[smarbaga]

so i thought i would try using the transformers from microwave ovens.
they put out the power just fine.
the primary gets the transformer real hot ( fry eggs) 250 degrees f.
this would keep the moisture out ( lol )
perhaps these transformers are ment only to be on for short periods of time say 3-4 minuits.
haven't checked the primary amps yet. but thats not adjustable.
i think if i can knock the primary voltage down about 10 or 20 volts they would run a lot cooler. (perhaps a single diode (120v x 0.707)
any ideas, sugestions, experiences with these beauties, as they are a dime a dozen.
thanks all.... :wave:

[gar]

050607-0715 EST USA

smarbaga:

Get a suitable Variac or equivalent, and ac ammeter. Primary only, no load on secondary.

Measure and plot current vs voltage from about 90 to 130 volts. Maybe start lower if there is an internal short. As you raise voltage stop and remove voltage if current rises rapidly.

A nominal maximum hot spot ( meaning internal, not the surface ) temperature should be less than 100 deg centigrade for 105 deg C wire. If higher temperature wire (meaning insulation rating) was used in the transformer design the design point would be higher.

You could be going heavily into core saturation ( should not be ) or have shorted turns.

Certainly the unloaded input current should not approach full load current.


Report back.

.

[smarbaga]

i have hacked the secondary off of 2 different microwave transformers and plugged them in to 120 vac, they both get real hot with or without a secondary winding, with no load, are these transformers suppose to be on for short periods of time only ?
the microwave ovens were working when i removed the tansformers.

[Al_The_Man]

If you can get hold of a clamp-on amp-meter and take a reading, it should show wether the current is high, there is usually certainly enough iron mass that sufficient inductance should not be a problem, there are others here that have used micro-wave transformers without indicating they had a heat problem.
I have seen articles though that some have removed the welded magnetic shunts on the side of the core as this apparently reduces the heating effect.
Al

[Al_The_Man]

I looked into this further and found that actually the magnetic shunts, which are found in a lot of high voltage transformers, actually act to increase the inductive reactance on the primary, so in removing them, it has been known to trip the supply breakers. It is just the the shunts offer poor regulation as far as the secondary is concerned, and might be the reason some have removed them.
It would appear because of this feature, a microwave transformers is not a good choice for a power supply.
Al.

[Al_The_Man]

After further checking it would appear to me that although there is alot of people using Microwave transformers, there are some drawbacks.
A microwave oven transformer has built in current limiting in the form of magnetic shunts between the primary and secondary, they are metalic inserts between the primary and secondary windings and some can be punched out and some are welded in, which offers a bit more of a challenge.
This way the transformer can be used as a normal transformer.
With the shunts in place if the secondary current exceeds a limit the flux is diverted back to the primary.
Also from some of the posts here, it would seem that the typical MW transformer is 1 turn/volt, this is compared to the regular power supply square lamination transformer of 4 to 5 turns/volt and 2 turns/volt of the more efficient Toroidal type.
This seems to indicate a low inductance to resistance ratio, This may explain why some have mentioned that their rewound transformer runs hot with no load.
AL.

[smarbaga]

i have not checked the primary resistance or the primary current draw yet on these microwave transformers (sanyo and ge ) but looking at the size of the wire i think they are probably sucking the amps (my guess is 10 to 12) thus heating the core. with no secondary its just an inductive reactor, i think 2 of them in series with long power runs would make good ones. i just had visitors for 2 weeks (inlaws) and had to put all this crap away but i will amp the things out tomorrow.also put a diode in line which should ( might) give me a quasi square wave input of .707 of the ac supply voltage, thus dropping the current.
if this is the case i will put a secondary winding on and begin to load it to see what the outcome will be.

[ViperTX]

Where is Bubba....he has a good handle on this.....Bubba....Bubba...where are ya??

