New Project: RCA 96B

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

DaveP

Well-known member
Joined
Nov 8, 2005
Messages
3,019
Location
France
Starting to think about next year's project.

I have been interested in the RCA 96A and 86A series as they are feedforward designs like the Group DIY 26C, which I hope will go up for auction soon.

I plan to make a simple  pre-war design like a modified 96A.  The final iteration of the old 6 pin 89 tube used in that was the octal 6K6GT, which is a low output tube, more in keeping with studio use.

The side-chain presents problems as the transformer is no longer available and in any case it is only fed by one of the signal phases.  The interstage is not really necessary nowadays either, but it was typical of pre-war designs; the price of copper must have been dirt cheap back then, as was labour.  I discovered with the 26C that the top end compression is very important so it is essential that the compressor has a full range up to 20kHz, this is much easier/cheaper with tubes than transformers.

I am guessing that the 86A was a later version of the 96A with uprated tubes, the component values remaining much the same.  Well I'll have to assume that they were as there are none shown on the 96A schematic.

I have designed an all tube side-chain to take the place of the transformer, an extra tube is a fraction of the cost of a transformer, I am guessing this was not the case before the war.

Anyway, I shall give it a go, I will probably sell it when its finished,  I plan to make vintage P2P tube designs for sale in my retirement in two years time, got to keep the brain active!

best
DaveP
 

Attachments

  • RCA 96A.pdf
    94.6 KB · Views: 128
Same as the 86, just use it as the model.  Only difference is the tube types and the outboard PSU.    I would have to call this a feedback limiter, like almost all vari-mu, the side chain signal comes from after the gain reduction stage, so it can only follow that which is already being reduced.  The BA-6 is the final RCA update.  The iron was definitely not cheaper than the tubes, looking at catalog prices. 

I'd be reluctant to pursue this design over the BA-6, if only because of the complexity of the stepped pots.  Those complexities are what RCA refined out. 

Maybexim missing something on the 26C too, it seems feedback side chain to me, side chain signal following gain reduction point.
 
Doug,
Point taken about the feedback. what I meant was that it did not take the sidechain signal from the output stage like most other vari-mu's, so it is possible to have an independent output control.

I won't be using the stepped pots, I will use my stock switched pads and a totally new side chain using both push-pull phases.

I have attached my first sketch, haven't sorted all the values yet.

best
DaveP
 

Attachments

  • BA-96B.pdf
    26.2 KB · Views: 68
Looks useful.  Maybe nit-picky, but without the interstage transformer B+ supply of the GR tubes, I consider it a different class of device.  There are some good older discussions here between PRR, Newyorkdave, and Larrchild about the differing effect of choke versus resistive loading of GR tubes, and I think it really affects the sound.  Both methods work, that's a different question. 
 
This is the thread:-

http://www.groupdiy.com/index.php?topic=24961.0

Copied it to word.......

57 pages to digest now!

DaveP

 
Since the fancy regulated PSU is the difference between 96 and 86, and I'm doubting the need for the regulated PSU here, this should really be the 86B.  That'll be more recognizable too.
 
Doug,

I think maybe your right, I only called it 96B because it will use the less powerful type of tube used in the 96A.  When it moves to the Lab I'll change the name.

best
DaveP
 
R2 can't be 5.6 ohms as indicated. 5.6_K_ may be right.

There's no "nFd" in vintage stuff. Won't sound vintage. Only mfd and mmfd.

Interstage: what happens on Attack? "CV" goes negative. 6SK7 current drops from 5mA to say 2mA. 6SK7 plate voltage rises from 135V to ~~206V. This is a rapid _71V_ positive-going transient to both grids of 6N7. It idles at about 2V 2mA (per section) bias. 71V up "should be" about 50mA (use 6N7 positive-grid curves and add cathode resistor drop). It can't be 50mA because 11mA would slam 6N7 plates from ~~200V down to ground. The 180V negative transient would shut-off 6K6es until the interstage R-C coupling networks discharge, many 500mS periods.

However when 6N7 grids rise more than 2V, the "infinite grid impedance" becomes about 1K. Then C1 C2 charge via R17 R17 through caps and grids. In this time no useful signal passes.

YES, such coupling was often done. It clips the initial transient. If not over-done, this avoids FCC tickets +and+ transformer expense.

In studios where everything is over-done, it isn't the same as transformer coupling.

An alternative (non-Vintage!) is to work the second stage as long-tail, so it can follow the common-mode transient without much distress.

The 6SK7 screen voltage "is" the basic reference for limiting level. If wall-voltage does not wander, or limiting level is trimmed for every track, it does not have to be regulated. If set-and-forget, with 10% (1dB) tolerance on limiting level and 10% line-sag on bad days, then screens should be regulated. Demand is small (but variable), a 100V (or 4*24V) Zener will do.
 
PRR,

Thanks for commenting.  Yeah, 5.6k typo.

I know, I will have to paint little mmF's on every cap or the audiophiles will hear the difference ::)

Re: the interstage.  So what you are saying is that the 71V DC transient would be cancelled in an interstage but not with cap coupling?  The time constant for a 22k plate resistor and a 0.47uF is about 10mS, so isn't this transient going to be still trying to fill the cap while the program will have already moved on?  The attack time of the 86A is supposed to be 1mS.

When I made the BA-6A I used this same 6SK7/22k/0.47uF combo and it seems to work ok.
Several other comps don't have interstages, Gates SA39 comes to mind, several other too.

What do you make of it?

best
DaveP
 
PRR,

Is this what you had in mind?

best
DaveP
 

Attachments

  • BA-96B-2.pdf
    26.6 KB · Views: 71
I think everyone in the eastern US is busy making storm preparations.

Good luck guys, hope its not too bad.

best
DaveP
 
I've finally got around to testing the GR on a pair of 6SK7's.

The conditions were B+ 250V, Vg2 supply regulated at 150V, plate resistors 22k, shared cathode resistor 500 ohms.  This biases the tube past the bend in the gm curve onto the straighter section (see data sheet)

The no GR quiesent voltages were Vp 160V, Vg2 100V, Vk 5.3V

The gain reduced from 14.4 down to 0.17 when the CV went from 0V to -39V, that is a GR of 38.5dB.

A chart is attached

best
DaveP
 

Attachments

  • 6SK7.pdf
    9.8 KB · Views: 21
Here are the actual figures

Vg1 Gain Loss GR
divider -dB
0.0 14.444 1 0
-8.9 5.555 2.600 8.30
-17.4 2.777 5.201 14.32
-23.0 1.388 10.406 20.35
-28.8 0.555 26.025 28.31
-36.0 0.277 52.144 34.34
-39.0 0.172 83.977 38.48

DaveP
 
I have a 96-a unit without power supply, and I look to build a new power supply can you help me?
In 96-a unit there a schematic but only for the unit.

I want to bring this 96-A back to life!
 
Copy the applicable portions of the 86 limiter, and you're there. Or use a regulated bench supply. The 86 manual info reveals all. The original supply would be extremely hard to duplicate, you'd need a custom power transformer and choke for starters.  I will be building a supply for my 96 using an International Power IHB/50-0.1 for B+, and would advise same for combination of practicality and closeness to original intent.

I'm collecting 96-A serial #s to figure out a rough idea of how many were made. Could I know yours? And is it a round VI meter or a rectangular VU meter?  Is the switch under the meter rotary or slide type?
 
I just need a schematic and information to build a 96-a PSU from scratch.

Im not a qualified electronic technician but I know a really good one they can help me.
 
Audio,
I have attached the 86 circuit, which is very close to the 96, just use this and adjust values to suit, which is what RCA would have done.
best
DaveP
 

Attachments

  • Rca86a1 Best Schematic.jpg
    Rca86a1 Best Schematic.jpg
    335.9 KB · Views: 58

Latest posts

Back
Top