available toroids

[Hul Tytus]

I've been looking for a toroid with a cross section area (one side) of
probably 2 or 3 square centimeters and an id of about 10 or 12 mm. The
intended inductor is about 14 micro henries and will handle between 40 to 80
amps in a flyback power supply at, again probably, 100 kc.
Fair-Rite, Micrometals, and Magnetics show, at the desired id, cross
sections typically less than 1 sqcent. Greater cross sections show greater
id's and consequently greater lengths. These toroids would result in windings
using only a portion of available space with the rest of the toroid doing
nothing but dissipating power.
Anyone know of some suppliers that have the bulky type of toroids that are
required in this case? Or, perhaps, another approach to limit the dissipation?

Hul

 

[amdx]

On 1/17/2020 9:10 AM, Hul Tytus wrote:

I've been looking for a toroid with a cross section area (one side) of
probably 2 or 3 square centimeters and an id of about 10 or 12 mm. The
intended inductor is about 14 micro henries and will handle between 40 to 80
amps in a flyback power supply at, again probably, 100 kc.
Fair-Rite, Micrometals, and Magnetics show, at the desired id, cross
sections typically less than 1 sqcent. Greater cross sections show greater
id's and consequently greater lengths. These toroids would result in windings
using only a portion of available space with the rest of the toroid doing
nothing but dissipating power.
Anyone know of some suppliers that have the bulky type of toroids that are
required in this case? Or, perhaps, another approach to limit the dissipation?

Hul

You might need to stack them. Or use a potcore.

Mikek

 

[Piotr Wyderski]

amdx wrote:

>  You might need to stack them. Or use a potcore.

If you consider ferrites at all, SFDT says that to handle 80A at 14uH
you would need U100/57:

https://www.newark.com/ferroxcube/u100-57-25-3c90/transformer-cores-u100-3c90-308/dp/24W7833?ost=u100%2F57&ddkey=https%3Aen-US%2FElement14_US%2Fsearch

There are no pot cores that big.

Best regards, Piotr

 

[Piotr Wyderski]

Hul Tytus wrote:

I've been looking for a toroid with a cross section area (one side) of
probably 2 or 3 square centimeters and an id of about 10 or 12 mm. The
intended inductor is about 14 micro henries and will handle between 40 to 80
amps in a flyback power supply at, again probably, 100 kc.

This is a 20.2uH@0A choke:

https://i.postimg.cc/C1D4nPxD/choke.png

for a 50A buck-boost @200kHz. 32mm OD on stacked high-quality 27mm
HiFlux cores.
I'm afraid that with your requirements it's not gonna happen.

Best regards, Piotr

 

[Hul Tytus]

Stacking might work. Certainly worth lookin into. The pot cores
of the 3 makers I checked had the same density/linear dimension
as the toroids.

Thanks, Hul

amdx <nojunk@knology.net> wrote:

On 1/17/2020 9:10 AM, Hul Tytus wrote:
I've been looking for a toroid with a cross section area (one side) of
probably 2 or 3 square centimeters and an id of about 10 or 12 mm. The
intended inductor is about 14 micro henries and will handle between 40 to 80
amps in a flyback power supply at, again probably, 100 kc.
Fair-Rite, Micrometals, and Magnetics show, at the desired id, cross
sections typically less than 1 sqcent. Greater cross sections show greater
id's and consequently greater lengths. These toroids would result in windings
using only a portion of available space with the rest of the toroid doing
nothing but dissipating power.
Anyone know of some suppliers that have the bulky type of toroids that are
required in this case? Or, perhaps, another approach to limit the dissipation?

Hul


You might need to stack them. Or use a potcore.

Mikek

 

[Piotr Wyderski]

Hul Tytus wrote:

> Stacking might work.

7x HF-106125-2. Not exactly compact.

Why a flyback in the first place? A half/full/push-pull bridge with a
current doubler looks more realistic. Or a planar core, say E43, with
the secondary made of copper a plate.

Best regards, Piotr

 

[Piotr Wyderski]

DecadentLinuxUserNumeroUno@decadence.org wrote:

> Stacked toroids.

A wrong solution to a wrong problem. This 80A flyback should not even be
considered seriously, to begin with.

Best regards, Piotr

 

Piotr Wyderski <peter.pan@neverland.mil> wrote in
news:qvtbgs$18di$1@gioia.aioe.org:

DecadentLinuxUserNumeroUno@decadence.org wrote:

I saw 45mm pot cores on that site. I think larger are out there.

If there are, Ferroxcube doesn't produce them. But come on, even
the best in class 3C92 material can handle 0.46T at 100C. This is
not a job for ferrite. HiFlux can handle 1.5T.

Best regards, Piotr

Stacked toroids.

 

[Piotr Wyderski]

DecadentLinuxUserNumeroUno@decadence.org wrote:

> I saw 45mm pot cores on that site. I think larger are out there.

If there are, Ferroxcube doesn't produce them. But come on, even the
best in class 3C92 material can handle 0.46T at 100C. This is not a job
for ferrite. HiFlux can handle 1.5T.

