On Sun, 15 Apr 2018 10:17:43 +0100, Roger Hayter wrote:

The Natural Philosopher wrote:

On 14/04/18 21:03, Johnny B Good wrote:
An effective way to avoid saturation issues

THERE ARE NO SATURATION ISSUES UNTIL CURRENT IS DRAWN

Which statement is total and utter ********. :-)


WE ARE NOT DRAWING POWER

Essentially true in this case but totally beside the point.


Absolutely. The only reason for using a transformer capable of
supplying significant power is to obtain one with proper mains voltage
insulation. Otherwise any small signal 50:1 transformer with a suitable
low frequency bandwidth would do the job.

Wow! Here's me thinking that it's only TNP that doesn't properly
understand basic electrical theory. :-)

Saturation issues have nothing to do with drawing power via a
transformer's secondary. The problem arises out of the magnetization
current increasing (due to either insufficient turns or else,
equivalently, excess voltage) to the point that the magnetic flux in the
core reaches a level that saturates the magnetic core material used.

There is an art to manufacturing an old fashioned wallwart transformer
as cheaply as possible such as those often hot running chinese wallwarts
where they've calculated not only the thinnest of wire to be used but
also the minimum number of turns required for the nominal voltage rating
without inducing excessive saturation effects. As a consequence, such
transformer designs leave no margin whatsoever for over-volting events in
the mains supply.

The heating effect in such transformers under no-load conditions is
typically a watt or two both from I squared losses due to the saturation
peaks causing transient reductions in the primary inductance which
results in corresponding current spikes and the increased hysteresis
losses in the core material itself.

In my search to find some figures on the magnitude of 'magnetisation
current' compared to the full load current, I noticed some interesting
animated graphs of core saturation effects in this wikipedia article

https://en.wikipedia.org/wiki/Transformer#Real_transformer.

Unfortunately, I wasn't able to find figures for transformer
magnetisation current which ISTR as being typically an order of magnitude
or so less than that due to their maximum load current ratings.

Using a pair of transformers wired as I suggested is an effective way to
eliminate any question regarding transformer saturation effects for the
sort of over-volting events being monitored. It's an arrangement that
allows for a 100% over-volting event to be measured without saturation
effects spoiling the integrity of the measurement of an event that would
instantly fry incandescent lamps and heater elements alike.

It doesn't matter that the transformers are only operating at half their
designed voltage rating. The arbitrarily low secondary voltage is simply
a faithful, low voltage, galvanically isolated replica of the mains
voltage being monitored anyway, the level of which will be calibrated so
that 240v corresponds to -10 or -20 dB of FSD in the audio recording
itself. Using an audio recording application to log the mains waveform is
just a quick 'n' dirty way to see whether or not there is anything of
interest to be logged in the first place.

--
Johnny B Good