balanced input stage

[Majestic]

Another interesting thread. (You guys rock! )

Without wishing to take any of the attention away from Sonix's last posting, which seems to be waiting for commment; I am really curious about Sch3mat1c's ( - hmm, must be an audio h4ck0r! ) point about how the first schematic can offer both an unbalanced and balanced inputs.

WTF, that first schematic has an unbalanced input...but no way to balance the signal when using it... omgHUH??

An externally hosted image should be here but it was not working when we last tested it.

Given this schematic actually comes from Svetlana themselves, I'm assuming they wouldn't have made such a fau-paxs?

So how does this work??
I've heard it's possible to theoretically balance a signal with just a couple of resistors, but never seen it done before. Is that what Svetlana are doing here? But then if it was that simple, why all the bother with imperfect phase splitters in our PP amps?!?

 

[zobsky]

sonix, .. you still haven't answered the question "why do you want a balenced input stage, .. for phase splitting in a PP amp or some other reason" , .. budget?

the responses and direction of this thread can vary markedly depending on the response.

PS. SY,.. i remember bothering you endlessly on this topic a few months ago, .. i will (eventually) get to build that 6bg6 PP amp, .. as time permits. your suggestions on balenced jensen input trannies are not forgotten (yet)

thanks, V

 

[sonix]

@ zobsky:
my preamplifier has exits XLR and I would like to have therefore entries XLR at the end amp. i like XLR

 

[Majestic]

...
i like XLR

Same here. I'm really fed up with RCA jacks.
I feel like using XLR's for all my normal unbalanced connections even!

 

[Sch3mat1c]

Re: Re: balanced input stage

I am really curious about Sch3mat1c's ( - hmm, must be an audio h4ck0r! )

point about how the first schematic can offer both an unbalanced and balanced inputs.
Given this schematic actually comes from Svetlana themselves, I'm assuming they wouldn't have made such a fau-paxs?

So how does this work??

The way I see it, the neon lights are either on all the time, to regulate cathode voltage, or they are there merely as a precaution to ensure it starts right and the voltages end up somewhat correct. The former doesn't make any sense in a balanced circuit. I guess it's because it's a direct-coupled circuit.. but I've made something more complicated (three 12AX7 triodes and half a 6SN7 in a circuit similar to Fred Nachbaur's quasi op-amp) without any parts to assure startup conditions. Basically it can't latch into another operating point because of designed-in stability, and I don't really see this one doing it either. But whatever, I'm droning on again. So anyway, the circuit is all balanced, as you might guess. I don't know what the output has in the cathode circuit but the other two stages (preamp and driver) are LTP. The negative supply voltage allows a vague approximation of a CCS for the cathode circuits of the tubes ('SL and 'SN), and cascading two will ensure relatively good rejection. (Think of it as a wrench on the end of a pipe. You can apply torque at the end, but if you don't push back on the pipe it'll be pulled out by the same force. If instead you balance this force by pushing with another wrench, there is no common mode force and the pipe stays put.)

Any resistance in the cathode circuit of a diff. amp (LTP) reduces the 'pushing' of the circuit to balance the force on that pipe. Ideally you use a CCS (pentode or transistor) so the circuit balances itself by sensing the unbalanced force and pushing back with another wrench of its own (in this case the other tube). If you have a balanced input then input controls both the wrenches. With unbalanced signals, the circuit gains control of the other one (indeed, note the switch that nulls input to that tube, putting things back on the behavior of the circuit instead).
But the cathode resistor is equal to the plate load resistor on the preamp stage so I doubt balance is very good there.

Anyway I've babbled on long enough here, you get the idea.

I've heard it's possible to theoretically balance a signal with just a couple of resistors, but never seen it done before. Is that what Svetlana are doing here? But then if it was that simple, why all the bother with imperfect phase splitters in our PP amps?!?

I don't think so, unless you want to do tricky floating power supply kludges (I've seen such a phase splitter design before). It always involves an active or active-ish device, tube or transformer. Resistors can't invert a signal. However, you can certainly get the common-mode junk from a balanced line using three resistors. Fine for de-multiplexing a stereo signal but no good for general audio use where common mode rejection means lower noise pickup off the cable run.

