HELP w/ RCA T28 Low Volume-- Bias Cell?
J Ventry-McGee
I am trying to get this fixed as an Xmas present for my neighbor.
Radio has had all paper and electrolytic caps replaced. Volume was
barely audible with ear to speaker. Disassembled and cleaned volume
control to discover an insect cocoon (sp?) wrapped around the slider.
Volume is now acceptable only at maximum volume. Tubes test ok on my
cheap but reliable tester; substitution has no effect. There is a
bias cell wired to the volume control which I imagine is now quite
dead.
Schematic here:
http://www.nostalgiaair.org/Resources/872/M0014872.htm
Could this be a possible cause of my problem? If so how do I replace
it? Other suggestions as to where to look?
THANKS
Jeremiah
Jeffrey D Angus
> There is a bias cell wired to the volume control which
> I imagine is now quite dead.
Yeah, that's it. Just put a little 1.5 volt battery in
it's place.
Jeff
--
�Egotism is the anesthetic that dulls the pain of stupidity.�
Frank Leahy, Head coach, Notre Dame 1941-1954
Pete Bertini
"J Ventry-McGee" <jven...@gmail.com> wrote in message
news:3d0b28b1-7175-418c...@u20g2000vbq.googlegroups.com...
Use a watch or computer cell, something that will last
longer than a few years. Also, in many cases the circuit runs just as
well without the battery.
Pete
jughead
up. If the cell's open it would cause low volume.
I always just got rid of the cell and replaced it with a cap (anywhere
from .001 to .0047 will do). And run a 10 meg resistor from grid to
ground. This allows the tube to self-bias a bit by space charge. Bill
(oc)
nesesu
Since it is for someone else, I would use Bill's method with the
resistor and cap since that will last forever and not need cell
replacement every few years.
Neil S.
J Ventry-McGee
you know how it turns out.
Meanwhile, why did they use a bias cell in the first place if a
reasonable substitute can be done with a resistor and a cap? I can't
imagine it would be a cheaper part since it would have to be made in
substantially lower quantities.
Thanks,
Jeremiah
jughead
>
>...why did they use a bias cell in the first place if a
> reasonable substitute can be done with a resistor and a cap? I can't
> imagine it would be a cheaper part since it would have to be made in
> substantially lower quantities.
>
Seems they cost 20 cents in 1936. In bulk to a mfr., probably
considerably less.
http://www.philcorepairbench.com/biascell.htm
Michael A. Terrell
The bias cell supplied a fixed bias voltage. The resistor doesn't.
--
Offworld checks no longer accepted!
HankVC
certainly is **a** problem, if not your primary problem. For test,
you can disconnect the bias cell and connect a 1.5 volt battery (AAA
is fine) in the circuit, negative end to the 6SQ7 grid.
Looking at the schematic, I see the diode plates on the next (phase
inverter) 6SQ7 are not used, and it's cathode is grounded. Here is a
very permanent fix to eliminate the bias cell permanently and leave
the circuit parameters identical to the original:
1. Move the diode pin connections (pins 4 and 5) over to the same
pins on the phase inverter tube.
2. Disconnect the cathode (pin 3) connection from ground on the first
6SQ7. Insert a 3900 ohm resistor parallelled by a 2 mfd capacitor from
cathode to ground. That should give you about 1.3 volts bias, which
should put about 110 volts on the plate. Positive lead on the cap
goes to the cathode terminal. A 1/2 watt resistor and 16WVDC or more
for the cap should be fine.
Worth nothing that these values are not overly critical---the 6SQ7
triode will peform about the same with any bias from around -1 to -3
volts. A quick look at the resistance coupled amplifier parameter
tables in an RCA Receiving Tube manual suggests that you'll get a
stage gain of 55-60, and that 2 mikes or more are sufficient to
prevent objectionable low-frequency rolloff. If anything, watch plate
voltage with that much gain---if it's much below 110, use a 4200 or
4700 ohm cathode resistor.
That sets up the first 6SQ7 as a self-biassed class A amplifier vs.
fixed-biassed. It's also a much more permanent and satisfactory setup
than trying to duplicate the original.
