[R-390] LINE /AUDIO POTS

[Craig]

Charles,

Most excellent dissertation. Most of the time I do agree to use the correct
part. Trust me on this Charles: If'en I ever find the issue/fix the "Too
Loud Amelco", I still have a bag of 2.5K audio/log taper pots for R104 &
R105. 

Good Night,
Craig

-----Original Message-----
From: R-390 [mailto:r-390-bounces at mailman.qth.net] On Behalf Of Charles
Steinmetz
Sent: Friday, March 18, 2016 8:15 PM
To: 390 list
Subject: Re: [R-390] LINE /AUDIO POTS

Steve wrote:

>If you need log pot response from a linear pot one method that works is 
>connecting a resistor that's 20% of the value of the linear pot between 
>the ground end solder lug and the center variable contact solder lug.

Jacques responded:

>Despite the trick presented by Steve works practically to "imitate" a 
>log potentiometer from a linear one, it cannot be used on R-390/390A 
>because the end to end value of both local and line potentiometers are 
>a part of the follower tube bias system.

The potential problem Jacques is alluding to is that when you add a low
resistance (let's say, equal to or less than the end-to-end resistance of
the pot) from the wiper to the CCW terminal, the total resistance from the
CW terminal to the CCW terminal now changes quite a lot depending on where
the pot is set.  (Whereas, when the load on the wiper is just a grid
resistor -- much larger than the end-to-end resistance of the pot -- the
total resistance from the CW terminal to the CCW terminal remains very close
to the pot's own end-to-end resistance, regardless of rotation.)  For some
hard numbers, two, 2.5k linear pots, each with a 500 ohm resistor from wiper
to CCW (20% of the pot value), would present a load to the driving circuit
of only ~208 ohms *if both pots were turned fully up*.  (Of course, almost
never would both pots be turned fully CW on an R390A.  In practice, the load
on the driving circuit would likely be no lower than 1k or so.)

This same issue arises if one substitutes audio-taper pots with values other
than 2.5k for R104 and/or R105.  In this case, the load on the driving
circuit would simply be the value of the two end-to-end resistances in
parallel, and would not change significantly with pot rotation.

So -- what is the reality of changing the value of R104 and/or R105?  I
simulated the V601B circuit (please refer to the schematic
below) to demonstrate the effect of various potentiometer values.

R607, R608, and R627 make V601B operate as an approximation of a current
source with respect to its DC bias conditions (recall that cathode-biased
triodes make pretty bad current sources).  This means that R104 and R105
(the Line and Local Audio potentiometers) have relatively little effect on
the bias of V601B.

The V601B plate voltage is approximately 200v.  With R104 and R105 = 2.5k
(net potentiometer resistance = 1.25k), the cathode voltage is ~ 10.5v.
Increasing R104 and R105 to 10k each (net potentiometer resistance = 5k),
the cathode voltage is ~ 17.7v.  And decreasing
R104 and R105 to 1k each (net potentiometer resistance = 500 ohms), the
cathode voltage is ~ 8.6v.  In all cases, V601B has from 180v to 195v of
voltage headroom, and has plenty of operating current to drive the output
amplifier grids.  All of the coupling capacitors (C602, C604, and C607) are
in high-impedance grid circuits, so the low frequency corner does not
change.

For all practical purposes, changing R104 and R105 from 1k to 10k has no
effect on the bias conditions of V601B.

There is one difference when you change the values of R104 and R105.  The
combined (net) load of R104 and R105 forms a voltage divider with cathode
resistor R627 at audio frequencies as well as at DC.  With R104 and R105 =
10k, the pots have about 5dB more audio voltage on them than with R104 and
R105 = 2.5k, so you would need to use a slightly lower setting of the
controls to get the same audio level.  Conversely, with R104 and R105 = 1k,
the pots have about 5dB less audio voltage on them than with R104 and R105 =
2.5k, so you would need to use a slightly higher setting of the controls to
get the same audio level.

Note that the effect mentioned in the first paragraph above -- the load on
the driving circuit changing with pot rotation if one uses a low-value
resistor from wiper to CCW of a linear pot -- makes that trick work even
*less* well than normal in the V601B circuit.  As the pot rotation
increases, just when you want the audio voltage on the wiper of the pot to
start increasing faster and faster, the attenuation due to the reduced value
of the pot load causes the audio voltage on the pot to *decrease* faster and
faster.

All of that said, I'm astounded at the amount of effort people on this list
devote to finding hacks to cobble up their radios in one butchery after
another.  Jeez, people, JUST GET THE RIGHT PART 
!!   This whole discussion will have some meaning in a distant future 
when there are no longer ANY audio taper pots available, anywhere.  But when
that time arrives, I suspect that boatanchor radios will have long ago
ceased to be useful for anything at all.

Best regards,

Charles