Tim

 

[benny]

hi,
i need a balanced input stage, like this:

An externally hosted image should be here but it was not working when we last tested it.

, for a 300b se amp like this:

An externally hosted image should be here but it was not working when we last tested it.

.
is this possible?
thanks
sonix

Hi,

to make this simple, i take it all you want somthing similar to the seccond schematic, only a ballanced design which you showed in the first schematic??? right???

if this is all you want, then yes, this is perfect for what you want. if the output from your source is already ballanced as would seem logical from the XLR output you say your preamp has, then this idea works fine. The amp is baisically just two seperate identical sides, right from the input. I don't know what all the discusion of needing input tranny's was about, but if your source is already ballanced, unless you need impedance matching, then you won't need an i/put tranny or anything... just plug your ballanced source in, and it should be fine... simple.

If you want to use the unballanced input, then maybe this is where an I/p tranny would be good, for use as a PI, and this would now give you a ballanced signal.

That's making it as simple as possible for you as i can, to save any confusion... hope your project turns out good for you whatever you decide to do.

cheers

 

[EC8010]

I wondered about those neons too, and came to the conclusion that they are there to prevent Vhk limits from being exceeded when HT is applied before the valves have warmed up. Used as thermionic Zeners, they would destroy any self-balancing action that the circuit might have.

EDIT: Balanced does not mean impedance matched, even though traditional telecomms used 600R.

For a proper explanation of balanced audio, have a look at the papers by Bill Whitlock etc at Jensen Transformers site.

 

[sonix]

@benny

to make this simple, i take it all you want somthing similar to the seccond schematic, only a ballanced design which you showed in the first schematic??? right???

it's right, but i would like to use only one 300b. is that possible and would it work?

 

[benny]

EDIT: Balanced does not mean impedance matched, even though traditional telecomms used 600R.

who said ballanced meant impedance matched???

i know i didn't. i said:

if your source is already ballanced, unless you need impedance matching, then you won't need an i/put tranny

this is saying, if the source is already ballanced, why use an input transformer.... unless of course, the impedance is too high and needs to be bought down... then this is a good reason to use an I/p tranny. I didn't say a ballanced input means the impedance is automatically right.

I will go and read those papers though, nothing wrong with reading as much as possible... always good to learn something new.

cheers

 

[EC8010]

I misinterpreted your punctuation to be a reference to the 600 Ohm legacy that's proving so hard to shake off. My apologies.

 

[SY]

I wondered about those neons too, and came to the conclusion that they are there to prevent Vhk limits from being exceeded when HT is applied before the valves have warmed up. Used as thermionic Zeners, they would destroy any self-balancing action that the circuit might have.

Indeed that's the case, as the reverse-biased diode indicates. And it can be deduced from the working voltages on the diagram, too. It's a traditional thing to do (see: Marantz) with large negative rails; this came up in a recent cathode follower discussion.

IIRC, the NE2H fires at about 65V, but it's been a while...

 

[benny]

I misinterpreted your punctuation to be a reference to the 600 Ohm legacy that's proving so hard to shake off. My apologies.

don't need to apologise. i realise what i had said had been misinterpereted... i was just making clearer what i said... i wasn't offended, don't worry.

@ sonix

it's right, but i would like to use only one 300b. is that possible and would it work?

yes, it's possible, but, you will want to use the seccond scematic anyway in this case because by using one 300B tube, your now using a SE config. Therefore it's a waste keeping the signal ballanced all through the preamp because you're using more parts/components than necesary. Best thing as far as i'm concerned is use an I/P coupling tranny... same idea as using a tranny as a phase splitter, only in reverse. This will turn ballanced signal into unballanced, which is what this amp will be as you want to use a SE config. Only advantage of keeping preamp ballanced and not coupling two sides together until just before output stage is you will still get common mode rejection of any noise picked up by the signal in the preamp stages, although i doubt there will be a lot if you're driving this from a preamp as the signal will already be well above the noise floor.

