[BrendaAnnD]

I'm down to trying to restore my Majestic 20 chassis. I spent two days piecing the cabinet back together (it came as a 3-D jigsaw puzzle, thanks USPS). Fortunately, the speaker is in near mint condition.

I got a new power transformer for the set, a perfect fit, and have finally tracked down all the parts that are inside those lovely "black boxes".

Most of the repairs look like they will be pretty easy, but I do have a question that I hope someone here can answer for me:

What can I use to replace the "power detector"/audio driver transformer? I have a couple transformers that work in older TRF sets, but this set doesn't use a first audio amp: it drives the output tubes directly from the 27 detector. The plate voltage on this tube runs to 255 volts, almost 5 times the normal plate voltage for a detector tube, and its resultant higher plate current.

Does anyone have any suggestions re this? Or is there a chance that one of the old TRF driver transformers will work?

Thanks in advance.

[Arran]

Without looking at the schematic it sounds similar to what many table sets and small consoles used in the early 1930s, they took the output from the second detector tube. I found the schematic on Nostalgia air, and it shows that there is a 25 K resistor between the full B+ line and the primary of that interstage transformer, so I think maybe the voltage spec was in error.
Assuming that the rest of what they quote for the second detector is true then the plate draws a current of 8 ma, so if there is a 25K resistor in there then it would have to drop 200 volts, therefore the plate voltage for that 27 should be 75 volts not 255 volts. See if there actually is a 25 k resistor in there and not a 2500 ohm one or something, either the 255 volt measurement is wrong, the current rating is wrong, or the value given for that resistor is wrong. As my math teacher used to say, "Ask yourself the question, does this answer make sense?".
Regards
Arran

[BrendaAnnD]

I have a copy of the Majestic schematic and setup. The tube is supposed to have 255V plate and draw 0.8 [not 8] mA, according to their paperwork.

[Mondial]

I would try the physically largest interstage transformer you have with a center tapped secondary. The quiescent plate current of the detector is very low since it is biased near cutoff, so this should not be a problem with any usual interstage transformer. If a problem were to occur, it would happen when the detector is being strongly driven, as when playing the radio loudly. A very strong signal into the detector might cause the plate current to rise to the point where you might saturate the magnetic core, but I doubt you would notice any distortion at normal listening levels.

The usual interstage/driver transformer ( for example 27 audio stage driving PP 45's) should work fine, since this is a more difficult situation compared to the detector driving the transformer directly. With the audio amp driver, the normal primary plate current would be about 5 ma, a level the detector driver would reach only at very high signal input levels

[Arran]

On the Nostalgia Air diagram it shows a decimal but it's after the 8 and not before it, so it reads as 8 ma or .008 amps, but it does show the 255 volts on the second detector's plate.

http://www.nostalgiaair.org/PagesByModel...039984.pdf

The N.A schematic looks like it came out of Riders, either that or a Beitmans book, so one of these schematics has a printing error. Does the set still have a 25,000 ohm resistor between the B+ line and the primary of the interstage transformer? If this is a weird interstage transformer it may be worth looking into the solid state substitute that they had published in the AWA Old Timer's bullitin.
Regards
Arran

[BrendaAnnD]

Yes, the 25K resistor is there (it's the only part still physically attached to the detector/power filter box.. ).

I have added the schematic. As this packet of information is directly from Grigsby-Grunow, I am tending toward trusting their voltage numbers. That detector tube is biased at -21V, so almost cut off (practically running class C?), so the plate current will be quite low.

I found a Stancor A83 at PTOP, which should work pretty well in this radio (and am I ever glad.. it wasn't looking too good for a while there..)

[BrendaAnnD]

I finally got in the parts for this radio (except the 45's). It's performing OK, as far as receiving stations, but the output is still pretty low. Distortion is low, up to the point on the volume control where the gain starts overloading the front end.

The only thing I can really see is that the +275 source is heavily loaded. It runs around 150-170 volts with the volume down and the tubes at cutoff. The only thing that brings it up is disconnecting the center tap on the output transformer, at which point it goes up to almost the full voltage from the rectifier. I've carefully checked my wiring (there were two wires switched between the 275 source and the main B+ source, but switching them to the correct connections helped very little.) and everything is OK along those lines now. The bias across the 700 ohm bias resistor is quite high, about 90 volts (which SHOULD cut the 45's off completely and allow the B+ on the plates to come UP!)

