On a sunny day (Tue, 2 Feb 2010 11:10:12 +0530) it happened "pimpom" wrote in :
The white cable, actuallty a 2 core mains wire in parallel to get less resistance, is in series with an electric heater that can do 1kW or 2kW. I do not remember exactly the voltages coming from the PB head, but it is rather high level. You can test any PB head on the bench with a signal generator and a piece of wire, even a few mA should get a signal. Th orientation of the wire relative to the head gap is important, if the wrong way then zero output. The LM324 is a amp followed by a peak detector + rectifier, and outputs 0-5 V.
Running the (closed source) software in verbose mode shows: # hcs -v Data read: Binary 144 137 49 35 voltage 2.90 2.76 0.99 0.71 hcs: finside_temp=17.06 fvin2=0.99 heater=0 heater_setting=2.00 heater_on_minutes=1, heater_off_minutes=0 heater_duty_cycle=100.00 ...
There are 4 analog values measured here, with a PCF8591, outside temp, inside temp, heater current, and altitude^H^H^H^H^H^H^H^Hsomething else. The software measures .99V from the LM324, and does a level compare with some preset values, and from that decides the 2kW element is on... And uses that then to calculate electricity use and cost. There is no need to do an exact analog measurement, as the heater settings result in is a fixed resistance. I have published the circuit here years ago, would have to open it up again to see what it was, and I am not going to, as it is in use. One of thsoe things you solder together one afternoon and then works for let's see: 12 years now?