Last Sunday the ADC1 net on the first SM1000 prototype went open circuit all of a sudden. After messing about for a few hours I lifted the uC pin for that net and soldered a fine wire to the other end of the net. That lasted a few days then fell off. I then broke the uC pin trying to put it all back together. So then I tried to use some Chip Quick I had laying about from the Mesh Potato days to remove the uC. I royally screwed that up, breaking several pads.
It’s been 7 years since my last surface mount assembly project and it shows!
However when the uC came off the reason for the open circuit became apparent. The photo below was taken through the microscope I use for surface mount assembly:
At the top is the bottom part of a square pad that is part of the ADC1 net. The track is broken just before the lower left corner of the pad. Many of the pads under the uC were in various stages of decomposition, e.g. solder mask and tinning gone, down to bare copper. Turns out I used too much flux and it wasn’t cleaned out from under the chip when I washed the PCB. For the past few weeks it was busy eating away the PCB.
Oh well, one step back! So this week I built another SM1000, and today I brought it to life. After fixing a few small assembly bugs I debugged the “switches and leds” driver and sm1000_main.c, which means I now have PTT operation working. So it’s normally in receive mode, but press PTT and it swaps to tx mode. The sync, PTT, and error LEDs work too. Cool.
Here is a picture of prototype number 2:
The three trimmers along the bottom set the internal mic amp, and line levels to the “mic” and “speaker” ports of the radio. The pot on the RHS is the internal speaker volume control. The two switches upper RHS are PTT and power. On the left is a RJ45 for the audio connections to the radio and under the PCB (not visible) are a bunch of 3.5mm sockets that provide alternate audio connections to the radio.
What next? Well the speaker audio is a bit distorted at high volume so I might look into that and see if the LM386 is behaving as specified. Then hook it up to a real radio and test it over the air. That will shake down the interfaces some more and see if it’s affected by strong nearby RF. Oh, and I need to test USB and a few other minor interfaces.
I’m very happy with progress and we are on track to release the SM1000 in beta form commercially in late 2014.