Richard I was a bit clumsily in my previous attempt so I will try again.??
All aluminum electrolytic capacitors are self resonate.? ?There is a frequency where the capacitor becomes an inductor.? C44 is such a capacitor.? The graph below is for a modern aluminum electrolytic capacitor in a round can with terminals on top.? The type of capacitor in question, found in the audio section of the SX-28 I believe, are twist lock.? In my search, I could not find any data on the self resonate frequency of this type capacitor package but I suspect that it is well within the midrange audio frequencies.? This self resonate frequency was not considered important in the late 1930's.? The sole electrolytic capacitor function at that time was as a power supply filter.??
I also suspect that electrolytic capacitors with low working voltages perform adequately at audio frequencies but those with high working voltages, not so much.? Their construction to withstand the high voltage made the internal parasitic components much larger and the self resonate frequency much lower.? C44 is rated at 300 volts and must filter audio frequencies to ground.? You contend that C44 is filtering the DC from the power supply on the other end of R38 which is 47k ohms.? That seems unlikely.? I contend that the job of C44 is removing midrange audio and it is doing this job poorly, even back in the day.
None of this means much now since it does not seem to be the fault in this SX-28 thread.? All of my observations were based on the reported 120 cycle hum.? It now seems that it is really 60 cycle hum.? I suspect that the SX-28 speaker is a modern type that has a frequency range that easily extends below 60 cycles.? The loud speakers back in the day were not so efficient.? The improvement noticed when the bass boost is out is the result of the lower voltage amplification in this position.? When the bass boost is in, the triode has higher gain so the hum is more noticeable.
The 60 cycle hum may be an artifact from the heater voltage or the relatively low value filter capacitors.? Increasing the filter capacitor value may reduce the hum but doing so will stress the power rectifier tube.? Large value filter capacitors demand a higher charging current which the 5Z3 cannot long survive.? This will also stress the power transformer because of I^2R losses.? If the source of the hum is the heater voltage, perhaps the three audio tubes can be selected for low hum? Regards, Jim
Logic: Method used to arrive at the wrong conclusion, with confidence.? Murphy
I suppose I have to reply to each comment. The plate load is the
impedance in the plate circuit between supply voltage and plate, that
develops the signal output. In the half of the 6SC7 that the bass switch
is in it is either the resonant choke or R-37 depending on the position
of the switch. R-38 is not part of the plate load because there is no AC
on it. It is bypassed to ground via C-44. C-44 is a 10uF electrolytic
but is probably effective over the entire audio range. If electrolytic
caps were as bad as you indicate no bypass cap could work. The voltage
at the junction of C-44 and R-38 is well filtered DC.
If you look up the circuit for a "floating paraphase" phase splitter
you will find exactly what is in the SX-28.
The term "load" may be confusing, I mean the impedance in the plate
circuit. In this circuit it is coupled to the grid of one of the 6V6
tubes via a capacitor C-45 which is connected to the grid resistor of
one 6V6 R-41. The actual load on the 6SC7 is the combination of the
plate load and the following grid load.
I still want to know what is actually in these receivers. The
presentation of the bass tone switch is exactly the same in both the
SX-28 and SX-28A and in the military manuals. I have seen cases of other
errors carried over from generation to generation of instruction manuals
or even text books but a real answer would be to look at an actual
receiver and see how its wired. From the curves in the books the
labeling means that the boost is ON when the switch is IN.
To clarify a previous remark, at the time lots of bass was
considered desirable for entertainment audio. Not necessarily good bass,
just lots of it. This was how juke boxes were designed. Probably also
intended for use with the phonograph input.
This still does not address the hum problem but I strongly suspect
that the increased low end gain is exaggerating a hum problem in the tube.
On 2/17/2025 7:56 AM, Jim Whartenby via groups.io wrote: Richard
Comments inserted in the text below.
Regards,
Jim
Logic: Method used to arrive at the wrong conclusion, with confidence.
Murphy
On Monday, February 17, 2025 at 03:09:24 AM CST, Richard Knoppow via
groups.io <1oldlens1@...> wrote:
Ch2 and C-43 form a parallel resonant circuit, a tank, if you will.
It is in the plate lead of the pre amplifier half of the 6SC7. When the
switch is in the position marked IN on the diagram it is shorted out by
the switch. The plate load of the tube is then R-37.
No, the plate load for the 1st triode section is R37 and R38.? The
voltage gain of the 1st audio stage is approximately the plate load
resistance divided by the cathode resistance or 147k / 1k? or 147 when
SW10 is in the IN position. All of those bypass capacitors that are
sprinkled throughout the audio amplifier cause a fast roll off of the
midrange audio so that the bass frequencies are emphasized because of
the higher gain.
