1:1 output transformer question

[Tyimo]

Hi Andrew T and Sajti!

Thanks for the help!

Let's get back for the original plan: to try a known capacitor coupled design with output transformer instead of the output cap.
Here is the reference JLH69 amp with dual supply rails. +/-22V

Please help me to calculate and design a suitable output coupling transformer for it!

Power = 20W
Voltage = 44Vdc?
Speaker = 8R


Greets:

Tyimo

 

[AndrewT]

I really think the expertise for this transformer design is more likely to be found in the tube amp category.

 

[Tyimo]

O.K. I would try to ask there too.

If I would know the amp's correct output imp and the transformator losses (typical is 10%) than I could calculate the
Secondary imp:
Zs=8/1.1= 7.2 R

and the

impedance ratio:
Z= Zp/Zs

The next would be the turns ratio:
N= square root Z

The voltage ratio:
Vs= Vp/N

To convert this to RMS

Vp = 1/2 Vp-p =?
RMS = peak * 0.707 =?
Power is therefore V² / 8 = ?

Am I right?

Tyimo

 

[AndrewT]

Hi,
the turns ratio is not the problem.
for matching input : output impedances the turns on the secondary are the same as the turns on the primary.

The problems as this non-tube guy can see are;
iron core area,
lamination thickness,
lamination type/specification,
copper gauge,
number of turns,
target frequency response,
interleaving strategy to extend the frequency response,
minimising low frequency distortion,
and I suspect a large number of questions I am not even aware of, never mind answers to.

 

[Tyimo]

Yes and thanks!

These are very good questions!

I will ask them.

Tyimo

 

[Sheldon]

Here is the reference JLH69 amp with dual supply rails. +/-22V

Please help me to calculate and design a suitable output coupling transformer for it!

Power = 20W
Voltage = 44Vdc?
Speaker = 8R

I think you are on the right track in looking at Susan's design. I'm by no means an expert, but see no reason why her basic sizing and turns calculations wouldn't work here. As said earlier, you just put the primary in place of the speaker and put the speaker across the secondary. You will need to make sure that the idle current through the transformer is low, but it needs to be low when connected directly to a speaker anyway. You can also use an E-core with a small gap (might even sound better with a little current, as you would move the crossover point away from the center of small signals). Susan mentions a gapped transformer on her site I think. For a test, I would take a similar sized core as she uses for her 35watt version and use a similar number of turns. Just wind both at the same time (bifilar) and use one for the primary and one for the secondary.

Having said that, I'm not sure what you gain in a design with bipolar supplies. Why put a transformer in there when you can drive the speaker directly? I could see it if you needed to impedance match a regular amp with say a 300 ohm speaker or with a 0.5 ohm array, or if you wanted, as you mentioned, to run at low voltage/high current.

Sheldon

 

[Tyimo]

I would take a similar sized core as she uses for her 35watt version and use a similar number of turns. Just wind both at the same time (bifilar) and use one for the primary and one for the secondary.

Thanks Sheldon! Good idea!

Why put a transformer in there when you can drive the speaker directly?

Sometimes direct coupled amps sounds even better with output caps......!
Maybe with transformer too or even better.....

I just one to here the sonic differences between direct coupling-caps coupling and transformer coupling .

Greets:

Tyimo

 

[Sheldon]

I just one to here the sonic differences between direct coupling-caps coupling and transformer coupling .

Fair enough. Go for it then.

Sheldon

 

[Shoog]

I would personally try a 110:110V Isolation transformer first. Of course these are not designed for the job and will normally distort badly bellow 60Hz. However my experience is that if you oversize your transformer by 4-5 times (120VA in your case), you will get good response down to about 10-15hz. DC should be zeroed. High frequency response is not really an issue with toroidals, though it would be with EI's.
If this fails to satisfy, then strip off the windings and use the core to start your own winding experiaments. This will probably work out as cheap as buying the core by itself, but could save you a hell of a lot of hassle.

Shoog

 

[AndrewT]

Hi,
an EI could be dismantled and all the Es aligned with each other and all the Is likewise, then reassembled with the bobbins in place and the Is gapped from the Es to achieve the DC tolerant core.
A toroid will not work with a DC bias through the windings.

 

[Tyimo]

However my experience is that if you oversize your transformer by 4-5 times (120VA in your case), you will get good response down to about 10-15hz.

Thanks Shoog!

 

[Tyimo]

Es aligned with each other and all the Is likewise, then reassembled with the bobbins in place and the Is gapped from the Es to achieve the DC tolerant core.

I have something like this. Thanks!

A toroid will not work with a DC bias through the windings.

Yes, but last time I was speaking about the dual rail supply-JLH69 without caps. In this case we needn't bias the amp through the windings like by Susan's follower.

Tyimo