• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

RCA MI-112182 Monoblock Tube Amps

It's quite a departure from a Williamson in many respects - the THD may be somewhat more even with the local feedback from driver to output stage - definitely worth setting the AC balance for lowest listening level distortion.

The 100kHz RC filters (470R and 3n3) may have suppressed some OPT related resonance, but there is no global feedback. That would likely need some HF squarewave response assessment.

It's good to see a diode buffer after the reg tubes on the 300V screen supply to allow 40uF decoupling.
 
On both of my MI-12182's I added 0.57uf of 1kv capacitance before choke and RC filter, connected to ground, to get rid of the mechanical hum emitted by the choke. I had to experiment with the value to minimize the hum while minimizing the B+ voltage increase.
upload_2023-5-11_15-12-51.png
 
Last edited:
That's a good move imho, as long as the 1kV cap rating is for full application of 1kVac. With good layout, that added cap can minimise the loop area of higher frequency rectification artefacts. Did you consider adding say three series 1N4007 in series with each of the 5R4 anodes, as a form of 5R4 arcing protection? It may be worthwhile doing a PSUD2 estimation of peak diode current with default filter caps after the choke, and with added cap scenario.
 
Interestingly, the schematic I have indicates a S5019 FW rectifier, which is a solid state 5R4 replacement. Indeed, S5019 is silkscreened on the amplifier chassis itself. However, the schematic also calls for a B+ delay relay, which my amplifier is not fitted with.

By a crazy coincidence I actually have a pair of original metal S5019s that came with a huge NOS parts lot I purchased. They are big -- about as tall as an actual 5R4 -- and quite hefty. I'll definitely try the amplifier with both.
 
Doing some more digging, some datasheets give this diode a really high Vf of 12 volts. I wasn't even aware conventional silicon diodes could have such a high Vf. Anyways I'm guessing they're thermally bonded to the case and that's why it's so large. Wish they were junk and I could cut one open.
 
What really amazes me is that RCA built anything with a 6550 tube! That is tantamount to admitting defeat to their vaunted 7027A tube (and of course, in the end it was defeated). RCA basically had to be nearly bullied into including the 6550 in their Receiving Tube Manual, and finally relented with (to the best of my knowledge) it first appearing in their 1971 RC-28 manual -- and this for an unbelievably popular tube since it first appeared in 1955. The competitive smugness between the tube manufacturers of that day was just incredible. They did the same thing with the 7591, which they copied to produce their 7868.........

This is not to say or imply, or even suggest anything I might personally have against RCA or their products. I have nothing against them at all. But the General -- being who he was -- was competitive virtually to a fault, so he would only relent if he was forced to if there was no other way out.

Very neat amps -- with their part making for a great chapter in the history of audio amplifiers.

Dave
 
So, I should make a disclaimer here: I'm not convinced that RCA actually made these amplifiers. First of all, the schematic I have includes lots of scrawled handwritten notes and revisions, which doesn't seem characteristic of anything approved by RCA brass. Even some portions of the actual circuit are hand-drawn. Also, the circular potted transformer is UTC and the chassis cut was clearly made in an amateurish fashion after the chassis was produced. Looks like someone used a hacksaw or something. Of course, the main factor that points to these being genuine RCA amplifiers it the silkscreened lettering and the fact an RCA product number, "MI-112182," is printed on the front. However, there's no RCA logo anywhere on the amplifier nor is there one of those paper RCA labels you often see on their theater amplifiers.

I wonder if it's possible that some third-party company bought unfinished MI-12182 chassis with power transformers and converted them in-house to run 6550s with a UTC output. It's possible the silkscreening was done as part of this conversion and perhaps they came up with "MI-112182" as an internal designation. It doesn't seem likely RCA would have been happy with another company hijacking their product naming scheme, but perhaps so few of these amplifiers were produced that they never got wind of it (or simply didn't care). The more I work on these amplifiers and read the schematic, the less I believe they were an actual RCA product. Also, the amplifiers themselves include lots of ad-hoc wiring and there's a turret strip near the input section that has nothing on it. More evidence IMO that these amps were a low volume product that wasn't engineered to normal RCA standards.
 
