on winding a transformer

I've got a toroidal core, and I'm planning on winding it with pulsed
12VDC on the primary, and want to get out 50V on the secondary, center
tapped so I can get +/- 25V.

The 12VDC will be provided from a 555 switching a power MOSFET at 10
kHz, 50% duty cycle.

I'm hoping to get a maximum of 100W out of the secondary, so the
secondary will be carrying 2 amps, and the primary will be carrying
about 8 amps. I'll be using 22 gauge wire. (A little thin, I know,
but it'll do at least for small scale tests at around 50W.)

12V / sqrt(2) = 8.5 V RMS. (Does this apply for pulsed square waves?)

50V / 8.5 = 5.9, so if the primary is of length X, the center-tapped
secondary should be of length 5.9X, right?

How many turns of primary are necessary for the transformer to work?
In theory, only one turn of wire would be necessary (and then I'd need
6 turns on the CT secondary), but I know this is hogwash. How do I
know how many turns I will need on the primary?

Thanks,

Michael

 

[Bob Eld]

<mrdarrett@gmail.com> wrote in message
news:bb341062-6a8f-4d0e-b4f9-7ee805e78362@s21g2000prm.googlegroups.com...

I've got a toroidal core, and I'm planning on winding it with pulsed
12VDC on the primary, and want to get out 50V on the secondary, center
tapped so I can get +/- 25V.

The 12VDC will be provided from a 555 switching a power MOSFET at 10
kHz, 50% duty cycle.

I'm hoping to get a maximum of 100W out of the secondary, so the
secondary will be carrying 2 amps, and the primary will be carrying
about 8 amps. I'll be using 22 gauge wire. (A little thin, I know,
but it'll do at least for small scale tests at around 50W.)

12V / sqrt(2) = 8.5 V RMS. (Does this apply for pulsed square waves?)

50V / 8.5 = 5.9, so if the primary is of length X, the center-tapped
secondary should be of length 5.9X, right?

How many turns of primary are necessary for the transformer to work?
In theory, only one turn of wire would be necessary (and then I'd need
6 turns on the CT secondary), but I know this is hogwash. How do I
know how many turns I will need on the primary?

Thanks,

Michael

What's with the RMS? Your primary drive is 12 Volts peak to peak. Your
secondary is 100 Volts peak to peak. The turns ratio is 12:100. It's not a
matter of length but rather number of turns.

Be sure to drive the transformer through a capacitor to keep ALL DC out of
it.

The number of turns depends on the area of the core and the maximum flux
density, Bm for the particular core. The core material must not have too
much loss at 10kHz.

You need to know something about the core, it's size and material.

 

On Jun 29, 2:27 pm, "Bob Eld" <nsmontas...@yahoo.com> wrote:

mrdarr...@gmail.com> wrote in message

news:bb341062-6a8f-4d0e-b4f9-7ee805e78362@s21g2000prm.googlegroups.com...



I've got a toroidal core, and I'm planning on winding it with pulsed
12VDC on the primary, and want to get out 50V on the secondary, center
tapped so I can get +/- 25V.

The 12VDC will be provided from a 555 switching a power MOSFET at 10
kHz, 50% duty cycle.

I'm hoping to get a maximum of 100W out of the secondary, so the
secondary will be carrying 2 amps, and the primary will be carrying
about 8 amps. I'll be using 22 gauge wire. (A little thin, I know,
but it'll do at least for small scale tests at around 50W.)

12V / sqrt(2) = 8.5 V RMS. (Does this apply for pulsed square waves?)

50V / 8.5 = 5.9, so if the primary is of length X, the center-tapped
secondary should be of length 5.9X, right?

How many turns of primary are necessary for the transformer to work?
In theory, only one turn of wire would be necessary (and then I'd need
6 turns on the CT secondary), but I know this is hogwash. How do I
know how many turns I will need on the primary?

Thanks,

Michael

What's with the RMS? Your primary drive is 12 Volts peak to peak. Your
secondary is 100 Volts peak to peak. The turns ratio is 12:100. It's not a
matter of length but rather number of turns.

secondary should be 50V p-p... or am I missing something here? +/-
25V after a bridge rectifier... as shown under "Power supply circuit
diagram" here: http://www.redcircuits.com/Page100.htm, to the right
of the transformer

Be sure to drive the transformer through a capacitor to keep ALL DC out of
it.

ok will do. But how do these guys get away without the cap?

http://www.repairfaq.org/sam/smpsbd.gif

from http://www.repairfaq.org/sam/smpsfaq.htm

The number of turns depends on the area of the core and the maximum flux
density, Bm for the particular core. The core material must not have too
much loss at 10kHz.

You need to know something about the core, it's size and material.

