Actions
Forum Thread Details
  • State Suggested Answer
  • Replies 20 replies
  • Answers 1 answer
  • Subscribers 359 subscribers
  • Views 5441 views
  • Users 0 members are here
Related

Polarity Sense Circuit

jw0752

In one of my recent builds it was necessary to come up with a circuit that would sense if the inputs were hooked up reverse polarity. I came up with this circuit:

 

image

I have tested this circuit to 100 volts DC on its Test Input terminals. If the Test Inputs are under normal polarity the output is LOW and if the Test Inputs are under reverse polarity the Output is HIGH. Unfortunately the Test Inputs will not sense a reversed polarity below 3.5 Volts. While these limitations were not a problem for my build I am curious if there are other better circuits to test polarity than what I cooked up here. Also if you see problems with the circuit above any critique would be appreciated. R1 limits the forward current through LED1 which acts like a zener and keeps the voltage Gate to Drain from exceeding breakdown limits. D1 is reverse biased under proper polarity. R2 keeps the P CH MOSFET off until there is a Negative Voltage applied at the gate by a reversed Test Input polarity.

 

Thanks John

  • 0 in reply to johnbeetem

    Hi John

    I like the thinking. I clicked on the link you provided and I was looking at my own behind. I actually tried to make the circuit originally with a comparator but I did not do a good job with the input circuit so I move on to a different approach. Since the inputs I was testing and the comparator had a common ground with the comparator my reverse polarity pulled the inputs of the comparator well below Vss. Now I am thinking that I can get around that problem with a split supply or a better design of the circuit that feeds the inputs of the comparator.

     

    My continued quest on this subject is motivated primarily by curiosity. The circuit in the DC Load is working well within parameters for my application. Frequently, however, an application catches my interest and I like to push it to see what the limits are and what other approaches there are to the problem.

     

    Thanks for your input and for reminding me of the unused LM339 sections.

    John

  • 0 in reply to Robert Peter Oakes

    Hi Peter,

    There are probably not many applications for this type of circuit or there would be more literature. The need I had in my DC Load Unit has been satisfied and right now I am just experimenting and playing with different ideas. Since I really do not have a pending need the ground rules are a little flexible. My imagination suggests that my goal would be a circuit that had minimal load on the circuit whether it was hooked up proper polarity or reversed. I would want it to cover a range of voltage from say 1 volts to 80 volts and it should be fast enough so that the inputs can be disengaged from the main circuit before any damage is done. Right now I have 4 or 5 new ideas to investigate thanks to you, , , , and .

    Thanks John

  • 0 in reply to D_Hersey

    Hi Don,

    I plan to follow up and do some more experiments with the Opto-isolator. I will incorporate your design ideas.

    Thanks John

  • 0

    I don't know what your application is for this polarity sense circuit, but if you're worried about DC power inputs being hooked up incorrectly you might consider using a Schottky Diode Bridge for your DC inputs.  That way it doesn't matter which way you hook them up.

     

    When touch-tone phones were first introduced in the USA, they required that the red and green telephone wires had the correct polarity.  However, since the polarity didn't matter for dial phones, lots of houses were wired incorrectly.  The touch-tone phone could receive calls OK -- the mis-wiring just made the keypad fail so you couldn't dial out.  Rather then try to fix everybody's wiring, phones now have diode bridges so it doesn't matter.

  • 0

    The FWBR trick is a good one.  So often the two drops it implies are an indifference.  Unit works plugged in either way.  If you can only tolerate one drop, use a single diode, works one way, doesn't the other.  The math fn the FWBR performs (in a perfect world) is 'absolute value.'  'Rectifier' seems so old-fashionably self-righteous.  Love those earlier terms.

  • 0 in reply to johnbeetem

    I remember the advent of touch tone. I was working for the Radio Shack and they were really big into aftermarket phones. We did a lot of phone service. In fact if you look at the picture of my shop the big white unit on the top shelf is a telephone tester. Today it is just about as useful as my BK tube tester and tube manual.

     

    Your suggestion for the bridge is a good one for many applications. With the schottky bridge we would only loose about 0.5 to 0.8 volts.

     

    Thanks John

  • 0 in reply to jw0752

    Hi John,

     

    I once took apart a very old touch tone phone (probably a trimline), it was interesting to see that they used analog tone generation circuitry! using pot core wound transformers for the tone generation : )

    I bet that rapidly drove the need to develop the digitally generated tones in more later phones : ) Those old phones must have cost a fortune, or be rented from the phone company I guess.

    So I can see that having a phone tester for troubleshooting phones must have resulted in very good business for Radio Shack!

  • 0 in reply to michaelkellett

    Hi Michael,

    I bread boarded the circuit that you supplied and ran it through some experiments. It does a very good job of detecting the polarity of the input. Even as low as 5 mV it will switch between HIGH for reverse polarity and LOW for correct polarity. I was making the problem too hard as usual. Simple and elegant. Thanks

    John

  • 0 in reply to jw0752

    I'm glad it works in practice as well as simulation. Sims are really useful but they never quite tell the full story.

     

     

     

    MK

  • 0

    Many op-amps and comparators have a common-mode range that extends a fifth of a volt below ground, nowadays.  This circuit could be accurate and fast, made this way without the need for a calibration step.  How low can the reversal go?  Is it clamped by a diode?