On Sat, 14 Apr 2018 09:28:08 -0700, jrwalliker wrote:

On Saturday, 14 April 2018 14:42:10 UTC+1, The Natural Philosopher
wrote:

Most small mains transformers are run very close to core saturation,
so any large transients are likely to be clamped by the transformer.

Oh dear. A pseudo authoritative statement from someone who doesnt know
the difference between voltage and current.

So do you disagree with my claim that small transformers are usually run
close to core saturation at their rated voltage?

Do you disagree with my expectation that grossly overdriving such a
small transformer (with the transient to be measured) will cause its
output to saturate and give an incorrect measurement result?

This
could be overcome by putting a resistor in series with the
transformer primary so that it is running at a small fraction of
mains voltage.
Have a load resistor on the secondary may improve the performance as
well.

Oh dear. Soemone who doesnt understand inductance either.

What exactly do you object to here? A resistively loaded transformer
looks like a resistor (apart from the leakage inductance) so in
conjunction with a series input resistor it will form a reasonable
potential divider thereby moving the operation well away from
saturation.

An effective way to avoid saturation issues when monitoring for
excessively high voltage excursions is to wire up an identical pair with
their 240v primaries in series and their low voltage (6 or 7 vac?)
secondaries in parallel (in phase current aiding) so as to help maintain
voltage balance between the primaries. Not only will this allow for 100%
excursions beyond the nominal 240vac supply, it should also allow high
frequency transients to be registered without molestation by saturation
effects. However, unless the transformers incorporate a 'grounded' inter
winding screen, high frequency transients can couple capacitively into
the secondary circuit, effectively magnifying their prominence.

If this is just a temporary setup, you can use a PC or laptop to record
the waveform using an audio recording application. Unless you need to
detect extremely high frequency transients, you can choose an 8 or 16 bit
mono sampling rate of 8 or 16 or 22.05 KHz to save disk space if planning
on logging for more than 12 hours worth (Heads Up! 24 hours in 16 bit
stereo at a 44.1KHz sample rate produces a 7GB file! DAMHIK, IJK).

If your main interest is checking for extreme voltage excursions rather
than high voltage spikes. adjust the recording level on the 50Hz
fundamental to -10dB FSD otherwise choose -20dB FSD if you're looking for
high voltage spikes (sub millisecond transients). A -10dB setting allows
you to see overvolting events just in excess of +200% and a -20dB setting
will let you identify spikes as large as 2.4KV without clipping (assuming
the transformer assembly can deal with such spikes).

A quick 'n' dirty way to check whether there were any 'events' worth
zooming in on, is to ask the audio app to calculate a normalisation
factor for the whole period. A normalisation factor of 9dB on a recording
level of -10dB FSD represents a 12.2% voltage boost. OTOH, a
normalisation boost value of 9.172dB represents a 10% overvolting event
on a -10dB recorded level.

HTH & GL!

--
Johnny B Good