Unwound the wire off the primary and know the exact length,L. Also the
bobbin dimensions of course. The wire I have around, is a slightly
smaller diameter ,d2, original d1. Would the rewind length , for the
approx same number of turns,be simply L * (d2^2)/(d1 ^2)ie shorter by
the ratio of areas or some other consideration like packing factor
change, assuming current carrying capacity is sufficient.
I know I should have counted off the turns but I assumed there would be
a "welded" mess of metal in the middle, and would have to rely on weight
of the wire.

N_Cook wrote:

Unwound the wire off the primary and know the exact length,L. Also the
bobbin dimensions of course. The wire I have around, is a slightly
smaller diameter ,d2, original d1. Would the rewind length , for the
approx same number of turns,be simply L * (d2^2)/(d1 ^2)ie shorter by
the ratio of areas or some other consideration like packing factor
change, assuming current carrying capacity is sufficient.
I know I should have counted off the turns but I assumed there would be
a "welded" mess of metal in the middle, and would have to rely on weight
of the wire.

Could you back-calculate the number of primary turns from the secondary
(which I presume is still in place)?

Wind on 10 primary turns but don't bother to replace the core plates,
then energise the secondary to some proportion of its working voltage
from a signal generator at a high enough frequency that the lack of an
iron core doesn't matter. From the voltage appearing across the
temporary primary you can calculate what nominal value of turns per volt
it would have when the secondary is running at its rated voltage.

From that, you can calculate the nominal number of turns required in the
primary. To allow for losses, you will then need to reduce the actual
number of turns by a percentage which depends on the type and size of
transformer. For units around 10 to 20 VA, the reduction could be as
much as 10% (especially if you are using slightly more lossy wire than
the original), but a 250VA transformer may need less than 5% reduction.

To avoid having to dismantle the whole thing and rewind it yet again,
err on the side of too many turns (which is what will happen anyway if
you use the correct length of slightly undersized wire). That way you
may be able to slip off a few turns until you get the correct number.
Another alternative is to leave a few tapping points and insulate-off
the ones you find you don't need.


--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

On 5/11/2014 7:13 AM, Adrian Tuddenham wrote:
N_Cook wrote:

Unwound the wire off the primary and know the exact length,L. Also the
bobbin dimensions of course. The wire I have around, is a slightly
smaller diameter ,d2, original d1. Would the rewind length , for the
approx same number of turns,be simply L * (d2^2)/(d1 ^2)ie shorter by
the ratio of areas or some other consideration like packing factor
change, assuming current carrying capacity is sufficient.
I know I should have counted off the turns but I assumed there would be
a "welded" mess of metal in the middle, and would have to rely on weight
of the wire.

Could you back-calculate the number of primary turns from the secondary
(which I presume is still in place)?

Wind on 10 primary turns but don't bother to replace the core plates,
then energise the secondary to some proportion of its working voltage
from a signal generator at a high enough frequency that the lack of an
iron core doesn't matter. From the voltage appearing across the
temporary primary you can calculate what nominal value of turns per volt
it would have when the secondary is running at its rated voltage.

From that, you can calculate the nominal number of turns required in the
primary. To allow for losses, you will then need to reduce the actual
number of turns by a percentage which depends on the type and size of
transformer. For units around 10 to 20 VA, the reduction could be as
much as 10% (especially if you are using slightly more lossy wire than
the original), but a 250VA transformer may need less than 5% reduction.

To avoid having to dismantle the whole thing and rewind it yet again,
err on the side of too many turns (which is what will happen anyway if
you use the correct length of slightly undersized wire). That way you
may be able to slip off a few turns until you get the correct number.
Another alternative is to leave a few tapping points and insulate-off
the ones you find you don't need.


What he said...but...
My experience has been that I can never hand wind the turns as tightly
as the machine. Wire not tightly wound can take up considerably more
space than you calculate.

On 11/05/2014 18:05, mike wrote:
On 5/11/2014 7:13 AM, Adrian Tuddenham wrote:
N_Cook wrote:

Unwound the wire off the primary and know the exact length,L. Also the
bobbin dimensions of course. The wire I have around, is a slightly
smaller diameter ,d2, original d1. Would the rewind length , for the
approx same number of turns,be simply L * (d2^2)/(d1 ^2)ie shorter by
the ratio of areas or some other consideration like packing factor
change, assuming current carrying capacity is sufficient.
I know I should have counted off the turns but I assumed there would be
a "welded" mess of metal in the middle, and would have to rely on weight
of the wire.

Could you back-calculate the number of primary turns from the secondary
(which I presume is still in place)?

Wind on 10 primary turns but don't bother to replace the core plates,
then energise the secondary to some proportion of its working voltage
from a signal generator at a high enough frequency that the lack of an
iron core doesn't matter. From the voltage appearing across the
temporary primary you can calculate what nominal value of turns per volt
it would have when the secondary is running at its rated voltage.

