# Transformer Wire Current Capacity

Discussion in 'Electronic Design' started by Rich Grise, May 5, 2006.

1. ### Rich GriseGuest

I was musing about my MOT, with 143 turns of either skinny #18 or fat #19
http://www.abiengr.com/~sysop/images/MOT-primary2.jpg
- it's probably some metric size - anyway, I go googling for "wire table"
and this is one of the hits:
http://www.thelearningpit.com/elec/tools/tables/Wire_table.htm . And I get
to #18, and it says, "Max Amps: 0.5390".

?????!?!?!?!??!!!!!! This MOT draws about 2.5 amps while IDLING! I read
somewhere in some electronics hobbyist mag decades ago, in an article
about transformer rewinding, that #22 was good for an amp in a
transformer. Proportionally (proportionately?), #18 should be good for
FIVE AMPS! So, I'm wondering where the guy got his figures, but I got so
het up I haven't even checked any of the other google hits, so I'll
probably be enlightened in no uncertain terms. ;-) Anyway, my interest
was I've got a potload of #24 wire from scrounged telephone trunk scraps,
and was wondering how many #24's I'd have to put in parallel to equal a
#18. ... Hmmm, according to the calculator, if I use these ludicrously low
numbers from the chart, 4. I'll have to see if I have that kind of
ambition[1] tomorrow or so. ;-)

Cheers!
Rich
[1] and patience! ;-)

2. ### Pooh BearGuest

That table looks wrong. According to them 10 gauge is only good for 3.4 amps.

What does ampacity mean btw ?

Graham

3. ### James WaldbyGuest

Although the ampere numbers are wrong, they might be internally
consistent; ie, 4 strands of #24 is right, to equal #18. Wire
size doubles (fairly closely) when guage decreases by 3, so #18
has 4 times the area of #24.
ampere capacity, I think

-jiw

4. ### ehsjrGuest

According to the NEC:
"The current, in amperes, that a conductor can carry continuously
under the conditions of use without exceeding its temperature
rating."

Don't know if "ampacity" means precisely that as used in
the table at the url Rich posted.

Ed

5. ### Paul Hovnanian P.E.Guest

Be skeptical of any tables or calculators that give wire ampacity
without specifying ambient temperature, maximum conductor/insulation
temperature, thermal resistance or at least specifications of the
conductor's installation, etc. Be doubly skeptical of a table that
begins with the word 'Gauge' misspelled.

6. ### Pooh BearGuest

Yes. In a transformer, the rating is going to be heavily dependent on cooling
which depends on construction.

Graham

7. ### Roy L. FuchsGuest

Ampacity of mag wire doubles every three gauges since circular area
does as well.

8. ### Roy L. FuchsGuest

You're kidding right?

9. ### Roy L. FuchsGuest

Yes. EACH wire out there is not alike.

A "single strength" mag wire cannot handle the same temp as a "double
strength" mag wire, or a High temp" mag wire.

There is a raw copper ampacity table, and from there on, as you have
said, there are a LOT of factors that go into a given wire
installation that determine what it can or should pass through it on a
continuous duty basis.

10. ### Pooh BearGuest

It's purely a US term.

Graham

11. ### Roy L. FuchsGuest

Mmmm... OK...

12. ### Michael A. TerrellGuest

The chart on my website doesn't list current ratings because there
are too many variables like insulation type, and how its used. For
transformers you have to decide how many circular mils per amp,
depending on how its to be used.

http://home.earthlink.net/~mike.terrell/RefAWG.html for the web page.

http://home.earthlink.net/~mike.terrell/AWG.PDF for a PDF you can print
out.

--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

13. ### Tom BruhnsGuest

My rule of thumb for line-frequency power transformers is "about" 500
circular mils per amp. As someone else posted, the rating depends on
transformer construction, wire insulation temperature rating, maximum
operating temperature, cooling options, ... But 500CM/A should be a
reasonable starting point. 18AWG is 1624CM; 19 is 1288. So 2.5A
sounds at least reasonable.

