Input transformers impedance questions

[Andypairo]

Hi all,
I'm intrigued by the use of transformers for line level, however I'm a bit puzzled about the impedance values quoted on manufacturer datasheets.

Usually line input transformers are quoted at some 10k impedance on the input circuitry side, but of course the impedance seen by that side is mainly the output impedance of the previous output stage reflected by the ^2 of the turns ratio.

So what the 10k figure refers to? A optimal value to avoid ringings due to parasistic elements?

I often see isolation transformers 600:600 Ohms used (with claimed great benefit) between CD and pre or pre and power amp... so what's the difference between using say a 600 Ohm trafo or a 10k one for input?
I just don't get it....


Cheers

Andrea

[Joel]

Hi Andrea,

"10k" is the optimal load on the secondary for best frequency response. The actual impedance seen on the secondary is much less than that.

Let's look at a specific example, the Jensen JT-11P-1:

Primary DCR: 1450 ohms
Secondary DCR: 1550 ohms
Turns Ratio: 1:1
Optimal Secondary Load: 10k ohms
Line Impedance: we'll say 600 ohms

So we get:
Zprimary=(11550*(1^2/1^2)+1450
Zprimary=13000 ohms

Zsecondary=(2050/(1^2/1^2)+1550
Zsecondary=3600 ohms

So, the input circuitry "sees" 3.6k ohms. The output stage of the previous circuit "sees" 13k ohms. But you can see how the transformer has no inherent impedance itself. It merely reflects the loads attached to it.

Joel

[Andypairo]

Hi Joel,
what if I use a 600 Ohm trafo (as most of the line output types) into a 10k impedance (or more) of the following amplifier?

Do I lose frequency response?
And if so, why using such a trafo as "standard" for line output if usually we have tens of KOhms input impedances?

Cheers

Andrea

[Joel]

No, you would not lose anything. What you describe, 600 ohms into 10k ohms is just fine.
In voltage transmission (as opposed to power transmission), you always want a low source impedance, and at least 10x that number for the receiving end. The higher the receiving end impedance, the less power is taken from the source ("loading").

You only lose frequency response when you increase the source impedance to such a point that it begins to interact with the capacitance of the transmission line and input of the amp.

Joel

[vax9000]

My understanding, the 10k primary winding might have a 1k DC resistance, so you can't use it as say, 600 Ohms if you change the secondary load accordingly. Also a 600 Ohm transformer can't be used as 10k, because the primary inductance may be too small. So you'd better use 10k as it is indicated.

[Andypairo]

Originally posted by vax9000
My understanding, the 10k primary winding might have a 1k DC resistance, so you can't use it as say, 600 Ohms if you change the secondary load accordingly. Also a 600 Ohm transformer can't be used as 10k, because the primary inductance may be too small. So you'd better use 10k as it is indicated.

Sorry but I don't get the point here.

I only knew about inductance requirements only if the trafo is also used as anode load for a tube, not in a input stage... am I missing something?

Cheers

Andrea

[Geek]

A simple rule-of-thumb I learned was the inductive reactance at the lowest expected frequency should be four times the desired impedance for HiFi.

Though modern circuit design techniques (and a wee NFB) can bring this closer to 1:1

[lndm]

So, (just to be difficult ) if I want to interstage couple a 6sn7 driver, and want to load it with 100k, to get 400k @ 20Hz I would want a 3000H winding. It would be nice .

A commonly used (more practical) value for a 6sn7 plate load is 400H, giving 50k @ 20Hz.

It is my opinion that such a winding ought to be referred to as 50k. Anything else would be misleading (for me at least).

[Geek]

Originally posted by lndm
So, (just to be difficult ) if I want to interstage couple a 6sn7 driver, and want to load it with 100k, to get 400k @ 20Hz I would want a 3000H winding. It would be nice .

Hehehehe

Parafeed an ungapped transformer with a core of the 50VA class and you have close...

[Shoog]

I just did a test on a mains toroidal of 50VA 110:115+115V.
With a 6Vpp input signal I got a nearly flat response down to 10hz with a -3db point at about 20khz. This is not a real world result as the only load was the scope and the only signal was my signal generator, still I was surprised.
This will be used as a parafeed phase spitting interstage transformer.

Shoog