LEDs and reverse polarity

[Ukuleray]

I'm trying to build a preamp for a piezo pickup in acoustic guitars.  The project is a mash-up of Mr. Tillman's preamp and Mr. Keen's polarity protector.  I plan to build two versions of this project.  The first will be powered exclusively by a 9V battery with a polarized snap.  The only time the polarity should ever be reversed is when a new battery is being installed and then it would be very brief and obvious that the snap doesn't fit onto the battery.  I've considered just using a diode to protect the preamp in that version.

The second version of the project will use a battery compartment that does not prevent accidentally inserting the battery backwards.  It will also have the option of a wall wart.  That version will definitely need the polarity protector circuit.

The problem is that if the assembly doesn't work, I cannot distinguish between a dead battery and blocked reverse polarity.  I came up with what I think is a solution using a two-color LED but being far from an electrical engineer, I'm not sure this is a good design:



D3 in the diagram is a two-color LED.  When the assembly is powered correctly as shown, the upper (green) portion of the LED will light but only when both plugs are inserted in the input and output jacks.  I'm not sure I like having D2 in the circuit and drawing power but I think it's necessary (or is it?).

When the assembly is powered in reverse, the lower (red) portion of the LED will light immediately to alert me to the problem, whether the cables are connected yet or not.  Again, that should be a brief situation so D1 drawing power for a few seconds isn't a big concern.

In normal operation, the red LED should be off and the green LED should be on unless the battery is dead.

Will this circuit work the way I'm thinking it will?  Is there a better way to go about this?  I'm not opposed to using two LEDs if that matters.  I would like to keep the number of components to a minimum and take up only minimal real estate.  I'm trying to learn and hope you'll share your thoughts on this project.

If you're not familiar with the Tillman and Keen circuits, here are a couple of links:

http://www.till.com/articles/GuitarPreamp/index.html

http://www.geofex.com/Article_Folders/cheapgoodprot.htm

Thanks!

Ray

[DavenPaget]

When it is reverse-biased you MUST put a 40-100k to be safe , use a waterclear .
http://www.tablix.org/~avian/blog/archives/2009/11/reverse_biased_led/

[Ukuleray]

Hi.  To be clear, I'm not trying to light the LED with the reverse polarity current.  I was hoping to use the diodes to redirect the reverse current into the red side of the LED as forward polarity current.  What I hadn't realized but have since learned is that an LED in a reverse polarity circuit will not only NOT illuminate but it WILL continue to draw current.  So with correct polarity, both LEDs would draw current although only the green LED would be on.  That's not what I want.   

I guess I need a more sophisticated design.  Since the Keen circuit recognizes and responds to reverse polarity, is there anywhere in that circuit where I could add an LED that would only be lit when the normal polarity is reversed?  In other words, can I put an LED backwards in his circuit so it will only light when the circuit is blocking reverse polarity and have it not even be exposed to any current when the polarity is correct?

[gritz]

Wire your led in reverse polarity with regard to the battery input with a signal diode (1N4148) in series and "pointing" the same way as the led to protect it. Add a series resistor (say 1k) to limit current. It'll light up if the battery clip is momentarily in contact with the battery in reverse polarity, but won't let out the magic smoke when everything is connected up properly.

I'm not overly sure why you want to do this though. There are plenty of ways to invisible protect a circuit from reverse polarity.

[earthtonesaudio]

I usually like polarity protection, but this seems overkill for a Tillman preamp.  The only vulnerability in that circuit are the polarized caps.

But if you truly desire wrong polarity indication, use a 2-leaded, "tri-color" led.  These usually light up red one way, green the other, and yellow for AC.  This only requires one resistor.

[Ukuleray]

Thanks guys.  Maybe this is a bit of overkill.  I think we're all saying the same thing, though, about the diodes and the LEDs.  That's what I tried to show in my original diagram.

The Keen circuit provides invisible polarity protection.  It's a simple addition to the Tillman circuit and it gets the job done.  If the polarity is reversed, the Tillman preamp won't come on and I was hoping for a way to easily tell what was wrong.  It could be a bad cable, a bad battery, a reversed battery or a circuit failure.  A pair of LEDs or a two-color LED would let me know if the battery was good and installed properly.  But maybe all of that isn't really necessary.  It's not an issue with a regular 9V battery snap.  It goes on only one way.  But I want to build this into an aluminum box with a battery compartment that I've discovered will allow the battery to be inserted either way.     The circuit is still protected, but I thought it would be helpful to easily be able to confirm that the battery is installed correctly.  OTOH, if I replace the battery every month or two, and even if I install it wrong every time, it's not like I have to deal with the problem every day.

Originally, I was planning to add a "power" LED between the Keen circuit and the Tillman circuit to remind me to unplug (turn off) the system to avoid draining the battery.  I may still do that.  It's pretty straight forward.  If it's on, the battery is good and connected correctly.  That got me thinking about the reverse polarity indicator.... and maybe I'm just getting carried away.   

Anyway, an LED wired "in reverse polarity with regard to the battery input with a signal diode (1N4148) in series and "pointing" the same way as the led to protect it." will draw current regardless of how the battery is connected, won't it?  In that case, I'm always wasting battery power and that's not efficient.  I was hoping this was going to be about adding an LED and a couple of diodes and Bob's your uncle.  But it doesn't sound like it's quite that simple.

Thanks for your input.  I feel like I at least learned something in the process.

Ray

[Ukuleray]

But if you truly desire wrong polarity indication, use a 2-leaded, "tri-color" led.  These usually light up red one way, green the other, and yellow for AC.  This only requires one resistor.

