For the last few days I have been helping Atcom debug the V1.3 Mesh Potato – this is the prototype of what will be the production Mesh Potato. Note the two printed omnidirectional antennas at one end of the PCB:
It also sports an integrated FXS port (this was a separate module on the V1.2 Betas). There was a bug with the FXS port that Atcom asked me to take a look at – no transmit or receive audio. After a few hours of work I found that the 48V DC was present, and Asterisk was detecting on/off hook events. This suggested that most of the FXS port was working OK, just no TX or RX audio. For example there was no dial tone in the phone and DTMF digits weren’t being detected by Asterisk.
I poked around with the oscilliscope and found nothing on the DRX net (audio signal from phone) from the FXS module. However when I looked at the DTX net (audio signal to phone) I saw this:
A “trinary” (3 level) signal! When I took the phone off hook and hit a DTMF key I could see the signal changing, a sign that the DRX signal was mixed up with the DTX signal. The “trinary” nature was due to two outputs driving the DTX net at the same time, e.g. a 0V and 3V3 averages to 1.65V. I guessed that this meant the DRX and DTX signals were mixed up. A little research of the earlier V1.1 design proved the point – the DRX and DTX nets had been swapped accidentally, and we had missed the subtle change during review. With some skillful PCB work Alen swapped the nets and reported dial tone re-appeared on his V1.3 PCB.
Apart from that the V1.3 design seems to be working OK. Next step is to get the Wifi side calibrated and test the printed antennas.