A DIP switch is a nice idea for higher volume. A DIP switch could be read directly as high-low and use I/O pins on the MCU. A DIP switch would need at least 2 pins, so ATtinyX
4 would be required. The current board uses all pins except the pin 1 on ATtinyX
5.
ATtinyX5 pins assignments on the controler are:
8 - VCC
7 - Analog temperature sensor
6 - Optocoupler for 12V input trigger
5 - Momentary button
4 - Gnd
3 - Control of MOSFET based switch for the external circuit up to 15A
2 - Control indicator LEDs (High/On and Low/Off states)
1 - Reset
Two options:
A. Update the constant in the code and update the MCU with the updated firmware.
B. The code uses MCU pin 7 through JP3 to read the voltage from the the precision LMT87 (
http://www.mouser.com/ds/2/405/lmt87-373050.pdf) temperature sensor. The code switches to the off state, if the sensor reports the voltage corresponding to temperature above 55C, which is defined through a constant. The 55C temperature threshold corresponds to 1888mV or lower on the JP3 as per the datasheet for LMT87. So at long as the supplied voltage is below the threshold, the controller will switch to the off (low) state.
On the physical side, the SMD version of LMT87 is small. I unintentionally flooded the pins and then used he solder wick to remove extra solder on the sensor mounting boards. Through hole TO-92 version of the sensor is available as well.
The PSU for the board is switching RAC01-05SC for 68% efficiency. The board has its own fuse as Amb recommends (
http://www.amb.org/forum/transformer-goof-t1786.html?hilit=fuse#p16532) . The holder is for Littelfuse TR5/TE5 fuses. These are neat. The fuse values go down to 0.05A! (
http://www.littelfuse.com/products/fuses/axial-radial-thru-hole-fuses/tr5-fuses.aspx)
I calibrated my ATTiny85 chips using Fluke 287 and the available code (
http://provideyourown.com/2012/secret-arduino-voltmeter-measure-battery-voltage/). The controller's code reads PSU voltage before taking a measurement from the sensor. Prior to calibration the temperature reading was >2.5C off. After calibration the reading is within 0.1C against the reading by Fluke 62.

You do not have the required permissions to view the files attached to this post.