Just scrapped another m.wave. Wasn't operating properly, was old, the
display was faulty and what was showing was meaningless. Also
sometimes it had started to microwave for no reason at all when the
door was closed. Was afraid it might become hazardous, e.g. operate
with door faulty or open!
So kept a few parts etc. including the heavy 115 to Hi-voltage mains
transformer.
But could someone remind me, again, why a m.wave oven transformer has
those 'shunts'  welded across the laminations. There is presumably
some DC (or pulses of AC) flowing in those windings?
Whereas audio AC transformers have non-magnetic separators between the
section of laminations to avoid saturation?
Does this mean that the transformer goes into magnetic saturation each
cycle or something? If so there might be magnetic losses and
generation of heat?
Thanks for any advice. TIA.

terryS wrote:
> Just scrapped another m.wave. Wasn't operating properly, was old, the
> display was faulty and what was showing was meaningless. Also
> sometimes it had started to microwave for no reason at all when the
> door was closed. Was afraid it might become hazardous, e.g. operate
> with door faulty or open!
> So kept a few parts etc. including the heavy 115 to Hi-voltage mains
> transformer.
> But could someone remind me, again, why a m.wave oven transformer has
> those 'shunts'  welded across the laminations. There is presumably
> some DC (or pulses of AC) flowing in those windings?
> Whereas audio AC transformers have non-magnetic separators between the
> section of laminations to avoid saturation?
> Does this mean that the transformer goes into magnetic saturation each
> cycle or something? If so there might be magnetic losses and
> generation of heat?

See:  http://wiki.4hv.org/index.php/Microwave_oven_transformer

"As it is intended to drive a capacitive load, the leakage inductance of 
the tranformer is deliberately increased by adding a small magnetic 
shunt between the primary and secondary coils. The inductance is roughly 
equal and opposite to the doubler capacitance, and so reduces the output 
impedance of the doubler. This specified leakage inductance classifies 
the transformer as non-ideal."

BE CAREFULL!!!

-- 
Virg Wall, P.E.

"terryS"
>
> But could someone remind me, again, why a m.wave oven transformer has
> those 'shunts'  welded across the laminations.

** As another poster says, it increases the leakage inductance.

> There is presumably
> some DC (or pulses of AC) flowing in those windings?

** There is no DC current and only fairly smooth AC.


> Whereas audio AC transformers have non-magnetic separators between the
> section of laminations to avoid saturation?

** Transformer lams are normally stacked as close and tight as possible, 
plus interleaved to avoid any air gaps.

Only inductors and single ended (valve) output transformers have deliberate 
air gaps in the cores.


> Does this mean that the transformer goes into magnetic saturation each
> cycle or something? If so there might be magnetic losses and
> generation of heat?

** Oven transformers operate under full load whenever the AC power is 
applied and have a fan to keep them cool.

 If operated with no load, the core will be driven into saturation.



.....  Phil

On Oct 28, 8:53=A0pm, "Phil Allison" <phi...@tpg.com.au> wrote:
> "terryS"
>
>
>
> > But could someone remind me, again, why a m.wave oven transformer has
> > those 'shunts' =A0welded across the laminations.
>
> ** As another poster says, it increases the leakage inductance.
>
> > There is presumably
> > some DC (or pulses of AC) flowing in those windings?
>
> ** There is no DC current and only fairly smooth AC.
>
> > Whereas audio AC transformers have non-magnetic separators between the
> > section of laminations to avoid saturation?
>
> ** Transformer lams are normally stacked as close and tight as possible,
> plus interleaved to avoid any air gaps.
>
> Only inductors and single ended (valve) output transformers have delibera=
te
> air gaps in the cores.
>
> > Does this mean that the transformer goes into magnetic saturation each
> > cycle or something? If so there might be magnetic losses and
> > generation of heat?
>
> ** Oven transformers operate under full load whenever the AC power is
> applied and have a fan to keep them cool.
>
> =A0If operated with no load, the core will be driven into saturation.
>
> ..... =A0Phil

Thanks much for the answers.
Yes the 'load' is some sort of single diode voltage doubling circuit
using, in this case, a 2500 volt 1 mfd capacitor by look of it.
So the magnetron presumably gets half cycle pulses of Hi Voltage DC?
When the 'other' high voltage half cycle is in series with the
discharging capacitor?
Yup will be very careful, have worked on high voltage transmitters
etc. and well aware of high voltage and up to kilowatt of RF
radiation. Yes there is/was or is usually a fan from which some of the
air is aimed at the transformer.
magnetic blocks between sections. Yes was thinking of single ended non-
push/pull audio transformers carrying for example some 20- 50
milliamps of DC plate current to a single pentode.
But come to think of it typical laminations are interleaved etc. to
obtain a good 'magnetic' circuit, right?
But are not laminations frequently insulated 'electrically' from each
other to reduce 'hysteresis' (induced currents?) through the metal
core?
Or am I confusing things here?
The term 'non-ideal' rings a faint bell from some almost 60 years ago,
so I'll think about that 'leakage inductance'  term. I only spent a
few days in the transformer department! They were just starting to use
'C' core transformers mainly for 50 cycle AC power IIRC.

