Toroid Core for LED V Boost

[Joerg]

Hello Jim,

Toroids are treacherous unless you can guarantee perfect balance...
any DC is death... that's why, back in the days when I was SMPSing, I
preferred pot cores (or E-I, or E-E) WITH air-gaps.

Yes, and that is a fairly little understood subject among engineers. A
saturated core behaves almost as if it wasn't there.

Most of the switchers I designed have a current mode loop which will
prevent such catastrophes. But I have seen designs without it that were
uncomfortably close to CCM. From there to kablouie it can be as little
as a few milliseconds. The resulting fireworks are usually quite impressive.

Regards, Joerg

http://www.analogconsultants.com

 

[Jim Thompson]

On Wed, 06 Apr 2005 21:31:21 GMT, Joerg
<notthisjoergsch@removethispacbell.net> wrote:

Hello Jim,

Toroids are treacherous unless you can guarantee perfect balance...
any DC is death... that's why, back in the days when I was SMPSing, I
preferred pot cores (or E-I, or E-E) WITH air-gaps.

Yes, and that is a fairly little understood subject among engineers. A
saturated core behaves almost as if it wasn't there.

Most of the switchers I designed have a current mode loop which will
prevent such catastrophes. But I have seen designs without it that were
uncomfortably close to CCM. From there to kablouie it can be as little
as a few milliseconds. The resulting fireworks are usually quite impressive.

Regards, Joerg

http://www.analogconsultants.com

"kablouie" reminds of another fun event in my life...

One of the first off-line switchers I was designing (for
OmniComp/GenRad, ~1978) was of great interest to marketing, because it
would significantly decrease the weight of their portable tester
concept, called the PSP (Portable Service Processor).

So all these marketing clowns come trooping into my laboratory when
they got wind that I was testing it.

So they crowd around my workbench.

I warn them that it's not been powered up before and they should stand
back.

They ignore me.

I plug it in.

KABLOUIE!

Flame blew out of the (ferrite) transformer, then it burned back along
the line cord, back toward the receptacle, in a way reminiscent of the
way you see bomb fuses burn, before the breaker finally let go.

After the marketing types regained their skin color, they accused me
of setting the whole thing up to scare them ;-)

...Jim Thompson
--
| James E.Thompson, P.E. | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona Voice480)460-2350 | |
| E-mail Address at Website Fax480)460-2142 | Brass Rat |
| http://www.analog-innovations.com | 1962 |

I love to cook with wine. Sometimes I even put it in the food.

 

[Watson A.Name - \"Watt Su]

"Jim Thompson" <thegreatone@example.com> wrote in message
news:nsm85154bj01k0tntqm0p2e7v8e7ljj89k@4ax.com...

On Wed, 06 Apr 2005 21:31:21 GMT, Joerg
notthisjoergsch@removethispacbell.net> wrote:

Hello Jim,

Toroids are treacherous unless you can guarantee perfect balance...
any DC is death... that's why, back in the days when I was SMPSing,
I
preferred pot cores (or E-I, or E-E) WITH air-gaps.

Yes, and that is a fairly little understood subject among engineers.
A
saturated core behaves almost as if it wasn't there.

Most of the switchers I designed have a current mode loop which will
prevent such catastrophes. But I have seen designs without it that
were
uncomfortably close to CCM. From there to kablouie it can be as
little
as a few milliseconds. The resulting fireworks are usually quite
impressive.

Regards, Joerg

http://www.analogconsultants.com

"kablouie" reminds of another fun event in my life...

One of the first off-line switchers I was designing (for
OmniComp/GenRad, ~1978) was of great interest to marketing, because it
would significantly decrease the weight of their portable tester
concept, called the PSP (Portable Service Processor).

So all these marketing clowns come trooping into my laboratory when
they got wind that I was testing it.

So they crowd around my workbench.

I warn them that it's not been powered up before and they should stand
back.

They ignore me.

I plug it in.

KABLOUIE!

Flame blew out of the (ferrite) transformer, then it burned back along
the line cord, back toward the receptacle, in a way reminiscent of the
way you see bomb fuses burn, before the breaker finally let go.

After the marketing types regained their skin color, they accused me
of setting the whole thing up to scare them ;-)

Any of them wet their pants?? ;-)

You gave them a demonstration of why the dept's high R&D budget is
justified.

...Jim Thompson
--

 

[John Woodgate]

I read in alt.binaries.schematics.electronic that "Watson A.Name -
\"Watt Sun, the Dark Remover\"" <NOSPAM@dslextreme.com> wrote (in
<115965dgmcden0f@corp.supernews.com&gt about 'Toroid Core for LED V
Boost', on Wed, 6 Apr 2005:

You gave them a demonstration of why the dept's high R&D budget is
justified.