When the switch is
in the OUT condition, the tank forms the plate load and R-37 is shorted
out.
What about R38?? In the out position, the voltage gain of the 1st triode
is approximately 47k / 1k or 47.? Figure 11 shows this gain reduction
when SW10 is in the OUT position so Figure 11 actually agrees with the
schematic.
Note that C-44 along with R-48 is a bypass filter for the B+ going
to both 6SC7 plates. Since the plate load with the switch in the IN
position is resistive there should be no frequency discrimination. In
OUT the plate load is a resonant choke (about 1100 Hz).
Again, what about R38?? It is still in series with the parallel
combination of CH2 and C43.? In the OUT position, the frequency response
of the 1st Audio is now relatively flat so there is no base boost
compared to what you see in the IN position curve.
I was surprised
the frequency is not lower but calculated it a couple of times. I have
not looked up the plate resistance of the tube. R-36 and R-37 are 100K.
C-44 is 10uF and R-48 is also 100K so there should be no audio at that
point even down to quite low frequencies.
You are considering the electrolytic capacitors to be ideal.? The
components in the late 1930s and early 1940s were far from that.? The
ESL and ESR were not controlled back then.? Modern components are much
closer to the ideal but still aluminum electrolytic capacitors start to
become inductive at midrange audio frequencies.? This is why they are
not recommended as coupling & bypass capacitors for high end audio.
They cause distortion.
As far as fidelity the SX-28 was intended to be a good fidelity
receiver for AM broadcasts. Like the Super-Pro, it is designed to have a
wide IF and, for the time, a relatively high quality output amplifier.
This begs the question, what is the speaker that is used with this
SX-28?? Could it simply be the lower gain of the 1st audio amplifier
when SW10 is in the OUT position that causes the normal 120 cycle hum to
disappear into the mud?? If so, then there is actually no problem in the
SX-28 audio amplifier.? We are chasing our tail.
While broadcast stations in the old days were required to have good
performance to about 10Khz (double the standard now) few receivers could
recover much beyond perhaps 4 or 5 Khz. Hallicrafters offered a "High
Fidelity" speaker for use with the SX-28, a bass-reflex made by Jensen.
The bass boost offered is, IMO extreme according to the response chart.
There is a crude high frequency control, all roll off, probably to
reduce the effect of static and other noise. A HF boost would have
required another stage of amplification. They knew how to do it, see the
old (third edition) of the RDH for some circuits.
I think the labeling of the BASS control in the handbook is an
error even though it got continued in the 28A.
If you consider that BASS really means BASS BOOST then what is happening
will make more sense.
Someone with an actual
receiver can determine it for us, I am just guessing from what's on the
schematic.
I am strongly suspicious of the tube, easy to prove by substitutingit.
On 2/16/2025 11:11 PM, Jim Whartenby via groups.io wrote:
So the hum in question might just be 60 cycle leakage instead of the
originally stated 120 cycle?? I guess that this is possible but what is
the explanation for the loss of the hum when SW10 is in the Bass
(boost)
IN position?
I believe that the SX-28 schematic is correct after all.? With SW10 in
the Bass (boost) IN condition, CH2 and C43 are shorted out and R37 &
R38
make up the plate resistance.? In the Bass (boost) OUT condition, CH2
and C43 are in circuit but only R38 is used as the plate resistance.? A
lot of control is accomplished by a simple SPDT switch.
It seems to me that the curves in Figure 11 of the manual are
believable
since the contribution at 1kc of CH2 and C43 are, as previously stated,
clearly evident in the Bass (boost) OUT curve.? What the
contribution of
R35 for the midrange frequencies in either curve is not mentioned but
clearly, the lower the resistance of R35, the lower the available
midrange audio frequencies.
Perhaps part of the problem is that the SX-28 was made a decade before
the high fidelity craze of the 1950s.? Today, over all tone is
considered to be composed of bass, midrange and treble frequencies.
Treble frequencies for the most part are not available in the SX-28
since the high audio frequencies are limited to perhaps 3 to 4 kc.
Midrange frequencies are fixed by C40 in the phase splitter and by C42
and the R35 pot in the 1st audio triode section of the 6SC7.? Since the
SX-28 is, in the end, a communications receiver, the loss of treble
frequency control is not an issue.
Jim
Logic: Method used to arrive at the wrong conclusion, with confidence.
Murphy
--
Richard Knoppow
Los Angeles
WB6KBL
SKCC 19998
|
Jim Whartenby