[Amps] Microwave Oven Autopsy

Wed Mar 15 22:07:54 EST 2006

Actually, the microwave oven transformers are true engineering marvels 
in which the $2 technology was delivered for $1 safely and effectively.  
Most of these transformers are ferroresonant types which rely on 
saturation to regulate the output voltage.  Ferroresonant transformers 
will run quite hot as the core losses are 2 to 4x per pound those of 
high quality linear transformers but they do the one thing that no 
linear transformer can do, deliver a constant voltage with no active 
devices. 

Tomm

Will Matney wrote:

>Marv,
>
>I can see what they're doing, probably running the transformer and tube maybe at 25% duty cycle by pulsing it. The thing is, even doing this, the transformer is going into saturation most likely. Hipersil or M-6 states that it can be ran up to 17 kilogauss (max), but they recommend 15 kilogauss. The reason being is that after 15 kg, the magnetizing current starts to raise rapidly, and that means the permeability will go to crap. That goes into what they call incremental permeability when the straight line starts to bend over into a knee and at some point in higher flux density, it will top out and begin a sharp fall. At the top of course is the highest permeability that can be reached but is well out of the linear region of either it or flux density. When they do the test to obtain a BH curve using a hysteresis loop tracer (The BH curve is inside 1/2 the hysteresis loop and divides it in half), and a scope, the curve pattern will go up linearly, then bend to a knee, keep goin
> g to a point until it takes a sharp nose dive back to the start then form the negative portion of the waveform. That sharp point is saturation. Believe it or not, it is around 20 kilogauss at this tip on a lot of material, but the permeability has fallen off so low that the primary looks almost like a short (but not quite enough current to be one) to the line because of the very high magnetizing current. They rate material at a point on down the curve, but above the start of the upper knee for maximum flux density. Also, design says your better to run a transformer at a flux density just below the knee. That is the linear region, and is just below where maximum permeability is, and that's below the curve where incremental permeability is. To find maximum permeability on a BH curve, you look and see there's two knees. There's a small lower knee where the curve starts, then almost a straight leaning line, and at last the line forms into a curve again being the upper knee. If y
> ou draw a straight line from the point where the BH curve starts (where the lower knee begins at 0), and go towards the upper knee with it until it becomes tangent with the curve, the tangent point is maximum permeability. What they've done is throw the book out the window over cost in my opinion. If they would have designed it to operate normally, a mans power bill would be less too!
>
>Best,
>
>Will
>
>
>*********** REPLY SEPARATOR  ***********
>
>On 3/14/06 at 8:35 PM wc6w at juno.com wrote:
>
>  
>
>>Hi Will,
>> Not certain but, I believe 1KW.  It had a 1uF cap which is typical for
>>that rating.  
>>
>> I didn't see a model number anywhere on the frame which was returned to
>>the "alley" after the autopsy; Then, claimed by the Morlocks.
>>
>>73 & Good afternoon,
>> Marv WC6W 
>>
>>
>>-- "Will Matney" <craxd1 at verizon.net> wrote:
>>Marv,
>>
>>20 kilogauss is even over the maximum flux density of Hipersil or M-6
>>which is 17 and recommends 15. M-2, the best is still 17 kg. I'd say that
>>transformer is running into stauration under full load, and runs hot as a
>>firecracker. In microwave ovens, even on high power, the tube is on for
>>maybe a minute, than off so long, and then on again. The lower powers,
>>this delay is even slower. I've noticed this while mine has been heating a
>>meal, listening to it kick on and off over the fan noise. Since they're
>>pulsing the power, they're giving the transformer time to cool down a
>>little with help from the fan. A 1-3/8 x 2-3/8 core is only 3-1/4 square
