Mosfet drivers prototyped

[Tage]

So I etched my board and wired up my dual bias supply with mosfet drivers. Once I got my bias voltage were I wanted it I checked the voltages @ my Mosfets. With my bias set to -55V I had -55V at my gates, -60 at my source and around -48 on my drain. do these voltages seem weird? I have proper voltages with them out of the circuit, +60V on the drains and bias voltage at the gates, then -160V at the source.

This is my schematic



Any feedback would be helpful

[tedmich]

you'll get more feedback if its not a .bmp file!

also its a little crowded layout

[defaced]

Drain voltage in all cases should be positive with respect to ground, not negative.

[Tage]

Yes my drain voltage is +60V, but once I have my mosfets in place I get a negative voltage showing when I check it at my drains, that is exactly what is puzzling me.

[Tage]

Is it possible that with my multimeter at 0V and with the other lead @ the drains I am simply observing the source voltage through the mosfet body diode? I have +190V on one side of R16 then on the other side I read ~-62V which is my source voltage.

Also I expected the source voltage to be lower than the gate voltage, but I have a reading around 2v higher? I am not yet coupled to the power tube grids though, but I don't see why that would change anything?

[loudthud]

Measure ohms Drain to Source on the FETs. You should see a diode reading or a dead short if you killed the FETs.

[Wombaticus]

Is it possible that with my multimeter at 0V and with the other lead @ the drains I am simply observing the source voltage through the mosfet body diode? I have +190V on one side of R16 then on the other side I read ~-62V which is my source voltage.

Also I expected the source voltage to be lower than the gate voltage, but I have a reading around 2v higher? I am not yet coupled to the power tube grids though, but I don't see why that would change anything?

Tage,

A couple of things here. First, as far as I can tell from your, how do I say this, individualistically rendered schematic, it looks as though you are aiming for a variant of the Universal bias supply in Merlin's power supply book. However, if so, I need to break the news to you that the book has a misprint, and the circuit won't work as intended. Please refer to the errata for that part of the book on Merlin's page for a corrected version: The Valve Wizard -Cathode Follower

Other than that, why bother with positive and negative rails? If you want a low impedance supply (I presume you are aiming for some variant of class AB2 or something?) why not just use a negative supply and P-channel MOSFETs, and be done with it? What exactly are your design goals here?

[Tage]

well I'll look into that bias supply, I thought something was fishy about it... i should be able to add that resister without any trouble.

I just wanted low impedance drivers, mainly to avoid grid blocking and to allow a small grid leak resister. as far as driving the grids positive and going class ab2 that would just be a side effect. As far as the design I was just basing most of it off of the Mosfet Follies stuff, which all include using N type Mosfets with positive and negative rails. Do you have a recommendation or example of the implementation of P type mosfets?

Also I checked my Mosfets and there is no shorted ones, my Body Diodes are fine.

[Wombaticus]

well I'll look into that bias supply, I thought something was fishy about it... i should be able to add that resister without any trouble.

I just wanted low impedance drivers, mainly to avoid grid blocking and to allow a small grid leak resister. as far as driving the grids positive and going class ab2 that would just be a side effect. As far as the design I was just basing most of it off of the Mosfet Follies stuff, which all include using N type Mosfets with positive and negative rails. Do you have a recommendation or example of the implementation of P type mosfets?

Also I checked my Mosfets and there is no shorted ones, my Body Diodes are fine.

Apologies for the misunderstanding, for some reason I thought this was intended to be just a bias supply (like for supplying the CT of a transformer phase inverter or somesuch) , but in fact the followers here are setup to drive the grids of the power tubes. Now that I see what you're really trying to do, then yes, it's appropriate to have N-channel devices and both a positive and negative supply. You're still going to have to fix the problem I pointed out with the bias discharge circuit though.

[Tage]

Yeah I'm gonna do that tomorrow and get back to you. I hope it helps, although I figure the diode kinda becomes redundant with the second resistor there is it not?

[Wombaticus]

...although I figure the diode kinda becomes redundant with the second resistor there is it not?

No, it's not at all redundant. The diode allows the cap to charge as quickly as it pleases, even while the new resistor (R3) allows a gracefully slow discharge at power down.

[Tage]

right that was the whole idea of the diode, I realy should re-read through that section of the book

[Tage]

So I added the bleeder resister, but it has no effect on my mosfet voltages. I can't figure out why I have 190V on the one side of R16, and then -45V on the other side/at the source of the mosfets. I know I have my pinout right, I just can't make sense of the voltage difference.

[Gingertube]

Check your pinout on the mosfet again - I think you may have your drain and source swapped.
When you measure the bulk diode with your multimeter and it is showing the forward biased (0.6V) condition then the diode cathode (negative) end (multimeter common or black lead) will be the drain and the anode (positive) end (multimeter red lead) will be the source.
Gate should be 2 to 4 volts positive with respect to the source.
Cheers,
Ian

[Tage]

Thanks Ian, I did check my pinout a few times, and it's correct.

SO my voltage at the source will be more negative than at the gate? I think I'm all right then, I was thinking I would have voltage loss resulting in a voltage decrease, regardless of positive or negative. so -44V at the gate wil result in -46 to -48V at the source?