Ian Iveson wrote:

Phil Allison" wrote

** Autos have less copper losses and much better bandwidth than
a
isolation type.

Sometimes. Is that what ** means?

Wanna compare figures on a 4:1 impedance step down toroidal
with a
160 VA size core ?

Read the other posts that I agreed with. 4:1 is easy. 400:1 is
nearer what most people would use here. Neither would 160VA be
appropriate in most cases. Care to cite a reasonable comparison?
Care to back up your statement about "much" wider bandwidth, in the
context of an OPT for a domestic valve amp? Can you think of any
other important parameters other than copper losses and bandwidth?

Ever wondered why no-one uses, or supplies, auto-transformers for
valve amp OPTs?

In my last post on the subject, where I described the use of a
symetrically balanced
OPT with a centre tapped secondary, with plate, ( or cathode currents )
flowing through the speaker winding is not done because
there really isn't great benefits to transformer efficiencies,
and the cost isn't seen as worth it.

But the other benefit of the fully isolated traditional OPT is that
you have a far safer amp, since lethal plate voltages are
unable to leap out and kill a child, adult, or a dog.

160 to 500 VA toroidals will have naturally low leakage inductances,
since the sections of windings can be wound over each other in close
proximity,
and equivalent to a wide traverse width in a plain bobbin wind up.
Such a plain bobbin wound tranny could be used for a low voltage
audio tranny, and the layers could be simply
interleaved to tightly couple the windings to extend BW
to 300 kHz if desired.
I have wound UL OPTs for 300 watt applications
using 1.1k to 5 ohms ratio, with 6S and 5P sections, to
get 20Hz to 300 kHz at full power.
If my trannies were used as balanced auto OPT,
the BW might extend slightly further.

In the case of a 300 watt amp the speaker secondaries
could have a CT, and be grounded, and the 38 vrms created to make 300
watts into 5 ohms
is an ideal amount of voltage which can be +/-19v either side of the CT,

and then fed to the cathodes of multiple OPV to provide enough cathode
FB for an acoustical connection to then be effective to achieve the
power of the tetrode,
but the low distortion and Ro of the triode.
So there thus need to be only a primary, and a secondary,
and the tube current circuit and speaker current circuit
"share" the same path, but in actual fact the signal current in the
shared
portion, ie, the speaker winding is equal to to the speaker current
MINUS the cathode signal current.
Winding losses will then be slightly lower than the two
separate windings with full isolation.

But the above idea is only of benefit with tubes like KT88/6550/EL34etc,

when you have perhaps a dozen involved; when just two OPV are used,
the speaker voltage isn't high enough to provide much useful
CFB to the OPV cathodes.

But 4:1 impedance ratio for a speaker matching application
is a neat way to achieve matching, and maintain the damping factor,
or improove it.

This idea makes sense if you have say 4 x "8 ohm" woofers
in parallel, and where 200 watts is required for a decent ceiling.
Many amps, including SS might baulk at 2 ohm or lower loads
trying to make 200 watts, but the matching transformer
allows the amp to think it has 8 ohms connected, and it may be able to
drive that easily,
but at the speakers, the voltage is halved, and current doubled, so you
get
200 watts into the speakers, so its a win-win situtation.

I have yet to see this application of auto trannies for speaker use
used commercially, because of beancounter concerns.

The other benefit with an auto tranny when using an SS amp
is that if the amp blows up, and a supply rail becomes directly
connected to the auto tranny, the rail fuse WILL then blow, because the
DC from the SS amp has an easy shunt path, instead of burning out the
voice coils
of the woofers.

So the VA of the auto tranny core and wind up details for speaker use
should be
carefully worked out to ensure the Bmax is less than 0.3 Tesla at 50 Hz,

to ensure Fsat is low. This applies regardless of the U of the iron.
During normal use, such a low B, especially with GOSS toroidals cores,
or other GOSS laminations, will give excellent low distortion
and low losses.

Patrick Turner.


cheers, Ian