[dconant]

Sorry Mike no commission. I'll likely be buying the 115 as soon as I can convince myself.

[dconant]

Can someone tell me if my 42 tubes are connected triode or pentode in fixed or cathode bias? How could I tell by looking at the schematic.

My somewhat educated guess is push pull cathode biased in pentode mode.

Any idea what the output watts would be?

[RodB]

Yes, you can look at the schematic. When the output tube screen grid is connected directly to the plate then it's wired as a triode. Cathode bias is developed by a cathode resistor of around 150-220 ohms usually bypassed with an electrolytic capacitor. Fixed bias is developed in the power supply usually with a voltage divider resistor connected to the center tap of the power transformer. Output power is found either by ohms law or looking at the tube manual for the output tube. The specs will list 2 or 3 configurations and give the output power for a specific configuration. In push-pull you double the value. For instance if it gives a value of 4.2 watts that would be 8.4 into the opt and 8 to the speaker. Transformers are about 85% efficient.

[dconant]

Thanks Rod. Since the screen grid is not tied to the plate then my tubes are pentode right? I see the candohm ties to the cathodes and the resistance will be 300 ohms from point 5 to 3 but I don't find the cap. Are the tubes cathode biased? Can you tell me what the Effective Load Resistance and power output is for my 42 tubes from the specs here https://frank.pocnet.net/sheets/093/4/42.pdf

I am trying to determine what output transformer to replace my bad one. I have a Hammond 125C but I am afraid the wattage is too small. It is an 8 watt transformer. I have a speaker that measures 5 ohms with a vtvm. Is that an 8 ohm speaker or 6 ohm speaker and can I use that transformer? I'm thinking I can use pins 4 & 6? I have tried it and it seems ok to me but I don't hear that well so it may not sound the best. Is it hurting the tubes?

[MrFixr55]

Hi DConnat, 

Yes, the 42s are in Pentode Mode, cathode bias.  I am using the schematic from the Philco Library, Link below:

Early Philco Schematics – Philco Library

A few things:

Switching or varying the B voltage to the screen grids was a relatively common method for controlling the sensitivity of a stage.  One use of a "local-distance" switch was to do just that, vary the screen voltage, thereby varying the gain.

If that cap pictured is actually #47, definitely get rid of it and replace it with a modern 0.001 uF cap.  The value of 0.001 uF is not unreasonable, but like Morzh says, I don't know how they developed it with that open metal frame job.  In old RCA sets like the Radiolas, open caps were good for about 0.00025 uF at the most.  Regardless, that one may have shorted, causing catastrophic failures.

A few warnings here:

When a circuit with power pentodes is fired up with the output transformer missing or open, the screen grids draw much current, often more than they can handle.  They will glow to an almost incandescent white, then burn up, ruining the tube.  In addition, the fried screen grid can short to the control grid and /or cathode, wreaking more havoc.  I hope this did not happen to you, especially if you have the globe style 42s (waaay cool tube, first "modern" pentode with indirect heated cathode, yet in a globe envelope).

Please note that on this radio, like many other older Philcos with a 3200 Ohm field coil, not all of the B+ current passes through the field coil.  the output stage, which comprises the bulk of needed B+ current comes right off the power supply, before the field coil.

Looking at the Candohm:  Note that Pin 3 is grounded, Pin 4 connects to the center tap of the HV Supply, Pin 5 connects to the cathodes of the 42s.  The 205 Ohms between the cathodes of the 42s and the HV return provide cathode bias.  The 95 Ohms between the HV return and chassis ground provides grid bias for the 42s through the interstage transformer center tap and provides grid bias for any other tube whose cathode is at chassis ground potential.  Any open here will greatly limit output power or cause total non-operation, but no collateral damage.

The other connections, B+ to the Field Coil to pin 1 of the Candohm (3200 Ohm),1 and 2 (900 Ohm, ad 2 of the Candohm to 3 of the Candohm at chassis ground (2700 Ohm) provide a "voltage divider" to provide a relatively predictable level of lower B+ voltage to the plates of the 1st Audio stage, the detector Amp (through another voltage divider) and to the plates of the IF, Osc / Mixer and RF tubes, from the junction between the field and Pin 1 of the Candohm; and to the screens of the IF, Osc /  Mixer and RF from pin 2 of the Candohm.  The majority of current drawn through the Candohm is from the speaker field. 

Please note that the dotted resistors and speaker are if this is a "221" version, employing 2 speakers.  The drawing can be most easily misinterpreted.

I don't think that the dotted line between pin 3 of the dotted Candohm, #46A connects to the line that it crosses.  I also do not think that it connects to resistor 41A as drawn.  I believe that Pin 3 of Resistor 46A is grounded.  The 2 field coils are in series from B+ to ground, forming the main voltage divider, therefore Resistor 41A is merely a dropping resistor.

Any opening of the string from Pin 1 to Pin 3 of the Candohm will break the voltage divider, causing insufficient current to the field coil and much higher voltages throughout the set, causing distortion.

I would have to think that most of these failures occurred from the radio being 93 years old.  It appears that open output transformers and open field coils are common.  

