vintage RCA amplifiers MI-12172-A (pics)

[Mailliw]

Hi,

I was referred to this forum from the good guys over at headwize.com.

I was given these rca amplifiers and would like to restore them for use as mic preamps. I have -zero- knowledge of tube circuits and just minor electronics knowledge in general.

Is there a schemtic for this floating around? Any help would be greatly appreciated!

more pics here:
http://s86953011.onlinehome.us/preamps/

Thanks,
William

 

[NewYorkDave]

I believe those are components of an RCA film sound system. If I come up with any info, I'll post it.

 

[NewYorkDave]

Well, I haven't turned anything up so far, except one post on Usenet from someone who was asking essentially the same questions you are. What I did glean is that the preamp uses a 12AY7 dual-triode with a 250:50K input transformer. It may be a fine mic preamp as-is. It may even be similar to the famed RCA BC-2B preamp, which uses the same tube and an input transformer with about the same turns ratio. There isn't much inside the box, so it's easy enough to trace out the circuit.

 

[PRR]

> I was referred to this forum from the good guys over at headwize.com.

Thread there, but I think the topic is more appropriate here. Re-posting my remarks from there:

> I am not sure how old they are.

I'm going to say 1956 to 1965. That's just a gut feel from living through that era and handling a LOT of resistors, caps, etc etc.... (gosh, looking at these pix I can smell the rotted caps, and how the solder-flux will stink when you replace them.) I could be off either way; especially RCA Audio disintegrated in/after that period and may have been tossing stuff together out of old parts years after the apparent date.

I'll also say that this is not some of RCA's "finest stuff". Someone at The Lab scored the custom mixer for American Tobacco Company from around the same time, and the difference between Langevin (and older RCA) pro-audio gear and this lightweight gear is large. But it may still be a ton of fun to restore and use. And there are people paying a lot for "vintage" gear. More if they know it works, especially if it is "Racked" (as-if relay-racks were the only good mounting).

I bet all the under-chassis cardboard-sleeve electrolytic caps will die the instant you apply power. Not a big deal to replace (but get a fan! Old flux stinks!). Other caps may fail. Contacts on connectors and switches are tarnished and will need cleaning. Tubes most likely work. The selenium rectifier in the power supply is astonishing, and I would not bet either way on it. No matter, if it smokes or croaks you can fake a heater supply enough to test everything else.

Added remarks......................

> I want to wire one for testing, but I am not exactly sure how to do it.

It is a pity the harness was cut, but there is enough left to give good clues.

I'd actually start with the power supply. And start cautiously. It appears to be in good condition, but old idle electrolytic caps may have deformed and may be open or short. You are supposed to put 117V wall power to the "power" "C" and "117" screws. (Line-cord may not have been grounded; if so it probably grounded to the overall case or rack.) I would get an old cheap lamp and wire it in series with the power supply C/117 screws. Wall-power comes out of the wall, through the lamp, through the power supply, and back to the wall. Start with a 60W lamp, so at worst-case there will be less than 60 Watts of excitement. (Yes, the fuse will protect against sustained shorts, but you could be replacing a lot of fuses before you find all the problems.) Remember there are lethal voltages on the "power" screws and also (we hope) at the "B" screws. Set a battery-power voltmeter to read 10V AC and clamp the probes under the two "6.3V" screws. If you have a second meter, set it for 500V DC and put it on the "B" screws. Put the supply on a flame-resistant surface, use a long extension cord so you can stand back at plug-in.

Ideally, at plug-in the lamp may flash to full brightness and then go dim or even dark. That means the caps took a charge and don't leak. The meter should show 6V to 7V AC. The B voltage should be a couple or three hundred volts.

If the lamp stays bright, meters read low, something is shorted. Either the blue-fin rectifier or the caps is my bet. I'd be inclined to sketch the circuit, and then rip all of that out. A couple silycon rectifiers can replace blue-fin, and modern Japanese electrolytics are much better than the old stuff. (Leave the blue-fin peeking out the holes just for looks. It is a distinctive historical artifact: end of the tube era, dawn of solid-state.))