"terryS"
 "Phil Allison"
>
> > But could someone remind me, again, why a m.wave oven transformer has
> > those 'shunts' welded across the laminations.
>
> ** As another poster says, it increases the leakage inductance.
>
> > There is presumably
> > some DC (or pulses of AC) flowing in those windings?
>
> ** There is no DC current and only fairly smooth AC.
>
> > Whereas audio AC transformers have non-magnetic separators between the
> > section of laminations to avoid saturation?
>
> ** Transformer lams are normally stacked as close and tight as possible,
> plus interleaved to avoid any air gaps.
>
> Only inductors and single ended (valve) output transformers have 
> deliberate
> air gaps in the cores.
>
> > Does this mean that the transformer goes into magnetic saturation each
> > cycle or something? If so there might be magnetic losses and
> > generation of heat?
>
> ** Oven transformers operate under full load whenever the AC power is
> applied and have a fan to keep them cool.
>
> If operated with no load, the core will be driven into saturation.
>

Thanks much for the answers.
Yes the 'load' is some sort of single diode voltage doubling circuit
using, in this case, a 2500 volt 1 mfd capacitor by look of it.

** The other "diode" is the magnetron itself.

So the magnetron presumably gets half cycle pulses of Hi Voltage DC?
When the 'other' high voltage half cycle is in series with the
discharging capacitor?

**  Yep.

But are not laminations frequently insulated 'electrically' from each
other to reduce 'hysteresis' (induced currents?) through the metal
core?

** There is a coating on the lams that provides adequate insulation for that 
purpose.

The term 'non-ideal' rings a faint bell from some almost 60 years ago,

** No transformer is " ideal "  - the closest among AC supply types is the 
toroidal.

so I'll think about that 'leakage inductance'  term.


** Silly name for a simple thing -  the simple fact that the coupling from 
primary to secondary diminishes at high frequencies just as if there was an 
inductor in series with the secondary.

With small E-core step-down types, that inductor might be 0.5mH.

With the large step-up job used in ovens the number is many Henries.



.....  Phil

On Thu, 29 Oct 2009 09:22:58 -0700 (PDT), terryS
<t...@nf.sympatico.ca> wrote:

>But come to think of it typical laminations are interleaved etc. to
>obtain a good 'magnetic' circuit, right?
>But are not laminations frequently insulated 'electrically' from each
>other to reduce 'hysteresis' (induced currents?) through the metal
>core?

Audio Xformers have the E's and I's going in the same direction, with
some paper to establish the gap width between the two.  

Power transformers are interleaved 1X1 or 2X2 to reduce eddy current
losses.  Lamination thickness also counts - high frequency power and
audio transformers either use very thin laminations or special
"boutique" iron to limit losses in both hysteresis and eddy currents.

The type of iron, and to a lessor extent how it is rolled,  determines
the hysteresis.  The iron should be "relaxed" to work in a
transformer.  You don't want it staying magnetized after the current
is at zero . . .  Rolling the iron sheets aligns the domains,
annealing afterwards relaxes them.

Grain oriented, silicon steel
-- 

"default"
>
> Audio Xformers have the E's and I's going in the same direction, with
> some paper to establish the gap width between the two.

** ONLY true for single ended, tube output transformers  !!!

 They are just like inductors and have to carry unbalanced DC current.

 All other audio transformers are made with interleaved lams closed up as 
tight as possible.



....  Phil

In article <p...@4ax.com>, default wrote:
>On Thu, 29 Oct 2009 09:22:58 -0700 (PDT), terryS
><t...@nf.sympatico.ca> wrote:
>
>>But come to think of it typical laminations are interleaved etc. to
>>obtain a good 'magnetic' circuit, right?
>>But are not laminations frequently insulated 'electrically' from each
>>other to reduce 'hysteresis' (induced currents?) through the metal
>>core?
>
>Audio Xformers have the E's and I's going in the same direction, with
>some paper to establish the gap width between the two.  

  Only ones that need to pass net DC in the primary for single-ended 
applications, most of which involve(d) vacuum tubes.

>Power transformers are interleaved 1X1 or 2X2 to reduce eddy current
>losses.

  That does not affect eddy current loss, but it does reduce magnetizing 
current and copper loss in the primary.

>  Lamination thickness also counts - high frequency power and
>audio transformers either use very thin laminations or special
>"boutique" iron to limit losses in both hysteresis and eddy currents.

  Thickness only affects eddy current losses.

>The type of iron, and to a lessor extent how it is rolled,  determines
>the hysteresis.  The iron should be "relaxed" to work in a
>transformer.  You don't want it staying magnetized after the current
>is at zero . . .  Rolling the iron sheets aligns the domains,
>annealing afterwards relaxes them.
>
>Grain oriented, silicon steel

  That is largely true.

  However, audio transformers needing to work at full power at bass 
frequencies do not need to be less lossy than power transformers.  One 
thing they do often need is windings interleaved in multiple layers to 
reduce leakage inductance, which affects performance at high frequencies.

  The need for such interleaved widnings is reduced if turns count is 
skimped in applications where low frequency performance is compromised, or 
if the application has less requirement for good high frequency 
performance (amplifiers intended mainly for voice applications, AM-only 
radios such as "5 tube" radios, and most tube TV receivers).

  There are audio transformers smaller than the smallest common sizes of 
power transformers.  They are mostly not fully functional at lower audio 
frequencies near and below power line frequencies.

 - Don Klipstein (d...@misty.com)