[smarbaga]

- the ge microwave transformer draws 12.5 ams when plugged into the wall, thats with no secondary winding.
- the sanyo with its untouched secondary still on the core draws 12.3 amps.
no wonder these things get so hot.
looks like about a 16 gauge wire on the primary
- guess they are only ment to be on for short periods at a time
- when i put a diode in series with the primary the current draw is the same.
i really thought that it would be x.707 less than the direct connection as the diode is only letting half the wave pass.
but the meter doesn't lie and the transformer still hummmmmmzzzz.

[Al_The_Man]

You should not use DC into the transformer as the Value of the inductive reactance (in Ohms) is greater when the current is alternating, all you have with a diode is pulsating DC .
The reason I believe that you saw a high current on AC is as I mentioned before, is the low Inductance/per turn of the winding. The greater the amount of inductance (L), the greater the opposition to current.
This opposition of current in an inductor by alternating current cannot be called resistance, The name given to it is inductive reactance, and once you add resistance to the formula the opposition is called Impedance (ohms).
The desirable effect in an inductor is to have zero or low resistance but high inductive reactance.
The bottom line is It looks like Microwave transformers are a poor choice for general power supplies.
Al.

[gar]

050619-1923 EST USA

smarbaga:

All ferromagnetic materials have the characteristic that as you increase the magnetizing force that the flux density falls away from a linear curve. Ultimately the flux density produced from the magnetizing force is no greater than if the magnetic material were removed and only air was present. The relative permeability is the permeability of any material relative to air. Thus the relative permeability of air is 1. For a discussion see "Electric and Magnetic Fields" by Stephen S. Attwood, 1949, p204. Also see chapter 13 p325 on Ferromagnetism.

There will be a hystersis curve for the magnetic core material. As the magnetizing force incresaes (this is the volt time integral of the input voltage) the incremental permeability decreases and the magnetizing current has to increase. When the incremental peremability drops appreciably this is called saturation. For some materials this change is slow and for others it is very sharp. Most cheap magnetic material will have a moderately soft hysteresis curve.

Given that you have a particular transformer, a fixed frequency sine wave excitation, then as you increase the applied voltage the input current with no load on the secondary will gradually increase and as you go into saturation the exitation current will increase rapidly. The more square the hysteresis curve the more sharp is this change in excitation current.

If you have access to a Variac you can run an experiment and plot input current to the transformer vs input voltage. As mentioned in a previous post. If there are no shorted turns or other load, then you will see a gradual increase in current with respect to voltage. As the voltage gets higher you will see a more rapid increase, and it gets worse the higher the voltage. A portion of the unloaded excitation current is called the magnetizing current.

In the design of a transformer or other ferromagnetic device in an ac application there is a trade off between a number of factors. One of those is cost. If you minimize iron and copper in making a transformer (lowers cost), then the magnetizing current will be higher. It appears this has been done in your microwave transformers. So you have a high magnetizing current which produces a lot of heat, but microwaves ovens are on only for a short time, and the transformer has a moderately long thermal time constant so the maximum hot spot stays within limit, but would not under continuous duty at full load.

I have run tests on two different small transformers. These both have a nominal 115 to 120 V 60 Hz rating. The Osborne 31537 is about a 500 VA unit manufactured about 30 years ago, the Signal Transformer is A41-175-24 which is 175 VA output and was manufactured in the last 10 years. I used a Powerstat variable transformer using its name plate calibration for setting voltage. These test points were 0 thru 140 in 10 V increments. The current was measured with a Fluke 87III so called True RMS multimeter.

Following is the raw data in Volts and Amperes:

Voltage......Osborne I......Signal I

---0--------0.000---------0.000
--10--------0.010---------0.002
--20--------0.015---------0.004
--30--------0.020---------0.007
--40--------0.028---------0.012
--50--------0.035---------0.015
--60--------0.043---------0.018
--70--------0.055---------0.023
--80--------0.072---------0.030
--90--------0.104---------0.048
-100--------0.180---------0.097
-110--------0.285---------0.159
-115----------------------0.207
-120--------0.441---------0.252
-130--------0.647---------0.375
-140--------0.942---------0.532

Next I normalized these to 1.000 at 120 V. The values follow:

Voltage......Osborne I......Signal I

---0--------0.000---------0.000
--10--------0.022---------0.007
--20--------0.034---------0.016
--30--------0.045---------0.027
--40--------0.063---------0.047
--50--------0.079---------0.059
--60--------0.097---------0.071
--70--------0.124---------0.091
--80--------0.163---------0.119
--90--------0.235---------0.190
-100--------0.408---------0.384
-110--------0.646---------0.630
-115----------------------0.821
-120--------1.000---------1.000
-130--------1.467---------1.488
-140--------2.136---------2.111

Note the extreme closeness of the two normalized curves even though the transformers were made years apart by two different manufacturers and are different sizes.