Best regards, Piotr

 

Piotr Wyderski <peter.pan@neverland.mil> wrote in
news:qvt9k7$vun$1@gioia.aioe.org:

amdx wrote:

 You might need to stack them. Or use a potcore.

If you consider ferrites at all, SFDT says that to handle 80A at
14uH you would need U100/57:

https://www.newark.com/ferroxcube/u100-57-25-3c90/transformer-cores
-u100-3c90-308/dp/24W7833?ost=u100%2F57&ddkey=https%3Aen-US%2FEleme
nt14_US%2Fsearch

There are no pot cores that big.

Best regards, Piotr

I saw 45mm pot cores on that site. I think larger are out there.

 

DecadentLinuxUserNumeroUno@decadence.org wrote in news:qvtdfl$1h82$1
@gioia.aioe.org:

You need a sine transformation into your

multipliers to pass some meat with the motion.

IOW, I agree with you. An 80 A flyback sounds like a bit much.

 

Piotr Wyderski <peter.pan@neverland.mil> wrote in
news:qvtcre$1drf$1@gioia.aioe.org:

DecadentLinuxUserNumeroUno@decadence.org wrote:

Stacked toroids.

A wrong solution to a wrong problem. This 80A flyback should not
even be considered seriously, to begin with.

Best regards, Piotr

Not the same but I made a 180kV supply once that was oil immersed.
+90kv - -90kV. We used multipliers.

Flyback relies on collapsing field. good to produce high voltages
with small push. You need a sine transformation into your
multipliers to pass some meat with the motion.

 

[Tim Williams]

Why does it have to be so small I.D.? It'll barely fit, hmm, maybe five
turns at the required ampacity?

You're welcome to ask mfgs for custom parts but don't be surprsed if they
roll their eyes at your quantity or budget.

There is a machinable powder material, that I'm familiar with through its
use in induction heating coils. Don't remember who makes it, but maybe it
has low enough losses you can make use of it.

As others mentioned, this is an X-Y problem. You should not be asking "how
do I make this fuckoff massive inductor?", when the obvious uestion is "how
can I avoid using a fuckoff massive inductor?"

You didn't mention voltage, but if it's modest (10-100V?), this is very much
in the realm of a multiphase, full wave, forward converter. Push-pull is
preferred at low voltages, full bridge in the middle, half bridge at high
voltages (say >800V). Phase interleave is preferred at low voltages. The
point is to reduce the peak current any given switch sees, keeping the
inverter impedance reasonable (above an ohm at least). Which makes layout
tractable, and keeps switching losses down. Phase interleave also saves on
ripple current seen by the input and output filter caps, another savings not
to be underestimated.

Resonant can also be considered at the power level, but it's probably not a
big deal because the inverter impedance will be so low, i.e., stray
inductance will dominate the dynamics, not switch capacitance. Well, an
inverted resonant might be okay, but current-fed inverters and upside-down
topologies aren't exactly beginners topics.

If isolation is not actually necessary, boost or SEPIC is perfectly
reasonable of course. Multiphase is still desirable for all the above
reasons.

With these changes, you should also find commercial offerings are very
available. You probably won't find transformers this size, or whatever
ratio you're looking for, but that's easy enough to custom order.

And yes, multiphase repeats a lot of stuff, it's a PITA to hand build.
Consider making a common module and ordering assemblies in quantity, then
make one common control/distribution board to plug them into. Much cheaper
when engineer proto hours are considered.

Tim

--
Seven Transistor Labs, LLC
Electrical Engineering Consultation and Design
Website: https://www.seventransistorlabs.com/

"Hul Tytus" <ht@panix.com> wrote in message
news:qvsipk$30i$1@reader2.panix.com...

I've been looking for a toroid with a cross section area (one side) of
probably 2 or 3 square centimeters and an id of about 10 or 12 mm. The
intended inductor is about 14 micro henries and will handle between 40 to
80
amps in a flyback power supply at, again probably, 100 kc.
Fair-Rite, Micrometals, and Magnetics show, at the desired id, cross
sections typically less than 1 sqcent. Greater cross sections show greater
id's and consequently greater lengths. These toroids would result in
windings
using only a portion of available space with the rest of the toroid doing
nothing but dissipating power.
Anyone know of some suppliers that have the bulky type of toroids that are
required in this case? Or, perhaps, another approach to limit the
dissipation?

Hul

 

[amdx]

On 1/17/2020 4:07 PM, DecadentLinuxUserNumeroUno@decadence.org wrote:

Piotr Wyderski <peter.pan@neverland.mil> wrote in
news:qvt9k7$vun$1@gioia.aioe.org:

amdx wrote:

 You might need to stack them. Or use a potcore.

If you consider ferrites at all, SFDT says that to handle 80A at
14uH you would need U100/57:

https://www.newark.com/ferroxcube/u100-57-25-3c90/transformer-cores
-u100-3c90-308/dp/24W7833?ost=u100%2F57&ddkey=https%3Aen-US%2FEleme
nt14_US%2Fsearch

There are no pot cores that big.

