Kagliostro 12AU7 push pull schematic

[sluckey]

I would not consider this push/pull output circuit as a voltage amp. If voltage was the design target there are much simpler ways to get voltage gain without using four triodes. Usually any circuit that is designed as a line driver will have a low output impedance (600Ω is common for audio) and be able to produce enough "current" to drive a low input impedance line receiver at the other end of the line. This line may be a hundred feet long, or a thousand feet long, or it may even run for a mile or more across an airport to connect audio from radio receiver/transmitter site to the air traffic control tower. Low impedance is desirable because it is less susceptible to noise pickup and crosstalk. A balanced line also provides common mode rejection that further reduces noise pickup. Usually a shielded twisted pair cable is used to connect the driver to the receiver.

[jjasilli]

If this is true, then the schematic depicts a voltage amp by definition.

isn't the OT the "shift" from V to I?  in my veggie brain if I provide big enough stable V outta the PA tubes then the step-down feature of the OT provides the needed I at the speaker.
 I get that P is P on either side of the = sign, (minus the engineers precision math:)

Yes, but many tube pre-devices use a line-out OT, or cathode drive, and are not considered current amps.  One reason is to avoid the hi DC on the plate (which is sometimes done with a "cap driven" circuit off the plate).  What I understand PRR & sluckey to be saying is that this device has line-out value impedance, but much higher than line value signal voltage out.  High signal voltage output could be justified by the voltage drop caused by the resistance of a long line being driven by the device, especially given unbalanced output. However hi current seems, at first, to be counterproductive as it would increase the voltage drop across the resistance of the long line. . .


The "high" current output must be needed to overcome the current limiting aspect of the resistance of a long line; and the high voltage is needed to accommodate the resulting voltage drop across that resistance. This leads to the conclusion that the thing is a Buffer; more specifically a Line Driver.

[shooter]

I would not consider this push/pull output circuit as a voltage amp.

I make a leap of faith to guitar amps, n standard OT's.  Any low power AC that needs to go miles should be digitized, packaged, then dealt with at the other end

[sluckey]

Any low power AC that needs to go miles should be digitized, packaged, then dealt with at the other end

It is. Telco provides all the magic in between. May be copper, fiber, or microwave. You just run your lines to the demarc on each end and pay the bill each month.

My FAA airport example was typical back in the '70s. That same audio (and control) between the radio receiver/transmitter site and the control tower is all handled with fiber today.

[PRR]

> whatever device it's feeding is expecting a voltage input, not a current input.
> The fact that it puts out current is beside the point.

A telephone line needs POWER. Not just Voltage.

Assume cable is 30pFd per foot. 5 miles, 25,000 feet, is 750,000pFd. This looks like 600 Ohms at 353Hz! By 3KHz, it is 60 Ohms. This C-only analysis is incomplete; such a long wire has significant inductance and series resistance. In the original 600 ohm world, lines had less C. Driving long lines is a special field. But it needs POWER amplifiers; voltage alone with incidental current will just get sucked down.

Note also that "line level" is specified in dBm, actual Power units.

I don't see how replace the line out tranny with a "speaker friendly" tranny changes anything fundamental. (Does work better for speaker!) You determine the load impedance magnitude. If it varies (speakers, long lines), you generally mark the lowest impedance (the hi-Z parts will take care of themselves). Whether that is 4 Ohms, 100 Ohms, 600 Ohms, that's what you design for.

Sheeze. How can a simple line-amp stir so much controversy?

[Tony Bones]

Assume cable is 30pFd per foot. 5 miles, 25,000 feet, is 750,000pFd.

Forgive my ignorance... wouldn't a twisted pair behave as a transmission line (I wonder where that term came from) with some characteristic impedance independent of length?

Otherwise, I understand the requirement to deliver power...

[PRR]

> wouldn't a twisted pair behave as a transmission line ..with some characteristic impedance independent of length?

Short, no. Long, yes.

The real fun is that many telco-length lines are between short and long. Or rather they act short for bass and long for treble. This was semi-understood before telegraphy: Telegrapher's equation. Put a dit in a long line and it comes out duuuhhhhh.

People who think a 30 foot wire is long never have to face these problem.

[shooter]

 

You determine the load impedance magnitude

sometimes the forest wins

is all handled with fiber today.

I just pitched about 100' of un-terminated fiber from old MRI's. I've moved to wireless, no grinding n polishing!

[Tony Bones]

Assume cable is 30pFd per foot. 5 miles, 25,000 feet, is 750,000pFd.

> wouldn't a twisted pair behave as a transmission line ..with some characteristic impedance independent of length?

Short, no. Long, yes.

The real fun is that many telco-length lines are between short and long. Or rather they act short for bass and long for treble. This was semi-understood before telegraphy: Telegrapher's equation. Put a dit in a long line and it comes out duuuhhhhh.

People who think a 30 foot wire is long never have to face these problem.

I would have guessed that 5 miles qualified as 'long.' I learned more than one thing today.

[kagliostro]

@ PRR

pʿejúta wicʿaša

 


Franco

[drgonzonm]

Does it really matter if the output from the circuit in question is  voltage or current?  As PRR pointed out, we have a floating balanced signal, of which we take advantage. 

Back in the days of playing with PID instrumentation, we turned a 1 to 5 volt signal into 4-20ma by using a 250 ohm resistor or was it the other way around?  Distances sometimes exceeded 1/4 of a mile. 

30 foot wire? (shh, we don't want audio engineers telling us, that unbalanced cables longer than 6 ft are problematic)    Then again why do we install some resistors at tube sockets?   

I don't know what kind of voltages were put on the old long distance telephone lines, (The old 625 pair cables), but I remember the voltage on the home phone was about 5 volts, and up to 80 volts, if phone was ringing. 
 

[PRR]

> I would have guessed that 5 miles qualified as 'long.'

20KHz is 9.3 miles wavelength. We need multiple wavelengths before the impedance wobbles settle down to the "characteristic value".

So yeah, 5 miles is on the edge of "long" for speech.