Using LED in place for zener fro mosfet static protection ok?

[liddokun]

I'm building a SHO, and I was doing some reading and it seems there's a version 1 and version 2 of the circuit, with version 1 using 2 diodes (first diode cathode to ground, anode to gate, second diode cathode to gate, anode to +9dc), and version 2 using a 9.1v zener connected between the source and gate (cathode to source) Both applications are used for static protection of the mosfet.

Anyways, my first question is, which method is better in terms of providing static protection? I hear some people prefer the zener method. The problem is, I've already bought diodes for verseion 1 circuit, and have no zeners.

Then I noticed that ROG's peppermill uses a red led in place of the zener for static protection. Does the led have the same effect?

http://runoffgroove.com/peppermill.html

Did a few searches, but coudln't find anything relevant. Perhaps my search terms weren't that great.

[WhenBoredomPeaks]

Real men don't use protection diodes, it is only for the weaklings.

[sorry]

[PRR]

> which method is better in terms of providing static protection?

Which is better, a belt or suspenders? As long as your pants don't fall down.....

The protection must pass a volt of signal and bias, but must clamp before ~~40V happens. (MOS spec-sheet may say 20V.) That's a lot of leeway.

In radio systems, the capacitance (nominal and variational) of the different diodes might muck-up tuning, but in guitar systems any diode is "small" compared to wiring and cable effects.

> ROG's peppermill uses a red led

This depends on the LED reverse breakdown being greater than the FET turn-on voltage (and also less than 20V-40V). Both parameters are loosely specified. FET turn-on "matters", so we have good numbers, albeit a wide range. LED reverse breakdown is often specified >5V, and almost always IS 5V-6V, but in many systems it is a "does not matter", and I have seen (cheap old) LEDs leak bad at 3V.

> version 1 ...diode ..ground.. to gate, ..diode ...gate ...to +9dc
problem is, I've already bought diodes for verseion 1 circuit, and have no zeners.

I don't see a problem, except you have to solder two extra leads. Clamping to ground and +9V is excellent protection. Using a 5V-20V Zener is also excellent, and saves one part and two joints. Using an LED is also excellent protection, just one part.

[liddokun]

Thanks a lot Paul. I guess I'll just stick with my two diodes.

I didn't mean to ask such an obvious  question, I suppose it just shows how new at this I still am. 1000 or so posts full of questions later in this forum, and still at it.
Hahaha.

Sorry if this comes up a lot!

[petemoore]

  Takes very little current to break over the 20v thick, glass gate.
 
    http://www.sciencemadesimple.com/eliminatestaticelectricity.html
  Couldn't find any numbers, I was looking for a static number greater than 20v potentail. 20,000v rings a bell though.
  We don't get much static and I tired of hunting Zeners and wanted the gate to have enough swing-room, so skipping the zener protection on the later Mosfet builds. Come to think of it, 'once in the circuit, you're ok' doesn't mesh with the scenario of the cable tip falls to create an arc just before landing on the synthetic carpet.

[80k]

was looking at the specs of my BS107 and it already has a protection diode integrated into the MOSFET. So I think it is optional with a lot of new MOSFETS?

[PRR]

> BS107 and it already has a protection diode integrated into the MOSFET

NO.

It has a drain-source diode. This is actually a "parasitic", not really wanted, but usually acceptable and not easily avoidable.

It does NOT protect the Gate and the hyper-thin "glass" that pete mentioned.

> skipping the zener protection on the later Mosfet builds.

This will work..... until the day it doesn't. Murphy's Law applies strongly. MOSFETs fail "for no reason". When you realize "static", and rig a test, often they don't fail when you think they should. Sell them to customers and they come back dead. So yeah, omit the protection, but keep spare MOSFETs and a hot soldering iron.

> 'once in the circuit, you're ok' doesn't mesh

That ill-logic actually works in radio tuners. The Gate is fed from 7 turns of #16 wire to ground or some substantial bias. Once you nail it in, it isn't going very far. You are often OK for MOSFETs buried deep inside a box, with multiple stages of amplifier in front.

Putting a MOSFET right on an external input is reckless. Say I had a dollar: if I leave it by my Front Door, someone will snatch it. Even if I "tie it down" with a long thin thread (or 1Meg resistor). If I tie it down with an inch of #16 wire, it's pretty safe. Or if I move it deep inside the house it won't be very exposed to outside influences.

Diodes are wise. Any of the usual plans work, and all nearly the same. 2 diodes is classic and obvious (but only for supplies under 20V). A single Zener works (even >20V supply), and saves soldering. The LED works 999 out of 1000, and 'round here, more geeks stock LEDs than plain or Zener diodes.

> I tired of hunting Zeners

Got transistors? The emitter-base junction of any common Silicon NPN is a 7V Zener.