If you have the single speaker model or are replacing both speakers of a 2 speaker model with PM speakers, definitely use the "121" option on the schematic for the Candohm.   In addition, it may be necessary to increase the capacity of some of the filter caps to reduce the hum introduced by the loss of the choke capacity of the field coil and loss of the hum bucking coil.

Keep us posted on your progress, please feel free to ask more questions, and good luck with your restoration!

[RodB]

Your 8 watt transformer is good, the 42s produce 6 watts and you won't be sending maximum volume to the speaker anyway. The 42 output impedance is 7000 so you can use the 6 ohm tap to the speaker, 2 & 5. The 42 cathodes are biased with the candohm but no bypass capacitor which will reduce the power output slightly.

[dconant]

Thank you Rod. 

Hello MrFixr55. Thank you for the long reply. My radio is the 121 version (1 speaker) and I am aware to use the 121 candohm (#46). Pin 3 of the candohm is connected to ground. The 2700 section was open so I replaced with a 10 watt 2700 wirewound. Alll the tubes in the radio are the globe tubes and they all test good. I just replaced #47 with a 630 volt .001.

I used another 2700 10 watt wirewound for the field coil replacement. Voltages on the tubes seem good to me. I have used 10MF for the filter caps. The 42 tubes are 200 volts on the plates and screen grids. Kind of low using a VTVM. B+ off the rectifier is 240 volts even with a different tube. Rectifier plates are 302 volts. Both 44 plates are 200 volts, screen grids are 75. Same for the 36 tube. Plate of the 37 (audio) is 100 volts. Plate of the 37 (Det Amp) is 50 volts. The other 37 plate is grounded. The 37(audio) is the only tube not low. Is this because I have 2700 ohms instead of 3200 ohm field coil replacement?

I'm not sure which schematic you are looking at since there are 2 for the 91 but the second one shows a switch #7, which my radio does not have. But there is a second switch on same post as the power switch that connects cathodes of both 44 tubes. There is also a 160K resistor between the two cathodes. When the switch is on it is a straight shot between the tubes. When off the resistor is between the cathodes. The effects the volume of the radio greatly. This does not show on any schematic I have seen.

I have included some chassis shots. Still need to derust the tube shields. I have them all.

Dan

[MrFixr55]

Really nice workmanship!!! Top of the chassis looks great.

Sorry about the long winded explanation, a bad habit of mine.  I was technical team leader on my job and was responsible for training people as well as dealing with the stuff others could not fix.

300VAC on the 80 plates sounds low, but if the radio works and the transformer is not hearting up, it may be OK.  42s are rated for 250V on the plates, and considering that the other tubes hooked by transformer to the same B+ Point, I think that you may be OK.  Other voltages sound OK.

I looked at the Schematic from the Philco Library and the following N/A one:
Philco Radio & Television Corp.: 91

Of the 3 or 4 N/A schematics, at least this one has voltages listed.  The 250V supply is about 25 or so percent low   Yes, the 2700 Ohm resistor instead of 3200 Ohm will put a bigger load on the power supply but not by much.  In these things, 20% is acceptable.  The plate (to center tap?) voltage of 300 V vs the listed 310 V is only about 3% off.

Sorry, I have been calling the wirewound unit a Candohm.  It is not.  The chokes look to be correct unless glommed a long time ago from another Philco.

Yes, changing the cathode bias of a tetrode definitely changes the gain of the tube.  In fact, this is done on several TRF sets and sets using Biased detectors.

Gotta run to Choir practice, keep us posted.

Run the radio, for a while, unplug and feel test the transformer and choke.  Unlike older sets like a 20, the 80 and 42s will get hot so be careful.

[dconant]

Thank you for the compliment. No problem with the long reply, I enjoyed reading it. I'll get a 3200 ohm resistor on my next order and see what that does. I still have cabinet work to do.

[morzh]

Dan,

beautiful job. It's a looker. It almost deserves not to be put in the cabinet and to be displayed by itself. With framed speaker next to it. On a shelf. Seriously. It will be more of a conversation piece than a 91 when fully done.
I am not sure the Mershons are authentic to this model, the clamps hinting the other type, but considering this was very soon after the 90, which could have them, it is justifiable. Plus, again, they look good.


I might've missed it but has it been receiving yet?

[dconant]

Thanks Mike, Yes it is a shame to put the chassis in a cabinet. The radio is playing quite nicely now. It sounds pretty good too. For the Mershons the part number in the schematic parts listing is the same as on the caps. Makes sense you might question them as they might have been secured to the chassis using the screw base.

[dconant]

I just saw a picture of a 90 chassis on eBay with the Mershons and see the mounting is as I would expect mine to be. Makes you wonder.

[dconant]

Been running the radio for three hours now. The power transformer has settled out at 140 degrees. The filter choke is about 115 degrees. I have an adjustable 20 watt resistor that I can set to 3200 ohms. I'll use that for the field coil resistor and see what the temps are after the set cools down. I suppose those 80 and 42 tubes are contributing to the power transformer heat since they are almost touching it.

[morzh]

What's the full value of that resistor?

[dconant]

20K 25 watt. I thought it was a 20 watt at first.