Balanced Operation Output Transformer Questions from N00b

[Oborous]

TubeCad Link

With regards to the above link, I noticed 'Balanced Operation' is an available option. Balanced operation, done well, seems/sounds better to me, but if I'm DIY'ing it, I don't understand how to do it well. And, Balanced Input always seems to loose something, so the way (to me) is to use the fully balanced operation.

Could people kindly explain how I can implement fully balanced operation (i.e. two seperate amplifiers in same chassis, using only one OPT). What my considerations are for OPT?

And if there are any additional safety considerations due to it being different than SE/PP.

Thx

[EC8010]

Balanced operation strictly refers to signal transmission between devices. In that instance, it refers to impedances (and signals) being equal, or balanced, from each leg to ground. If you want to find out more about balanced operation, there are some very good papers by Bill Whitlock at the Jensen site (transformer manufacturer).

[Oborous]

Hi,

Thanks for the response.

What is the correct terminology for an amplifer that takes XLR input and has an amp for one signal line and a seperate amp for the other signal line? By signal line, I mean either wire of the two wires that is equal impedance to ground.

What I want is an amplifier that takes a balanced input amplifies both signals and outputs a balanced (now amplified) signal.

I know I could use two amplifiers, and input one signal line to each of them, and bridge the output... but I'm wondering if I can use just a single output transformer, and how I'd go about that.

I'll try to draw a diagram....

[Oborous]

My attempt at drawing.

[EC8010]

The whole point of a balanced signal is not to separately amplify each leg but to only amplify the difference between the two legs. In this way, hum can be rejected. One way effectively uses an amplifier for each leg with some form of cross-linking to provide the rejection. This is complex and not often worthwhile. The other method is to use a transformer at the input and the output and use a single amplifier internally. This is what is used in studio audio equipment. What you have drawn is effectively a push-pull amplifier (valve term) or a bridged amplifier (solid state term).

[David Davenport]

Originally posted by Oborous

What I want is an amplifier that takes a balanced input amplifies both signals and outputs a balanced (now amplified) signal.

Hi Oborous,

That is a differential amplifier. An example of one can be seen at:

http://www.raleighaudio.com/figure_18.htm

Dave

[Oborous]

Originally posted by EC8010
One way effectively uses an amplifier for each leg with some form of cross-linking to provide the rejection. This is complex and not often worthwhile. The other method is to use a transformer at the input and the output and use a single amplifier internally. This is what is used in studio audio equipment. What you have drawn is effectively a push-pull amplifier (valve term) or a bridged amplifier (solid state term).

Thought bridging was bridging two amps together for additional current. And, thought Push-Pull is when the amp splits the signal and passes off part of it to one tube, and the other half to another tube.

But, with David Davenport's post, can I utilize a P-P Output Transformer, and wire it as per the link? Or can you have an optimized output tranny made for a 'differential' amp

Originally posted by David Davenport

That is a differential amplifier. An example of one can be seen at:

http://www.raleighaudio.com/figure_18.htm

Dave

Excatly what I'm looking for. Have to figure out the OPT, Have you tried anything like this?


Thanks All!

[David Davenport]

Originally posted by Oborous



Have you tried anything like this?

Hi Oborous,

Yes, of course. That schematic is from a series of articles originally published in AudioXpress magazine. You can read the articles online at: http://www.raleighaudio.com/articles.htm

Dave

[EC8010]

Oborous: No, bridging puts two voltage sources in series to double their voltage. If you use the same load resistance as with a single amplifier, it now requires twice as much current from each amplifier. Many amplifiers can't supply this, so this is why you will find that amplifiers that can be switched to bridged mode do not produce the quadrupling of power that they theoretically should.

I really think you need to do some reading around to find out why balanced audio is used and what advantages it confers before fretting about the practicalities of how it might be implemented. Go to your library and try reading some of the studio engineering books (not the ones about how to operate the equipment - their technical content is frequently incorrect), Any of the books by John Watkinson or Francis Rumsey are good.

[Oborous]

Originally posted by EC8010

I really think you need to do some reading around to find out why balanced audio is used and what advantages it confers before fretting about the practicalities of how it might be implemented. Go to your library and try reading some of the studio engineering books (not the ones about how to operate the equipment - their technical content is frequently incorrect), Any of the books by John Watkinson or Francis Rumsey are good.

Yes, thanks for the advice for the books. I'm trying to figure out how to calculate out the required values for various topologies of valve amps. However, I've found the terminology is very variable on the web.

I picked differential amp as I assumed the output transformer is the same as a P-P amp, but wanted to make sure.

I've been reading some articles, but am getting confused by what's available on the web. I want to hunt down some -old- books on these things, as so much of the web is only listing what's 'in vogue'.

ETF Presentation

I assumed all differential amps would be like the above link called Differential Parallel Feed (the bottom of the article), and can't figure out why someone would want either the "Western Electric Parallel Feed" or "Differential (Forced Balance)".

Atma-sphere Papers with Book References

Lists some further books for reading, and due to the age, I'm assuming they're going to actually have the different topologies. Any comments on further books?

My background is heavy in statistics, and that's why I'm questioning these things. I don't understand P-P, as when the signal is passed off to the other tube, it just seems like a bad place for random noise. But then, I also wonder why people don't use solid silver wire, about 14 ga to minimize the random noise calculations... or some cooling in valve amps. I also don't understand SE amps, you inherently need two wires to pass a signal, so why not have balanced signal and amplify both signals, rather than having all signal on one wire. You wouldn't have to amplify the signal nearly as hard and more effort generally causes more random variations, due to magnitude differences.

So, I'm trying to learn about topologies that make statistical sense to me, so I can comprehend with what's going on in there, and then figure out the other topologies.

DC going through an output transformer just seems... wrong to me somehow.

Thanks for your patience.