833A (GU-48) Amp

[pre65]

So, in simple terms, what would be needed to make the output impedance lower.

Or is there no simple answer ?

[Nick]

Depends what you regard as a simple answer. I guess while still using a simple catjode follower you need a higher gm valve, given that the output impedance of a cathode follower is 1/gm.

Simple way is a mosfet source follower.

[pre65]

Depends what you regard as a simple answer. I guess while still using a simple catjode follower you need a higher gm valve, given that the output impedance of a cathode follower is 1/gm.

Simple way is a mosfet source follower.

Thanks Nick, I will read up on that. Better get a new box of headache pills first though.

[pre65]

So, would the mosfet source follower be used instead of the EL34 cathode follower, just trying to get a broad picture to start with.

[pre65]

I found this article most interesting. Am I on the right lines ?

http://www.tubelab.com/powerdrive.htm

The writer has used it on 833a, he's the guitar playing nutter with the hairy chest.

[Nick]

Yep, thats the sort of thing. mosfets have higher transconductance, so the outout impedance is lower.

But at some point you fiond you have made a solid state amp.

[Mike H]

What if the EL34 was converted to tetrode operation?

[Nick]

no idea.

[Mike H]

Actually no, looking at it again it's tied to DC as it is.

Can't simulate the offset, but what I have noticed is, that when the 833 grid stops conducting on the neg half cycles the EL34 current "flat lines" at 28 mA and won't go any lower.

Actually I say can't sim it, but can just about see it here:

[Mike H]

And then compared with EL34 current (in pink):

[Mike H]

What if the EL34 was converted to tetrode operation?

Actually no, looking at it again it's tied to DC as it is.

Yes we can! Just AC couple the screen to cathode. Doh!

[Mike H]

Well LTspice reckons it might have possibilities.

[Mike H]

BTW all the green plots are 833A grid, AC only (as if relative to 0V)

Only not sure about R7 value as depends on EL34 screen current

[pre65]

Thanks Mike, I will give it a go.

How important is getting the fixed bias to -26.95V, and is the test point (and 50n cap/1M resistor) necessary for other than as a test point ?

I will measure the EL34 anode voltage and adjust R4 to suit the new voltage requirement. Actually, the bit of alloy that R4 sits on needs to be more of a heatsink, so that needs doing at the same time.

[Mike H]

Adjust to get the required 833A biased current. 168mA? So 10V across R5?