Odd Output Transformer Needed

[Kasanay]

I got a crazy idea to build this little headphone amplifier from a 1960 something magazine article that I stumbled upon (attached).
Problem is, I have now rounded up nearly all the parts except the elusive output transformers!
The article says I could use a Triad S-13 X. There is nothing special about the S-13X except they are rated at 25,000Z to 4Z. They need to have a turns ratio of about 85:1
I have found a few equivalents by checking various cross-reference charts. The transformers are quite small - only a couple of watts of power. Any of the following should work:

Triad S-13X
Triad S-13Z
Sancor A-3857
Stancor A-3327
Merit A-2999
Merit A-2937
Allied 6W47HF

There is 1 of these up for bid on ebay and I may just buy it, of course, I'll still need to find a second.

Anyone know where I can get a pair of these?

Thanks ,
Joe Y

 

[Joe Roberts]

What is the impedance of your headphones?...Probably not 4 ohms these days.

if you want to present a 25k nominal load to the 6AU6, recalculate the needed turns ratio accordingly. This should get you down into the range of more common transformers.

 

[BinaryMike]

A good-quality OPT with 6K or 6.6K primary and 8-ohm secondary specifications should work nicely with 32-ohm load, which is typical but not universal for headphones these days. Whaddaya got?

 

[Kasanay]

What is the impedance of your headphones?...Probably not 4 ohms these days.

if you want to present a 25k nominal load to the 6AU6, recalculate the needed turns ratio accordingly. This should get you down into the range of more common transformers.

50 to 600 ohm phones should work out.
I did not do the calcs. These were done in 1961.
I believe the the intention is to present a load MUCH higher than 25K.

 

[Kasanay]

Thanks for the comments but I am not looking to redesign this circuit at this time. I just want to locate the spec'd OPTs.

Thanks,

 

[BinaryMike]

It's really too bad that Robert Voss didn't explain his design in more detail. A close examination of the text and schematic leaves me suspicious that it was all a seat-of-the-pants project, probably using whatever components happened to be lying around. The high OPT turns ratio is a particular problem.

Despite your wishes, Kasanay, I propose changes. Specifically, I would strap the 6AU6 as a triode, reducing its anode resistance to about 6250 ohms, and select an OPT primary in the 5K~10K (to 8 ohms) range, which reflects a 20K~40K load to the 6AU6 with a 32R headphone load. An operating point around 12.5mA at 250V (3.1W anode dissipation) should yield excellent overall performance. See the very attractive triode curves in the 6AU6 datasheet! The cathode resistor should be about 330R. Output power is reduced, but that's immaterial or even advantageous with headphones. Distortion may be reduced so much that negative feedback can be omitted. A rough calculation shows that overall gain is near unity for the proposed output stage and OPT with bypassed cathode resistor, which puts about 30mW into 32R headphones with 1Vrms at the grid. That's more than enough power for many headphones, so you may find that the input stage can be omitted. Try it.

 

[Kasanay]

It's really too bad that Robert Voss didn't explain his design in more detail. A close examination of the text and schematic leaves me suspicious that it was all a seat-of-the-pants project, probably using whatever components happened to be lying around. The high OPT turns ratio is a particular problem.

Despite your wishes, Kasanay, I propose changes. Specifically, I would strap the 6AU6 as a triode, reducing its anode resistance to about 6250 ohms, and select an OPT primary in the 5K~10K (to 8 ohms) range, which reflects a 20K~40K load to the 6AU6 with a 32R headphone load. An operating point around 12.5mA at 250V (3.1W anode dissipation) should yield excellent overall performance. See the very attractive triode curves in the 6AU6 datasheet! The cathode resistor should be about 330R. Output power is reduced, but that's immaterial or even advantageous with headphones. Distortion may be reduced so much that negative feedback can be omitted. A rough calculation shows that overall gain is near unity for the proposed output stage and OPT with bypassed cathode resistor, which puts about 30mW into 32R headphones with 1Vrms at the grid. That's more than enough power for many headphones, so you may find that the input stage can be omitted. Try it.

Mike,
That is a really constructive reply. I appreciate you sharing your thoughts.
I dabble a fair amount ( enough to put something like this amp together) in this audio hobby but I am not a seasoned expert - still learning way more than teaching.
I have few other irons in the fire right now but I'll try to sketch up the circuit as you recommend. It will be a while before I could start building. I'm sure I'll have a ton of questions.
First question: in your comment about omitting the input stage, are you thinking this could be a 1-tube amp?

Thanks
Joe Y

 

[BinaryMike]

First question: in your comment about omitting the input stage, are you thinking this could be a 1-tube amp?

Correct. One tube per channel anyway. I would either breadboard it or build it with space to add an input gain stage, just in case headphone sensitivity isn't a good match. This post shows one way to construct a breadboard.