The two-color LEDs I had in mind were all three-lead devices which complicated things.  I just found some two-lead two-color LEDs that simplify the situation immensely.  That would give me the power-on indication I want and the occasional reverse polarity indication without (presumably) wasting power on an unlit but still in-circuit device.  Good call.  Thanks!

[PRR]

> I'm not opposed to using two LEDs

It would make the intent clear at a glance... no asking others what this does or will this work.



Try this and see if the reverse voltage in-circuit is over 8.9 volts, implying nearly-no waste current in the un-lit LED.

You don't have to worry over-much about reverse voltage on LEDs.

If you did, nobody could DIY a PC... since the connectors are poorly marked the front panel LEDs always get reverse voltage until trail-and-error straightens them out.

The reverse breakdown is at least 7V and on modern ones often close to 50V.

Your plan works. Compared to my plan it is one hole cheaper. My plan allows running the OK lamp stage-dim and the "Whoops!" lamp way-bright.

> I'm not sure I like having D2 in the circuit and drawing power

What "drawing power"? You pick your LED current. You are going to spend that much of 9V. There's 1.7V across a red LED. Your diode adds another 0.6V. The resistor must blow-off the excess 6.7V. Without the diode that resistor would have to blow-off 7.3V. Total power drain the same either way, and based on a suitable LED current and the fact that battery voltage is far above what the LED needs.

[Ice-9]

+1 for PRR's above diagram, except i would leave out the "Green/Happy" LED. you only require the Red/Backward LED in to warn the battery is reversed. No need for the green one to constantly draw your battery down.

[DavenPaget]

+1 for PRR's above diagram, except i would leave out the "Green/Happy" LED. you only require the Red/Backward LED in to warn the battery is reversed. No need for the green one to constantly draw your battery down.

He wants a "notification LED" 

[earthtonesaudio]

I like this idea too, having separate LEDs - in separate places, with their own label "backward!" "happy", is crystal-clear.  Protecting the LED from a large reverse bias voltage may be overkill but if you put the LEDs in parallel* (while keeping their directions as shown in Paul's drawing), you can omit one of the current limiting resistors.  The trade-off here is you can no longer easily get two different brightness levels.  On the other hand, resistors are cheap, and obvious indication is always nice, so maybe Paul's way is the best.

*like this

[Ukuleray]

>

Thanks, Paul.  Your comments helped a lot.  Ultimately, I think yours is the best approach.  The only question I have is about the diode in your drawing.  It's there to provide reverse polarity protection for whatever is connected to the right of the drawn circuit, right?  It has nothing to do with the LEDs?

[Ukuleray]

Wire your led in reverse polarity with regard to the battery input with a signal diode (1N4148) in series and "pointing" the same way as the led to protect it. Add a series resistor (say 1k) to limit current.

Do I need the extra signal diodes to protect the LEDs?  The only reason I had diodes in my design was to accomodate using a single resistor which needed to go to + in one case and - in the other depending on battery polarity.  With separate resistors and LEDs, I wouldn't have thought I needed diodes.

[Ukuleray]

I like this idea too, having separate LEDs - in separate places, with their own label "backward!" "happy", is crystal-clear.

Agreed.  As much as I like the coolness of a single 2-lead LED, I think separate LEDs is the better way to do this.  Not only can I control brightness and current drain with separate resistors, but I can also control where the LEDs are placed in the circuit (before and after the switches in my drawing).

Protecting the LED from a large reverse bias voltage may be overkill...

Do you mean by adding diodes to the LEDs and resistors?  Under what circumstances could a large reverse bias voltage get to the LEDs and what would be the result?

[gritz]

Leds are sensitive to being exposed to reverse voltage - datasheets generally give 5 volts as a maximum. Just a few days ago I accidentally killed a white led by applying 9 volts "backwards" - and that was through a current limiting resistor! The most elegant solution imo (as has been mentioned already) would be to connect the two leds in inverse parallel (i.e. nose to tail), sharing a common series resistor. A red led's forward voltage is about 1.8V, and for green it's about 2.2V (depending on current and whatnot), so each will protect the other from excess reverse voltage and untimely demise. If you're set on using the above diagram, then you'll need protection diodes (and that extra resistor). Hope this helps.

[PRR]

> the diode in your drawing.  It's there to provide reverse polarity protection for whatever is connected to the right

Right.

It is semi-customary to have a diode to protect the circuit. I hadn't realized you planned R.G.'s dandy trick circuit instead of a dumb diode. Whatever.

Do note that MANY 9V transistor circuits are not very prone to reverse-polarity disasters. I admit that predicting reverse failures requires mind-flopping analysis which is "more costly" than a 1N4007 or even an R.G.'s RPP circuit.

[PRR]

> datasheets generally give 5 volts as a maximum

Yes, but that's an old-old spec based on most LED circuits using little to no reverse voltage, and seems to be out of sync with present production.

OTOH, there is your dead White LED. It is very hard to picture how it would die with 9V though any reasonable limiting resistor. Maybe white has failure modes not found in good old Red and Green.

[gritz]

> datasheets generally give 5 volts as a maximum

OTOH, there is your dead White LED. It is very hard to picture how it would die with 9V though any reasonable limiting resistor. Maybe white has failure modes not found in good old Red and Green.

The failure mode was me not wearing my glasses! It was a pretty low value resistor too - 470R, although in theory it shoudn't have exceeded the dissipation whatnot. When I refitted the led the right way round it it up dimly for a moment and then karked. Leds are not zener diodes!
You may well be right about white ones being extra fussy.