HomeRoast Digest


Topic: Way OT - Transformer Performance was: Dimmer plus diode caution (2 msgs / 126 lines)
1) From: Jeffrey A. Bertoia
Deward
dewardh wrote:
<Snip>
Thanks
<Snip>
I was not promoting the use of a small power transformer for DC/Pulsed 
applications.  But you can easily
determine it's approximate capabilities.
A 'typical' 60 Hz stacked core, small power transformer is rated for 
exactly two HCs at full voltage.  The
fully derated (at least 50% voltage derating by manufacturer) of any 
power stacked transformer core is
V*(1/f/2) or in the US 120V*(8.33mS) or approximately 1 volt-seconds. 
 This means that you can apply
1 volt for one second to the primary (while the secondary is open) 
without the magnetizing current exceeding
a multiplier of it's rated value (normally 2).
A stacked high permeability core has an exponential volt-second decay 
but due to the long time constant,
on the order of minutes, it does not assist you in limiting the 
saturation.This is the reason that the manufacturer
derates the volt - second rating of the core by atleast 50%.  Consider 
the worst case scenario, the transformer
just completes a positive half cycle is turned off and then is turned 
back on quickly (less than 1 minute) catching
another positive half cycle.  You get some saturation but you probably 
don't exceed maybe 10-20 times the
magnetizing current.  Fortunately this is somewhat self correcting 
because the effect of bouncing off the non-linear
region of the hysterysis curve and applying a half cycle in the other 
direction is to help center the retained flux.
It has been a long time sine I looked at the deep saturation 
characteristics of a stacked core transformer but if
memory serves....  Beyond the 2 half cycle point saturation current 
(current not coupled to the secondary)
increases exponentially.  So at the next HC you have many times the 
rated load current or hundreds of times the
magnetizing current.
For the big invertors that I have designed the load is a large (600V 
1200A primary) wound hypersil steel core
transformer.  These are much less forgiving than stacked core 
transformers....  While writing the frequency
determination algorithms we recorded primary currents of more than 
10,000 Amps after 2 consecutive
half cycles. (:-))
<Snip>
Well I think the earlier discussion shows that without flyback diodes 
and other methods of flux depletion
(BTW - Also required for pulse transformers in half wave applications) 
they are fairly useless for DC.
Phase control is another issue entirely.  Transformers are rated 
primarily three ways:
1 - Flux in the core (see above)
2 - Current in the windings (relates to heat)
3 - Maximum frequency - as frequency goes up magnetizing current goes up 
(relates to heat)
I could probably dig out the fourier analysis but put simply:
Phase control increases the effective frequency of the waveform, this 
translates to more core
losses and more heat.  Therefore the maximum current must be reduced in 
order to maintain
the same rise over ambient at full load.  Generally the 3db point of a 
power transformer is
about 180 Hz.  So about half the power of the third harmonic goes into heat.
If you run the numbers you find out that a 20% current derating is more 
than sufficent in
phase control applications....  This place physics works for you :-) 
 Because as
you decrease the phase delay and the waveform becomes more continuous 
the harmonic
content of the waveform is reduced until at a full sin wave it is of 
course the fundamental.
Simply put as current goes up losses, due to the higher frequency, go down.
<Snip>
You are not getting away with anything.  The MIL Spec
reliability calculations allow usage of these types of calculations
and deratings without reliability penalty and it's really
all about cost and reliability.
jeff
<Snip>
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

2) From: Rick Farris
That's because the dimmer is lopping off equal amounts of both the positive
and negative portions of the waveform.  The important part here is that the
average voltage is zero.
-- Rick


HomeRoast Digest