>>inches and is way to small for 1 KVA, maybe 500 watts at the most. What
>>size of mocrowave was this in watts?
>>
>>Best,
>>
>>Will
>>
>>*********** REPLY SEPARATOR  ***********
>>
>>On 3/13/06 at 6:32 PM wc6w at juno.com wrote:
>>
>>    
>>
>>>Hi Will,
>>> No holes in the lams except the one for the secondary termination rivet.
>>>
>>> The numbers are worse than you imagined.
>>>
>>> The core cross section is 3.5cm x 6 cm.
>>>
>>> The primary is wound with (approx.) 103 turns #14;  .33 ohms DCR, 34 mHy
>>>inductance.  Even with the mag. shunts removed, it takes quite a bit of
>>>idle current.
>>>
>>> The secondary is wound with #27;  95 ohms DCR.
>>>
>>> The above numbers figure out to about 40 Watts coppper loss, and over
>>>20K gauss in the core which is likely another 40 Watts loss at 1KW.  Eeek!
>>>
>>> Inspires a new meaning for ICAS:  Inappropriate Consumer Appliance
>>>Service.
>>>
>>> I think if I was going to rewind one for something I'd add at least 20
>>>turns to the primary and only run the thing at 750 watts or so to stay
>>>within "amateur" ICAS.
>>>
>>>73 & Good morning,
>>> Marv WC6W
>>>
>>>-- "Will Matney" <craxd1 at verizon.net> wrote:
>>>Marv,
>>>
>>>Well, they're getting cheaper, no screws now! Does the lams still have the
>>>screw holes or are they solid? A 1/8" fillet weld wouldnt be too hard to
>>>cut out. The problem with welds here is causing eddy currents that are
>>>higher than normal thus running the losses up.  At one tune per volt and
>>>about 12 kilogauss, it would use about a 5 square inch core and be rated
>>>at about 1KVA or just under. I'd say though they're running it higher than
>>>12 kilogauss, maybe 15-16. The easiest way to tell the turns per volt is
>>>if there is any room between the coil and the cores outside legs, snake in
>>>some wire, and do a 10 turn winding on the outside of the coil. Any size
>>>small wire will do. Then, power up the primary and read the voltage of the
>>>10 turn coil. Take that reading and use it to figure the turns per volt
>>>with the line voltage. I use 10 turns so it's easier to figure it math
>>>wise. At least on this one, the secondary connection to the core was
>>>easily cut, some aren't. The shunt can be 
>>>driven out with a hammer and a block of wood or a punch. The shunt was
>>>used as a current limiter type of arrangement if I recall, or something to
>>>that effect. On a power transformer for an amp, it would not be needed or
>>>wanted. It's a shame how they've cheapend everything to cut manufacturing
>>>costs using welding on a core.
>>>
>>>Best,
>>>
>>>Will
>>>
>>>
>>>*********** REPLY SEPARATOR  ***********
>>>
>>>On 3/13/06 at 12:50 AM wc6w at juno.com wrote:
>>>
>>>      
>>>
>>>>Hi Will,
>>>>  This one has four 1/8" wide welds down the opposing sides.  And yes,
>>>>        
>>>>
>>it
>>    
>>
>>>>has the shunts but, they are encased in transformer paper so, I'd guess
>>>>that they could be easily driven out without disturbing anything else. 
>>>>        
>>>>
>>No
>>    
>>
>>>>screws what so ever.  The lams are even welded to the mounting frame in
>>>>four spots.  The "cold" end of the secondary was soldered to a lug which
>>>>was riveted to the lams -- I necessarily disconnected that wire for the
>>>>hi-pot check.
>>>>
>>>>  It looks like it's wound at about 1 turn per volt.   As the core
>>>>wouldn't practically dissassmble, the only easy rewind would be for a
>>>>filament transformer.  The secondary could be sliced out by a careful guy
>>>>and rewound with a bundle of large wires in parallel making a KW filament
>>>>transformer... for say a 4CX5000/10000?   Or twenty 813's?  :-)
>>>>
>>>>73 & Good afternoon,
>>>> Marv WC6W
>>>>
>>>>**********************************
>>>>
>>>>
>>>>-- "Will Matney" <craxd1 at verizon.net> wrote:
>>>>Marv,
>>>>
>>>>Did you check to see if one of the windings was connected to the core,
>>>>        
>>>>
>>and
>>    
>>
>>>>if the core had a shunt in it? That is what most have that I've seen.
>>>>Another thing that really ticks me about their manufacture, and a few
>>>>other transformers too, is they weld the lams together. They put a weld
>>>>right down one side of the lams, front to back, one bead about 1/4" wide.
