Mosfet cathode follower

[lowell]

I'm trying to work this out for the 6V6 grid drive in a 5F1 champ amp. It's cathode biased so it's a bit tricky and I'm not completely clear on RGs setup for this.

The image from Geofex is here: http://www.geofex.com/Article_Folder...ly/mosfol5.gif

Why the need for -V (shouldn't the grid of the 6V6 be 0v?) and what value should the -V be?

Finally how would the 5v zener be implemented? I'm not clear on that because if a 5V zener was in series with the gate tapped off the B+ then the voltage would be B+ minus 5V right?

[tubeswell]

When I did an IRF820 source follower with a BFPR cloney thing a little while back, I put a 100R in series with the gate (from the preceding stage's plate - for a direct-coupled follower stage) and a 12V zener between the gate and the source (banded-end pointing towards the gate). Was this enough? Seems to still be working (touch wood).

[R.G.]

Why the need for -V (shouldn't the grid of the 6V6 be 0v?) and what value should the -V be?

The V- is needed because the signal applied to the grids must swing both positive and negative around the DC idle voltage, which is zero volts. The V- value in this circuit only needs to be a little more than the cathodes ever get positive. Call it -20 to -30V and you're probably good.

Finally how would the 5v zener be implemented? I'm not clear on that because if a 5V zener was in series with the gate tapped off the B+ then the voltage would be B+ minus 5V right?

I'm a little confused about that. Do you mean how should the +5V power supply be done?

The +5V is the main DC bias setting voltage for the MOSFETs, and the MOSFET sources set the DC level for the tube grids. The tube grids would be 0V in the normal circuit, so they need to get back to 0V in the MOSFET circuit too. The +5V supply makes up for the fact that the MOSFETs need a couple of volts positive on the gate with respect to the source, and so to get the source to 0Vdc, you need the gates to be +2 to +5V dc to get the sources up to 0Vdc.

There's nothing magic about the +5V. It could easily be +6, +7, etc. But each MOSFET needs a different bias pot because two MOSFETs of the same type will in general NOT have the same Vgs voltage to turn on, so each is probably going to be slightly different. So each MOSFET needs to be tweaked a little to get the tube grids to be 0Vdc at idle.

[lowell]

Tubeswell,
Like this? No coupling cap from the PI?

[tubeswell]

yup - for a DC coupled pair (and the source had the load resistor between the source/diode junction and the ground) - I did use an IRF820, which is a shoe-in substitute for a pre-amp triode (so apart from the zener and the gate series resistor, everything else is as per a DC coupled cathode follower stage).

[lowell]

Ok that seems simple, thanks. What value load resistor for the source did you use? So eliminating the coupling cap from the PI makes this simpler.

[R.G.]

Hang on.

I may be confused, but I don't think direct coupling from the plate of the PI will work.

Are you two guys talking about different circuits?

[tubeswell]

220k on the source and 220k on the plate of the DC-coupled triode stage in front of it. The HT for both stages is about 420VDC. I used a 0.1uF 630V coupling cap from the source to the next part of the circuit

[tubeswell]

I'm talking about using an IRF820 for a source follower, in place of a triode CF stage, on a DC coupled pair (which in my case is in a trem circuit. but I don't know how significant that is for the purpose of setting up any old a mosfet source follower - I'm just relating my experience) Seems to work okay to-date (touch wood)

[lowell]

That's a very different setup there. It's not direct coupled to a grid from the source. Ok so back to where I was at. RG I'm still not clear on the zener setup. Could you be so kind as to sketch this for me?

[R.G.]

I'm talking about using an IRF820 for a source follower, in place of a triode CF stage, on a DC coupled pair (which in my case is in a trem circuit. but I don't know how significant that is for the purpose of setting up any old a mosfet source follower - I'm just relating my experience) Seems to work okay to-date (touch wood)

OK. THat's what I thought. That's the source follower setup from my original "MOSFET Follies" article, and it does indeed work for a follower coupled to a plate, and then capacitor coupled to a following stage from the MOSFET source.

lowell is doing something different, though. He's trying to use MOSFET followers to drive more DC into the grid of a power tube, giving a class AB2 operation instead of AB1, the difference being that a low impedance DC drive can drive the grids positive even when they start drawing current. The use of a MOSFET to drive the grid means that the change into grid conduction when the output tube grid goes positive cannot pile up charge on a capacitor and start the bias shift that leads to grid blocking. The MOSFET gate at the input never starts conducting on any signal level below destruction, so it cannot start bias-shift blocking. And the extra current to the tube grid drives the output tube smoothly into grid conduction.

I'm still not clear on the zener setup. Could you be so kind as to sketch this for me?

Sure. The illustration has now been updated with the three zeners I'd use. Two are gate-protection zeners on the MOSFET gate-source so that transients can't pop the typically 20V max of the gate oxides. The other is the "5v" zener, used to make 5V so the bias pots can set the MOSFET gates to put the sources just at zero volts.

Actually, now that I type that, I think there is a bug there. I think the source resistors for the MOSFETs have to go to a voltage below ground. Lemme think about that for a few minutes. I'll be back with more thoughts.

[Bruce / Mission Amps]

Anyone know why I can NEVER open any attached pdf files here... my only option is to safe it in a directory and then go find it to reopen it with Adobe?
On other websites, the pdf files seem to auto open.

[lowell]

RG any progress regarding the Rs to ground or a negative voltage?

[Steve Conner]

Off the top of my head I'd say you need the negative voltage: at least as far below ground as the cathode is above ground.

[lowell]

Ok right on. Why is that? Is this not going to bias the following stage cooler?