understanding the millenium...

[Toney]

I have built the millenium2 with a 2n7000 and it works great...
I think its an ingenius solution.  Really clever.  Well, if you're me thats really clever anyway.
So rather than just build it, I'd really like to understand it. I have read GEO on this three times and searched..... Could be too many beers over the years, but I have struggled a bit on a couple of points...

Here's how I understand it (or think I do)

In the millenium 2, a mosfet is turned on via slight leakage of a 1n914 diode through its junction.
This small positive leakage is sufficient to "gate up" the fet and complete a pathway to ground from the fet's source to it's drain and then (through the appropriate resistor) to the led, thus lighting it.
This condition occurrs when the "control" connection is put to the circuit's ground .... and works to switch on the gate, when there is appropriate dc resistance to ground.

So without the dc RESTISTANCE ...if the control was going straight to ground .... the gate would just not switch.  

Is this correct?

And , could anyone help me understand HOW the second protection diode
actually protects the fet?

many thanks....

[R.G.]

You don't need much if any help. Your understanding is correct. I'll toss in a couple of things.

This condition occurrs when the "control" connection is put to the circuit's ground .... and works to switch on the gate, when there is appropriate dc resistance to ground.

So without the dc RESTISTANCE ...if the control was going straight to ground .... the gate would just not switch.
Is this correct?

To be perfectly pedantic, I would have said "works to switch off the gate", since the voltage is in the direction that makes the drain-source channel quit conducting. But yes, it's correct.

And , could anyone help me understand HOW the second protection diode actually protects the fet?

The fragile part of the FET is the gate. The gate of a MOSFET is insulated from the channel of the MOSFET by 20 volts worth of high-quality glass. Corning Glass says

Glass has such a high dielectric strength that it is difficult to measure, and the results can be interpreted with the least certainty. Various techniques have been tried, including testing under oil and using special sample shapes designed to prevent flashover around the sides of the sample. When it is important to know this property, it is best to measure it under conditions that approach actual service condition as closely as possible.

So it's fairly amazing that they can get the insulating layer that thin, but that's the case.

If you get the gate of the MOSFET more than about 20V higher or lower than the source, it ruptures the insulating layer and the MOSFET becomes a Darkness Emitting Diode (a DED).

The bottom diode actually works with the top diode to protect the MOSFET. The power supply that the MOSFET works from sets some limits and (presumably) can conduct static electricity away from the MOSFET if it gets the chance. The diodes provide the chance. The top diode acts like a current source when reverse biased, but if the gate of the MOSFET is accidentally pulled up above the power supply, the diode is then forward biased and conducts any amount of current into the power supply, holding the voltage on its anode - and the gate of the MOSFET - to no more than one diode drop above the power supply, hence saving the FET.

The bottom diode is normally reverse biased, so when a voltage transient in the negative direction comes along, it forward biases and clamps the voltage to no more than a diode drop below ground, again saving the FET.

The bottom protection diode is one of those things that you can get along without until you really need it, like insurance. If you never reverse a battery or never touch the inside of your pedal until you're discharged, it won't matter. It's there for accidents and forgetful owners.

Some MOSFETs have an internal protection diode there, which is nice. I believe that the 2N7000 does, the BS170 does not, but having an extra does not hurt.

It has to be low leakage, though, because otherwise you're leaking out the current that that would otherwise pull up your gate.

There really ought to be a 10K or so in series with the gate, but that hasn't seemed to matter in practice.

[Toney]

If you get the gate of the MOSFET more than about 20V higher or lower than the source, it ruptures the insulating layer and the MOSFET becomes a Darkness Emitting Diode (

So, the purpose of the second diode is to parallel the gate to the source (less one diode drop) in the case of any transient "spikes"....so it stays within th 20v range...  ah, so.

This has been my personal rubik's cube for the last day or so...

thank you.