4/3/2014— Jon McKay
Tessel version nine arrived at our office yesterday. It came in yesterday straight out of Worthington Assembly’s ovens.
Each new hardware revision means a whole slew of tests to make sure we fixed any known problems and haven’t created any new ones.
Eric is testing out how robust the board is after merciless shortings on both USB input and external power lines. He had designed new power circuitry to try and prevent the board from being fried in the event of a short and preliminary results are looking promising.
I have a test rig to put our power switching circuitry through the hoops. Tessel can be powered off of USB or external power (up to 15V). We want users to be able to program Tessel over USB, then detatch out the USB cable and have their program continue running off of external power (if it’s connected) without a hiccup.
Yes, the header on the old relay modules were assembled backwards and yes we are embarrassed. This relay module is plugged in upside down, and if you do that with the production models they won’t work.
My test rig (above) has both USB and an external LiPo battery connected to Tessel at the same time. The USB power is intercepted by a debugging circuit board (below) that allows us to splice off USB power into a relay connected to Tessel.
That allows Tessel to tell the relay to either connect the USB circuit or disconnect it. That is, Tessel is testing its own power regulation circuitry by switching its input power between USB and the LiPo. If the switch fails, Tessel will stop blinking lights because the program is stored in RAM and will be erased on power failure. Tessel testing itself is a beautiful thing.
It both terrifies me and excites me that this revision of the board seems to be working to spec (knock on wood). If all goes well, we can be set to start up production early next week.