They would interpret the incident as showing that the department was
staffed by dangerous incompetents and should be closed down immediately.
--
Regards, John Woodgate, OOO - Own Opinions Only.
There are two sides to every question, except
'What is a Moebius strip?'
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

 

[Michael A. Terrell]

John Woodgate wrote:

They would interpret the incident as showing that the department was
staffed by dangerous incompetents and should be closed down immediately.

Marketing, or R&D? ;-)

--
Former professional electron wrangler.

Michael A. Terrell
Central Florida

 

Jim Thompson wrote:

On Wed, 06 Apr 2005 21:31:21 GMT, Joerg
notthisjoergsch@removethispacbell.net> wrote:



Hello Jim,



Toroids are treacherous unless you can guarantee perfect balance...
any DC is death... that's why, back in the days when I was SMPSing, I
preferred pot cores (or E-I, or E-E) WITH air-gaps.


Yes, and that is a fairly little understood subject among engineers. A
saturated core behaves almost as if it wasn't there.

Most of the switchers I designed have a current mode loop which will
prevent such catastrophes. But I have seen designs without it that were
uncomfortably close to CCM. From there to kablouie it can be as little
as a few milliseconds. The resulting fireworks are usually quite impressive.

Regards, Joerg

http://www.analogconsultants.com



"kablouie" reminds of another fun event in my life...

One of the first off-line switchers I was designing (for
OmniComp/GenRad, ~1978) was of great interest to marketing, because it
would significantly decrease the weight of their portable tester
concept, called the PSP (Portable Service Processor).

So all these marketing clowns come trooping into my laboratory when
they got wind that I was testing it.

So they crowd around my workbench.

I warn them that it's not been powered up before and they should stand
back.

They ignore me.

I plug it in.

KABLOUIE!

Flame blew out of the (ferrite) transformer, then it burned back along
the line cord, back toward the receptacle, in a way reminiscent of the
way you see bomb fuses burn, before the breaker finally let go.

After the marketing types regained their skin color, they accused me
of setting the whole thing up to scare them ;-)


You gotta love it!

If you didn't set it up, you *should* have.

Ed

...Jim Thompson

 

Joerg wrote:

Hello Jim,

Toroids are treacherous unless you can guarantee perfect balance...
any DC is death... that's why, back in the days when I was SMPSing, I
preferred pot cores (or E-I, or E-E) WITH air-gaps.


Yes, and that is a fairly little understood subject among engineers. A
saturated core behaves almost as if it wasn't there.

Most of the switchers I designed have a current mode loop which will
prevent such catastrophes. But I have seen designs without it that
were uncomfortably close to CCM. From there to kablouie it can be as
little as a few milliseconds. The resulting fireworks are usually
quite impressive.

Worse might be when the insulation on the windinginside the
L melts slowly and you can't see the problem. I won't say
"give me catastrophic failure every time..." but sometimes it's
helpful to see at a glance what died.
Ed

Regards, Joerg

http://www.analogconsultants.com

 

[Jim Thompson]

On Thu, 07 Apr 2005 14:31:26 GMT, ehsjr@bellatlantic.net wrote:
[snip]

Worse might be when the insulation on the windinginside the
L melts slowly and you can't see the problem. I won't say
"give me catastrophic failure every time..." but sometimes it's
helpful to see at a glance what died.
Ed

I've melted a few bobbins in my day.

Finally sat down and ran the math... up to the 9th harmonic (of 20KHz)
produces a heating term due to skin effect.

Switched to Litz. With a few outputs as tape.

...Jim Thompson
--
| James E.Thompson, P.E. | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona Voice480)460-2350 | |
| E-mail Address at Website Fax480)460-2142 | Brass Rat |
| http://www.analog-innovations.com | 1962 |

I love to cook with wine. Sometimes I even put it in the food.

 

[Joerg]

Hello Jim,

I've melted a few bobbins in my day.

Finally sat down and ran the math... up to the 9th harmonic (of 20KHz)
produces a heating term due to skin effect.

Switched to Litz. With a few outputs as tape.

Quite a few are all wound as tape.