I'm open to any suggestions of what I might try.. any chance the output tube sockets are leaking?

Thanks in advance.

[codefox1]

Good chance you have a carbon trace on one or both sockets if everything else checks out OK. Happens quite a bit on the old sets, and difficult to find. Replace 'em before you fire up the set again. If the 45's are good, would be a shame to lose them.

I'm sure someone has some to contribute or barter, try the classiied section, need to know existing dimensions too.

[Mondial]

Do you measure any positive voltage on the grids of the 45's? Since the center tap of the driver transformer secondary is grounded, there should be very little DC voltage at the grids unless the 45's are gassy or the sockets have leakage.

If you remove the 45's, can you measure any DC on the grid contacts of the socket? If there is no DC, this would tend to eliminate leakage from the sockets.

My guess is that one or both of the 45's are bad and drawing excessive current. Normal bias should be 45 V, so if you are measuring 90 V the tubes are drawing twice their normal current. As you said, with 90 V bias, the tubes should be cutoff, and drawing very little current, unless the center tap of the driver transformer secondary is not grounded. If the center tap is floating, the grids could assume a positive voltage and cause excessive current regardless of the bias developed across the filament return resistor.

[Arran]

What does this set have for power resistors? Are they chassis mounted candohm type resistors with multiple taps? I was thinking that it could be causing some problems, shorted or open sections. Even the ceramic coated ones mounted on a metal rod were trouble, and G.G liked using those. One bad habit some old time repairmen had was to parallel open sections of those big resistors with replacements, of course if they decided to fix themselves you could end up with a something like a 150 ohm section becoming 75 ohms.
Regards
Arran

[BrendaAnnD]

The set uses a huge (about an inch diameter) wirewound resistor that mounts to the chassis by two metal tongues that fit into each end of the resistor. All sections of the resistor check fine.

[Arran]

Is there more then one of these larger wire wound resistors? Like one for a B+ divider and another for the bias?
Regards
Arran

[BrendaAnnD]

It's all in one big resistor. If you look at the schematic above, you'll see it's of two sections, basically bisected by the volume control.

[BrendaAnnD]

I've now finished everything on this radio, except for clearing up the problem in the audio power output section. This includes the two 0.1uf caps in the IF cans. Replacing those did nothing for the voltages, but did seem to help the IF slopes. Once doing an IF alignment, the sensitivity is actually quite good.

Now for the output issue:

This evening I checked this whole section thoroughly. The input to the grids from the driver transformer is quite clean until the volume is turned up far enough to overload the input (RF) signal. Unfortunately, that's where the normality stops. Going to the output tubes themselves, I find that there is far too much current being drawn. Inserting a DMM into the center tap connection of the output transformer, I get 120mA of current draw, where the total according to the service literature should be 56mA. This draws the voltage down to ~110V on each plate, where it should be 275. The cathode to ground is ~91V, where it should be 45. No grid to ground voltage, other than the AC from the signal. It seems to me that there should be some tiny DC voltage there, too, given the DC resistance of the driver transformer. But perhaps not.

Experimentation shows that if I put a series resistance in the center tap to the output transformer, it lowers the voltage to the plates, but the voltages to all other points on the chassis did come up to near normal.

I've tested DC resistance (leakage) on the output tube sockets, and find no significant leakage (actually, immeasurable, as my DMM registers only to 20MOhm, and I get no reading (OL) on the meter from grid to cathode, grid to plate, or plate to cathode. (this with removing the wires from the grids to the driver transformer and the plates to the output transformer). Despite this, I am getting some ceramic sockets to replace the two phenolic sockets.

Also have some 45's on their way, hopefully these will help. (Just for giggles, I put in a pair of matched 2A3's, it made little or no difference.. too much plate current I'm sure).

[Arran]

I don't think that the sockets are the problem, unless the leakage is from the plate pin of one of the #45s to ground in some way. I would try a substitute bias resistor in place of the one in the big voltage divider, maybe its a problem that only shows up under load? Since the bias section goes from the center tap of the #45 filament winding to ground it's easy to try. What about the output transformer itself? Shorted primary? Leakage from primary to the core? It could also be that the set doesn't like the substitute audio interstage transformer, normally they are a 1:3 ratio but back then the factory could have had one custom wound.
Regards
Arran