Thoughts about tube output stages...

[Steve]

Hi all,
Never worked much on tube circuits, though lately my interest has
increased for this old technology.
I thought some questions could clarify a few things:

1. The bias current through the tubes, this is pretty much as the idle
current in a solidstate design, right?
2. How is the idle current set?
3. As far as I can see, the idle current flows through the output
transformer, why doesn't this saturate the transformer?
4. What function does the screen grid have? Some design apply full HT
to these grids while others apply HT through a resistor. Do these have
an influence of the gain in the tube and/or the idle current?

Thanks, I'm sure I'll come up with some more questions, hope it's ok
to ask.

Regards
Steve

 

[Bruce Varley]

Hi all,
Never worked much on tube circuits, though lately my interest has
increased for this old technology.
I thought some questions could clarify a few things:

1. The bias current through the tubes, this is pretty much as the idle
current in a solidstate design, right? Yep, analogous.
2. How is the idle current set?

By the bias on the grids. Typically, there'll be a DC negative voltage set
to the grids via resistors, or through the centre tap of the driver
transformer. It's common to have a pot to adjust this. Normally there'll be
a few specified conditions in the data for the tubes, typically they'll be
'class AB', with some standing current, and 'class B' with significantly
less. As you get closer to class B linearity becomes more of an issue.

3. As far as I can see, the idle current flows through the output
transformer, why doesn't this saturate the transformer?

Not a problem for a pushpull stage, for single ended the tranny has to be
designed for it. Single ended power amps are very inefficient, you'd rarely
have one above a few watts IME.

4. What function does the screen grid have? Some design apply full HT
to these grids while others apply HT through a resistor. Do these have
an influence of the gain in the tube and/or the idle current?

Generally the screen supply will be lower than the anode voltage. Once
again, the specs provide the correct value. A screen supply above what's
specified can damage the tube, because the grids are relatively fragile. A
correctly set up tetrode or pentode will have higher gain than an equivalent
triode.

There are unofficial variants. I remember driving a pair of 807s in a
modulator with the input on the screen grids and the control grids fixed to
control the standing current. It really smoked.

HTH

 

[Tim Wescott]

Hi all,
Never worked much on tube circuits, though lately my interest has
increased for this old technology.
I thought some questions could clarify a few things:

1. The bias current through the tubes, this is pretty much as the idle
current in a solidstate design, right?
2. How is the idle current set?
3. As far as I can see, the idle current flows through the output
transformer, why doesn't this saturate the transformer?
4. What function does the screen grid have? Some design apply full HT
to these grids while others apply HT through a resistor. Do these have
an influence of the gain in the tube and/or the idle current?

Thanks, I'm sure I'll come up with some more questions, hope it's ok
to ask.

For a complete exposition on the function of the screen grid you need to
find an old ARRL handbook, RCA tube data book, or other text from the day.

Basically, a screen grid screens the cathode from the plate, so the
cathode current isn't affected by the plate voltage. For most of the
tube's operating region most of the electrons are captured by the plate
even if its voltage is lower than the screen's, so the effect of the
screen grid is to significantly raise the plate resistance. This also
increases the "power sensitivity", or the ratio of power out to voltage
in, significantly over a triode. The downside is that you must pay
attention to the care and feeding of the screen.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Posting from Google? See http://cfaj.freeshell.org/google/

Do you need to implement control loops in software?
"Applied Control Theory for Embedded Systems" gives you just what it says.
See details at http://www.wescottdesign.com/actfes/actfes.html