Question about IRF820, R.G? (2020)

[Derringer]

Hi R.G. and whomever else may be able to help me here.

I've bulit a tube pre for recording purposes. I built in an irf820 as a source follower after the final tube stage to drive the output transformer.

The voltage at the anode of the last tube stage, it's 12au7 cascode design, is about 150v. The Ht there is about 275 volts.
I ran a 240R stopper from the anode to the gate of the IRF820, drain to the Ht 275volts, source has 100K to ground and measures about 125 volts.

I then did a 100r build out resistor from the source to a 1 uf cap, 2m2 r to ground after the cap, then to the primary of the 15K/600r OT.

My question is, should I consider a lower source resistor? I'm not sure how the source resistor, the 100K not the 100r build out resistor, affects the function of the buffer. Lower will make the buffer use more current, right? But how would I figure out how much current I really need there to adequately drive the transformer?

thanks

[R.G.]

Driving a transformer single ended through a capacitor might not be the best way to get your signal out. The drive is going to be asymmetrical, as the MOSFET pulls up really well, but can't drive the transformer down towards ground. The source resistor pulls it down, but that's a high 100K (as you correctly worry about). Most single ended setups use the DC in the transformer to drove the opposite direction when their driver turns off.
In turn, this puts requirements on the transformer's construction and ratio. You don't mention what the characteristics are for the transformer. Did you have a specific one in mind or on hand?
You're correct in assuming that things will work better if you lower the source impedance, but you can eat a LOT of power in that resistor.
You'd be better off considering something like a constant current device in the source or a single ended push-pull pair. A CC instead of a resistor for the source of the MOSFET will give you much more symmetrical drive and release many of the requirements on an output transformer.

[Derringer]

The transformer is an Edcor WSM
https://www.edcorusa.com/wsm_series

It can handle 0.5 watts and 10Vrms


The design already has two tubes. I'd like to make this "transformer drive" section as simple as possible, so that's why I went with the mosfet. Is there a solid state CC or SEPP you'd recommend I investigate?

[PRR]

> The voltage at the anode of the last tube stage, ...is about 150v.   ... to the gate of the IRF820, ...source has 100K to ground and measures about 125 volts.

This is already fishy. The IRF gate should be about 5V higher than the source. So we expect 145V. Not 125V. I wonder if your IRF has already died and its source is hanging by the gate-source protection diode. Or meter loading or mis-reading.

I'm less worried about resistor loading. It is inefficient but you already picked a supply FAR higher than your signal, it will pull-down fine. However 1mA is awful light for an affordable 10k(??) transformer at few-Volt output. On paper it swings nearly 1mA*10K= 10V which is ample, but the "10k" drops in the bass. _I_ would pencil 10mA and see if I can stand the heat. About 3 Watts? You have a couple 12A?7 so already 4W in heater heat. I might compromise at 5mA and see how it swings. Almost a Watt in MOSFET and resistor so you want heatsink and power resistor.

[Derringer]

Thanks, PRR

I did a lot of futzing with things yesterday, biasing in my gain stages and values in my power supply.

Here are my current voltage readings of the IRF. It has a 50K source resistor now (pair of 100K paralleled)
Drain - 310 volts
Gate - 133.5 Volts
Source - 133 Volts

So, I'm still not getting that  5v difference, but it is passing signal, cleanly, Will they still do that if dead?

"However 1mA is awful light for an affordable 10k(??) transformer at few-Volt output. On paper it swings nearly 1mA*10K= 10V which is ample, but the "10k" drops in the bass. _I_ would pencil 10mA and see if I can stand the heat. About 3 Watts? You have a couple 12A?7 so already 4W in heater heat. I might compromise at 5mA and see how it swings. Almost a Watt in MOSFET and resistor so you want heatsink and power resistor."

Let me see if I follow. The output transformer is 15k : 600r.
The power supply is beefy, so I have room to play. This was an old P1Ex build from ax84 that I have made into this preamp.
The PT is a Hammond 270Dx. Mine is labeled to output 90 ma.

When you say to compromise at 5ma, is your meaning to have the source resistor be chosen so that it is dropping enough voltage to yield 5ma? As in ... if I have 133 volts at the source and use a 22K resistor source to ground, it will dissipate about 1 watt at 6 ma?



Also, fwiw, I was getting clipping, but it wasn't coming from the IRF. My gain stages were producing too much gain. I increased some cathode resistors, removed a 22uf cathode bypass cap, lowered a resistor in the power supply to get higher HT, and allof that really cleaned things up. It sounds pretty good so far recording electric guitar. Through the scope, I am getting an unclipped signal, for the most part, unless I really push the input signal and/or max out the gain control on the build. I did notice on my scope that there's a slight attenuation in high frequencies from maybe 6K or so on up. I'm not sure what's causing that because I assume it could be input or output transformers, something in my gain stages or maybe the IRF?

Any further insight you can give me is greatly appreciated!

Thanks again.