What kind of cores are commonly found in pc power supplies? that's
where I got it...

thanks

Michael

 

[Phil Allison]

<mrdarrett@gmail.com>

Be sure to drive the transformer through a capacitor to keep ALL DC out
of
it.


ok will do. But how do these guys get away without the cap?

http://www.repairfaq.org/sam/smpsbd.gif

from http://www.repairfaq.org/sam/smpsfaq.htm

**See:

" Note that the transformer, T1, is a special type which includes an air gap
in its core (among other things) to provide the inductive characteristics
needed for operation in flyback mode. "


...... Phil

 

[Bob Eld]

<mrdarrett@gmail.com> wrote in message
news:bb24b823-bbb1-440b-ac7d-84ca0fc11e24@y22g2000prd.googlegroups.com...

On Jun 29, 2:27 pm, "Bob Eld" <nsmontas...@yahoo.com> wrote:
mrdarr...@gmail.com> wrote in message


news:bb341062-6a8f-4d0e-b4f9-7ee805e78362@s21g2000prm.googlegroups.com...



I've got a toroidal core, and I'm planning on winding it with pulsed
12VDC on the primary, and want to get out 50V on the secondary, center
tapped so I can get +/- 25V.

The 12VDC will be provided from a 555 switching a power MOSFET at 10
kHz, 50% duty cycle.

I'm hoping to get a maximum of 100W out of the secondary, so the
secondary will be carrying 2 amps, and the primary will be carrying
about 8 amps. I'll be using 22 gauge wire. (A little thin, I know,
but it'll do at least for small scale tests at around 50W.)

12V / sqrt(2) = 8.5 V RMS. (Does this apply for pulsed square waves?)

50V / 8.5 = 5.9, so if the primary is of length X, the center-tapped
secondary should be of length 5.9X, right?

How many turns of primary are necessary for the transformer to work?
In theory, only one turn of wire would be necessary (and then I'd need
6 turns on the CT secondary), but I know this is hogwash. How do I
know how many turns I will need on the primary?

Thanks,

Michael

What's with the RMS? Your primary drive is 12 Volts peak to peak. Your
secondary is 100 Volts peak to peak. The turns ratio is 12:100. It's not
a
matter of length but rather number of turns.


secondary should be 50V p-p... or am I missing something here? +/-
25V after a bridge rectifier... as shown under "Power supply circuit
diagram" here: http://www.redcircuits.com/Page100.htm, to the right
of the transformer


Be sure to drive the transformer through a capacitor to keep ALL DC out
of
it.


ok will do. But how do these guys get away without the cap?

http://www.repairfaq.org/sam/smpsbd.gif

from http://www.repairfaq.org/sam/smpsfaq.htm



The number of turns depends on the area of the core and the maximum flux
density, Bm for the particular core. The core material must not have too
much loss at 10kHz.

You need to know something about the core, it's size and material.


What kind of cores are commonly found in pc power supplies? that's
where I got it...

thanks

Michael

Phil above has it right. That gif is a flyback design. The primary with the
air gapped core forms an inductor designed to handle the DC component.

Interestingly some torroids can work in this mode, most cannot. The ones
that can have a powdered iron core. The minute spacings between the iron
grains form the equivalent of an air gap.

Again, you have to know what your core is to determine how to use it. I know
you got it from a power supply but that doesn't mean much without some
knowledge of the circuit.

It is possible to measure the core if you have a scope and a few components
by winding a test winding on it then measuring the B-H curve. Then, by
applying DC, determine how the curve responds and what happens to the
saturation characteristics.

But, that's part of a course in magnetics and probably a bit beyond the
scope of what you are trying to do.

I tell you to add a capacitor to the primary to keep DC out of the core.
That is the safest thing to do, but your drive circuit has to both source
and sink current. You can't do this with a single FET. You'll probably have
to use a half bridge or what amounts to a two switch forward configuration.

I think the turns ratio of 12:100 is correct. Remember you will have +/- 6
volts on the primary and +/- 50 volts center taped on the secondary. That's
6:50 or 12:100, same thing.

 

[Paul E. Schoen]

"Bob Eld" <nsmontassoc@yahoo.com> wrote in message
news:CrX9k.6561$L_.111@flpi150.ffdc.sbc.com...

Phil above has it right. That gif is a flyback design. The primary with
the
air gapped core forms an inductor designed to handle the DC component.

Interestingly some torroids can work in this mode, most cannot. The ones
that can have a powdered iron core. The minute spacings between the iron
grains form the equivalent of an air gap.

Again, you have to know what your core is to determine how to use it. I
know
you got it from a power supply but that doesn't mean much without some
knowledge of the circuit.

It is possible to measure the core if you have a scope and a few
components
by winding a test winding on it then measuring the B-H curve. Then, by
applying DC, determine how the curve responds and what happens to the
saturation characteristics.

