an old amplifier with a strange schematic

[birt]

you might remember there was a topic about an old military amp i have. i finally succeeded in tracing the schematic (i left out the 12V part with the vibrator).
but what i found is kind of a mystery to me.

-there's a very basic input stage
-a rotary switch as a simple yet very effective tone control
-another triode gainstage

except for the floating cap on the first cathode i get all this. it's at that weird transformer where it gets blurry for me. it is completely covered and only has those wires coming out.
it gets signal from the anode as well as the cathode but only feeds one 5881. and then there's a gigantic choke (the OT is as big as an 100W and the choke is just a little bit smaller) doing something i don't get...

i'm trying to learn here but google doesn't give me any clues to what's going on in this thing.

[Shrapnel]

It looks as if it is using a transformer as the PI. It has been done before, but it is an expensive way to drive the PA stage. I'm not sure as to why the other leg of the transformer's primary is connected to the Cathode instead of to ground, and the other to the plate through a cap.

[Heinz]

It seems that there is a missing connection from the transformer to the grid of the second power tube. I don't think the amp would work with an open power tube grid.

Generally, using an interstage transformer has some advantages:

• higher output power. The driver stage can drive the grids positive and squeeze more power from the power tubes.
• no blocking distortion. A cap coupled power stage will experience blocking distortion when the grids are driven positive.
• increased reliability. The fragile PI tube can be dropped and replaced with a sturdy transformer.

It does make some sense in the military context, where cost is not a problem and power/reliability are the main objectives.

The huge choke could be some sort of frequency corrected feedback, but I'm not quite sure.

[PRR]

Schematic is clearly not right.

Choke grounded at both ends is useless.

The other 5881 grid must go somewhere.

The grid-driving transformer is typical for this age and application. Taking it back to the cathode is not wrong. However there's too much that "is wrong" to know what's really going on.

[birt]

i guess this schematic makes more sense.
but i still don't really understand why there are so many wires at that interstage transformer, why would you need more than 2 primary and 3 secondary?

[FYL]

Here's the schemo for a 50W model - 807-powered.

[birt]

that's a whole different animal it seems. what i have should be closer to the CP20 model.

[birt]

i replaced the filtercaps to be able to test it properly with 10uF/450V. then i had to turn it off quickly because they were getting 500V... i don't have any 500V caps. is there any way i can test it then? 2 caps in series in those 3 spots? then they are only 5uF...

i could bring down mains voltage with the variac but then i don't get full voltage on the heaters. or i set the PT for 245V instead of 225V. i wonder if that will be enough.

[Tiny_Daddy]

Hey Birt, it's time for some 500V caps! Doug sells them. Get the 22uf 500V Illinois for $3 and fifty cent!
I don't understand the "black wire to PT" connection though....

[Shrapnel]

I'm not sure of the black wire to transformer either, BUT I have a speculation as to the extra wires from the interstage transformer to the plates. I'm not sure what the resistances are, let alone impedances, but It looks as if it may be a negative feedback loop, per side.  As your schematic doesn't show any NFB circuitry anywhere else, this just might be the case.

[birt]

I don't understand the "black wire to PT" connection though....

me neither :p

as for the caps, i will probably get them here in europe. its faster and cheaper for me :s

[kagliostro]

"black wire to PT"

I think that can be a PT shield connection

Kagliostro

[PRR]

The CT on your HV is not hard-grounded as your interpretation shows. It is the "black wire to PT" which goes to the choke. The power supply is Choke-Input. Because the choke was not rated for >500V, it was put in the groundy side of the supply.

Its DC resistance adds to the 5881s' self-bias.

The extra driver transformer leads are taken directly from the 1938 paper.

http://pmillett.com/tubebooks/tubedata/beam_power_tubes.pdf

[birt]

i found some 20uf 550V caps in my drawer :D
as soon as the power tubes are in the amp the B+ drops a lot. more than 100V.

with a guitar on the mic input i get decent tone but with the volume more than halfway it suddenly gets very distorted and gated. it can't handle bigger input signals either. i should probably use the PU input to prevent overload. and maybe a beefier rectifier. right now i'm just happy all the main parts seem to work as they should because i won't keep it as is.. i'm thinking B15 at this moment. i'm considering to keep the transformer PI though.

[birt]

i have tried disconnecting the .1 caps at the power tubes and connecting the orange and green wires of the interstage transformer (so the 6K5's and choke were not connected anymore). both didn't result in a volume boost, only different hum levels. i will probably just use the tube PI for my B15 build.

the choke sees only 12V in this circuit. that seems like a really low voltage for such a big choke. would it be of any help somewhere in the B15 circuit or isn't it really usefull in any other circuit?

[sluckey]

the choke sees only 12V in this circuit. that seems like a really low voltage for such a big choke.

That's just the dc voltage drop across the choke due to coil dc resistance. All B+ current flows thru the choke and that's the important thing.

[birt]

i don't see how all B+ current could flow trough the choke the way it is used in the circuit?

edit: i think i'm starting to get it now... the CT of the high voltage is normally connected to ground as the ground reference for the B+. in this amp it's connected to ground through that big choke so the choke. so i disconnected the choke from ground and the B+ dropped to about 50Vdc (measured to ground) because it's got no reference point anymore. the 30K resistor then starts boiling because all current the amp wants goes trough that resistor without the choke.

but why is the first 8uf connected to that ct instead of ground? and why are the 6K5 resistors not connected to ground as the drawing PRR posted suggests? the purpose of that 30K-12uF is still a mystery to me

[sluckey]

Think of the PT and 5U4 as a battery. The cathode of the 5U4 is the positive terminal. The center tap of the HT winding is the negative terminal. All that current thats flowing in the B+ rail has to come from somewhere. That would be the center tap. You have the choke in series with the center tap, therefore all that current flows thru the choke.

Putting the choke (or a resistor) in the center tap lead is a cheap and useful way to develop a negative voltage that can be used for grid bias.

[birt]

i measured the CT dc voltage in reference to ground and its at -53V.

so i can use this for a fixed bias right?

but i was thinking if i want more voltage across the power tubes i could use this -53V as a reference point for the cathode bias instead of ground. if the voltage across the powertubes is for example 400V and i connect the cathoderesistor to -53V the i get 453V across the powertubes right? and a big resistor in series with the choke lets me set the anode voltage anywhere i like. or am i missing something here?

[HotBluePlates]

... if the voltage across the powertubes is for example 400V and i connect the cathoderesistor to -53V the i get 453V across the powertubes right?

Yes, as long as you really meant that there is 400v plate to cathode, and that when you measure from plate to ground, you get 400v plus whatever voltage is across the cathode resistor.

Regardless, moving the cathode resistor from 0v to -53v will put 53v more across the tube, less whatever voltage the cathode resistor drops due to extra current.

[PRR]

> i can use this for a fixed bias right?

No, it is self-bias.

The total drop is 53V at normal current and zero at zero current. And "current" is 99% power-tube current. It "is" a cathode-resistor, just wired funny, and also doing some filtering.

> with a guitar on the mic input i get decent tone but with the volume more than halfway it suddenly gets very distorted and gated. it can't handle bigger input signals either.

So don't turn-up more than halfway.

Actually: that mic in is probably scaled for maximum 5mV-20mV mike level. Guitar levels run 20mV-200+mV. It really isn't made to take guitar. Try the grammyfone input.