It's possible to measure the average temperature of the wire in
operation by measuring the wire DC resistance. You can either compare
that with the resistance measured at some known temperature (e.g. room
temp), applying the known temperature coefficient of resistance for
copper, or put the transformer (non-operating) into an oven and raise
the temperature until you reach the same resistance. The latter method
is more difficult but more accurate. That won't tell you the peak
temperature in the winding, though, which will be somewhat higher than
the average.

Some good things to know about AWG numbers:
-- 3 gauges doubles (or halves) the cross-sectional area
-- Therefore, 6 gauges doubles (or halves) the diameter
(This is approx., so you'll be slightly off if you try to
use it to go from AWG0000 to AWG48!)
-- To a good approximation, copper wire resistance at room
temperature is 10^(0.1*AWG# - 1) ohms per 1000 feet.
(So 10AWG is 1 ohm per 1000 ft, 20AWG is 10 ohms, ...).
The approx is almost right on at 10AWG, 5% low at 40AWG.
-- Exactly, diameter(mils) = 5 * 92^((36-AWG#)/39)

You'll notice that the gauge# behaves like dB: 3dB doubles the power;
6dB doubles the voltage.

Cheers,
Tom

14. ### Rich GriseGuest

Probably not in this table - it starts increasing at #18, which doesn't
make any sense at all.

Thanks,
Rich

15. ### Ralph MoweryGuest

From the charts I am seeing it looks like you are right on with the wires.
The chart specifics 700 cm/amp, but states that from 500 to 1000 are being
used. The #18 is listed for 2.3 amps using 700 cm/amp.

I would guess that it depends on what the transformer is used for. If for
long term use such as in a VCR or computer monitor probably the larger
wire would be used and if for a short term use like a printer where it may
use a small amount of current all the time but large amounts for a very
short time the smaller wire may be used.

16. ### joseph2kGuest

At 60 Hz only up to about #2/0 (assuming copper). At higher frequencies the
transition gauge is smaller, above 1GHz it is very thin indeed.

17. ### Paul E. SchoenGuest

I did some calculations for wire size vs power dissipation and made an
Excel
spreadsheet which has my estimation for ampacity of wires as well as bus
bars. When wires are bundled together (as in a transformer or motor
winding), ampacity is related to cross sectional area, but for single
conductors, it is more related to the surface area for cooling effect. For
this example, #22 wire should handle 1.9 amps, and #18 will handle 4.7
amps, based on NEC guidelines for single conductors with ordinary PVC
insulation. So, in a tight bundle, 2 or 3 amps should be fine, and even 5
amps should be OK with high temperature insulation.

For really large conductors, especially high frequencies, copper tubing is
ideal. Run chilled liquid coolant through it and you can get a lot more
current for the same amount of copper. This is being done in some large
distribution transformers.

Paul E. Schoen
www.pstech-inc.com

18. ### Tim WilliamsGuest

I've ran oh, 200 or 300 amperes RMS through some 1/4" copper tubing, er,
about 8' of it). It stays as cool as, well, whatever the temperature of the
water is you pump through it! ;o)

After ten or twenty minutes at that power level, you could make tea with the
water though...(it wasn't going through a radiator or anything, just dumping
into, and pumping out of, a milk jug, so, yeah).

Tim

19. ### Roy L. FuchsGuest

Do you wind transformers for use at GHz frequencies?

20. ### Tom BruhnsGuest

Why yes, I do sometimes. But even at the frequencies used by modern
switchers the skin depth becomes an issue. At 1MHz in copper, it's
about 2.6 mils. For a winding that needs to handle several amps (and
therefore need to use wire with an effective diameter considerably more
than 5 mils), it's worth thinking about. It's not unusual to see Litz
wire being used in such transformers.

Cheers,
Tom