Hank
jughead
between a direct jumper instead of the cell, and the resistor/cap. Tin
ear maybe. (: But did the resistor/cap anyway because it seemed more
kosher. oc
jughead
>
> That sets up the first 6SQ7 as a self-biassed class A amplifier vs.
> fixed-biassed. It's also a much more permanent and satisfactory setup
> than trying to duplicate the original.
>
Wonder why it wasn't engineered that way in the first place, with
those unused diode plates sitting there looking forelorn. But most
sets employing bias cells did not have that option. oc
HankVC
cheaper than using the other 6SQ7 as the detector and installing a
resistor and cap in the first 6SQ7 cathode circuit.
Actually, it's pretty rare to see a 6SQ7 set up with genuine grid bias
in a home-entertainment radio. There was a lot of hullabaloo about
"contact potential" in the AA5-type circuit (large grid and plate
resistors, zero bias) which, as I recall was more-or-less settled in
Karl Spangenberg's "Vacuum Tubes" (McGraw-Hill, 1948).
Hank
jughead
electrolytic vs. a bias cell.
Richard Knoppow
"jughead" <oldco...@gmail.com> wrote in message
news:8172e046-4b27-4bc0...@e4g2000prn.googlegroups.com...
> On second thought, maybe it was a cost issue; a resistor +
> a small
> electrolytic vs. a bias cell.
>
Bias cells were sometimes used in equipment meant to
run on batteries to reduce the voltage drop needed for
biasing. Most bias cells were mercury types which would last
for essentially their shelf life because virtually no
current is drawn. That can be years for a mercury cell. As
has been posted its possible to eliminate them. Since this
is not a battery set its difficult to know why the bias cell
was used.
--
Richard Knoppow
Los Angeles
WB6KBL
dick...@ix.netcom.com
J Ventry-McGee
volume, perhaps with slight distortion. Hank, thanks for engineering
a nice solution. My supplies are low, however-- the smallest
electrolytic I have in 160 wv is 10 mfd. I also have several 4.7mfd
but only at 50 wv. Will the 10 work or will that be too much? I
don't know where I can get one locally at short notice, radio shack
rarely has anything over 50 wv. If only I had been alive for the age
of the neighborhood electronics store...
Jeremiah
J Ventry-McGee
>6SQ7. Insert a 3900 ohm resistor parallelled by a 2 mfd capacitor from
>cathode to ground. That should give you about 1.3 volts bias, which
>should put about 110 volts on the plate. Positive lead on the cap
>goes to the cathode terminal. A 1/2 watt resistor and 16WVDC or more
>for the cap should be fine.
Also, pardon me for being a noob, but could you explain the bias
calculation if its not too much trouble? I've done my share of rote
capacitor replacement but haven't really done any sort of
engineering. I really want to learn the mysteries of electronics.
Thanks!
Jeremiah
Ken
cathode end. The grid resistor puts the grid at ground potential and the
cathode is at some pos value. This makes the grid negative with respect
to the cathode, just what the tube needs. Ken
HankVC
about the capacitance is "enough"----it has a major effect on
low-frequency response. The 2 mike number I gave you is "more than
enough."
Actual working voltage across that cap is only about 1.5 volts, so you
do not need as much as 50 volts---a 16 volt cap is fine.
I know the feeling about local parts availability. I think that all
the local Wratte Shaque has are cell phones, fuses, batteries, and TV,
with some computer stuff. No components at all. I have to maintain
inventory with orders from Mouser.
Hank
Richard Knoppow
"HankVC" <hvan...@nyx.net> wrote in message
news:12609271...@irys.nyx.net...
degenerative feedback via the cathode resistor by by-passing
the AC to ground. If the capacitor is not large enough
there will be some roll off of the bass due to the feedback
below some point. The calculation is that at the frequency
where the capacitive reactance equals the value of the
cathode resistor the gain will be down 6db. Cathode by-pass
caps are usually chosen to limit the low frequency responce
and get a little better stability without the need for very
large filter caps in the power supply. In this case, if the
cap is too small there will be a bit of low frequency
roll-off, if too large nothing very bad will happen.
