From: Xfmr Guy on
On Tue, 10 Aug 2010 12:43:38 -0700 (PDT), Nunya <jack_shephard(a)cox.net> wrote:

<SNIP>
>
> Not only that, but using weight as the rule is not going to be
>very accurate either.

It's just about as accurate as an estimate based on length and/or weight
measurements can be.

The power handling capability of a transformer is directly proportional to the
WaAc product; see page 4.3 of:

http://www.diegm.uniud.it/mattavelli/elettronica_industriale/Magnetica/CoreDesign.pdf

Since the WaAc product has dimensionality (length)^4, and weight has
dimensionality (length)^3, the power handling capability of a transformer should
be proportional to (weight)^(4/3).

If weight vs. power handling is plotted on loglog paper, you will get a scatter
plot along a straight line with a different slope than if you had plotted
WaAc vs power handling.

A plot of weight vs power handling is not quite as good as a plot of WaAc vs
power handling for a number of reasons.

The ratio of WaAc to core volume is the same for all transformers with the same
aspect ratio; that is, for all EI core transformers with a "square stack", that
ratio will be the same no matter how big or small the transformer. If the core
weight vs power handling for a bunch of transformers having the same aspect
ratio is plotted, the plot should look the same as a plot of WaAc vs power
handling, just with a different slope.

But, transformers having an "oversquare" or "undersquare" stack will have a
different ratio of WaAc to volume, and the plot of WaAc vs power handling will
not map to a plot of weight vs power handling in the same way as for a "square"
stack.

Another source of error with a weight vs power handling plot is that the
measured weight of a finished transformer isn't just the weight of the core; the
weight of the copper, insulation, and mounting hardware, is also included.

Nevertheless, a plot of transformer weight vs power handling is a good start for
determining the power rating of a transformer. To refine the result, a
temperature rise test should be performed.

A fairly extensive discussion of this problem, with plots of a large number of
commercial transformers is found at:

http://forum.allaboutcircuits.com/showthread.php?t=38273


From: Xfmr Guy on
On Wed, 11 Aug 2010 10:25:37 +1000, Grant <omg(a)grrr.id.au> wrote:

<SNIP>
>
>I been wondering lately how to tell the power split between two
>windings on a nice looking transformer in the 'junk box'.
>

A well designed transformer should have the same current density in all the
windings. It's not too hard to show that given a couple (or three) windings,
with one winding carrying a current I1 and with an output voltage V1, and a DC
resistance R1, and similarly, a second winding with I2, R2 and V2, the current
density in winding 1 is proportional to (I1*R1)/V1, and in winding 2, it's
proportional to (I2*R2)/V2.

Then the currents in the two windings (for equal current densities) are related
like this:

I1/I2 = (R2*V1)/(R1*V2)
From: Robert Baer on
Scott wrote:
> I have a toroidal power transformer that I'm considering using for a
> power supply project. It has two secondaries -- 18v-0-18v and 0-48v.
> Part number is 7-40-0011 (already tried googling it, no luck). I'm
> pretty sure I purchased it from one of the online surplus places,
> maybe All Electronics, probably about a decade ago.
>
> What's the best way to determine how many amps the secondary is rated
> for? Can I just throw a resistive load on it an measure the current?
> or does this risk damage?
As in any "unknown" transformer, start by determining the approximate
power rating from the cross-section; all else then falls in place.
From: Robert Baer on
Jan Panteltje wrote:
> On a sunny day (Tue, 10 Aug 2010 08:32:06 -0700 (PDT)) it happened Scott
> <smbaker(a)gmail.com> wrote in
> <d721ef17-ccf7-44b4-9cf2-b6e83883da51(a)t5g2000prd.googlegroups.com>:
>
>> I have a toroidal power transformer that I'm considering using for a
>> power supply project. It has two secondaries -- 18v-0-18v and 0-48v.
>> Part number is 7-40-0011 (already tried googling it, no luck). I'm
>> pretty sure I purchased it from one of the online surplus places,
>> maybe All Electronics, probably about a decade ago.
>>
>> What's the best way to determine how many amps the secondary is rated
>> for? Can I just throw a resistive load on it an measure the current?
>> or does this risk damage?
>
> Wire diameter?
Not a very good indicator..wire could be undersized for the MFG
rating, or oversized, like i do: circular mils is current rating in
milliamps.
From: Grant on
On Tue, 10 Aug 2010 17:45:44 -0700, John Larkin <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote:

>On Wed, 11 Aug 2010 10:25:37 +1000, Grant <omg(a)grrr.id.au> wrote:
>
>>On Tue, 10 Aug 2010 09:29:53 -0700, John Larkin <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote:
>>
>>>On Tue, 10 Aug 2010 08:32:06 -0700 (PDT), Scott <smbaker(a)gmail.com>
>>>wrote:
>>>
>>>>I have a toroidal power transformer that I'm considering using for a
>>>>power supply project. It has two secondaries -- 18v-0-18v and 0-48v.
>>>>Part number is 7-40-0011 (already tried googling it, no luck). I'm
>>>>pretty sure I purchased it from one of the online surplus places,
>>>>maybe All Electronics, probably about a decade ago.
>>>>
>>>>What's the best way to determine how many amps the secondary is rated
>>>>for? Can I just throw a resistive load on it an measure the current?
>>>>or does this risk damage?
>>>
>>>This will get you started.
>>>
>>>ftp://jjlarkin.lmi.net/XfmrScatter.JPG
>>
>>Thanks for that. Quite useful, though our 50Hz xfmrs would be a
>>little heavier, I think.
>
>I think all these were 50/60 Hz transformers. All the ones we buy are.
>
>>
>>I been wondering lately how to tell the power split between two
>>windings on a nice looking transformer in the 'junk box'.
>
>You could measure the resistance and the open-circuit voltages. You
>could assume that each winding could be loaded to, say, 90% or 95% of
>its open-circuit voltage.
>
>>
>>Xfmr looks around 120VA with 9V and 33V secondaries.
>>
>>I was thinking measure no load, then load to 5% or 10% voltage
>>drop? Work from those results.
>
>Oops, yes.
>
>>
>>Tim's resistance measure would shed some light too.
>>>
>>>
>>>Or just load it progressively and see how hot it gets.
>>
>>Good double check of the other methods. Bit slow?
>>
>>This xfmr is a good quality E+I one with earth screen, got the
>>outside shorted copper turn and iron turn to reduce external
>>field, switch and fuse in a partial enclosure. Destined for
>>a bench power supply, I think. Once I have an idea where to
>>set the power limits.
>>
>>
>>Using SLA batteries for testing stuff has produced some fun
>>moments, watching the magic smoke escape ;)
>>
>>http://grrr.id.au/image/74HC74-let-go-magic-smoke-with-a-bang.jpg
>>
>
>I don't suppose it still works.

Nah, everything connected to 5V blew up, except a TL431 'cos it had
a current drop resistor ;)
>
>>That was an odd situation where a line of five 'HCMOS chips
>>started emitting smoke from their middle pins, until the one
>>in the centre one popped its top. Didn't blow the 35A battery
>>fuse.
>
>Wirebonds make nice detonators, but don't pass a lot of I^2*T. The
>first atom bombs used exploding wires. Later ones used gold thinfilms.
>
>I spent the weekend blowing up DC/DC converters, not too bad at only
>$4 each.
>
>I unpotted one, but lost the transformer in the process.
>
>ftp://jjlarkin.lmi.net/VASD1_top.JPG
>
>
>Next project is to breadboard a current limiter circuit, and blow them
>up together.

Better you blowing them up than your customers :) Enjoy!

Grant.