Regards, Joerg

http://www.analogconsultants.com

 

[John Popelish]

"Watson A.Name - \"Watt Sun, the Dark Remover\"" wrote:

Well, now that you own up to experimenting with this, what,
quantitatively, was the result? If we're going to go busting skulls,
I'd like to see the results before I decide to get invoived. Give us
some before and after figures to get an idea of what the benefits and
drawbacks are. After all, I've glued my fingers to stuff with CA crazy
glue a few times and I don't particularly like peeling off a layer of
skin every time. ;-)

It's been a few years, and I was playing with some boost converter
designs. All I remember is that before I broke and glued the cores (I
used 5 minute epoxy, not instant glue, with a rubber band around the
toroid, cleaned up with sand paper, after ward) they were saturating
and killing the efficiency. There was a big spike of current in the
charge phase that did little for the dump phase. After being gapped,
(I had to approximately double the number of turns to get the same
inductance) the current rose much more linearly during the charge
phase and ramped down more linearly during the dump phase and the
overall efficiency improved quite measurably, in spite of the higher
copper losses (which were a tiny part of the total losses). If you
want to compare apples to cracked and glued apples, you need to make
two inductors with the same core and the same inductance (one gapped
and one ungapped) and make comparative measurements with them plugged
into the same circuit.

I don't want to spoil all your fun.

--
John Popelish

 

[Michael]

John Popelish wrote:

"Watson A.Name - \"Watt Sun, the Dark Remover\"" wrote:

Well, now that you own up to experimenting with this, what,
quantitatively, was the result? If we're going to go busting skulls,
I'd like to see the results before I decide to get invoived. Give us
some before and after figures to get an idea of what the benefits and
drawbacks are. After all, I've glued my fingers to stuff with CA crazy
glue a few times and I don't particularly like peeling off a layer of
skin every time. ;-)

It's been a few years, and I was playing with some boost converter
designs. All I remember is that before I broke and glued the cores (I
used 5 minute epoxy, not instant glue, with a rubber band around the
toroid, cleaned up with sand paper, after ward) they were saturating
and killing the efficiency. There was a big spike of current in the
charge phase that did little for the dump phase. After being gapped,
(I had to approximately double the number of turns to get the same
inductance) the current rose much more linearly during the charge
phase and ramped down more linearly during the dump phase and the
overall efficiency improved quite measurably, in spite of the higher
copper losses (which were a tiny part of the total losses). If you
want to compare apples to cracked and glued apples, you need to make
two inductors with the same core and the same inductance (one gapped
and one ungapped) and make comparative measurements with them plugged
into the same circuit.

I don't want to spoil all your fun.

--

This discussion has picqued my interest. I gotta fidn some more "donuts" and
have at it.

 

[Watson A.Name - \"Watt Su]

"QuantSuff" <uce@nospam.ftc.gov> wrote in message
news:1111498713.4e7698a737b9d77bc8a57f9b0e2a2094@teranews...

"Watson A.Name - "Watt Sun, the Dark Remover"" <NOSPAM@dslextreme.com
wrote
in message news:113vfmod40vn3a@corp.supernews.com...

(snip)

What I would really like to know is what were the parameters for the
.19" (ID) cores that All Electronics sold, but which are now out of
stock. I think someone said they were charcoal colored. It would
be
informative is someone with a bunch of them wound ten turns of, say,
24
gauge wire on a few and measured the inductance, so we could get a
good
idea of their permeability.

The last batch of the .19" I ordered were thinly-coated ferrite and
gives a
whopping 3uH per turn^2 - 10 turns of #28 (heavier wire will bind in
the
tiny hole) gave around 270uH. I stacked 2 of then together and got a
3000uH
inductor with just 16 turns!

The .30"-ID one is powdered iron, and is about 55uH with 10 turns.

QS

That's great info. I'm trying to figure out which it is. GO to this
URL and click on material 77, for example. It gives a chart of the
different sizes of cores. It looks like the FT37 size has the .19" ID
hole. But I'm not sure how the Al value corresponds.

http://www.bytemark.com/products/ferrmat.htm

 

[Watson A.Name - \"Watt Su]

"John Woodgate" <jmw@jmwa.demon.contraspam.yuk> wrote in message
news:zw4YkWB6YMVCFweN@jmwa.demon.co.uk...

I read in alt.binaries.schematics.electronic that "Watson A.Name -
\"Watt Sun, the Dark Remover\"" <NOSPAM@dslextreme.com> wrote (in
115965dgmcden0f@corp.supernews.com&gt about 'Toroid Core for LED V
Boost', on Wed, 6 Apr 2005:

You gave them a demonstration of why the dept's high R&D budget is
justified.

They would interpret the incident as showing that the department was
staffed by dangerous incompetents and should be closed down
immediately.

Something in there about biting the hand that feeds it, I think.
Flickin' marketing types think they're god's greatest gift to mankind.
but they wouldn't be there if it wasn't for the engineering types.

--
Regards, John Woodgate, OOO - Own Opinions Only.
There are two sides to every question, except
'What is a Moebius strip?'
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk