GEOFEX - Cheap and Good Polarity Protector

[alparent]

OK. I was thinking about this over the weekend.

The big(est) deal about simply using a diode is the 0.6v it sucks out! Right?

Then I thought! Why not use a LED instead (I'm going to put one in anyways!) OK not that simple 'cause I need a resistor so I don't kill the LED. But then after I don't have enough voltage to run my effect!!! Even worst then using a diode.

But then I thought, why not use a transistor in there? On one side I have the 9v + LED + resistor that feeds the trans. If the polarity is good the LED lights up and latches the transistor. The transistor acts like a switch that is feeding the un-altered 9v to the effect

(Don't know if my explanation is clear?)

What the best way to go about doing this? What's the best trans. to use? Or maybe this idea is not as cleaver as I think it is? Please educate me  
 

[PRR]

> about simply using a diode is the 0.6v it sucks out! Right?
> Why not use a diode instead....

? ? ?

> Don't know if my explanation is clear?

Somethng is not getting through. Do you have a crayon?

Do you mean "_LED_ instead"?

The diode designer gets to pick a Reverse Voltage Breakdown. It can be 500V for heavy rectification, 50V for small-signal. When "no" reverse voltage is expected, the usual doping gives a 5V reverse breakdown.

So if you put 9V into a 5V breakdown junction, you still have reverse 4V to threaten your circuit.

> why not use a transistor in there?

Like R.G. did? Maybe with Merlin's simplification? Maybe with the LED in the base path, but you still need the 50V diode. That gives 0.6V+1.7V+0.6V= almost 3V drop, and significant turn-on current variation as the battery drains from 9V to 6V.

> If the polarity is good

The brain-busting question is: If the polarity is WRONG, what happens?

We amplifier designers have one quadrant of the transistor equations memorized, the only "useful" one. But when working with reverse voltages, you need at least the diagonal quadrant at your fingers.

[alparent]

Thanks you for educating me...........I think?

Somethng is not getting through. Do you have a crayon?

Is this condescension or I'm I getting the wrong impression?

Isn't this forum suppose to be about learning?

[tomfoy]

I built the "cheap and good" circuit as presented on R.G. Keen's site (also depicted on the first page of this thread in simulation), but it didn't work for me. I got .6 something reverse voltage; heated up my TL072 pretty good. I had it connected to two 100k resistors in series, forming a voltage divider for the op amp input, to give me a buffered virtual ground.

Sorry I don't have the skill to correct the problem. It's a great idea, though. I'm using a series Shottkey instead.

[DiscoVlad]

This is kind of related, I didn't see the need to start a new topic when this is still being worked on...

This is the supply/biasing for an opamp big muff that I'm currently building. Given that the original didn't have any reverse polarity protection I thought it would be a good idea to put something in.

Q1, and R18 are the actual reverse polarity protection, while D7 is there for protecting the mosfet from high gate voltages. The BS250P can handle up to 45V so at least for battery powered pedals is optional (ignoring ESD... probably at my peril ).



The BS250P MOSFET isn't horribly expensive (RS Components part 215-6688) like it used to be, and the three parts fit into a 0.3" square on the board, current with reversed supply is in the micro-amps region if at all.

It's not perfect (nothing is) but for battery powered stomp boxes is IMHO the best solution.