[coffeedj]

I have been wondering why there are not more designs using plate chokes
instead resistors for driver stages.

Part of the answer is surely that there are not as many choices for tubes
with a low Rp, but that can't be the only reason. Another part of the
answer is probably cost.

However, from the design side it seems like you get the best of both worlds
if a plate choke is used properly. You could eliminate an extra gain
stage--or keep from running at the rails, which is a great plus for NOS
tubes.

It is true that you won't get a DC to daylight response--but the output
transformer pretty much eliminates that possibility anyway.

Thoughts from the group?

[Patrick Turner]

coffeedj wrote:

I have been wondering why there are not more designs using plate chokes
instead resistors for driver stages.

Part of the answer is surely that there are not as many choices for tubes
with a low Rp, but that can't be the only reason. Another part of the
answer is probably cost.

However, from the design side it seems like you get the best of both worlds
if a plate choke is used properly. You could eliminate an extra gain
stage--or keep from running at the rails, which is a great plus for NOS
tubes.

It is true that you won't get a DC to daylight response--but the output
transformer pretty much eliminates that possibility anyway.

Thoughts from the group?

Using plate chokes was a favourite way to save power use from
batteries used in mobile gear in 1950, but the cost and weight and
bandpass
restrictions plus iron caused distortins make choke loading
too hard for most makers.
Chokes and transformers were mainly used in 1930, but then resistances
and caps became cheap.

If you want a wideband blameless high Z dc supply element, use a CCS.
They work better than a choke or resistor.

Patrick Turner.

[John Byrns]

In article ,
Patrick Turner wrote:

If you want a wideband blameless high Z dc supply element, use a CCS.
They work better than a choke or resistor.

There is at least one case where CCSs don't work better than a choke,
and that is if you need to develop the maximum possible voltage swing
with a limited B supply voltage.


Regards,

John Byrns

--
Surf my web pages at, http://fmamradios.com/

[Patrick Turner]

John Byrns wrote:

In article ,
Patrick Turner wrote:

If you want a wideband blameless high Z dc supply element, use a CCS.
They work better than a choke or resistor.

There is at least one case where CCSs don't work better than a choke,
and that is if you need to develop the maximum possible voltage swing
with a limited B supply voltage.

Well true, so double the B+ to give the required headroom.

Some HV rated TV transistors van swing HV, but perhaps having an active
tube CCS is better.

If one is really determined, one can also use a resistance in series
with a choke.

Consider an EL84 in triode used as a driver to a following output tube.

To get a large Vswing without a supply of 600V with a transistor based
CCS,
a 450V supply could be used with a choke of 50H at 20mA,
and also have a 7.5k anode supply R.

The choke may be 50H at 50Hz, but the iron µ will be a lot less at 1kHz,
and L may be lower,
but as F is higher, then ZL is still a very high ac load at 1 kHz.
But at very low F the ZL becomes low enough to shunt the 2k of Ra at a
pole at 6.3Hz,
giving -3dB in the response and +45D phase shift.
Both are unwanted if the amp has GNFB and so the 7.5k
isolates the low ZL at LF and the lowest load becomes 7.5k, even at DC.
At extreme HF, the choke's shunt C of perhaps 300pF would cause a pole
and roll off but
with the 7.5k, the roll off is prevented and mimimum load is 7.5k.
So the bandwidth is greatly extended and phase shifts almost eliminated
by using an R load.
For most of the audio band the combined R&L load is a very high value,
so virtually no
power is lost in the dc carrying load, so there is minimal tube THD.
And this also allows a lower value of following cap coupled grid bias R
value
to ensure the output tube grids don't runaway with +ve grid voltages.

So using 33k for Rg bias is OK, and at 1khz, dc load 200k total,
so real ac load on the EL84 = 28k, and triode gain at Ia = 20mA = 18
approx,
and THD 1% at 75Vrms output, or 0.1% at 7.5Vrms.

But within this THD there will be some iron caused THD.
if measured where the 7.5k is connected to the choke, it looks
disturbingly high,
but because the 7.5k and triode Ra from a resistance divider, THD is
well below tube distortions, although is is 3H++ high order junk.

So using best winding techniques are called for, and perhaps
50% GOSS and nickel core material.

Patrick Turner.




Regards,

John Byrns

--
Surf my web pages at, http://fmamradios.com/

[Ian Bell]

coffeedj wrote:

I have been wondering why there are not more designs using plate chokes
instead resistors for driver stages.


I don't know about driver stages but plate chokes are common enough in pro
audio tube preamps. The classic V72 preamp (used in the EMI consoles used
to record the Beatles) uses a pair of pentodes with the second one having a
plate choke. Designed over 60 years ago it uses global NFB and its sound is
much revered. Enthusiasts build replicas of them to this day.

Ian

[robert casey]

It is true that you won't get a DC to daylight response--but the output
transformer pretty much eliminates that possibility anyway.


The inductance of the choke would cause various phase shift issues,
especially if the amp uses global negative feedback loops. But if it's
a push-pull amp, the choke could become a transformer with a
centertapped secondary to feed the output tubes' control grids. That's
your inverter stage. So if you're willing to deal with the phase shift
issues, you could do it. Especially if you forgo the global feedback
loop. There was a guy in Japan that liked to use interstage
transformers in his tube amps, and said that it made for great sound.
Never heard one myself, though.