The way the harness was cut, I assume you can't figure out how the PS was wired to the modules, so we have to reverse-engineer. Take a 12AY7 preamp and get a 12AY7 pin-out diagram. Two wires run from the module connector to pins marked "H" on the 12AY7, and nowhere else. This is heater power and seems to be 6.3V. Either AC or DC will work, and in a pinch a 5V 2A wall-wart will heat-up the tubes so they work OK (not quite as good as 6.0V-6.3V, but it will give gain and pass signal). Two more wires run from module connector to the octal socket under the grey can, and (probably) nowhere else. These are your mike input connections. One of those black stripe caps goes from a 12AY7 plate or cathode to one of the module connector pins, that would be signal output. One module connector pin runs to the outer shell of the big aluminum can, that is "B-" and also probably output signal return. Another module connector pin runs to an inner terminal of the big aluminum can, possibly through a few-K resistor (looks like a 50K 1W); that is "B+". B+ was probably around 200V-250V, though it will work down to 100V (reduced overload level) and up past 300V (limited by cap ratings).

There is a switch on the volume pot. It seems to run to a bunch of wires in heat-shrink tube, which is anachronistic (heat-shrink came later than amps like this). Seems like a later modification. Are they all like this? Is there any clue where they once went? How many clicks on the switch?

 

[PRR]

This is another (older, grander) RCA mike amp. But it has a 10-pin connector, as do yours. I dunno if it is the same connector, but you might study the pinout for similarity. RCA could change pinout on a whim, or keep the same/similar pinout through many product lines and many years. And on the whole, RCA has been less fashionable than some other brands, so there is less collected information.

11-12 go to the grey transformer (on the mike-amps; the mixer and other odds/ends will differ).

6-7 go only to 12AY7 pins.

9 and/or 10 go to big filter cap shell and several other places.

5 goes through a resistor to an inner lug on the filter cap.

Your amp appears to lack the second transformer T-2, so the connection on 3-4 will not be the same. It has several other differences, like a 12AY7 twin-triode instead of two pentodes. Yours is intended for a system that is all inside the box, no external patching, and probably low-cost. This other is a general purpose workhorse for huge systems with long lines and patchbays. Yours seems to have a gain control in it, this other would sit remote from the controls.

The other one is Broadcast, where you live or die by audio reliability and flexibility. Yours smells more like an occasional-use auditorium PA mixer: 6-in, tone, volume, no patching, and the power amp in the same rack (no long line from mixer to power amp). Of course both preamps are obsolete in their original markets: we can buy fine PA mixers for chump cash, and low-end broadcasters use slightly better modern stuff. So the value of this is as "flavored" mike channels or possibly (with much work) as a small 6/2 location stereo mixer with "flavor" and "glow".

 

[Mailliw]

Hey,

I tested the power supply following the method you outlined. I had a 60W bulb in series with the PS C/117 screws, a digital multimeter on the B screws (set at 500VDC) and an analogue meter on the 6.3V screws (set at 10V AC).

When I first plugged it in, the bulb went slightly dim and then quickly faded out. The 6.3V screws measured just slightly below 6 volts and the B screws went to 100 volts within a few seconds and slowly kept rising. Over the next few minutes it reached 150 volts and kept going, but its rate of ascent decreased and decreased, where at 150 volts, it would take 10-15 seconds to increment one more volt. It appeared as if it would keep climbing forever, albeit very slowly.

Without the bulb wired in, the same phenonmea occured. However the 6.3V screws now measured 6 volts exactly, whereas before it was just under 6 volts.

After I unplugged the PS, I tried to get a voltage reading off the capacitors, but nothing measured. Shouldn't they have a residual charge left in them, or are they set up to be bled after the wall power is disconnected?

As for your question about the pots on the voltage amplifiers, the inner-most knob clicks twice, ie it has 3 positions.

 

[CJ]

The slowly increasing voltage might be the filter caps reforming. Or maybe something going on in the power transformer core.

 

[PRR]

> the B screws went to 100 volts within a few seconds and slowly kept rising. Over the next few minutes it reached 150 volts and kept going, but its rate of ascent decreased and decreased, where at 150 volts, it would take 10-15 seconds to increment one more volt. It appeared as if it would keep climbing forever, albeit very slowly.

I would expect over 150V (I could be wrong). However I see 450V caps, so it probably should be over 300V no-load.

It should go up as fast as the meter can move: slow-rise either means a high-resistance rectifier (possible on selenium) or very leaky caps (highly likely).

Don't leave it that way for more than a few minutes (I should have said that before....). Very leaky caps will get hot, and may burst before they re-form (they may never re-form well). Bursting is more frightening than dangerous, but a real mess to clean up. I see no good reason to keep them. They were cheap, are now old, are not visible, and are cheap to replace.

I really can't make out the circuit from these pictures. The way the two cardboard caps are wired suggests a voltage doubler (but why 2*450V?). If so, the middle lug on the blue-fin selenium rectifier would be an AC input, but it seems to run through the red-red-blk 22Ω resistor to a junction of three resistors that look like DC filtering.

Unless you can figure out the circuit better, I'd suggest just replacing the caps. 450V, over 33uFd, hi-temp.... try http://www.digikey.com/ part P5951-ND $2.35 (yes, this is 47uFd, but better, and oddly cheaper than the 33uFd 450V parts). These are far smaller than the originals, really much better. They are differently leaded: you will have to extend the leads to reach where they need to go. They may fit the original clamps, but I think clamping electrolytics is bad (I've seen them crushed and shorted). For testing, just get them connected. BTW: a wire on the lug top-right of ps3_640.gif is soldered to the isolated lug but appears to be touching the grounded lug. That can't be correct. Be sure it clears.

You may have other leaks in the two big aluminum cans. But I think with the two cardboard caps replaced, you should get large and fairly solid voltages on the two underchassis caps.

I've suggested you buy $5 of caps but DigiKey has $20 minimum. List the volts and uFd of ALL your caps, all modules, including the 2 to 4 caps inside each aluminum can. That should put you over a $20 minumum, and I would assume that most of these caps are now sick and need replacing. And I suspect you can replace them all with two types: a 30-50uFd 450V like that P5951-ND which is $20.11 for 10, and a 220uFd 25V 105F like P12383-ND $2.16/10.

Theory: in general, amps have a high voltage supply that needs filtering, and some low-voltage points that need bypassing. In a tube amp like this, we might start with 350V DC and run through several stages of resistor capacitor filtering to get 150V of very clean DC at the input stage, with high-level stages tapped along the way. So you need a bunch of caps from 200V to 400V, and often about 30-50uFd, and not particularly critical. More V and more uFd is better, until you go to extremes. 450V 47uFd will replace all the power caps. There are also some cathodes that need bypassing, and only need to be 2V or 5V or so. The 450V will work at first, but is extremely over-rated (and overpriced) and may de-form sooner than if worked near its rating. And cathode caps were usually "too small" for economic reasons: I'd slap in 220uFd 25V for any 5uFd-150uFd low-volt cap, especially since they are only $0.22 each.

> After I unplugged the PS, I tried to get a voltage reading off the capacitors, but nothing measured. Shouldn't they have a residual charge left in them, or are they set up to be bled after the wall power is disconnected?

Both. There is a bleeder (I think), but if they are leaky they will self-discharge real fast.

> The slowly increasing voltage might be the filter caps reforming. Or maybe something going on in the power transformer core.

CJ- shut yo mouth! First, I have never seen that on a power tranny, even when feeding through a lamp. Second, we really HOPE that big iron is good: it would not be easy or cheap to replace. One reason I like the 60W lamp in series is that a transformer fire is very unlikely no matter how many rectifiers and caps have to be burnt-out before it all gets happy again. Caps and rectifiers we can replace cheap. Hammond may have a "similar" transformer, but you know it won't be quite right, especially the mounting feet. So don't jinx it!

 

[NewYorkDave]

I would never power up old gear (or any other potentially faulty gear) without my homemade lightbulb-based current limiter. A Variac is nice-to-have, to ramp up the voltage gently, but the lightbulb limiter is essential. I used to have one with several lamps that could be switched in series or parallel for a different currents. But nowadays I use a simple single-socket setup consisting of a "floodlight" type fixture screwed into the side of a two-gang electrical box, with a single-gang outlet and an on/off switch mounted on the plate. I just plug in different lamps as needed.

 

[CJ]

Mouth shut! :razz:

 

[Mailliw]

Hi,

I sketched, to the best of my ability, a schematic of the power supply:

Next thing I am going to do is find out all the capacitor values from the amplifier boxes (like you said) for my digikey order. Should I add silicon rectifiers to my order, to replace the selenium one?

 

[SSLtech]

The 10-pin pinout is exactly the same as the five BA-11's I was just working on here away from home.(The BA-11 is longer and has output trannys though)

I recapped everything using the same caps PRR has suggested and the voltages look better. I'm also staring down the barrel of an old Fended Deluxe which has also dried up the old, cheap main power caps.

Replace them.

Keith

 

[PRR]

Well, it is a voltage doubler, though it is a mystery why it uses "30uFd 450V" caps when each cap can only get half the output voltage and the output voltage can not be more than 450V, the rating of the output filter. A lowest-cost design would use 250V in the doubler. Maybe they had a crate of these 450V, maybe these have been replaced once already (repair shops usually stocked-up on 450V caps).

Also there should not be more than 400V on the 150K 1 Watt(?) bleeder, which with the 27K matches the 450V rating of the output caps.

300V output leads to a more conservative design: half the resistor power rating, 2/3rd the cap voltage rating. So I would not expect a lot more (though I did have an RCA record player that ran a full Watt in a 1-Watt resistor.) In heavy-duty pro preamps, B+ was often held to 250V to stay well clear of the tubes' 300V rating, but this looks like it is all resistor coupled, tube voltage maybe 2/3rd of supply voltage, so 300V is fine and 450V not too daring. But stuff like this can also run on 150V just fine (lower overload level), so that would not be a surprise either.

The 150K, 27K, and 56K do two things. The 150K 27K bleeds about 2mA, so it won't stay charged forever if the amplifiers are disconnected. And the 56K and cap-section go to a transfomer wire, which has to be the center-tap of the heater winding. When all is well, the heater screws have 6.3V AC across them, and also about +50V DC to ground. Ideally heaters are perfectly insulated from the signal-handling Cathode, but nobody's perfect. A small positive voltage on the heater tends to make any leakage flow to the heater, rather than to the cathode. Also we know you have a Cathode Follower, with cathode well above DC ground, and holding the heater positive reduces stress on its heater insulation.

Are you looking to "restore" or "improve" this? Because there isn't a lot of hum filtering there. Estimating the rejection in each amplifier, you may have a signal/hum ratio of only about 70dB-80dB. Acceptable in 1960 PA systems, poor in modern 16-bit 96dB S/N recording systems. Doubling all the caps, to say 100uFd, would get another 24dB of hum rejection and put you in the modern ballpark.

> Should I add silicon rectifiers

I just don't know. Why don't you get some just in case? 1N1007 is the general purpose 1 Amp diode up to 1000 Volts (get a 10-bag, you can use 1N1007 in place of any 1N100x-series diode, used in many many DIY projects). DigiKey has 1N4007DICT-ND $1.36/10 but only 1,077 in stock today. If they run out, they have tons of 1N4007GDICT-ND $2.48/10.

 

[Mailliw]

Hi,

I calculated all the cap values, they are:

Voltage Amplifiers (6)
---

1x .1 MFD, 200 V D-C, +/- 10%, molded paper

2x .1 MFD, 400 V D-C, +/- 10%, molded paper

Mixer-tone control (1)
---

2x .022 MFD, 400 V D-C, +/- 10%, molded paper

1x .0047 MFD, 800 V D-C, +/- 10%, molded paper

1x .047 MFD, 400 V D-C, +/- 10%, molded paper

1x .01 MFD, 400 V D-C, +/- 10%, molded paper

1x .001 MFD, 800 V D-C, +/- 40%, molded paper

Mixer-cathode follower (2)
---

1x .1 MFD, 200 V D-C, +/- 10%, molded paper

1x 25 MFD, 25 V D-C, paper

1x 5 MFD, 150 V D-C, paper

Are you looking to "restore" or "improve" this?

I'm interested in how these would work for vocals or guitar. I don't want to put too much money into them without knowing the result.

Does it make sense to "restore" the power supply and one (or two) of the voltage amplifiers, and then try it out to see if I like it (keeping in mind the hum could futher be reduced). I don't mind making an initial investment if its only 20-30 bucks, and then a bigger investment later, if I am satisfied and desire more out of the amplifiers.

Thanks

 

[PRR]

> Does it make sense to "restore" the power supply and one (or two) of the voltage amplifiers, and then try it

I would bet most or all of the paper caps are still good. And I would keep them for initial tests. Caps affect sound a little and these are not the super-virgin neatsfoot-oil caps that AudioPhiles crave, but they won't suck and they may have a Vintage sound that would be lost with too-good caps.

I think you have forgotten the big aluminum can-cap on top of each voltage-amp? I'd guess 2*40uFd 450V, something like that. And that's a 50:50 bet to be bad.

Yeah, I'm trying to get power and one mike-amp up and running: that should be enough for a test and see where you want to go. If you are in the over-kill school of thought, you might over-size the power caps now, rather than go back and do it later. OTOH I think I mis-estimated the ripple: it won't be as bad as I said with all stock-value parts, will be quite usable. And if after trials you decide it could be cleaner, that's not a big chore. (Also when you run just one module the ripple will be much lower than with all 8 or 9 modules loading the supply.)

 

[Mailliw]

the aluminum cans on the voltage amplifiers are:

20 MFD, 350 V / 25 MFD, 10 V

I sort of overlooked the big aluminum ones, they look very durable

 

[Mailliw]

I drew up a schematic for the amplifiers below.

I noticed that 2 pins on the 10-pin connector had snipped wires still soldered to them. Also I discovered that all 3 positions on the pot-switch are soldered to the same pin on the back connector. The two snipped wires used to belong to the 2 remaining switch positions?

Thanks,
William

 

[NewYorkDave]

I took the liberty of creating a schematic based on your beautifully-drafted wiring diagram. Many of us will find this format easier to follow.

It all looks pretty straightforward. The only big mystery is the function of that 0.1uF cap connected to the input transformer. Do you read any continuity between pins 4 and 7, or between either/both of those pins and the other windings?

One thing that's immediately obvious is that these preamps are for use as part of a system (e.g., a mixer) as opposed to stand-alone. The output is taken right off the 12AY7 plate and the source impedance is too high to interface to the "outside world", as it were.

 

[PRR]

> overlooked the big aluminum ones, they look very durable

Nice cans often hide cheap junk. They may be fine; but if not, they'll suck all the B+ away from the tubes. And possibly burst and spew paper-muck, though the large 51K resistor will probably burn-up first.

> I drew up a schematic for the amplifiers below.

Connector marked:

Also: grounding the transformer primary center-tap is good 1950s radio-console practice, but incompatable with Phantom Power, and not needed if you do not work inder a 50,000 Watt AM transmitter. Sometime in this project, certainly before you try to add Phantom power, cut the wire to pin 6 on the transformer. If you do have a large radio station nearby, replace the wire with a 1nFd capacitor to Chassis Ground. But most of us just let the primary float and it works fine.

Re-rendered in conventional form:

The 300K 51K 100K feedback network around the first tube is so unusual that I had to model it to see what it does. It forces the input impedance to about 50K on the secondary (250Ω at the primary). This defies recent convention: we normally work a mike into a "high" impedance like 2K. Matching was sometimes done in the old days (though RCA's Olsen argued against it about the time this amp was built) and seems to be coming back in fad as "variable impedance". The feedback around the capacitors also tends to bump-up response around 20Hz, probably compensating small but cumulative bass-loss in the transformer and coupling/bypass caps. It also limits gain in the first stage: it can swallow the output of the hottest mike and musical source without great distress.

I don't have a simulation model for 12AY7, but it will give results midway between 12AT7 and 12AU7:

150 ohm source:

With 12AT7
total gain:
75dB, 1Meg load;
69dB, 33K load (Zout ~ 33K);
63dB, 10K load (-3dB at 37Hz)
input Z: 250 ohms

With 12AU7
total gain:
64dB, 1Meg load;
60dB, 33K load;
58dB, 18K load (Zout ~ 18K);
54dB, 10K load (-3dB at 54Hz)
input Z: 370 ohms

The unloaded gain seems enormous; but in the orginal app it would feed some kind of 6-in mixer with some nominal impedance and at least 15dB loss. For speech PA use this gives a couple volts at the mix bus, a nice value. The gain pot is surely tapered so you can reduce gain 30 or 40dB if needed.

Using these amps solo: the high output impedance means they should not drive more than about 10 feet of output cable or the highs will roll-off. (Not a problem on the original harness!) And many modern line inputs are only 10K impedance: gain and output level is reduced significantly. However if the line-in input sensitivity is set to give loud signal from a home hi-fi CD player, this amp can make similar levels even into a 10K input.

So it will work as a modern standalone preamp, but keep it close to the board or recorder. And you want to triple the size of the output cap to maintain 20Hz response in low-Z loads.

 

[Mailliw]

Thanks Dave! I really appreciate all the help you guys are giving me.

As for the input transformer, my test showed no continuity between any two pins.