From that, you can calculate the nominal number of turns required in the
primary. To allow for losses, you will then need to reduce the actual
number of turns by a percentage which depends on the type and size of
transformer. For units around 10 to 20 VA, the reduction could be as
much as 10% (especially if you are using slightly more lossy wire than
the original), but a 250VA transformer may need less than 5% reduction.

To avoid having to dismantle the whole thing and rewind it yet again,
err on the side of too many turns (which is what will happen anyway if
you use the correct length of slightly undersized wire). That way you
may be able to slip off a few turns until you get the correct number.
Another alternative is to leave a few tapping points and insulate-off
the ones you find you don't need.


What he said...but...
My experience has been that I can never hand wind the turns as tightly
as the machine. Wire not tightly wound can take up considerably more
space than you calculate.

At least the replacement wire is thinner and so more space to sloppy. I
will use a manual coil winder but I never seem to get the traverse rate
just right , but better than by hand

On 11/05/2014 15:13, Adrian Tuddenham wrote:
N_Cook wrote:

Unwound the wire off the primary and know the exact length,L. Also the
bobbin dimensions of course. The wire I have around, is a slightly
smaller diameter ,d2, original d1. Would the rewind length , for the
approx same number of turns,be simply L * (d2^2)/(d1 ^2)ie shorter by
the ratio of areas or some other consideration like packing factor
change, assuming current carrying capacity is sufficient.
I know I should have counted off the turns but I assumed there would be
a "welded" mess of metal in the middle, and would have to rely on weight
of the wire.

Could you back-calculate the number of primary turns from the secondary
(which I presume is still in place)?

Wind on 10 primary turns but don't bother to replace the core plates,
then energise the secondary to some proportion of its working voltage
from a signal generator at a high enough frequency that the lack of an
iron core doesn't matter. From the voltage appearing across the
temporary primary you can calculate what nominal value of turns per volt
it would have when the secondary is running at its rated voltage.

From that, you can calculate the nominal number of turns required in the
primary. To allow for losses, you will then need to reduce the actual
number of turns by a percentage which depends on the type and size of
transformer. For units around 10 to 20 VA, the reduction could be as
much as 10% (especially if you are using slightly more lossy wire than
the original), but a 250VA transformer may need less than 5% reduction.

To avoid having to dismantle the whole thing and rewind it yet again,
err on the side of too many turns (which is what will happen anyway if
you use the correct length of slightly undersized wire). That way you
may be able to slip off a few turns until you get the correct number.
Another alternative is to leave a few tapping points and insulate-off
the ones you find you don't need.



As that procedure has some suck-it-and-see I think I'll calculate via
area ratios and add arbitrarily 20%. Wind that whole length on ,
counting turns, and then try with some medium frequency drive, sans
iron, and then count off the error. Even if that +20% is too little
there will be space to do a well insulated splice of some more turns

I think I'll try sans iron and try mains with just the "E" plates
interleaved, as easy enough to do, and compare the results

It turned out ,for same voltage out and for my ,in effect, manual
winding, the same length of wire required , going down from .32mm to
..28mm. Using just the Es minus 1, and no Is , magnetising not complete
and could only run the variac up to 70 per cent or so, but gave the
voltage ratio at 50Hz. Next time I'll probably add the Is as well, as
its easier inserting the Is than the Es. Tip for final E insertion,
force a razor blade in between the interim last E and the bobbin, then
insert the last E between the razor, bent a bit, and the in place E.

If you can, I would add a few more turns to account for the voltage drop that you will get when the transformer is under an actual load.

On 05/26/2014 09:23 PM, wrote:
If you can, I would add a few more turns to account for the voltage drop that you will get when the transformer is under an actual load.

Any Radio Amateur Handbook from before the modern era will have
complete detailed instructions for rewinding transformers. fyi

On 27/05/2014 22:35, dave wrote:
On 05/26/2014 09:23 PM, wrote:
If you can, I would add a few more turns to account for the voltage
drop that you will get when the transformer is under an actual load.

Any Radio Amateur Handbook from before the modern era will have
complete detailed instructions for rewinding transformers. fyi

There is some good information here also:
http://ludens.cl/Electron/trafos/trafos.html

On 6/1/2014 3:17 AM, Chris Jones wrote:
On 27/05/2014 22:35, dave wrote:
On 05/26/2014 09:23 PM, wrote:
If you can, I would add a few more turns to account for the voltage
drop that you will get when the transformer is under an actual load.

Any Radio Amateur Handbook from before the modern era will have
complete detailed instructions for rewinding transformers. fyi

There is some good information here also:
http://ludens.cl/Electron/trafos/trafos.html


That's a great site, work your way back to the home page, he has a lot
of neat articles.

Mikek