>>>>That makes them a bit*h to use for a rewind. A person with a mill, or be
>>>>good with a disc grinder can remove the weld. Also, the shunt needs to be
>>>>driven out with a hammer. This is really good though as it gives you more
>>>>vertical window area for a new coil. What gets me though is why weld them
>>>>and still use screws to hold the lams together?
>>>>
>>>>Best,
>>>>
>>>>Will
>>>>
>>>>*********** REPLY SEPARATOR  ***********
>>>>
>>>>On 3/12/06 at 11:57 PM wc6w at juno.com wrote:
>>>>
>>>>        
>>>>
>>>>>QST
>>>>> Yesterday, I acquired a 2004 vintage Sharp microwave oven, from the
>>>>>"alley exchange", which was light (weightwise) enough that I thought it
>>>>>might contain a switching power supply.
>>>>>
>>>>> Upon examining the unit, I found that the incredible lightness was due
>>>>>to its construction with frame metal of soda can thickness, and that it
>>>>>contained a conventional transformer manufactured by one Digital Power
>>>>>Communications Co, Ltd.  
>>>>>
>>>>> There was an article in QEX about 10 years ago that discussed the use
>>>>>          
>>>>>
>>>of
>>>      
>>>
>>>>>these microwave oven transformers in plate power supplies.  It advised
>>>>>against using them in a conventional full wave bridge but, rather used
>>>>>them in an unorthodox half wave connection.
>>>>>
>>>>> Examining the transformer at hand, it appeared that the insulation was
>>>>>uniform over the full length of the secondary.
>>>>>
>>>>> I disconnected the "low side" secondary lead from the frame and
>>>>>hi-potted the secondary (also the primary just for science...) up to
>>>>>          
>>>>>
>>4KV,
>>    
>>
>>>>>referenced to the frame, with zero leakage.  
>>>>>
>>>>> This suggests that one of these transformers might be employed in a
>>>>>"normal" connection for a low voltage (2500V) KW input amplifier.   Or a
>>>>>pair of identical units in parallel for a 2KW input amp.   
>>>>>
>>>>> Caution, this suggestion MAY NOT hold true for other makers of this
>>>>>style of transformer.  
>>>>>
>>>>>73 & Good afternoon,
>>>>>  Marv WC6W  
>>>>>
>>>>>
>>>>>P.S. -- The magnets from the magnetron assembly make great refrigerator
>>>>>magnets!
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>*
>>>>>
>>>>>
>>>>>________________________________________________________________________
>>>>>Try Juno Platinum for Free! Then, only $9.95/month!
>>>>>Unlimited Internet Access with 1GB of Email Storage.
>>>>>Visit http://www.juno.com/value to sign up today!
>>>>>
>>>>>
>>>>>_______________________________________________
>>>>>Amps mailing list
>>>>>Amps at contesting.com
>>>>>http://lists.contesting.com/mailman/listinfo/amps
>>>>>          
>>>>>
>>>>
>>>>_______________________________________________
>>>>Amps mailing list
>>>>Amps at contesting.com
>>>>http://lists.contesting.com/mailman/listinfo/amps
>>>>
>>>>
>>>>
>>>>________________________________________________________________________
>>>>Try Juno Platinum for Free! Then, only $9.95/month!
>>>>Unlimited Internet Access with 1GB of Email Storage.
>>>>Visit http://www.juno.com/value to sign up today!
>>>>        
>>>>
>>>
>>>_______________________________________________
>>>Amps mailing list
>>>Amps at contesting.com
>>>http://lists.contesting.com/mailman/listinfo/amps
>>>
>>>
>>>
>>>________________________________________________________________________
>>>Try Juno Platinum for Free! Then, only $9.95/month!
>>>Unlimited Internet Access with 1GB of Email Storage.
>>>Visit http://www.juno.com/value to sign up today!
>>>      
>>>
>>
>>_______________________________________________
>>Amps mailing list
>>Amps at contesting.com
>>http://lists.contesting.com/mailman/listinfo/amps
>>
>>
>>
>>________________________________________________________________________
>>Try Juno Platinum for Free! Then, only $9.95/month!
>>Unlimited Internet Access with 1GB of Email Storage.
>>Visit http://www.juno.com/value to sign up today!
>>    
>>
>
>
>
>_______________________________________________
>Amps mailing list
>Amps at contesting.com
>http://lists.contesting.com/mailman/listinfo/amps
>
>  
>