Hi to all,

someone wound me a se-transformer for my 2a3 amp.What i have is a
bobbin and the c-cores.
How can i find the right air gap for the cores?

Knows someone a simple measurement to determin the gap?

I have voltmeter,scope and wavegenerator available.

Thank you very much

frank

Frank,

Simply take some paper with varying thickness' and stack them up to 0.010
(0.25mm). Insert them into the "c" cores. This is a good starting point. The
ideal won't be far from this value. The target is to get the inductance as
flat as possible with varying levels of dc current through the primary. The
2A3 takes about 50 mA. So your transformer should be flat in inductance from
0-50 mA.

If you could tell me the core cross sectional area, the core material (with
its' laminate thickness), and the number of primary turns, I can calculate
the ideal thichness of the gap.

There is an empirical way of accomplishing this with simple voltmeter
readings. It is a triangulation method that determines inductance. It is too
complicated to describe here but if you are intereested I can send you a
scan of the pages.

Carroll


"frank" wrote in message
m...
Hi to all,

someone wound me a se-transformer for my 2a3 amp.What i have is a
bobbin and the c-cores.
How can i find the right air gap for the cores?

Knows someone a simple measurement to determin the gap?

I have voltmeter,scope and wavegenerator available.

Thank you very much

frank

frank wrote:

Hi to all,

someone wound me a se-transformer for my 2a3 amp.What i have is a
bobbin and the c-cores.
How can i find the right air gap for the cores?

Knows someone a simple measurement to determin the gap?

I have voltmeter,scope and wavegenerator available.

Thank you very much

frank

I assume you don't know if there will be a sufficient amount of
inductance
at LF.
Set up the transformer with an initial gap of 2 sheets of paper
from your notebook to each side of the C-core joins.

Set up the amp to run at just under the full power into the rated load
value
of dummy resistor load at 1 kHz.
Then wind down the frequency, keping amplitude constant to the 2A3 grid.

At some F below 100 Hz,
With too little paper in the "gap" there will be sudden distortion and
pulses of current in the tube during part of the sine wave cycle.
Increase the paper layers until the tranny does not cause current pulses

from saturation until 20 Hz.
The reponse should still be very near that at 1 kHz.
this means the amount of inductance is sufficient to
cause no loss of LF response.
But saturation might still begin at 10 Hz, sen as sudden gross
distortion on the waveform on a CRO.
Add slightly more paper, until current pulses don't occur as F is reduce
from 100 to 1 Hz.
You may find the -3 dB response is at say 16 Hz,
which is due to the gain in the map reducing 3 dB when the primary
inductance
has become a low enough impedance to cause a 3 dB loss.
In this triode case there is a 3 dB gain loss when the reactance in ohms

of the inductance equals the load plus the Ra of the tube in parallel.
So ir Ra = 800 ohms, and RL = 4k, RA = 666 ohms, and if the
-3 dB point is 16 Hz, then Lp = 666 / ( 6.28 x 16 ) = 6.6 Henries.

This isn't a lot of inductance.
The gap has the effect of reducing the effective permeability, u, of the
iron,
from perhaps 10,000 for grain oriented silicon steel C-cores,
to perhaps only 500. The gap limits the magnetisation of the core by the
DC flow
in one direction, with the added flow of signal current, which
demagnetizes the core
in one direction, and increases the magnezation in the other, and we
have to try
to prevent running the tranny so that the wanted voltage level tries to
exceed
the limitation of the ability of the iron to magnetize.
Its possible that your tranny will allow a full power level signal
without
much core distortion seen on the CRO trace, right down to 10 Hz, full
power,
in which case the tranny has good bass response.
But many SE OPTs start distorting badly, ie, more than 5% from the core,

at 40 Hz, near full power, and no matter where one sets the gap,
and such a tranny is probably shelved for another lower power project.

Its probably better to have a slightly bigger gap than required.

One can measure the inductance with DC by removing the secondary load,
and placing a 10 ohm R between anode and OPT primary, and by
measuring the transformer primary current and dividing that into the
voltage across the primary, all done away from full power or distortion,
one can measure the impedance of the primary
with a certain gap. As the gap is increased from zero, the inductance
rises, then at some gap you have to find experimentally, the L is at a
maximum,
before slowly declining as yet more paper sheets are inserted into the
gap
using the same test conditions.
I prefer using an inductance of 10% less than maximum,
by increasing the gap past the peak L.
This allows for the vaguries of DC transient events, or an increase in
Ia at some time.

Insufficient inductance is seen as a -3dB point which is above 20 Hz.
So it would be possible that after tuning the gap for slightly less
than maz L, the -3 pole is at say 30 Hz, but no matter what signal
is fed to the tube less than tube overload, ie, clipping, saturation
won't occur.
We want the tube to overload agaed of the core saturating.

I make no apologies on behalf of the God Of Triodes,
who deliberately made setting up SET OPTs difficult
to make sure DIYers had a tough walk towards getting fidelity
from SET amps, which are heaven on earth when they are done right.

Patrick Turner.

Gentleman,

thank you very much!!

First i have to translate and than to understand!

frank