antwih]

If your antenna is disconnected from your chip via dry solder, open circuit or component failure; your system will no longer be able to emit rf at 2.4ghz.

Have you probed the circuit to see if you have any open connections or shorts?

When you designed your system did you account for the environment when choosing your components? The failure might be due to component ratings.

Anything that disconnects the chip from the antenna will be seen in the transmission and reception performance. Your device might still be transmitting because could be on a different frequency now since the 2.4ghz radiating element is not in the path.

[Roger Clark]

Thanks Michael

I've checked the board with a usb microscope, and I can't see any dry joints or shorts.

IMHO a short is unlikely as the board initially worked, but when I put it in its plastic enclosure and left it outside (it measures temperature and pressure and other environmental factors), for a few days, I noticed it had stopped transmitting.

Its likely it got reasonably hot in the sun, perhaps 60 or 70 deg, but AFIK that's with the operating spec, and also I would expect it to recover.

The strange thing is that its only the RF that's failed, the rest of the MCU and all the peripherals are still working fine.

I didn't solder the board myself, it was made by a design house in China, and seems quite well soldered (for hand soldering with an iron)

I may run my reflow air tool over it and see if that helps, otherwise I think its going in the bin ;-)

[Michael Antwih]

Do a continuity check first within your RF path. First make sure you can go from series component to series component and then check all shunts to ground to make sure there aren't any failed components. Are you able to share an image of your RF section? Or you can just place it in the bin but it would be nice to know the failure so the next board up doesn't encounter the same issue.

[Roger Clark]

I probably won't put it in the bin. Just in the stack of boards that don't work any more (mainly not nRF51's)

If I get chance I'll try to check the continuity, but I now have 200 machine made boards, which are just the same, so I will try putting on of those in the same position and see if that fails as well.

[ovrebekk]

If everything works except the RF it is often a crystal issue. The radio will not work without an accurate 16MHz clock source, but all the other parts of the system will.