Plot this data on linear graph paper and it emphatically shows the saturation effect. When graphed the curves look even closer than looking at the raw data.

The specifications for the signal transformer are 115 V 60 Hz input, and 28 V @ 6.25 A output. So output is 175 VA. At the input this is 1.52 A. Thus, the magnetizing current is 0.207/1.520 = 0.14 or 14% of the full load current excluding the magnetizing current. The percentage is somewhat less on the larger Osborne transformer.

This information is valuable for anyone on any transformer because the general characteristics are the same.

In the early 50s Attwood was head of Electrical Engineering Department at the University of Michigan. His book or a similar one is too technical for most of you but there are useful elements to you in there. On p333 is a graph showing magnetizion curves for various materials.

Hopefully I have given you some insight into the excitation current aspect of transformers or similar devices.

The microwave oven transformer might work well down around 80 to 100 V input if the problem is not a shorted turn.

.

[smarbaga]

ty gar for the perdinant info and data on the flux density vs voltage of transformers.
and thank you all for your inputs in this microwave transformer matter,
if there was a cheap way to get 80v ac from the wall i would try it.
i did think that a single diode would work. but when i put the diode on 1 side it blew my 15 amp breaker.
i don't want a 24v dc power supply at say 10 amps (240 watts) thats going to cost me 12.5 amps (1440 watts) of electricity on my electric bill anyways.
i am just a little crazy, but not all the way nuts.....

[Al_The_Man]

I would be more interested in seeing a comparison of primary inductance, (secondary open) for a microwave transformer and a 'regular' power transformer of equal core dimensions.
Al

[gar]

050620-0549 EST USA

Al:

Because the inductance of an iron core coil is non-linear you will get different answers depending upon the excitation voltage when the measurement is made.

If the same core material is used in both transformers you compare and the core shape and area are the same, then at low excitation levels the transformer with fewer turns on the primary will have a lower inductance.

In a transformer that is overdriven well into saturation the primary heating is coming from the highly peaked current occuring from saturation and near the voltage zero crossing.

Connect a scope to simultaneously display the input voltage and current, and synchronize on the voltage waveform. Now increase the voltage and watch that peaking increase as you go into saturation.

To determine where a safe input voltage is for a transformer with standard core material the volt-amp curve is probably most useful means. Then use a criteria that the excitation current under no load should be about 10% of full load. Larger transformers would want to work at a higher efficiency than this.

If you compare coils with different core materials, then inductance may be of no value for comparison.

There are some core materials with low permeability at low excitation, then increase to a high level before dropping down in saturation.

.

[gar]

050621-0801 EST USA

Added a load resistor to the Signal transformer. The secondaries are connecfted in series and the load resistor is 10.0 ohms. Now the data looks like:

Following is the raw data in Volts and Amperes:

Voltage......Osborne I......Signal I.........Signal + 10 ohm load

---0--------0.000---------0.000---------0.000
--10--------0.010---------0.002---------0.043
--20--------0.015---------0.004---------0.102
--30--------0.020---------0.007---------0.155
--40--------0.028---------0.012---------0.210
--50--------0.035---------0.015---------0.271
--60--------0.043---------0.018---------0.329
--70--------0.055---------0.023---------0.385
--80--------0.072---------0.030---------0.447
--90--------0.104---------0.048---------0.505
-100--------0.180---------0.097---------0.570
-110--------0.285---------0.159---------0.636
-115----------------------0.207
-120--------0.441---------0.252---------0.714
-130--------0.647---------0.375---------0.812
-140--------0.942---------0.532---------0.920

