Compressor Timeline!

[PRR]

> regular (unnamed; probably the RCA Photophone unit?) use in motion picture studios for several years, and that the delayed introduction of limiters into broadcasting is a mystery to the author.

I should be asleep, but:

IIRC, the WE sound recorder had benign overload; RCA dodged the patent with something like a ribbon mike which clashed badly on overload. So RCA *needed* a limiter for film recording. WE had one just because WE was thorough.

OTOH, radio didn't need limiters. Well into the 1930s the standard receiver used a power-law detector. The distortion was already bad at 30% modulation, though it was a long way from quasi-clipping. So actual peaks tended to be 10dB down from 100% modulation. And if they got a bit higher, the receiver sounded a little sour, not nasty.

And there were no loudness wars. There was little money in the racket, and not many stations in each area.

Mid-1930s the diode detector became common. Now radio could modulate 80+% cleanly. Also radio prices fell and fell. Listenership rose. This opened up advertising dollars. More stations came on the air. Having saturated the city, broadcasters went looking in the growing suburbs, meaning they needed more range but were already being limited in power. Now loudness had to be optimized to hold and grow audience and revenue.

 

[gary o]

In America, the first compressors I'm aware of all follow the lead of the Western Electric 110A, which showed up in 1937.  It was immediately joined by the RCA 96-A, the Collins 26C, the Gates 17-B, and shortly thereafter the Wilcox 57-D in 1938.  

Post-war, the Collins 26C was re-worked to be the 26W.  The Gates 17-B morphed through several revisions ending with the 28-CO's exit from the market after 1949.   The RCA 96-A lived on with a vastly simplified power supply as the 86-A.  The WE 110A was replaced with the 1126A.  

The RCA 96-A appears to be the first one using vari-mu tubes.  The Gates 17-B used some vari-mu in conjunction with lightbulbs (the first professional opto-limiter?)  as variable resistors in a bridge network.  The WE 110A and the Collins 26C also both used bridge networks with varying network arms to achieve gain reduction.  The Collins used tubes as network arms, the WE 110A appears to use Selenium rectifiers in each arm(!).  

The earliest article I've seen about broadcast limiting is a DIY article in Electronics Magazine, June, 1936, which I think pre-dates commercially available products.
The earliest mention of lightbulb limiting or expansion I've seen is in Electronics Magazine,  also somewhere around 1934.  I'll correct when I find it again.

Hi everyone ....Doug Im intrigued by these really old limiters are there schematics or anymore info about these, thanks.

 

[stitch-o]

 

[emrr]

A 1960 article written by an engineer at WLW in Cincinnati states:

"To the best of our knowledge, WLW was the first broadcast station to use an automatic amplifier; this was as early as 1935.  This is substantiated by our early patent position in this area."

No mention of the patent number nor any other references is made.  Any idea what this is referring to?

Joe

Thanks Joe, any reference there?  Where can the article be found?  If they were NBC, they may have had one of the early RCA/NBC network units which were not offered for sale.  With a patent reference, they likely had a home made unit.  It is well known that there were home made units in the field prior to the commercial offerings.  There's an AES article from 1950 that is written by an NBC engineer which mentions, in only vague detail, the use of tungsten bulbs as early as 1929, and the various in-house units built from the mid-1930's. 


Article and simplified circuits:

http://www.davidgleason.com/Archive%20Communications%20Magazine/Communications%20December%201937.pdf

The 86-A and the 96-A are the same thing, the 86-A uses a simplified PSU. 

 

[emrr]

IIRC, the WE sound recorder had benign overload; RCA dodged the patent with something like a ribbon mike which clashed badly on overload. So RCA *needed* a limiter for film recording. WE had one just because WE was thorough.

OTOH, radio didn't need limiters. Well into the 1930s the standard receiver used a power-law detector. The distortion was already bad at 30% modulation, though it was a long way from quasi-clipping. So actual peaks tended to be 10dB down from 100% modulation. And if they got a bit higher, the receiver sounded a little sour, not nasty.

Mid-1930s the diode detector became common. Now radio could modulate 80+% cleanly. Also radio prices fell and fell. Listenership rose. This opened up advertising dollars. More stations came on the air. Having saturated the city, broadcasters went looking in the growing suburbs, meaning they needed more range but were already being limited in power. Now loudness had to be optimized to hold and grow audience and revenue.

Very very good points, thanks!

 

[lassoharp]

A 1960 article written by an engineer at WLW in Cincinnati states:

"To the best of our knowledge, WLW was the first broadcast station to use an automatic amplifier; this was as early as 1935.  This is substantiated by our early patent position in this area."

No mention of the patent number nor any other references is made.  Any idea what this is referring to?

Joe

Thanks Joe, any reference there?  Where can the article be found?  If they were NBC, they may have had one of the early RCA/NBC network units which were not offered for sale.  With a patent reference, they likely had a home made unit.   It is well known that there were home made units in the field prior to the commercial offerings.   There's an AES article from 1950 that is written by an NBC engineer which mentions, in only vague detail, the use of tungsten bulbs as early as 1929, and the various in-house units built from the mid-1930's.   


Article and simplified circuits:

http://www.davidgleason.com/Archive%20Communications%20Magazine/Communications%20December%201937.pdf

The 86-A and the 96-A are the same thing, the 86-A uses a simplified PSU. 

Thanks for posting this - excellent.




. . . . . $2000.00 for an 849A. ouch.

 

[gary o]

Thanks for posting PDF Doug for some reason my reader says its unreadable probably my program out of date ......Bugger  :'(

 

[lassoharp]

??

In 2009ish dollars


sorry - a bit off topic

 

[PRR]

> Communications magazine Dec 1937, feature on broadcast limiters

http://www.americanradiohistory.com/Archive-Communications-Magazine/Communications%201937%2012%20December.pdf

 

[PRR]

> Communications magazine Dec 1937

I sense a press-release. Wireless World, Dec 1937, also has coverage of the WE 110-A.

www.americanradiohistory.com/Archive-Wireless-World/30s/Wireless-World-1937-12.pdf

> WE 110A appears to use Selenium rectifiers

Selenium was hot-stuff in that day; but this seems to be early Varistors, Carborundum (Silicon carbide), one of the fringe semiconductors.

"The units shown on the diagram as X1 are Varistors (a term coined by the Bell System to denote an element whose resistance changes markedly with either voltage or polarity -as the copper-oxide rectifier and the carborundum resistance employed as X1 in this unit. Thyrite is a third material of this nature), the resistance of which is a function of the voltage to which they are subjected. As will be seen, each side of the balanced network contains one Varistor in series and one in shunt connection. This arrangement maintains the terminating impedances at fixed values while varying the loss through the network. The loss through this network is controlled by a DC potential applied in the manner to be described in the following paragraphs."

V1 boosts/buffers to V2 rectifier, C1/C2 hold cap, R4 decay resistor. V3 takes this to make the current to drive the Varistors, apparently differentially to preserve circuit impedance.

"The constants of R4 and its shunt condenser are selected as a compromise to give an insertion time of about 20 milliseconds and  a removal time of 250 milliseconds for position 1 on the diagram (the normal position of switch Di). To provide for those users who desire slightly faster  operation, position 2 will give operation intervals of about one-half those in position 1."

 

[AusTex64]

some people think chandler is equally as dispicable as behringer, its not like they are the black sheep of audio or something.

dave

I understand Chandler has some kind of agreement or permission to build the EMI gear, whereas Uli Behringer stole Marvin Caesar blind, to the point Caesar sued them for infringement.