Seeing as you're running from a preamp, why not use just a 300B from the power section, and just one preamp stage as a driver for the 300B???? With signal form preamp, it will already be relatively hot, and by the looks of this design, it was made to be used with lower level inputs, such as that from a mic (not entirely sure, i'm not familiar with 6SL7 tubes, so i don't know what kind of amplification/headroom you get out of these)... by using two stages, chances are the preamp signal will get amplified enough by the first stage enough to drive the seccond stage into clipping easily, seeing as there's no atenuation of the signal between them. If you do use the 2 stages, you might find it a good idea to add in a volume control between the first and seccond stage.

anyway, hope this helps.

cheers

 

[Sch3mat1c]

Re: Re: balanced input stage

but if your source is already ballanced, unless you need impedance matching, then you won't need an i/put tranny or anything... just plug your ballanced source in, and it should be fine... simple.

But it makes a nasty kludge and is blatently no good for long runs, where the noise cancelling effect of balanced shines. By taking the signal off just one (+ or -) line you amplify both the common mode (noise) and audio. Of course, since this is a home situation the chances are that balanced was used in the first place only as a novelty, on runs between amps no longer than maybe 12', something easily served with shielded patch cables. So the point of using balanced is moot, there will be little noise pickup and it is safe to use it in an unbalanced way. Of course, it defeats the balanced output on the signal source and one can no longer say his system is balanced, resulting in a mental degredation of sound quality.

Tim

 

[DrDeville]

For a proper explanation of balanced audio, have a look at the papers by Bill Whitlock etc at Jensen Transformers site. [/B]

Nice tip!

http://www.jensen-transformers.com/apps_wp.html

Thanks EC8010!

Best,

George Ferguson

 

[Majestic]

Re: Re: Re: balanced input stage

... resulting in a mental degredation of sound quality.

(LOL!)

 

[fdegrove]

Hi,

Of course, since this is a home situation the chances are that balanced was used in the first place only as a novelty, on runs between amps no longer than maybe 12', something easily served with shielded patch cables.

That plus the fact that some preamps unashamedly sport XLR outputs on unbalanced outputs as well.
So you'd better make sure the output is actually balanced and if so it's always interesting to know how it was done....

Active or xformer (which could well be wired as unbalanced to balanced anyway, but that's not a bad idea per se)...etc.

Cheers,

 

[benny]

Re: Re: Re: balanced input stage

But it makes a nasty kludge and is blatently no good for long runs, where the noise cancelling effect of balanced shines.
Tim

i said this about turning ballanced into unballanced:

Best thing as far as i'm concerned is use an I/P coupling tranny... same idea as using a tranny as a phase splitter, only in reverse. This will turn ballanced signal into unballanced, which is what this amp will be as you want to use a SE config

By using this before whatever stage you are going to turn the ballanced signal into SE/unballanced, you are still keeping the benifits of the ballance you had.

That plus the fact that some preamps unashamedly sport XLR outputs on unbalanced outputs as well.

that's a very good point... you'd hate to spend money on this ballancing, only to find you're not gaining anything at all....

cheers

 

[Sch3mat1c]

Ya. An input transformer is always the absolute best passive solution, if you want to go all-out and throw in two tubes I'd use a LTP with pentode tail. The electronic equivalent except it has some gain to boot.

TIm

 

[Steve Eddy]

Ya. An input transformer is always the absolute best passive solution, if you want to go all-out and throw in two tubes I'd use a LTP with pentode tail. The electronic equivalent except it has some gain to boot.

If you want some gain to boot, you can do that with a transformer too. In fact the transformer I specified previously has about 1.5dB of voltage gain whereas a more typical input transformer (such as the 11P-1) will have a loss of about 3dB.

You can even go a bit further and have ALL of the amp's voltage gain done passively using transformers.

se

 

[Sch3mat1c]

Actually I meant gain in general, that is, power gain, which may or may not be needed.

Tim