vintage RCA amplifiers MI-12172-A (pics)

[PRR]

> big mystery is the function of that 0.1uF cap connected to the input transformer.

Is it really connected to the transformer? Or does it go with a dummy-plug for line-in use?

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

ANY pins? 5, 6 and 8 should all be a few dozen ohms apart. 2 and 9 should be about 5K apart.

"Continuity testers" often disregard connections over a few ohms, use an ohm meter. And digital meters are often baffled by high-inductance windings like 2-9: use an analog ohm meter. But best to not use an ohm meter (or continuity tester) at all: the DC test current can magnitize small transformers and reduce their performance. Mark the tranny you tested. Don't test any more until you can test with audio. If it passes signal well, we don't care what the ohms may be. If you ohm-tested and then it does not have as much bass as the others, you'll want to demagnitize it.

 

[Mailliw]

I did use a digital multi-meter to test the input transformers (yes 2 :\ ), so maybe thats why! I've marked them for later on, like you suggested.

About my digikey order; I'm getting lost in all the discussion about which caps I need to buy.

For the power supply underside cardboard capacitors, I'm getting: 47uFd, 450V (P5951-ND). These can also replace the .1MFD 400V, and the .1MFD 200V caps in the voltage amplifiers?

I'm mostly confused about the aluminum can capacitors, I am subsituting these multi-unit capciators with 2 individual capacitors, and wiring them underneath?

What I believe now:

voltage amplifiers:

old ---> new
2x 0.1uFD, 400V ---> 2x 47uFd, 450V
1x 0.1uFD, 200V ---> 1x 47uFd, 450V
1x 20uFD, 350V ---> 1x 47uFd, 450V
1x 25 uFD, 20V ---> 1x 220uFd 25V

power supply:

old ---> new
2x 30uFD, 450V ---> 2x 47uFd, 450V
3x 40uFD, 450V ---> 3x 47uFd, 450V
1x 20uFD, 50V ---> 1x 47uFd, 450V

That makes a minimum order (to restore the power supply and one voltage amplfier) of:

10x 47uFd, 450V
1x 220uFd 25V (but wise to buy ~10, considering how inexpensive they are)

I'm sorry that you have to explain this to me more than once.

Take care,
William

 

[PRR]

> I'm getting: 47uFd, 450V (P5951-ND). These can also replace the .1MFD 400V, and the .1MFD 200V caps in the voltage amplifiers?

No!

The big 47uFd caps are in the power supply, where a milliAmp of leakage is tolerable. (When I say that old caps have gone "leaky", I mean many-many milliAmps of leakage, enough to drag the voltage down and overheat the cap.) These are electrolytic: they leak a little but are cheap.

The 0.1uFd caps drive the tube grids, very high impedance, where a microAmp of leakage is a problem. These are oiled-paper (these days: plastic) insulation. Low leakage, but expensive and large in large sizes.

47uFd 400V in modern plastic is about 20 cubic inches and $70; in electrolytic is 1 cubic inch and $2.31. The electro will leak about 1.3 milliAmps per cap; we can tolerate 10*1.3mA= 13mA leakage in this 50mA power supply (and that 1.3mA/cap is a max spec: ten caps will probably leak less than 13mA unless they ALL barely passed inspection).

I suspect most of your 0.1uFd caps are good, but there may be failures. As long as you have a box coming, get a 10-pack of E4104-ND Panasonic E-series poly caps, 0.1uFd 400V, $4.98/10. (If you are cheap and just want to get one amp going, worst-case you need 2 at $0.58 each.)

You have some odd value caps in the tone control. You may not have need of just one tone control; anyway these caps are very likely to be good, since they probably live with zero voltage and are good solid oiled-paper caps.

> the aluminum can capacitors, I am subsituting these multi-unit capciators with 2 individual capacitors, and wiring them underneath?

Yes. This is an ugly hack. If you could find these caps today, -with- those insulated mounts, you'd get another can-cap. Shop around: guitar amp specialty houses sometimes have a limited selection. A 40-40-40/450V may even be available and suited for the power supply. Is that cap insulated from the chassis or not? If not, you might find an exact replacement. But if it is insulated from the chassis, then you need to buy a new insulator, or very carefully remove the old cap without breaking the fragile insulator. (Or use a fret-saw and file to make a new insulator from un-clad circuit board stuff.)

The 20/350+25/20 in the volt-amp is an odd combination. Trying a 20-20/350V will work for a few months and then the 350V unit working at 1 or 2 V will fail. And a 25u/25V or even a 220u/25V of today is SO small that you can just tack it across the 1K cathode resistor.

And even the larget caps in modern electrolytics are SO much smaller that you can hide them all under the chassis, avoid all hassle with obsolete can-caps on fragile insulators.

You will have to disconnect the center lugs on the can-caps (you can leave the shell connections). And they probably used those lugs for support. In the voltage-amp, you have three resistors to the 350V cap lug: get skinny cutters in there, cut the lug under the resistors, and lift the resistors-junction up a few millimeters so it can't touch. I bet the stiffness of three short resistor leads will be ample support. Add a new-cap from that junction to the can-cap case.

A really industrious radio-restorer would take the can-cap out, yank the insides out, fabricate a new bottom lug-plate, and hide the new caps inside the old shell. Unless you want a museum-quality restoration, that's insane.

 

[CJ]

http://www.tubesandmore.com/

:guinness:

 

[NewYorkDave]

PRR,

I don't know what flavor of SPICE you use, but here's a link to 12AY7 and 12AV7 models for 3F4 SPICE. I haven't used the AY7 model, but the AV7 model from the same author is excellent.

http://www.groupdiy.com/index.php?topic=3792

 

[Mailliw]

Hi PPR,

Thanks for clearing that up. A quick read about capacitors at Phil's Old Radio page helped too

Thanks,
William

 

[PRR]

> 12AY7 and 12AV7 models for 3F4 SPICE

Thanks, but "too much trouble" for Mailliw's situation. I figure he should get some power up and try 'em. If we wanted exact answers, there are loads of precision mike preamps around; the charm of simple tube amps is their imprecision.

I just got curious where they were going with that feedback network. It looked like it reduced input impedance, damped the tranny, and limited gain and gain variation. And a mental estimate suggested Zin about 1/2 of 100K, simple matching. But I didn't realize that the bass-corner was bumped-up. A little more with 12AT7 than with 12AU7, so with 12AY7 it will boost 1 or 2 dB around 20Hz, and there is surely 1 or 2dB loss at 20Hz elsewhere.

For that matter, you could probably swap-in 12AU7 to 12AX7 and get various gains, bass-bumps, and overload points. The gain won't vary a LOT because of feedback on the first stage and heavy loading on the second stage, but it will give different flavors.

 

[Mailliw]

I re-capped the powersupply, 6 caps in total. Took two nights, I went slowly.

Now the B+, B- terminals give a nice 215 +/- a few volts !!! WOOO

Tomorrow night I will re-cap a voltage amplifer and then test it out.

What is the best way to test one of the voltage amplifiers, could it damage a mixer?

Thanks!!

 

[PRR]

How odd. I forgot all about this for a month, until this morning, thought of it again tonight, and just found your note as I'm going to bed.

> could it damage a mixer?

Most mike-amps: no. But I have a bad feeling on this one. Let me sleep on it until noon.

In fact: while you are re-capping, confirm if there is a cap that goes from one of the tube-pins to the back connector. There "should" be (and I'll look at the pix from work), but I wanna be sure RCA didn't do something cheap.

Do you think the harness is complete enough to power-up one volt-amp? Can you figure out how to connect something to the signal output? To the mike input?

 

[Mailliw]

[quote author="PRR"]
Connector marked:

[/quote]

confirm if there is a cap that goes from one of the tube-pins to the back connector

Is that .1 MFD cap coming from the 'out' pin what you mean?

Can you figure out how to connect something to the signal output? To the mike input?

I clearly see where to put B+, B-, 6V (x2), GND, but as for input, output...

Input: If I use a mono, two-wire connector from the mic, does one wire goto pin IN- and one to IN+?

Output: Which two pins, OUT and GND?

 

[PRR]

I'm assuming you ultimately want a standard 3-pin XLR input. And if you use these stand-alone, you may want to put the XLR on the module or on a length of wire for easy connection to modern mikes.

If your test-mike is 2-wire, be sure it is a low impedance mike. These are most commonly supplied with 3-wire connections, but if yours is 2-wire then strap XLR pin 1 to one of the other pins (I think pin 2 is conventional but opinions vary, it works either way).

The output is much higher impedance than many modern mixers' line inputs. It will work, but gain may be low and bass may be weak. Try it. Also try feeding the Tape or Aux input on an old hi-fi amp and speaker: that may be a more happy impedance and a better representation of how it sounds.

 

[Mailliw]

Ok, I re-capped a volt-amp and wired it up, with an xlr input and a 1/4" output.

(btw I didn't change any of the .1MFD molded-plastic caps, just the 2-cap can)

For my test:

input: shure 544 dynamic mic
output: amp (AUX) to speakers

It worked but not well. There was MASSIVE hum, twice as loud as the mic itself. Also when plugging the output 1/4" into an adapter (which converted it to RCA plugs for the amp), the microphone made a very high pitched sound as the tip of the 1/4" touched the sleeve of the adapter. Some sort of feedback loop?

Any ideas?

Thanks!

 

[Mailliw]

The hum is coming from the power-supply, somewhere the chassis ground and the signal ground (B-) are shorted, so I am going to try and fix that now.

The voltage that was appearing at the mic input is gone after I rewired the connector.

I'll update soon.

 

[Mailliw]

Hi,

The hum goes away if I short B- and chassis GND. Hardly any noise at all with my speaker+amplifier test setup.

I thought it was the power supply, because a bad insulator was shorting B- and GND when I wiggled the big can-cap, and the hum would disappear. Took me a while to figure that out!

But... is this normal? Do I want B- and GND to be one?

Later

 

[PRR]

> Do I want B- and GND to be one?

Yes. But.

In a large system, they should come together at ONE point. The case grounds intercept buzz and crap and divert them to the system rack. The signal-common and power-common (same thing in this system) should reference the same point, but must not carry crap intercepted by the cases. The link between chassis and power/signal common may be in that harness, or may be in the main chassis connectors that seem to have been cut-off and lost to history.

For testing, put a strap in the power supply from chassis-ground to B-. If you rack all six preamps together, you might bring all output signal-commons to a metal panel, and strap that to the power supply chassis ground and all the modules' chassis-grounds. If you just rack a couple with a new power supply, just tie the chassis ground and power/signal-common together in the power supply.

> plugging the output 1/4" in..., the microphone made a very high pitched sound as the tip of the 1/4" touched the sleeve of the adapter.

Ouch. Has that been fixed with B- and GND connected together?

It isn't clear just what the feedback path is. But with just the tip of the output touching a semi-grounded point, and the signal ground floating, something fed-back and oscillated. And powerfully enough that oscillations passed through the input transformer into the mike, which worked as a speaker. It seems unlikely that it made enough power to damage a dynamic mike (they can typically take many milliWatts of power as "speakers"), and it would not harm a condenser, but I would worry about ribbon mikes. I hope it was just floating-grounds.

BTW: if the modules are open-bottom as shown, they surely sat on a metal shelf to shield the bottom. You might need to slip a cookie-sheet under for testing to get lowest noise/buzz. This sheet should contact chassis ground: just setting the module on the sheet might be enough, but for best results it should have a solid bond-wire to chassis ground.

 

[Mailliw]

shure unidyne --> RCA preamp --> m-audio audiophile 2496

gone

 

[JustinS]

Hey, nice sounds!!

Justin

 

[Mailliw]

Hey PRR,

I have a lot of free time now, so I want to work on this some more...

The output is much higher impedance than many modern mixers' line inputs. It will work, but gain may be low and bass may be weak. Try it.

You're right, I have to turn the gain up a lot (near full). I'm interested to hear how it sounds better matched for my solid-state mixer. From what I've read, this is done with a transformer?


Thanks

 

[PRR]

> I have to turn the gain up a lot (near full).

Turn it up. To full if need be. In tubes, gain is expensive: they rarely gave you more than you need, you can't waste any gain in turned-down pots.

> how it sounds better matched for my solid-state mixer. From what I've read, this is done with a transformer?

If your inputs are 20K, this is already "matched". If they are 10K, a 20K:10K transformer "matches" and gives about 3dB more level. That's not much, not worth the hassle/cost/bulk of a transformer. And it isn't easy to build a good 20K transformer.

As I think I said a page or two back: this lightweight module doesn't have the oomph to drive high levels or long lines. It was made to drive a high-impedance summing network a few inches away.

How can we suck more output from it? Hmmmmm.... 12AY7 is a good tube but not made for "oomph" and not one I have worked with much.

If you roughly double the current in the second stage it plots like this:

That's 35V p-p in 10K load, gain is about 10 or so, which is not bad for a teeny low-noise triode. (OTOH, it may not be a lot better than the original values.)

Parts changes to arrive at this point:

You might try a 50uFd 6V cap across the cathode resistor: it increases gain but also distortion.

 

[Mailliw]

Thanks for the great information. Since I won't be using all 6 amplifiers at once, I will modify at least one this way for variety.

Two other questions, if you don't mind:

I've recapped 3 amplifiers now, and they all have different gains. I recorded some music coming out of headphones and noted the peak level during a part of a song.

#1: -13 dB
#2: -17 dB
#2: -29 dB

What sort of variation would they have considered healthy back in the day in its original application?

I kept the same tube and input transformer going throughout the test (wasn't sure if the input transformer would have an effect on gain, but did it anyway since its socketed and easy to swap).

Last question:

I snapped a resistor in one of the amplifiers. If I were to replace it, or if I was going to modify an amplifier it like you suggested in the previous post, would I go after similar parts (carbon resistors?) or simply get the cheapest modern equivalent?

Thanks PRR!