On Wed, 14 Jul 2004 19:56:52 -0700, Roy Lewallen
wrote:

Welcome to the world of transformer design.

What you need to do is make the impedance of the winding considerably
higher than the impedance of the load. For example, the winding
connected to the 32 ohm load should have an impedance that's several
times 32 ohms. A factor of 5 or 10 is generally adequate to keep the
disturbance of the winding to a tolerable level, so shoot for about 160
- 320 ohms of impedance. You'll need to consult the data for the core
you're using to determine how many turns that will require. Do this
calculation for either winding -- the other will come out the same if
the impedances are matched.

When you're dealing with audio and ordinary cores, the winding impedance
will primarily be reactive, and the winding inductance will be fairly
constant with frequency for a given number of turns. Consequently, the
winding impedance (reatance) will be directly proportional to frequency.
That means you need to do the calculation at the lowest audio frequency
you intend to pass through the transformer. Capacitive coupling between
the windings and other effects tend to interfere with proper transformer
action at high frequencies, so high frequency performance degrades if
you use way more turns than needed. Truly high fidelity audio
transformer design is something of an art -- the audio output
transformer was often a major limiting factor in tube-type audio
amplifier performance.

Thanks, Roy. Looks like there's rather more to it than I imagined. :-(
I wonder if Reg has written a program to design impedance
transformers? :-)
I woke up this morning, having given it due thought overnight and
thought I'd got it cracked. But that was until you said the primary
inductive load Z should be many times the design load Z. Oh well,
back to the drawing board, I guess. Looks like it's gonna take a lot
more wire and a lot more turns than I'd thought. OTOH, I could just
stick an emitter follower stage on the end of the amp chain and have
done with it. :-/
--

"What is now proved was once only imagin'd." - William Blake, 1793.

OTOH, I could just
stick an emitter follower stage on the end of the amp chain and have
done with it. :-/

Paul-

Now that you have a good grip on the problem, consider one more idea - that
most audio amplifiers have more gain than they need, and operate with the
volume control set at a fraction of full range.

If you were to use a series resistor (27 or 56 Ohms?) and NO transformer, would
the output stage have sufficient power to produce an acceptable audio level
with the volume control advanced?

73, Fred, K4DII

On 15 Jul 2004 11:03:36 GMT, (Fred McKenzie) wrote:

OTOH, I could just
stick an emitter follower stage on the end of the amp chain and have
done with it. :-/

Paul-

Now that you have a good grip on the problem, consider one more idea - that
most audio amplifiers have more gain than they need, and operate with the
volume control set at a fraction of full range.

If you were to use a series resistor (27 or 56 Ohms?) and NO transformer, would
the output stage have sufficient power to produce an acceptable audio level
with the volume control advanced?

I instinctively doubt it, Fred, but I'm in a position to give it a try
so I shall duly do so! I'll report back the results later.
BTW, this isn't a conventional hi-fi type audio amp; it's solely been
designed for picking up very faint, distant sounds. I was intrigued by
the design so set out to further examine how it worked. I've spiced it
and have now built it; the last piece in the jigsaw is the Z-matching
problem.
--

"What is now proved was once only imagin'd." - William Blake, 1793.

I, and I'm sure quite a few others, would have recommended a simple
transistor stage as a solution except for your request in the original
posting about not wanting to suffer too much power loss. If your only
concern is the amount of audio power you get out, and you're not
concerned with overall power consumption, then an active stage is
certainly an alternative. The type of circuit you choose would depend on
how much efficiency you need and how much audio output power you
require. An emitter follower is the simplest but least efficient; a
comlementary-symmetry stage would be more efficient but more complex.

Roy Lewallen, W7EL

Paul Burridge wrote:

Thanks, Roy. Looks like there's rather more to it than I imagined. :-(
I wonder if Reg has written a program to design impedance
transformers? :-)
I woke up this morning, having given it due thought overnight and
thought I'd got it cracked. But that was until you said the primary
inductive load Z should be many times the design load Z. Oh well,
back to the drawing board, I guess. Looks like it's gonna take a lot
more wire and a lot more turns than I'd thought. OTOH, I could just
stick an emitter follower stage on the end of the amp chain and have
done with it. :-/

Oh, if it's something you built anyway, you might consider using
something like an LM386 for the audio output stage. Besides providing
considerable gain (which might eliminate some other stages), it'll drive
an 8 ohm earpiece with no problem. And it's reasonably efficient.

Roy Lewallen, W7EL

Paul Burridge wrote:

I instinctively doubt it, Fred, but I'm in a position to give it a try
so I shall duly do so! I'll report back the results later.
BTW, this isn't a conventional hi-fi type audio amp; it's solely been
designed for picking up very faint, distant sounds. I was intrigued by
the design so set out to further examine how it worked. I've spiced it
and have now built it; the last piece in the jigsaw is the Z-matching
problem.

On Thu, 15 Jul 2004 10:33:08 -0700, Roy Lewallen
wrote:

Oh, if it's something you built anyway, you might consider using
something like an LM386 for the audio output stage. Besides providing
considerable gain (which might eliminate some other stages), it'll drive
an 8 ohm earpiece with no problem. And it's reasonably efficient.

Well using an op-amp did cross my mind, Roy, but it's cheating in my
book. I like to use discretes so I get a better idea of what's going
on and it's more of a challenge to get right!
--

"What is now proved was once only imagin'd." - William Blake, 1793.

On Thu, 15 Jul 2004 10:30:02 -0700, Roy Lewallen
wrote:

I, and I'm sure quite a few others, would have recommended a simple
transistor stage as a solution except for your request in the original
posting about not wanting to suffer too much power loss. If your only
concern is the amount of audio power you get out, and you're not
concerned with overall power consumption, then an active stage is
certainly an alternative. The type of circuit you choose would depend on
how much efficiency you need and how much audio output power you
require. An emitter follower is the simplest but least efficient; a
comlementary-symmetry stage would be more efficient but more complex.

THanks again. I note your comments but will try a couple of other
things first, whilst the initial assembly is still in one piece. One
such idea is the doubling or trippling of the supply voltage alone.
Does anyone know by how much it's safe to increase the 1.5V supply to
the mic insert and whether that might also cause distortion?
--

"What is now proved was once only imagin'd." - William Blake, 1793.

On Thu, 15 Jul 2004 19:03:47 +0100, Paul Burridge
wrote:

THanks again. I note your comments but will try a couple of other
things first, whilst the initial assembly is still in one piece. One
such idea is the doubling or trippling of the supply voltage alone.
Does anyone know by how much it's safe to increase the 1.5V supply to
the mic insert and whether that might also cause distortion?

I've taken it up to 3V and there's a marked improvement in gain, at
any rate. I can live with it now. I'll still try the other ideas folks
here have had, though!

--

"What is now proved was once only imagin'd." - William Blake, 1793.