PWM compressor from 1952 and other tube limiter stuff

[Heikki]

I thought this might be interesting (PDF page 55)
https://ia902300.us.archive.org/24/...smpte-motion-imaging-journal_1952-02_58_2.pdf

 

[gyraf]

Darn, what a find! I want to try that..!

 

[zamproject]

YESSS !

Big fan of PWM comp, at least the early 70' Studer implementation (have 5 stereo modules loaded in my desk )

Since I dive in tube 2 or 3 years ago (with a crazy 50' syth restoration) I was hardly thinking of a tube PWM comp prototype/experimentation... I start nothing yet but like @gyraf I want to try that
At least there is something to start from, and not totally from scratch

Tks for the find @Heikki

Cheers
Zam

 

[zamproject]

After a quick read, the proposed pulse generator seem very low at 45Khz
maybe a typo for 450 ? or they don't care that much about intermodulation attenuation at 45Khz by that time ?

Implementation is interesting with the pushpull cathode. I had other way in mind with a remote pentode

 

[gyraf]

..they aim for double audio bandwidth to have room to implement aliasing filter. Quite good explanation in the text imo..

/Jakob E.

 

[zamproject]

I read more since, and yes the 45kHz is confirmed, they talk about 30Khz min (with BW at 15khz) as pulse integration at 22us.
How do they have enouth room for the filter ? there is no info of the filter used, I wonder how they do with only twice BW to not cut significantly at 15k or not having enough attenuation at 45k ?!?
The picture of the unit show 3 trafo (so probably interstage) and way more tube than basic drawing, maybe they used high order active filter ?

IIRC all PWM comp that I look at have at least 10x BW pulse

 

[Heikki]

I'd rather not make a new thread, so I'll put this here.

Dynamic Performance of Peak-Limiting Amplifier (PDF page 207)
https://worldradiohistory.com/Archive-IRE/40s/IRE-1947-11.pdf
At the end of the article there is described a feedforward limiter with power amplifier is driving driving the rectifier to create control voltage. Has anyone seen a schematic for this limiter?

This excellent performance is attributable to a unique
circuit design wherein the automatic-gain-control voltage
is a function of the input signal voltage, rather than
of the output signal voltage. The control-voltage gen.-
erating section of the amplifier incorporates large powertype
amplifier and rectifier tubes in low-impedance circuit
arrangements, resulting in extremely fast develop
ment of the automatic control voltage. Another major
factor contributing to the exceptional performance
shown in Fig. 7(a) is the use of a special time-delay network
in the signal channel just ahead of the point where
gain reduction takes place. This network acts to delay
the signal by approximately 80 microseconds, and the
gain is already reduced by the required amount upon
the arrival of the peak at the point where gain control
is efected. A close examination of Fig. 7(a) will reveal
that automatic gain reduction occurs near the point q
in time, whereas the 14-db peak does not arrive until a
!apt- time at point p.

The 10,000-cycle performance shown by Fig. 7(b)
again ilustrates the extremely short efective attack
time of the amplifier. Even at 10,000 cycles, the 14-db
peak which arrives at the point p in time is efectively
prevented from exceeding the maximum steady-state
value. Fig. 7(b) indicates excessive gain reduction (overcontrol)
for the first cycle after the arrival of the peak.
Over-control which persists for so short a time as shown
in Fig. 7(3) (approximately 100 microseconds) cannot
be perceived by a listening test, and is certainly preferable
to under-control, since it renders overmodulation of
subsequent equipment impossible.

It can be observed from Fig. 7 that the "thump" component
of the amplifier is of very small amplitude, an
additional design feature of this unit.
The amplifier described above was developed by E.
E. Schroeder of the CBS-Chicago technical staf, under
the direction of J. J. J3eloungy, and has been used at
station WBB M since 1945. Additional amplifiers, based
upon this development, are in service in other Columbia
Broadcasting System stations and are also available
commercially from a well-known manufacturer.

EDIT: GE BE-5-A seems to fit the description

 

[emrr]

I'd rather not make a new thread, so I'll put this here.

Dynamic Performance of Peak-Limiting Amplifier (PDF page 207)
https://worldradiohistory.com/Archive-IRE/40s/IRE-1947-11.pdf
At the end of the article there is described a feedforward limiter with power amplifier is driving driving the rectifier to create control voltage. Has anyone seen a schematic for this limiter?

EDIT: GE BE-5-A seems to fit the description

That GE makes the most sense of anything I can think of. They are cool sounding units, alignment has to be "by the book" or the meter reading bears no relation to gain reduction.

 

[Heikki]

Perhaps one of the first dynamic processor ever used.

A compandor system has been in service on the New York-London long-wave radiotelephone circuit since about July 1, 1932.

https://worldradiohistory.com/Archive-Bell-System-Technical-Journal/30s/Bell-1934c.o.pdf
In the compressor tubes were used as a shunt resistance of a H-pad and expander used the vari-mu arrangement we are used to seeing. The text does mention.

The circuits as shown use variable shunt control for the compressor and variable series control for the expandor. Either or both may be changed to the other by inverting the polarity of the control potential and properly designing the rectifier characteristics of the control circuits.

 

[richiyobs]

Hey super Heikki
Are you going to design a new interesting device?
Hope so....
Following
Best

 

[Heikki]

Are you going to design a new interesting device?

No, nothing new coming except improvements on some of my older designs.