Oh, and I forgot the most obvious thing: the (basically accurate) schematic I have doesn't say RCA anywhere. Instead, it says "The Providence Company" in the bottom right-hand corner. The address included is in a neighborhood that doesn't look like it was ever industrial or commercial.
 
That being said, RCA did use 6550s in ppp in its massive MI-9289C 200 watt monoblocks, so your comment stands!
 
AlexHempel said:
Doing some more digging, some datasheets give this diode a really high Vf of 12 volts. I wasn't even aware conventional silicon diodes could have such a high Vf. Anyways I'm guessing they're thermally bonded to the case and that's why it's so large. Wish they were junk and I could cut one open.
Click to expand...

probably a bunch of standard diodes in series to make the PIV. Silicon junctions are right around 0.6 volts, so a 12v Vf implies about 20 junctions in series.
 
gadget73 said:
probably a bunch of standard diodes in series to make the PIV. Silicon junctions are right around 0.6 volts, so a 12v Vf implies about 20 junctions in series.
Click to expand...

Hmm, but if that were the case, would that "part" have its own JEDEC designation and be available in a catalog? I did find the 1N1239 in some scanned vintage semiconductor catalogs as a standalone part. Perhaps it's standard diode junctions in series encased within a package to make a single part?
 
most likely lots of junctions in one package, basically the same construction as selenium rectifiers. Those only had a lot of forward drop because they had lots of junctions in order to get useful PIV ratings. You can find modern high voltage rectifiers with a Vf that is some multiple of 0.6 for the same reason. Modern silicon you can get 1kv out of a single junction, in 1960 that was probably not so much the case.
 
Been working on one of these for a while and finally got it sorted. Gonna run tests tomorrow. As you can see in the before pictures, tons of connections were badly spliced with thin, loosely fitting plastic, and I had to redo all of those.

I also re-configured the mains wiring, which was questionable (they put the fuse on neutral and the switch on hot...). Replaced the main B+ filter cap with a big KEMET DC link capacitor using a PCB I designed. The B+ supply is 600 volts, and I'd rather avoid using series electrolytics. The other huge axial film cap is output decoupling for the gas regulated screen supply -- I actually bought it on accident years ago and have been looking to use it in something.

I left in all the original signal path resistors, which is perhaps a controversial choice. However, this amp isn't hifi -- it doesn't have gnfb, and I have determined the output transformer is a UTC CVP-3, which is a universal PA transformer. I'd rather keep these amps as historical pieces sounding as close to original as possible, which includes the carbon comp resistors. I did measure every single one of them, and they are all within spec, so I got lucky in that respect. If they cause too much noise, I can always replace them later.

A few other major changes include removing the selenium rectifier (of course) and changing the 1/4 bias test jack configuration. They are shorting jacks and were originally configured to measure current directly -- when the plug is inserted, the shorting connection is broken and you measure the current with an ammeter. Of course, this means when there is no jack inserted the output tube current flows through the shorting connection, which I don't really trust. So I changed them to shunts, with hand-selected 10 ohm carbon comp resistors in parallel with the shorting connections. Break it and you measure voltage across the resistor. This way nothing should really happen if the shorting connection goes spotty, as there will still be a low impedance path for current flow.

You might notice the 100 ohm resistor hanging loose at the left hand side. That was part of the resistor cap combination in parallel with the choke (discussed earlier in the thread). I want to run tests with and without it to determine whether it's really necessary, as there's not really a great way to mount the capacitor.

IMG_7412 crop.JPG

IMG_7786_crop.JPG

IMG_7416.JPG

IMG_7789.JPG
 
Last edited:
The fuse on one side and switch on the other was common in the era of non-polarized cords. 50/50 which side either was actually going to be on anyway.

Much neater with modern smaller parts inside
 
Definitely, except this thing came in a rack-mountable chassis and was meant for permanent installation. Also it's got hot, neutral, and ground marked on the screw terminals (no line cord), so there's less of an excuse in this instance.
 
You must log in or register to reply here.
Back
Top Bottom