But, that's part of a course in magnetics and probably a bit beyond the
scope of what you are trying to do.

I tell you to add a capacitor to the primary to keep DC out of the core.
That is the safest thing to do, but your drive circuit has to both source
and sink current. You can't do this with a single FET. You'll probably
have
to use a half bridge or what amounts to a two switch forward
configuration.

I think the turns ratio of 12:100 is correct. Remember you will have +/-
6
volts on the primary and +/- 50 volts center taped on the secondary.
That's
6:50 or 12:100, same thing.

This sounds a lot like the same project that was discussed by the OP at
length March 22 ("Need a big transformer") and March 31 ("transformers in
parallel"). If the OP simply needs +/- 24 VDC at 2 (or even 5) amps, there
are easier and better ways to get that than tinkering with parts from an
old PC power supply. Unless you need something with exotic specifications,
you can always buy it for less than you can build it (unless your time is
worth nothing or it is mostly for learning). And for learning, it is best
to build basic circuits until you understand them, and there is also a lot
you can do by just running simulations. Much better and safer than blowing
up parts in a breadboard.

I bought a 200 watt dual output +/- 43V 2A switcher originally designed for
an audio amp for about $22 on eBay, and it is a well-made device that will
meet my needs if I ever decide to build a high-power op-amp or audio
amplifier.

Paul

 

On Jun 29, 10:32 pm, "Paul E. Schoen" <pst...@smart.net> wrote:

"Bob Eld" <nsmontas...@yahoo.com> wrote in message

news:CrX9k.6561$L_.111@flpi150.ffdc.sbc.com...





Phil above has it right. That gif is a flyback design. The primary with
the
air gapped core forms an inductor designed to handle the DC component.

Interestingly some torroids can work in this mode, most cannot. The ones
that can have a powdered iron core. The minute spacings between the iron
grains form the equivalent of an air gap.

Again, you have to know what your core is to determine how to use it. I
know
you got it from a power supply but that doesn't mean much without some
knowledge of the circuit.

It is possible to measure the core if you have a scope and a few
components
by winding a test winding on it then measuring the B-H curve. Then, by
applying DC, determine how the curve responds and what happens to the
saturation characteristics.

But, that's part of a course in magnetics and probably a bit beyond the
scope of what you are trying to do.

I tell you to add a capacitor to the primary to keep DC out of the core.
That is the safest thing to do, but your drive circuit has to both source
and sink current. You can't do this with a single FET. You'll probably
have
to use a half bridge or what amounts to a two switch forward
configuration.

I think the turns ratio of 12:100 is correct. Remember you will have +/-
6
volts on the primary and +/- 50 volts center taped on the secondary.
That's
6:50 or 12:100, same thing.

This sounds a lot like the same project that was discussed by the OP at
length March 22 ("Need a big transformer") and March 31 ("transformers in
parallel"). If the OP simply needs +/- 24 VDC at 2 (or even 5) amps, there
are easier and better ways to get that than tinkering with parts from an
old PC power supply. Unless you need something with exotic specifications,
you can always buy it for less than you can build it (unless your time is
worth nothing or it is mostly for learning). And for learning, it is best
to build basic circuits until you understand them, and there is also a lot
you can do by just running simulations. Much better and safer than blowing
up parts in a breadboard.

I bought a 200 watt dual output +/- 43V 2A switcher originally designed for
an audio amp for about $22 on eBay, and it is a well-made device that will
meet my needs if I ever decide to build a high-power op-amp or audio
amplifier.

Paul

Yep, same project, and learning is the main objective (along with
powering an LM3886 someday).

I already built a boost power supply with a 555 switching the same
inductor, pulsing at 100 Hz. It provided about 100VDC (!!), powered a
4W 120V night light, and heated my TIP31As (but not my IRF530 - guess
the R_ds_on, was lower than the resistance of the TIP31A).

I've learned loads since the boost converter project, and I'm still
learning.

I'm using a 12V battery as a starting point - safer than a first
homemade switcher powered by the mains.

Thanks,

Michael

 

[Eeyore]

"Paul E. Schoen" wrote:

This sounds a lot like the same project that was discussed by the OP at
length March 22 ("Need a big transformer") and March 31 ("transformers in
parallel"). If the OP simply needs +/- 24 VDC at 2 (or even 5) amps, there
are easier and better ways to get that than tinkering with parts from an
old PC power supply. Unless you need something with exotic specifications,
you can always buy it for less than you can build it (unless your time is
worth nothing or it is mostly for learning). And for learning, it is best
to build basic circuits until you understand them, and there is also a lot
you can do by just running simulations. Much better and safer than blowing
up parts in a breadboard.

Totally agreed.

I bought a 200 watt dual output +/- 43V 2A switcher originally designed for
an audio amp for about $22 on eBay, and it is a well-made device that will
meet my needs if I ever decide to build a high-power op-amp or audio
amplifier.

