1938 RCA 55-B bridging amplifiers

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

emrr

Well-known member
Joined
Apr 12, 2006
Messages
8,655
Location
NC, USA
Last spring I got around to restoring a couple of RCA 55-B’s, circa 1938.  Partial cleaning done, more to do. 

50054902516_edf827b78f_c.jpg


Cosmetic and ergonomic design by industrial designer John Vassos, well known for many other famous Art Deco designs. He headed RCA’s design department for several decades, and was responsible for the famous ‘mushroom’ knob first found here. 

The basic circuit was used from 1934 through 1950 for program amp use in various rack mount and console versions.  40-C, 40-D, 84-A/B, 55-A, 55-B, 83-A/B/C, and the 76 console program amp, all the same family.  Most versions have two or even three single ended stages, an interstage transformer, then a push-pull output stage that can deliver 1W into 600 ohms, or feed a speaker.  Look at the program amp in the 76 series consoles for an example; this is a short-loaded version of the 76 console program amp.    There are very similar amps in the Photophone lines also, though very obscure in comparison.

The earliest versions of these amps use 77’s and 89’s, mid period may be 1603’s instead of 77’s, then they moved to octal 1620 and 6F6.  It's interesting the pretty obscure 89 tube was used rather than the more common 42, same shift from the 96 to 86 limiter.  Virtually all 89 or 89Y you encounter are in US Navy contract boxes, and I'm not aware of anything outside of RCA broadcast that uses it. The vast majority are 89Y's, which are a low loss socket for RF use.  The earlier 59 tube is somewhat similar to the 89. 

As a side note to the puzzling aspect of the 96 limiter preceding the 86 limiter, the 80 series was the ‘Economy’ line (83-A/B/C, 84-A/B, and thematically, the 76-B/C/D consoles), there was a ‘Standard’ line (40-C, 55-A), and this 55-B is the ‘Deluxe’ line product along with the 40-D.  Mostly the same circuits and transformers in various cosmetic packages. The post-war BA-xx series is the new-design extension of the Economy line, with the old Standard packages and the Deluxe line dropped. 

This version is a 2 stage amp, it has just one single ended stage which is fed from a transformer secondary connected pot.  The input transformer is unique here, it has 600 or 20K input options, mainly intended for 20K to bridge and split main program lines for distribution.  It demands a matching load on the 50K secondary (provided by the pot), rings like crazy as soon as you lighten the secondary load.  Specs say it’ll take a +46VU input signal with the bridging connection……I’m not buying that the Daven pot will though! 

The chrome plated chassis was advertised as corrosion resistant, but after 80 years of neglect and tobacco smoke, most have some degree of crazing and pitting.  A fold down front door gives tube access from the front, and the chassis is hinged to the rear of the rack shelf for access to the guts from the back, ideally it didn’t have to be removed from the rack for service.  The chrome louvres on the front are totally overbuilt - individual plated pieces with brass rod and spacers threaded through, tightened with hidden nuts then screwed to the door and its mounting hinge.  Taking those apart for cleaning is very time intensive. 

50054902721_73cc3842c4_c.jpg


50054902961_1dc640932a_b.jpg


50054903191_03e95e66e7_c.jpg


Excluding a bleeder resistor and a pair of interstage secondary load resistors, all other resistors are multi-chamber wire wound types, so noise contribution equivalent to modern types.  The Economy and BA-xx lines went with carbon resistors, where the Standard and Deluxe used custom wire wound types in almost all audio positions. 

Some of the electrolytics still measure ok for capacitance, but all very high for ESR.  The one film coupling cap ‘worked’, but ESR was so high it actually rolled off some treble, that I’ve never encountered otherwise.  I found raising the value from 0.5 to 0.69 removed a 16Hz bump, making these flat to 12Hz.  These were turned on continually from 1939 through about 1989, so pretty impressive life performance overall.

50054902221_dc6dc6edd7_c.jpg


They measure great and match very well, I’ve been using them as make-up amps after passive mix bus summing.  -1dB is roughly 10Hz and 16kHz, very flat in between. On a few occasions I’ve used them as preamps, though power supply hum residuals would not suggest ribbons and quiet sources.  They are fine for lead vocal with a condenser.  In the dedicated program amp and console versions the power transformer or entire PSU is external, the two stage 55-B and the 80 series versions are the only ones with an entirely internal supply. 
 
Wow! thanks for posting :)  What a beauty that is.  Regardless of function, I just wanna own one or three as artwork.

50 years continual use is very impressive.  Considering the era these were born in and the limited range of all domestic systems then, so is that frequency response.  I bet they sound glorious. 

Off to look for/at a schem now.  I may be some time as this could turn into a cool rabbit hole :D 



 



 
Winston O'Boogie said:
  What a beauty that is.  Regardless of function, I just wanna own one or three as artwork.

Considering the era these were born in and the limited range of all domestic systems then, so is that frequency response.  I bet they sound glorious. 

First time I saw something from this series I experienced unavoidable gear lust for sure.

All the early ‘30’s pro gear I’ve heard has been satisfactorily ‘full range’.  Considering the number of series amplifiers in any system back then, they had to be to maintain something ’higher than’ the consumer receiving end. Everything from then we hear today is diminished sonics from the capture mediums. 

Check the 76 console manuals online and imagine some less parts.  Most of the relevant manuals aren’t out there, the 40-C version may be in tech docs, it’s in a few old books. 
 
EmRR said:
Considering the number of series amplifiers in any system back then, they had to be to maintain something ’higher than’ the consumer receiving end. Everything from then we hear today is diminished sonics from the capture mediums. 

True enough.  I forgot to account for the accumulated loss thru successive pre and program amps so, makes perfect sense. 