Best regards, Piotr


I saw 45mm pot cores on that site. I think larger are out there.

They do get big, I have some Feroxcube 66/56 3C81 potcores, They are
2.5" diameter and 2.1" tall.A sub L is 18,200.
Mike

 

[Jeroen Belleman]

On 2020-01-17 23:12, Piotr Wyderski wrote:

DecadentLinuxUserNumeroUno@decadence.org wrote:

I saw 45mm pot cores on that site. I think larger are out there.

If there are, Ferroxcube doesn't produce them.

Ferroxcube produced some 240mm OD ferrite toroids for
me, so they make some pretty big rings. No, I didn't
want them for switching power supplies.

Jeroen Belleman

 

[Hul Tytus]

Thanks Piotr - generally ferroxcube appeared to be a supplier to check.
Having a specific target on their web site (or Newark's) is convenient.

Hul

Piotr Wyderski <peter.pan@neverland.mil> wrote:

amdx wrote:

??You might need to stack them. Or use a potcore.

If you consider ferrites at all, SFDT says that to handle 80A at 14uH
you would need U100/57:

https://www.newark.com/ferroxcube/u100-57-25-3c90/transformer-cores-u100-3c90-308/dp/24W7833?ost=u100%2F57&ddkey=https%3Aen-US%2FElement14_US%2Fsearch

There are no pot cores that big.

Best regards, Piotr

 

[Hul Tytus]

Mike - good to hear there are such beasts.

Hul

amdx <nojunk@knology.net> wrote:

On 1/17/2020 4:07 PM, DecadentLinuxUserNumeroUno@decadence.org wrote:
Piotr Wyderski <peter.pan@neverland.mil> wrote in
news:qvt9k7$vun$1@gioia.aioe.org:

amdx wrote:

????You might need to stack them. Or use a potcore.

If you consider ferrites at all, SFDT says that to handle 80A at
14uH you would need U100/57:

https://www.newark.com/ferroxcube/u100-57-25-3c90/transformer-cores
-u100-3c90-308/dp/24W7833?ost=u100%2F57&ddkey=https%3Aen-US%2FEleme
nt14_US%2Fsearch

There are no pot cores that big.

Best regards, Piotr


I saw 45mm pot cores on that site. I think larger are out there.


They do get big, I have some Feroxcube 66/56 3C81 potcores, They are
2.5" diameter and 2.1" tall.A sub L is 18,200.
Mike

 

[Hul Tytus]

Thanks again Piotr. I'll take a look. As to the use of the flyback,
it is simple. The objective is charging a lead acid batter of 175
volts from a 120v ac line with a nearly continuous 20 amps without
a large cap on the the rectified line.

Hul

Piotr Wyderski <peter.pan@neverland.mil> wrote:

Hul Tytus wrote:

Stacking might work.

7x HF-106125-2. Not exactly compact.

Why a flyback in the first place? A half/full/push-pull bridge with a
current doubler looks more realistic. Or a planar core, say E43, with
the secondary made of copper a plate.

Best regards, Piotr

 

[Hul Tytus]

Thanks, I'll keep my eyes open & check the other types.

Hul

DecadentLinuxUserNumeroUno@decadence.org wrote:

Piotr Wyderski <peter.pan@neverland.mil> wrote in
news:qvt9k7$vun$1@gioia.aioe.org:

amdx wrote:

??You might need to stack them. Or use a potcore.

If you consider ferrites at all, SFDT says that to handle 80A at
14uH you would need U100/57:

https://www.newark.com/ferroxcube/u100-57-25-3c90/transformer-cores
-u100-3c90-308/dp/24W7833?ost=u100%2F57&ddkey=https%3Aen-US%2FEleme
nt14_US%2Fsearch

There are no pot cores that big.

Best regards, Piotr


I saw 45mm pot cores on that site. I think larger are out there.

 

[Sergey Kubushyn]

Hul Tytus <ht@panix.com> wrote:

Wow, that's a lot for flyback... Did you already make its clamping network
and decided how are you going to deal with heat? What frequency is it
supposed to run at?

175V at 20A is just mere 3.5kW, child's play for a flyback, right?

Man, for such power you need a full bridge...

Thanks again Piotr. I'll take a look. As to the use of the flyback,
it is simple. The objective is charging a lead acid batter of 175
volts from a 120v ac line with a nearly continuous 20 amps without
a large cap on the the rectified line.

Hul

Piotr Wyderski <peter.pan@neverland.mil> wrote:
Hul Tytus wrote:

Stacking might work.

7x HF-106125-2. Not exactly compact.

Why a flyback in the first place? A half/full/push-pull bridge with a
current doubler looks more realistic. Or a planar core, say E43, with
the secondary made of copper a plate.

Best regards, Piotr

---
******************************************************************
* KSI@home KOI8 Net < > The impossible we do immediately. *
* Las Vegas NV, USA < > Miracles require 24-hour notice. *
******************************************************************