Do you still have the item number ? I'd be intruiged to just look at that.

Graham

 

[Eeyore]

mrdarrett@gmail.com wrote:

Yep, same project, and learning is the main objective (along with
powering an LM3886 someday).

I already built a boost power supply with a 555 switching the same
inductor, pulsing at 100 Hz. It provided about 100VDC (!!), powered a
4W 120V night light, and heated my TIP31As (but not my IRF530 - guess
the R_ds_on, was lower than the resistance of the TIP31A).

I've learned loads since the boost converter project, and I'm still
learning.

I'm using a 12V battery as a starting point - safer than a first
homemade switcher powered by the mains.

Flyback converters are a completely different fish from forward converters.

Power Integrations (powerint.com ?) has a wealth of application notes obviously
targeted at their own parts yet which are incredibly useful for learning about
switching supplies.

Graham

 

On Jun 30, 1:32 am, Eeyore <rabbitsfriendsandrelati...@hotmail.com>
wrote:

"Paul E. Schoen" wrote:
This sounds a lot like the same project that was discussed by the OP at
length March 22 ("Need a big transformer") and March 31 ("transformers in
parallel"). If the OP simply needs +/- 24 VDC at 2 (or even 5) amps, there
are easier and better ways to get that than tinkering with parts from an
old PC power supply. Unless you need something with exotic specifications,
you can always buy it for less than you can build it (unless your time is
worth nothing or it is mostly for learning). And for learning, it is best
to build basic circuits until you understand them, and there is also a lot
you can do by just running simulations. Much better and safer than blowing
up parts in a breadboard.

Totally agreed.

I bought a 200 watt dual output +/- 43V 2A switcher originally designed for
an audio amp for about $22 on eBay, and it is a well-made device that will
meet my needs if I ever decide to build a high-power op-amp or audio
amplifier.

Do you still have the item number ? I'd be intruiged to just look at that.

Graham

Probably similar to this one:
http://cgi.ebay.com/200W-Switching-Power-Supply-Dual-43V-2A-Audio-Amp_W0QQitemZ150201799823QQihZ005QQcategoryZ36323QQcmdZViewItemQQ_trksidZp1713.m153.l1262

Cheers,

Michael

 

On Jun 29, 10:32 pm, "Paul E. Schoen" <pst...@smart.net> wrote:

"Bob Eld" <nsmontas...@yahoo.com> wrote in message

news:CrX9k.6561$L_.111@flpi150.ffdc.sbc.com...





Phil above has it right. That gif is a flyback design. The primary with
the
air gapped core forms an inductor designed to handle the DC component.

Interestingly some torroids can work in this mode, most cannot. The ones
that can have a powdered iron core. The minute spacings between the iron
grains form the equivalent of an air gap.

Again, you have to know what your core is to determine how to use it. I
know
you got it from a power supply but that doesn't mean much without some
knowledge of the circuit.

It is possible to measure the core if you have a scope and a few
components
by winding a test winding on it then measuring the B-H curve. Then, by
applying DC, determine how the curve responds and what happens to the
saturation characteristics.

But, that's part of a course in magnetics and probably a bit beyond the
scope of what you are trying to do.

I tell you to add a capacitor to the primary to keep DC out of the core.
That is the safest thing to do, but your drive circuit has to both source
and sink current. You can't do this with a single FET. You'll probably
have
to use a half bridge or what amounts to a two switch forward
configuration.

I think the turns ratio of 12:100 is correct. Remember you will have +/-
6
volts on the primary and +/- 50 volts center taped on the secondary.
That's
6:50 or 12:100, same thing.

This sounds a lot like the same project that was discussed by the OP at
length March 22 ("Need a big transformer") and March 31 ("transformers in
parallel"). If the OP simply needs +/- 24 VDC at 2 (or even 5) amps, there
are easier and better ways to get that than tinkering with parts from an
old PC power supply. Unless you need something with exotic specifications,
you can always buy it for less than you can build it (unless your time is
worth nothing or it is mostly for learning). And for learning, it is best
to build basic circuits until you understand them, and there is also a lot
you can do by just running simulations. Much better and safer than blowing
up parts in a breadboard.

I bought a 200 watt dual output +/- 43V 2A switcher originally designed for
an audio amp for about $22 on eBay, and it is a well-made device that will
meet my needs if I ever decide to build a high-power op-amp or audio
amplifier.

Paul

Thanks for the info. That is a cheap find on eBay. I'll keep it in
mind, though I think +/- 43V would shorten the lifetime of an LM3886
(not to mention my brother-in-law's speakers).

I'll experiment on the lower power end to minimize possible damage
(and to avoid inadvertent in-house pyrotechnics).

Michael