EmRR said:
Check the 76 console manuals online and imagine some less parts.  Most of the relevant manuals aren’t out there, the 40-C version may be in tech docs, it’s in a few old books.

Got it.  I looked at the 76 console schem and have a good idea of things now.  I think I'd rather have the pre octal valves though like yours is, just to make finding good sockets more difficult for myself. 
I've no idea if you found a difference with what you have compared to 1620's & 1621's?  But a lot of times I found those truly ancient valves can kick some serious butt. 
 
Good question on the tube type comparison. I haven’t found any deficiencies. The input tubes had been changed to 1620 actually, and one had 6V6’s in the output stage. They seemed fine, but I didn’t do in depth comparisons. Yeah, it was a bitch finding those sockets.  3 transformers tend to dominate sonics anyway, so long as tubes are reasonably similar and in the same environment.  It could be interesting to now make some octal adapters and do comparisons.
 
More on the chrome louvres.  You can't really get them clean without disassembly.  Overnight in lacquer thinner has proven the most effective and gentle method, then chase that with WD-40 wipe down.  This gets into the same territory as car and motorbike restoration.

51145020843_145c62cf5c_b.jpg


51144799191_49456a2865_b.jpg


51145022088_243f979320_b.jpg


51145578984_6f74333d5c_c.jpg


51145954460_b2bbce4b58_c.jpg


51145024383_115f99ccd8_c.jpg


51145580904_d8cdc1ed4b_c.jpg


51145028403_0ddd2b4452_c.jpg


51144804466_8b3d0e42dc_c.jpg


51145587524_8e4aeb4579_c.jpg
 
mjrippe said:
Wow, you were not kidding about the cigarette smoke!  Reminds me of the dingy green Pultec at the first studio I worked at  :D

Yep once you get this far back in time everything looks like this.  Usually some topsoil involved too. 
 
It is a beauty, great restoration!

It is a generic question but I guess it applies to this amp too. Amps from this era seem to use unshielded cases besides the frame and faceplate. The intended use was to put these in walls of racks. I'm sure it helped with heat and all but not so much with interference and noise? PSUs were remote I'm guessing but my very limited knowledge tells me these open frames did not perform well in terms of if and noise. I actually plan to build a few circuits this classic way but can't stop wondering if it worths the effort instead of a "standard" rack case.
 
EmRR said:
Usually some topsoil involved too.

Hahaha  :D
No joke on the restoring/cleaning of those chrome grills being akin to car or bike resto. work.
Great to see the innards stripped down like that, and really great to see that RCA went the whole hog with the Deco design - cost be damned. 


Janos: audio transformers on top are shielded in their own cans. All wiring and passives are in the bottom of the case which is shielded. 
Yes, the audio tubes (77 & 89) are dual ended with the control grids brought out on top, but these types were usually connected via shielded cable to a shielded top cap connector.  As were these.
I'm guessing Doug or a previous owner had to replace the one that's now white ceramic.  I've only ever seen  used top caps that were shielded and most were cruddy, rusty, and very old.  New ones are ceramic or Chinese plastic.  Unless someone knows otherwise.

Still, I doubt Doug has issues with it or he'd have fixed it.

Edit:  bringing the control grid out to the top like this was beneficial in certain cases as hum pickup from within the valve and associated wiring could be much lower.
This continued through the octal era and into the 9 pin miniature era for certain valve types.
A UK example of a nine pin from the early '50's built this way is the Brimar 6BS7 small signal pentode which was intended for A/ amateurs who might be somewhat sloppy in their socket wiring.  But also B/ lab instrument makers who were building devices to pick up very small, sensitive signals.





 
 
Yes it’s fully shielded.  At least triple shielding on transformers, power included, and you have to realize the outer case of can caps are used as shields between elements on many amps.  Here there are no tube shields because this is for high level audio which swamps most/all noise.  In the higher gain amps and preamps the tubes all have external shields.  The use of metal tubes in later pieces, inherently shielded. This construction is no different in most ways than a Gates Stalevel, and definitely more shielded.  I would argue many of these pieces are better shielded than most modern pieces enclosed in an overall box.
 
Winston & Doug,

Thanks for the explanation! I noticed the heavy shielding on the transformers and the component arrangement. WE / RCA and other engineers and designers knew what they were doing, no doubt.

What I really like about these units is the seemingly easy access to components and wiring. As much as I kinda hated the crammed build space and tubes hanging out with my "historical look" LA-2A, I'll attempt to build my diy amps this way where there is space to prep the layout better :) Probably it will help prototyping as well.

Thanks again!

J.
 
Winston O'Boogie said:
I'm guessing Doug or a previous owner had to replace the one that's now white ceramic.  I've only ever seen  used top caps that were shielded and most were cruddy, rusty, and very old.  New ones are ceramic or Chinese plastic.  Unless someone knows otherwise.

bringing the control grid out to the top like this was beneficial in certain cases as hum pickup from within the valve and associated wiring could be much lower.

Good eye, the smaller 6J7 cap was there and that's what I had to replace it with.    I searched forever and finally found a stash of NOS metal grid caps, those don't exist in any new version.  Soon to be updated here.

I think besides hum pickup being reduced with a grid cap, inter-electrode capacitance is also lower. 
 
EmRR said:
Good eye, the smaller 6J7 cap was there and that's what I had to replace it with.    I searched forever and finally found a stash of NOS metal grid caps, those don't exist in any new version.  Soon to be updated here.

I think besides hum pickup being reduced with a grid cap, inter-electrode capacitance is also lower.

Cool, thanks for the clarification on no new available metal top caps.  I gave up looking.
Yes I'm sure you're right on the lower inter-electrode capacitance. 

For all the hassle of the odd-ball sockets and searching for NOS grid caps, I still have a feeling I'd prefer having those valves  :)

 
Back
Top