Sensor circuit?

[metertech58761]

Can anyone tell me how this particular circuit is supposed to work? I THINK it is some kind of temperature sensor but I have no way to verify.

The output goes into a 6821 PIA, and the software reads the pin it's connected to and proceeds accordingly.

There are two different arrangements, depending on whether the board is part of the load management terminal (full circuit) or the field test unit (the left 3 components omitted, which apparently renders the circuit inert).

[SeanB]

Monitor for the 12V circuit, the Jfet is normally pinched hard off if the 12V is present, but if it is not there the Jfet turns on, pulling the comparator input to below the reference voltage, and thus pulling the output low. Little bit of hysteresis provided by R42, R43 and C28, so it switches cleanly as the voltage on the Jfet gate drops, as it has a RC filter on it that takes roughly 80 seconds to either charge or discharge, so the microcontroller can see, based on the state of the pin at power up, so if it takes roughly 2 minutes to go high, it is the load management unit, and if it is high on power on, or after roughly a second, it is a field test unit. Likely they all power up as field test for 2 minutes or so, before changing operation mode, probably to allow for calibration or adjustment.

[metertech58761]

With that, I would assume that it's a soft-start delay, likely to allow the transmitter and receiver daughterboards to warm up and stabilize before acting on commands.

There are two 12V supplies for some reason, one powering the I/O and transmitter daughterboards, the other powering the receiver daughterboard and bulk supply for the 5V regulator. This circuit sits on the receiver daughterboard.

The code for the field test unit does not check the output of this circuit at all, so eliminating those three parts allows that circuit to be "parked" with minimal modification.

OTOH, the load management unit apparently caches incoming data until the pin changes state.

[metertech58761]

Sorry for the 'necromancer bump', but I finally have a bit more insight on this circuit.

It seems there are two power outage monitors in the code - one to capture 'blinks' and one to capture power outages lasting over two minutes.

Am I correct in thinking that this would be the circuit to capture the longer outages?

[metertech58761]

Much appreciated. This helped confirm one part of the code and changed my understanding of another part.

As much as I really ought to just set aside this 'reverse engineering' project and work on other, more productive pursuits, I keep coming back to this thing and hacking away at it in a 'thousand cuts' approach.