Borislav,
Actually, to demagnetize one, you should be able to do it by just applying line
voltage to the primary. It is possible for any transformer to get magnetized
according to which part of a cycle the current was removed from the primary.
According to what I've read, using half wave rectifiers can also produce this
at times. I always recommend using a full wave bridge for any rectification.
When a transformers core becomes magnetized, the next time voltage is applied
to the primary, the inrush current will be a good bit higher than normal until
the cycle switches and the core levels out. Generally, that's all it takes to
cure the problem.
It could be a problem about the frequency and the meter your using, I'm not
sure. Really, running 200 Hz on a 60 Hz core will give you way more iron than
needed. At 200 Hz, the core would probably be 1/4 the size. The core would
probably give a higher inductance reading than it would at 60 Hz. To be honest
though, that should not vary that much between the two, linear or not. The
reason being is that 60 Hz would be at the top of the BH curve right where it
starts from a line to a curve where the permeability is the highest (the way
they first designed it for max efficiency). 200 Hz would be much lower down on
the straight linear portion of the line where the lower flux densities are.
Using a core for 60 hz at 200 Hz then means that the flux density will be very
much lower because of the increased amount of iron.
To check for a shorted shield, just measure the primary current of each
transformer. If one is higher than another, then something is wrong (something
drawing more current than the other). The measurement needs to be made with all
the secondary leads unconnected. Each may avry a little amount normally, but
not by a large amount. The shields current draw from the primary (if shorted)
can be calculated by wattage in = wattage out x 1.10 for a 90% efficiency. This
efficiency can change from say around 70 to 90 %. It's according the the iron
type used in the lams.
Best,
Will
*********** REPLY SEPARATOR ***********
On 6/30/06 at 11:08 PM Borislav Trifonov wrote:
>The thing is, these were initially identical transformers, I just took
>them apart and replaced the secondaries. There's no difference between
>the cores.
>
>Since they do draw the same current from the mains when secondaries are
>open, I guess can't be the shield shorted.
>
>I don't see how gapping could be the issue, since the laminations are
>interleaved, Es and Is switching place every other layer; the gapping
>varies a bit this way due to me manually reassembling them, but again,
>putting a squeeze in the vice makes very little difference, 5% not 100%.
>
>I'm interested in your partially magnetized core suggestion. How can I
>check for that, or demagnetize it?
>
>
>Will Matney wrote:
>> Borislav,
>>
>> I know you've answered some of the following questions below, but I'm
>adding all I can think of to help. Does both transformers have the same
>amount of iron in them? Is the laminations the same size? Are the
>laminations the same thickness? Are you sure both have the same number of
>turns in the secondary? Is the iron type different between the two
>possibly (this really shouldn't make a huge difference in the inductance
>though unless one has something added in the iron which impedes the
>magnetic flux adding more reluctance).? Transformer iron doesn't vary as
>much as does ferrite, or iron powder where you have different AL factors
>unless it may be something like Permedure, etc. Could one core be
>partially magnetized? If the shield were shorted of course you would have
>a low voltage, high current short for what ever one turn equals out to in
>voltage by using the turns per volt figure. When you applied current to
>the primary, you would sure know it as it would be a big hum and
>should blow a fuse, etc. I would definately double check this! Any short
>generally acts this way.
>>
>> What controls the inductance is the number of turns and the amount of
>iron in the core in square centimeters or square inches. A different core
>material could possibly cause it if one type impeded the flux more than
>another (greater reluctance). A gap will effect it some more because of
>flux impedance (reluctance), but really it's used to keep a choke from
>saturating at high current levels, or where you have DC mixed with AC, not
>for transformers. If ones gaps a transformer, it needs to be a very thin
>gap like 0.002" or even thinner as in a C-core type. These are actually
>ground and lapped so they fit really close. Check all this over, and let
>me know what you find. I'll do my best to help.
>>
>> Best,
>>
>> Will
>>
>>
>> *********** REPLY SEPARATOR ***********
>>
>> On 6/30/06 at 9:10 PM Borislav Trifonov wrote:
>>
>>> I had rewound the secondaries on two 950 W power transformers a few
>>> months back, both identically, each with four identical windings side
>by
>>> side. I hadn't touched the primaries. There is a layer of aluminum
>foil
>>> as an electrostatic shield between the primary and secondaries, not
>>> making a complete turn, with drain wire. Yesterday I tried measuring
>the
>>> inductances of their windings with my DMM, and to my surprise one of
>the
>>> transformers measures more than twice as high as the other one, on all
>>> its winding... WTF!
>>> The meter uses 200 Hz in the range I was measuring with, so it gives me
>>> a much smaller number than the actual inductance as the laminations are
>>> designed for 60 Hz operation. Nonetheless, the huge difference between
>>> the two transformers is consistent across all windings, same ratio.
>>> Moreover, the transformer with the lower inductance buzzes the outer
>>> magnetic shielding more when powered, indicating more leakage (I know
>>> the cores do not saturate as I get fine sine waves on the scope for
>both
>>> of them). Measurement of current through shorted secondary when
>>> powering the primary through a ballast is the same for both
>>> transformers, and both draw the same current from mains when
>secondaries
>>> are open. All secondary windings produce the right voltages, and drop
>>> the same under heavy load.
>>> Yet, the measurement difference and the buzzing difference clearly
>>> indicate something is wrong with one of the transformers. I was
>>> thinking partially shorted winding, but then the voltage output would
>be
>>> changed. It's possible the electrostatic shielding foil between the
>>> primary and secondaries is shorted, though I'm pretty sure I had the
>>> ends of the foil not touching each other (and poking with a needle from
>>> the side and shorting the foil on the other transformer didn't seem to
>>> create a difference anyway).
>>> I considered gapping between the Es and Is of the transformers.
>However,
>>> since I assembled both manually (interleaved, of course), I'd expect
>the
>>> variations in gapping throughout the layers on each transformer to
>>> average out to similar values for both transformers, I'd say up to 1/10
>>> mm; putting them in the vice to squeeze Is towards Es does increase the
>>> meter's measurement about 5%, but the same amount on both transformers,
>>> so the ratio remains the same.
>>> Well, I'm running out of ideas here. To start to take apart the more
>>> buzzing/lower inductance (I guess more leaking) transformer, I'd have
>to
>>> also take the other one apart simultaneously to do comparisons so I
>know
>>> when I've reached the trouble spot. This especially sucks since
>>> squeezing in the heavy gauge wire in the amount of space available was
>>> very difficult work when I had put these together.
>>> Help!
>>> _______________________________________________
>>> Amps mailing list
>>> Amps@contesting.com
>>> http://lists.contesting.com/mailman/listinfo/amps
>>
>>
>>
>>
>>
>
>_______________________________________________
>Amps mailing list
>Amps@contesting.com
>http://lists.contesting.com/mailman/listinfo/amps
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|