In some high quality audio circuits the cathodes are
delibrately left unbypassed in order to reduce distortion of
the stage by a little. The price is some loss of gain. I am
not quite sure but think the RCA tube handbooks have the
formulas for calculating the values of both coupling and
by-pass caps for audio circuits buried in them.
The tube manual will also provide some information on
the bias calculation but its just the drop across the
resistor due to cathode current. Why RCA did not used this
simple bias method is beyond me. Bias cells are really a
hangover from the days of filamentary cathodes where they
made biasing circuits simpler and cheaper.
--
HankVC
tell me why Hallicrafters used an externally-excited 6SA7 mixer----and
chose to use another 6SA7 as the oscillator to drive it. That in the
SX-28, which was Hallicrafters' attempt (along with the S-27/S-36) to
go first class in construction. And I think of another RCA design
from about the same period as this one---a seven tube AM-SW console.
Tube 1 was a 6SK7 TRF preamp resistance-coupled to a 6SA7. Following
the IF, one 6SQ7 had only the diodes connected as detector/avc.
Another 6SQ7 following had the diodes grounded as the AF voltage
driver---but in the traditional AA5-type circuit with the cathode
grounded. That drove a 6K6 audio output tube. I thought somebody had
pulled a fast one on somebody when I rang out the wiring until I was
able to find a Rider's reference to the set and discover that lo and
behold, RCA built it just as I found it. Even Hallicrafters, cheap as
they were with components and construction, wouldn't have pulled a
stunt like that. At least it had a speaker field coil in the B+
filter, so was hum-free. But a single 6K6 isn't going to rip the cone
out of a 12 inch floor console loudspeaker.
Hank
HankVC
values.
First of all, when dealing with the transfer characteristics of a
vacuum tube, the classical method of finding operating points is to
draw a load line on the chart of curves giving plate current and
voltage for various grid voltages. That load line represents what's
left for the tube after deducting drop across the plate load resistor.
The end points are:
1. Resistor in an open circuit. Current is 0, voltage is B+.
2. Plate terminal shorted to ground (all B voltage dropped across the
resistor). Voltage is 0, current is calculated by Ohm's Law.
A line between those two points shows plate current and voltage at
each point where it crosses a grid voltage line.
In this case, the schematic lists the quiescent (no signal) plate
voltage as 109V, B+ as 250V, grid voltage as -1.1V, and tube (cathode)
current as 330 microamps. Some Ohm's law calculations and locating
that operating point on curves for a 6SQ7 show that the numbers
cross-check reasonably well (they don't match exactly, but the points
that each gives is in a small group, and in the middle of class A
country for the tube). I didn't draw a load line, but the points
plotted are all on or near the load line for 470K and 250 volts B+.
Self-bias uses the drop across the cathode resistor to establish the
initial grid bias. Since this is a triode, there is no screen current
to consider. The target is "about 330 microamps" and "about 1.25
volts" Crunch that through Ohm's Law, pick the nearest standard value
in the series 3.3, 3.9, and 4.7, and I got 3.9K.
Now, for a bypass cap, the objective is to give a good AC ground at
low audio frequency. I skipped calculations and looked at the
resistance coupled amplifier value charts in the RCA RC-14 Receiving
Tube Manual for an audio bypass cathode cap across a 3900 ohm cathode
resistor, and got 2 mikes from the table. The RCA tables are based on
100 Hz low frequency cutoff (3 db down, the point at which the human
ear detects "quieter"). That's typical for a radio of this era.
You can run through the calculation of capacitive reactance with the
formula Z = 1/(omega*C) where omega is 2*pi*F. Z in ohms, F in (watch
out! farads), F in cycles-per-second.
Using the R + jX form, you can plug in 3900 for R, and -Z for X; find
the resultant vector, and you've got the phase angle and length
(effective impedance). That's a starting point for determining the
maximum impedance tolerable for the low-frequency dB point. I used
the table value as a shortcut to doing a bunch of arithmetic----after
all, why reinvent the wheel?
Hank