Next I normalized these to 1.000 at 120 V. The values follow:

Voltage......Osborne I......Signal I.........Signal + 10 ohm load

---0--------0.000---------0.000---------0.000
--10--------0.022---------0.007---------0.170
--20--------0.034---------0.016---------0.400
--30--------0.045---------0.027---------0.620
--40--------0.063---------0.047---------0.830
--50--------0.079---------0.059---------1.070
--60--------0.097---------0.071---------1.310
--70--------0.124---------0.091---------1.530
--80--------0.163---------0.119---------1.770
--90--------0.235---------0.190---------2.000
-100--------0.408---------0.384---------2.260
-110--------0.646---------0.630---------2.520
-115----------------------0.821
-120--------1.000---------1.000---------2.830
-130--------1.467---------1.488---------3.220
-140--------2.136---------2.111---------3.650

Notice how the load resistor straightens the curve.

At 120 vac in the secondary has an output voltgae of 27.46 volts with the 10.0 ohm load. This is 75 watts or about 1/2 the rating of the transformer.

Note also that the VA input to Signal at 120 vac no load is about 30 VA. Most of this is reactive current. With the 10 ohm load the VA input is about 85 VA.

[gar]

050622-0623 EST USA

Following is the previous data plotted:

Also at 120 V in with the 10 ohm load the transformer is about 95% efficient.

[gar]

050622-1100 EST USA

smarbaga:

Is there any way you can get a Variac so you could run a current vs voltage experiment?

Lacking that with no secondary load connect the primaries of two identical transformers in series across the 120 line and measure the current. Also measure the voltage across each primary.

.

[smarbaga]

gar
- the graph is very interesting, thank you, i am guessing that u (oops) you are a school teacher
- my variac only goes up to 8 amps and it looks like these (oops again) those) transformers will draw to many primary amps for what i need it for anyways
- my target voltage is around 25vac @ 5 amps.
- so i am guessing almost 2 amps input @ 120vac
- perhaps i will make a car booster/batery charger out of them
- maybe i can find 3, 200 watt dohnuts on e_bay
thanks again, i still think the single diode thing should work tho.

[Al_The_Man]

i still think the single diode thing should work tho.

The reason that you probabally blew the fuse with a rectifier, it is not desirable to pass DC through the transformer winding, as apart from removing the AC component, DC has the effect of lowering the impedance, there used to be a use for this as in magnetic amplifiers, that basically had a primary, a secondary and a third winding that was fed with pure DC, As the DC current was increased, this lowered the ability to transfer energy from primary to secondary, hence amplifier effect.
Going back to the microwave transformer with magnetic shunts, a similar arrangement is used in stick welding transformers, where it would be dangerous to allow full current to flow with a shorted stick condition.
The magnetic shunt has the effect you would have if you had an increasing series impedance in the primary as the current increases, this reduces the ability to transfer energy to the secondary as current increases.
Al.

[gar]

050622-1335 est usa

smarbaga:

Not a school teacher.

Your 8 amp Variac is fine to use for the experiment. You will not need to exceed its limit because a lot of information can be obtained from a curve up to that point.

I believe that you said that you had removed the secondary on one of the transformers. This means there can not be a shorted turn(s) from the secondary. If there were a single shorted turn on the primary it would not produce the very high no load current you see. It would take a substantial number primary turns being shorted to alter substantially the number of primary turns and also produce a shorting load to get this high primary current.

My current impression is that the microwave transformer designers are running far into saturation to lower their cost. But if you can run the experiment on your transformer, then maybe we can get a better idea of what is happening.

Also anybody else that has used a microwave transformer might provide some feedback.

Generally with enough information you can get good correlation between theory and practice.

A magnetic shunt should not lower the primary inductance, but it will increase the leakage inductance between primary and secondary. That is why it helps limit output current.

In the equivalent circuit for a transformer there is a series inductance between primary and secondary and it is this that is called the leakage inductance. It gets its name from the fact that all the flux lines do not encompass both primary and secondary. Those that do not encompass are leakage.

.