Mosfet protection diode...

[lapsteelman]

On most boost circuits I see the diode (usually zener) go from the gate to the source... on others I see it go from the gate to ground.  Is one method better than the other?

[antonis]

Protection diode is for Gate-Sourse reverse current flow (n-channel) prevention..

In case it's orienteted from Gate to GND, we're talking about grounded Source configuration..
(no resistor between Source & GND)

[R.G.]

Antonis has it right - place the zener between gate and source. It's gate to source voltage you want to keep from getting too big.

[rankot]

Protection diode is for Gate-Sourse reverse current flow (n-channel) prevention..

In case it's orienteted from Gate to GND, we're talking about grounded Source configuration..
(no resistor between Source & GND)

Is it really necessary to use that kind of protection in pedals?

[iainpunk]

yes, static electricity builds up on hands, and can easily fry some mosfets

cheers

[antonis]

Is it really necessary to use that kind of protection in pedals?

IMHO, for 9V single supply, we only have to beware of transients and ESD..
In general, Zener voltage rating should be lower of particular V_GS absolute max rating..

edit: Iain types faster ..

[iainpunk]

Is it really necessary to use that kind of protection in pedals?

IMHO, for 9V single supply, we only have to beware of transients and ESD..
In general, Zener voltage rating should be lower of particular V_GS absolute max rating..

edit: Iain types faster ..

don't sweat it, i was the quickest in typing class. (that's 12 years ago)

cheers,

[rankot]

Is it then necessary to put them only on input and output stages, or every MOSFET stage must be protected?

[lapsteelman]

Is it then necessary to put them only on input and output stages, or every MOSFET stage must be protected?

In many MOSFET boost pedals the input is coupled through a single capacitor to the gate so It seems it would be fairly susceptible to static damage... farther down the line of the circuit perhaps not so much?

[R.G.]

MOSFET gates and channels (the conductive part) are separated by a layer of high purity glass that is 20-30V thick. That is, more than the max voltage, and the gate insulation ruptures from voltage stress, and the MOSFET is no longer functional.

If your circuit is not capable of causing more than the rupture voltage in stages other than the input or output stages, you don't need protection diodes on the "inside" devices. Unfortunately, this requires that you know whether you can get bigger voltages than are listed on the datasheet for your MOSFETs between gate and source. Well, OK, maybe you can just depend on being lucky.   

Input MOSFETs are open to the worst sort of abuse. The voltage on the input of a pedal or amp connected to a phone jack cord can be anything that the cord tip can connect to. Static electricity zaps on the end of the cord are probably the worst. This can kill MOSFETs outright, but it will also make bipolar transistor bases noisier too. This is one reason old solid state gear is noisy - the inputs have been zapped too many times and now are permanently noisy. New transistors on the inputs cures this, mostly, and a protection diode reverse across the input transistor base-emitter is highly effective in permanently preventing this from recurring.

Any circuit with inductors in it needs some consideration as to whether the inductor can ever generate more than about 7V if connected to a bipolar base, or 20V if connected to a MOSFET. If so, you need a protection diode for the bipolar or a zener for the MOSFET.

[Rob Strand]

Is it then necessary to put them only on input and output stages, or every MOSFET stage must be protected?

You need to protect if that part is exposed to ESD.    For input and output stages it's wise to include protection.  Typically the output is protected with diodes or zeners on the output line; it's better to have a resistor between the output line and the protection devices.   On the output side, a gate-source zener could be considered optional.    You might consider protection on the gates of MOSFETs which connect to pots and switches, especially if you have metal knobs, as an ESD zap on the front panel can find its way to the MOSFET gate.

[R.G.]

I suppose the simple way to look at it is - if it has a connection off-board, protect it.

[anotherjim]

You can generate "wrong" voltage on inputs within the circuit in some circumstances.
Consider part of the circuit at 4.5v bias feeding on to the next stage which is ground referenced via a DC blocking cap.
The cap will have +4.5v stored across it. If something happens to drop the input side of the cap to 0v, the output side must go negative to -4.5v. You'd expect it to discharge via the surrounding circuit and do no harm, but are you sure?

[R.G.]

You're right A.J. The original source of my advice on not breaking bipolar base-emitters actually recommended putting a reverse protection diode on every base-emitter where the diode's reverse capacitance didn't disable the circuit, as in RF work. It also included the advice that circuit capacitors could store enough voltage to cause temporary breakdowns. Coupling caps was one of these; emitter bypass caps was another.

Pedal work with 9V power is pretty safe with typical 6-7V BVbe devices today, but there are instances where it can happen.

Of course, it's possible to go overboard with this. I put reverse protection diodes on bipolar inputs these days as a matter of course, and give at least a cursory look for the idle voltage of emitter caps and coupling caps. But I don't put them on every single device.