Mystery RCA Mixer-Amp to Mic Preamp?

[doorunrun]

Hello all,

I've had this old RCA Mixer-Amplifier stored in the garage for over 20 years. Unfortunately squirrels took up winter residence years ago and made a mess of the chassis.




The best I can make from its inspection plate is MI-13115. Can't find any reference to it. The closest reference is a RCA MI-1109a with some pics in this thread:
RCA Vintage Tube Mixer - 4 X 1 Mic preamps
I can't find any reference to that number either.

I've got it in my head to rebuild it as a single channel ribbon mic preamp. It uses 6SJ7 pentodes for the mic inputs, a 6SF5 triode intermediate stage and 6J5 triode to drive the output transformer (not marked). The input transformers are XT-2452 that have received a low rating here.

I have an old console stereo chassis with power transformer & choke that I think might serve as a test bed for this theoretical mic preamp; I'll only need to pulls 1 input transformer, the output xformer and some tubes from it.

It might be an exercise in futility and I value the advice given here.

I'm drawing out a schematic and so far it looks pretty straight-forward.

Thanks!

 

[emrr]

The output xfrmr will be marked on the terminal board. I would assume this to be a PA product. May be cool. I don't think you need to pull anything unless you just don't like this chassis. You will have best noise with outboard power. Just disconnect 3 of the input stages, remove anything that's mixer related, like unneeded series resistors. The 6SF5 is likely there for mixer gain make-up, I'd switch to a lower mu tube like another 6J5/6C5. If the 6SJ7's are pentode connected, again you probably won't want all that front end gain, go to something easily pluggable like....6C5/6J5. Or strap the 6SJ7's triode.

 

[doorunrun]

The output xfrmr will be marked on the terminal board. I would assume this to be a PA product. May be cool. I don't think you need to pull anything unless you just don't like this chassis. You will have best noise with outboard power. Just disconnect 3 of the input stages, remove anything that's mixer related, like unneeded series resistors. The 6SF5 is likely there for mixer gain make-up, I'd switch to a lower mu tube like another 6J5/6C5. If the 6SJ7's are pentode connected, again you probably won't want all that front end gain, go to something easily pluggable like....6C5/6J5. Or strap the 6SJ7's triode.

emrr,
Thanks for those great ideas! The 6SF5's are pentode connected. The terminal board on the output xformer is covered a bit by the chassis. I'll try and get a closer look at it.
With a lower mu tube at the front end would you suggest keeping the intermediate stage (6SF5) or bypass it and go directly to the 6J5 output?

It would be simpler keeping the PS outboard for proof of concept and make some wiring adjustments in the original chassis. All the caps and resistors should go.
There are two multi-section electrolytic cans that each contain four 25uF@400V caps and appear to be used to stabilize B+ for each of the 4 mic inputs. I haven't checked the variable pots values. I want to get the schematic drawn before picking it apart.

Thanks again!

 

[doorunrun]

The output transformer # is 900542-501. I see it gets a mention on a topic here about the Ba2A.

BTW, I ordered a CD-ROM version of Ryder's PA Equipment Manual from eBay. Price is reasonable, could not find an online version. There are a lot of RCA MI's listed but none of what I'm looking for. I thought it might be good to look over some of those circuit designs if some small changes need to be made with mine.

 

[doorunrun]

This is a good representation of the mixer-amplifier's input stages:

I'm learning Scheme-It from Digi-Key so we're both a little rough around the edges.

I pulled out a couple of the .22uF and .05uF caps to get a clearer view of the wire routing. They all tested well in my component tester.
I also removed one of the multi-stage electrolytics (4@10uF-400V) and all sections were bad, no surprise there.

More documentation ahead...

 

[doorunrun]

This is a good representation of the unit's driver and output sections:

I'm not sure how common it was to add this break-out point to allow some sort of tone control.

 

[emrr]

I'm not sure how common it was to add this break-out point to allow some sort of tone control.

Not common, but not wild and crazy either.

 

[doorunrun]

I've rewired the unit so only one input stage feeds the driver output. All the existing carbon comp. resistors are still in place; only the caps were replaced.
I have some tube voltages to share. For testing purposes I'm using a SMPS-based source for B+ and Filaments.
B+ voltage: 220V
Filament: 6.3VDC.
The input stage is showing these voltages:
Plate: 24.62V
Screen: 23.03V
Grid: 0V
Cathode: 0.728V

...mixer-amplifier's input stage:
View attachment 113084

The next two stages have these voltages:
Driver: 6SF5
Plate: 151.2V
Grid: 0V
Cathode: 1.418V

Output tube is now 6C5:
Plate: 211V
Grid: 0V
Cathode: 7.6V

...driver and output sections:
View attachment 113134

I have a jumper tying the two sections together. Pots are set about midway. I do not have the Volume Indicator pot wired to the output.
The only dynamic testing so far has been to connect a SM58-like dynamic mic to the input and measure output with a Simpson 260 voltmeter set on AC.
The result was lots of output! The meter doesn't show a measurable noise level which good at this stage of testing.

I plan to put a 600 Ohm terminating resistor on the output and use my 'scope to look at the output signal for problems. I'd like to build a simple attenuator so I can connect my simple function generator to it for more testing.

Do these voltages seem reasonable? I appreciate any advice/modifications before going too much further.

 

[doorunrun]

I've rewired it to some degree. Tip of the hat to Doug, @emrr, for his advice on triode mode for the 6SJ7 and his version of the MI-12241 schematic. Here's the current set up:

The 'front-end' is close to the MI-12241 design save for the plate and cathode resistor/cap. I wanted to minimize rewiring and they were what I had on-hand. The next two stages follow the original MI-13145A design. I like having the 15k resistor in series with the OPT.

It has got reasonable gain; maybe a little too much. With VR1 open about 3/4 it starts to motorboat. That may be a result of the SMPS power supply. What little testing I've done with my ribbon and dynamic mics the gain pot sits about 1/3 the way up.

I have B+ set for 220V and reading these voltages:
V1 Plate: 50V V2 Plate: 107V V3 Plate: 153V
V1 Grid: 0V V2 Grid: 0V V3 Grid: 0V
V1 Cath: 1.95V V2 Cath: 1.08V V3 Cath. 5.5V
Which all seems pretty reasonable. I should probably read over the Class-A R/C values for 6SF5 and 6C5 and compare.

Cheers!

 

[emrr]

Cool.

First glance:
R3 is pretty low. 56K would be pretty low too but is indeed seen in old designs. Most modern would put at least 100K there.
R5 - can't be doing anything good. If it's fixing something, it's a band-aid not fixing the actual thing.
Take C3 way up in value, see if that better controls motorboating. That could also be layout. C2/5/7 probably better staggered at differing values if any are to stay so low. Common to just put 100 most places these days for better linearity....until it collides with motorboating. At last probably 100 at C7.

 

[doorunrun]

Doug,
Thank you for the comments!

R3 is pretty low. 56K would be pretty low too but is indeed seen in old designs. Most modern would put at least 100K there.

Here I was trying to match the resistance of the B+ resistor string in the MI-12241 pre...

R5 - can't be doing anything good. If it's fixing something, it's a band-aid not fixing the actual thing.

One thing I noticed after recording some of my narration is I'm over-driving somewhere.
Meaning, I'm testing using a little Mackie mixer using its Line input. The channel gain is all the way down ( +4 dB/unity). VR1 was set to produce adequate peaks on voice. Listening and looking at the recorded waveform it's obvious audio is getting limited somewhere. Backing off and recording at a lower level seems fine. Not sure where it's happening at the moment. I suspect the Mackie is getting overloaded.

Take 2c# way up in value, see if that better controls motorboating. That could also be layout. C2/5/7 probably better staggered at differing values if any are to stay so low.

2c# ? Probably C2, right?

Common to just put 100 most places these days for better linearity....until it collides with motorboating. At last probably 100 at C7.

Those 27uF @ 50V caps are there just to match the originals.

A little earlier in the process I tried bypassing the V2 stage connecting R5 to the grid of V3. The level into the Mackie was a lot lower, as you can imagine. At that point I thought it better to keep the V2 stage.

Thanks again!

 

[emrr]

Fixed it, C3.

trying to match the resistance of the B+ resistor string in the MI-12241 pre...

not a thing. only the last one after the last filter cap matters. 82K in the MI-12241.

I would remove R5 entirely. It looks like a mixing resistor in a thing that's no longer a mixer.

Do you have a 620-1K range resistor strapped across the output? That will knock max output down and give better low end response.

You might try R1 10K directly to an additional 10mfd/output trans, add 4K7+ between that and C3/R6. Might kill/improve motorboating.

 

[doorunrun]

OK, understand about the 82K and removal of R5.

Do you have a 620-1K range resistor strapped across the output? That will knock max output down and give better low end response.

Yes, it's a Dale 604 1% left over from my days in radio.

You might try R1 10K directly to an additional 10mfd/output trans, add 4K7+ between that and C3/R6. Might kill/improve motorboating.

I think I understand that. I'll draw it out; sort of resistor isolation between plate loads?

Or, just change R5 to 150K? Maybe I'm missing the point (advantage?) of having two resistors in series in the plate circuit.

Thanks!

 

[emrr]

It's just to add another filter cap with isolation between the stages.

 

[MidnightArrakis]

I've rewired it to some degree. Tip of the hat to Doug, @emrr, for his advice on triode mode for the 6SJ7 and his version of the MI-12241 schematic. Here's the current set up:
View attachment 115572
The 'front-end' is close to the MI-12241 design save for the plate and cathode resistor/cap. I wanted to minimize rewiring and they were what I had on-hand. The next two stages follow the original MI-13145A design. I like having the 15k resistor in series with the OPT.

It has got reasonable gain; maybe a little too much. With VR1 open about 3/4 it starts to motorboat. That may be a result of the SMPS power supply. What little testing I've done with my ribbon and dynamic mics the gain pot sits about 1/3 the way up.

I have B+ set for 220V and reading these voltages:
V1 Plate: 50V V2 Plate: 107V V3 Plate: 153V
V1 Grid: 0V V2 Grid: 0V V3 Grid: 0V
V1 Cath: 1.95V V2 Cath: 1.08V V3 Cath. 5.5V
Which all seems pretty reasonable. I should probably read over the Class-A R/C values for 6SF5 and 6C5 and compare.

Cheers!

Can anyone here provide the CINEMAG and JENSEN equivalent transformers for both T1 and T2?

Does anyone here have a recommended B+ power-supply schematic for this circuit?

I am assuming that VR3 is a trimpot?

If XLR connectors are used for both the input and output connectors, would their Pin-1's connect to the GND connection shown as TB??.

Would VR1 be either an "audio-taper/logarithmic-taper" or a "linear" taper pot?

Does anyone have any idea as to when all of the circuit component values may be finalized?

I have a ROYER SF-12 stereo ribbon microphone that such a circuit as this might be fun to use with!!!

/

 

[heatwalk]

C2/5/7 probably better staggered at differing values if any are to stay so low. Common to just put 100 most places these days for better linearity....until it collides with motorboating. At last probably 100 at C7.

Was the reason legacy designs used 20/40uF on cathode bypass largely attributed to space confinement with the multicans used back then? Larger caps give better bass linearity but there was physical limitation? Or were OG designers watching out for motorboating?

And regarding staggering of values - why would this be more beneficial?

 

[doorunrun]

Can anyone here provide the CINEMAG and JENSEN equivalent transformers for both T1 and T2?

I can't, sorry.

Does anyone here have a recommended B+ power-supply schematic for this circuit?

From what I can tell most preamp of this nature a 250V supply. To get it up and running I got a SMPS through Amazon: https://www.amazon.com/gp/product/B0992GCCSK/ref=ppx_yo_dt_b_asin_title_o04_s00?ie=UTF8&psc=1
It's not the greatest product and needed some component reworking. In the long term a tube-based linear PS is probably best.

I am assuming that VR3 is a trimpot?

In my case it's a LARGE trimpot, 36k Ohm ~5W pot with a 1/4" slotted shaft. Overall diameter is 1 1/4"
I wouldn't think you'd need one that large. I don't plan on using one for the rebuild.

If XLR connectors are used for both the input and output connectors, would their Pin-1's connect to the GND connection shown as TB??.

This mixer-amp used a barrier strip on the rear panel so really the numbers are meaningless. I used the Scheme-It app from DigiKey to draw the schematic and there are some limitations to the some of components used. Sorry for the confusion!

Would VR1 be either an "audio-taper/logarithmic-taper" or a "linear" taper pot?

I'm thinking I might be able to build a stepped attenuator to replace VR1. Since it's used for level setting I think linear would be OK.

Does anyone have any idea as to when all of the circuit component values may be finalized?

I have a ROYER SF-12 stereo ribbon microphone that such a circuit as this might be fun to use with!!!

/

Thanks for your interest and questions! You might want to consider the MI-12241 design that @emrr wrote about which is simpler (two tube).
https://groupdiy.com/threads/rca-mi-12241-preamp-kind-of-like-a-ba-2a-with-eq-addition.16210/
I'm trying to leverage what came with this mixer-amp which is essentially the transformers and tubes.

Good luck!

 

[MagnetoSound]

Lowering R6 to 56k or even 47k will reduce gain at V2.

Definitely get rid of R5.

 

[doorunrun]

Here's a revision based on suggestions by @emrr & @MagnetoSound as I understand them.

I've highlighted areas where I'm not sure of correct values. I appreciate your thoughts and suggestions!
Thanks!

 

[rock soderstrom]

Here's a revision based on suggestions by @emrr & @MagnetoSound as I understand them.
View attachment 115615
I've highlighted areas where I'm not sure of correct values. I appreciate your thoughts and suggestions!
Thanks!

You want to supply V1 with a voltage that is as clean as possible, so I would make the following change, which will give you more ripple filtering for V1.

The additional 1M resistor will keep the potential stabil on V2 grid even if VR1 gets old and corrosive. Not absolutely necessary, but good practice.