Balanced Tube Headphone Amplifier?

[Steve Eddy]

I don't understand. If the center tap of the secondary is grounded, how can it be floating?

Are all your grounds in the schematic taken to the same point?

se

 

[SpreadSpectrum]

I'm with Sodacose on this one, I don't know how a winding whose center tap is grounded can be considered floating.

 

[Steve Eddy]

I'm with Sodacose on this one, I don't know how a winding whose center tap is grounded can be considered floating.

Sorry. Misread the schematic. I thought the headphone output was taken between the top of the secondary and the center tap and the bottom not connected. It's the headphone driver that will float.

Why is the secondary center tap tied to ground? Safety reasons?

se

 

[Steve Eddy]

Wait a minute. If the top of the primary is 16 ohms, how can the 4 ohm tap give you balance if it's tied to ground? Half of 16 is 8, not 4.

se

 

[SY]

Impedance goes as the square.

 

[Steve Eddy]

Impedance goes as the square.

Derp!

The embarrassing thing is that I knew that. Just didn't follow it through. Just focused on there needing to be an equal number of turns on either side of the center tap.

Thanks for the smack upside the head.

se

 

[Sodacose]

Ok, good. I was thinking really hard about this for a minute there. Glad I'm not crazy.*


*Or at least conclusive evidence has not yet been furnished to that effect.

 

[Steve Eddy]

Ok, good. I was thinking really hard about this for a minute there. Glad I'm not crazy.*

My apologies.

se

 

[Steve Eddy]

One question though. Why is the primary's center tap on the input transformer tied to ground?

se

 

[Sodacose]

Balanced input (again at request of a friend). In practice I'd wire to allow balanced or unbalanced.

 

[SY]

Actually, I think your balance will be better if you don't ground the center tap- if you do, balance is only as good as the winding symmetry.

 

[Sodacose]

Which center tap are we are king about now? Input?

I could use a ground lift switch as well I suppose?

 

[SY]

Output, sorry for my lack of clarity.

 

[Sodacose]

Isn't the secondary floating in that case? I've always read to avoid this because the secondary could charge to a voltage potential. With speakers I might not worry, but when the load attached is going on the user's head...

 

[SY]

If that worries you, run a large (megohms) resistor from each side to ground.

 

[Sodacose]

OK, that makes sense. How does the secondary charge up BTW? Capacitively?

Not having to ground the secondary center tap means there may be a lot more options for output transformers. BUT, now I'd be dividing the a2a by 4 instead of 2, right? So I'll have to find something with a fairly high primary z.

Thanks for the help, SY!

 

[SY]

CT or no will make no difference in the reflected impedance because you're not changing the relative turns ratio between primary and secondary. See, I'm full of good news today!

I've seen a few handwaving explanations of the charging, but haven't seen any data showing that it happens, much less the cause.

 

[Sodacose]

Thanks! Maybe grounding both phases through a single resistor would offer the best balance?

Have to admit this balanced stuff is kind of fun even if it doesn't have a practical benefit.

 

[SY]

If you use the meg+ resistors, it won't affect the balance if you think about it. What's 2R in parallel with 1M? What's 2R in parallel with 1.5M? The only thing they do is drain off any stray voltage.

Life's easy.

 

[Sodacose]

When you put it like that it really does make the whole balanced output idea sound a little silly. Also solves the reflected load thought pretzel I was working on.