Polarity Sense Circuit

without an OP-Amp or other transistor your limited by the Diode voltages and also the VGS Threshold of the FET so without making it much more complex it seems you hae a good solution or what you needed, A logic level FET may improve things a bit

Peter

I hadn't thought about a logic level FET. I bet that would bring me down another volt or two. I will have to order one up as I only have them in N channel.

Thanks John

How about this:

You can tune R4 to get the threshold in the right place - it won't be that stable because it depends on Q1 Vbe and gain and stuff which are temperature dependent.  You could use a single rail comparator chip with rail to rail input capability.

MK

Why not use an opto-isolator?  You could shunt the input diode with a reversed signal diode. 

Hi Michael,

I will breadboard the circuit tonight and play around with it to see what it will do. Thank you for the nice suggestion.

John

Hi Don,

I actually tried to get an opto-isolator to work. It was doing the job but I could not get it to work over a wide input voltage range. It might be that a better selection of components would have fixed the limits. The opto-isolator I was using had a 1.7V max 20 mA LED. I put diodes in parallel with the OI's LED to shunt current when the voltage exceeded the 1.7 volts and as I mentioned it was working but if I had it current limited for any reasonable voltage level ( think 30 V) then there would not be enough current to turn on the LED of the OI at 5 volts and lower. In any case I will probably revisit the idea and do some more experiments. Thanks for your idea and input.

John

Sounds like a good use for a spare LM339 comparator : http://www.element14.com/community/message/157410/l/lm-339-comparator-question#157410

The best polarity-sense circuit for this depends a lot on how much input impedance you want on the test inputs.  If you'd like them to be high-impedance, a comparator IC is a good way to go.  If you need electrical isolation and don't mind a high current draw, then an opto-isolator is good.

so a few ground rules

No Battery ??

minimal load on circuit...

any more ?

John

The question I would ask is "What are the Inputs"

If you are trying to ensure your battery discharger is correctly connected, then the loading of the inputs are fine.

However if its something high impedance then its possibly going to drag the inputs down.

It's a very interesting issue, and the only other thought I have is a DC-DC inverter chip.

They have a wide range at the bottom end (from 0.7v), but might struggle at the upper end.

It could allow a two fold approach ... something for the bottom and something for the higher voltages.

The best I saw for a 12v system was a diode feeding the coil of a relay, and the contacts applied the voltage if it was correctly connected.

Mark

There are a few ways to go with that one, John.  As I understand it, you want a sensitive circuit, but you want it to not be vulnerable to overload.  The simplest thing to do is clamp your input with a backward Zener and a resistor.