MrAl
Flashlight Enthusiast
Let\'s Get Cracking!! (Inductors)
Hello again,
As the subject title suggests,
I decided to put one of my ideas for building inductors to the test.
Since different kinds of toroid cores are found in the surplus market
at prices almost ten times less than a single 'store-bought' inductor,
wouldnt it be nice to be able to use one in a dc converter configuration?
This is especially true if you already have some cores laying around
not being used for anything.
Unfortunately, dc to dc converters often have to run with relatively
high dc current and small toroid cores with high permeability saturate
quite easily, meaning the toroid cores, as is, wont work. Sure, there's
bound to be a larger toroid out there that will work at the required
dc current but even on the surplus market the larger cores are usually
more money, and, after all, why would we want to use a big clunky
core if we can get away with using a smaller, more compact core?
Power supply engineers have been doing this very thing for decades
(although using toroids is still a relatively new technology when
compared to E cores and the like) by inserting a small air gap
in the magnetic path of the core. The small air gap makes the core
look much bigger than it really is, and so fools it into working
with higher dc current, even though the size doesnt change much.
There's a bit of a problem, however, that must be overcome when
using run of the mill surplus toroid cores. That is, in order to
get a precision gap cut into the tough, brittle material that the
toroid is made of you'll need to purchase a diamond tooth cutting
machine that can cut gaps as small as 0.005 inches into cores made
from materials as hard as ceramic. Im sure the purchase price is
well within anyone's budget, he he. But if you're working with
a low budget like i am, you dont have $50,000 to shell out just
to make one 0.005 inch cut in a toroid core costing 25 cents just
so you can save three dollars from not having to buy a ready made
inductor.
Ok, so we have a problem -- we have a small toroid core that doesnt
have an air gap, and we'd like to add a small air gap so we can use
it with a converter that carries significant dc current.
Machines that will cut a gap into such cores are very expensive, so
we need an alternate strategy to get a small air gap into the core.
Here's where it gets interesting...
If the core is wrapped in a soft white cloth or white paper napkin
and placed into a vise, gentle but firm pressure will cause the
toroid to crack into four approximately equally sized pieces.
The core, when glued back together, exibits all the properties of
a core with a small air gap, because now it has four tiny air
gaps which make the core much more useful in dc applications.
The only problem remaining then is what kind of glue to use.
Since the gap has to be very small in most cases when the core
is small (maybe 0.5 inch OD) Super Glue is the first that comes to
mind. When the glue is applied and the core is reassembled the
thin layer looks like it's around 0.002 to 0.003 inches wide,
which is just about right. Since there will be four such gaps
in a core with four separate pieces, the total gap length in
the core will be about 0.008 to 0.012 inches. This works out
pretty well when the core has to carry dc current. The
down side to using Super Glue is that the useful temperature
range probably goes down quite a bit, although i havent found
any real data on the max temperature of cured Super Glue yet.
For anyone disliking this drawback, i've found a glue called
"PC11" which is like a thick epoxy which wont drip. This glue
works up to about 90 deg C when cured. The only drawback to
using this glue is that the gap may come out somewhat larger
than with the Super Glue, because more material will be inside
the gap once the core is glued back together. Also, this
glue takes at least 12 hours to cure, meaning turnaround time
for constructing a core will be much longer than with Super
Glue, which takes about 1 minute or so to harden.
Ok, so now that we have a method that allows us to gap small
toroid cores, what exactly are the effects on the properties of
the resulting inductor? Do we really get something more useful?
Some tabular data is in order here, in order to compare properties.
The most important properties are:
1. The resulting inductance
2. The max dc handling capability
so these properties are shown here for cores made from two types
of materials: one made from high perm (mu) material and the other from
much lower perm material.
The format for the tables below is:
The header "mu= " shows the permeability for that core, and
the first entry is the gap length in inches (stepped every 0.002 inches),
the second entry is the inductance in Henries,
the last entry is the maximum dc current handling capability at
Bdc=2500 gausses.
For example, for the material with mu=8030 and a gap of 0.002 inches
the inductance is 234uH and the max dc current is 0.418 amps.
mu=8030
0.000 0.003267 0.030
0.002 0.000234 0.418
0.004 0.000121 0.805
0.006 0.000082 1.193
0.008 0.000062 1.581
0.010 0.000050 1.968
0.012 0.000041 2.356
0.014 0.000036 2.744
0.016 0.000031 3.131
0.018 0.000028 3.519
0.020 0.000025 3.907
mu=830
0.000 0.000338 0.289
0.002 0.000144 0.676
0.004 0.000092 1.064
0.006 0.000067 1.452
0.008 0.000053 1.839
0.010 0.000044 2.227
0.012 0.000037 2.615
0.014 0.000032 3.002
0.016 0.000029 3.390
0.018 0.000026 3.778
0.020 0.000023 4.165
From simple inspection of the two tables above, we can see some
very interesting properties about cores with and without gaps...
1. Notice how the high mu material (8030) with no gap can only
handle 30 milliamps dc current? Gee, not very useful.
As the gap goes up, so does the dc current handling ability.
This is basically because the gap makes the core look like it has
a longer magnetic path, and a longer magnetic path means the
core can handle higher dc currents.
2. For two cores of equal size, one with high perm and one with
low perm, the high perm core has higher inductance. This can be
found by looking at the two cores with 0.000 gap length (no gaps).
3. As the gap is increased in both cores, the inductance becomes
very much the same even though the perm of each core is very different.
Looking at the inductance for the two cores with gaps of 0.020 inches
one has L=25uH and the other has L=23uH. Thus, the inductance is
almost the same for both cores. This means the gap begins to take
over control of the inductance level, not the permability of the
core! This is a very useful property indeed, because surplus cores
from the same batch might have perm's that vary as much as 2:1 even
though they are binned the same. Without a gap, this could change
the inductance by a factor of 2 in either direction. With a gap
this variation is almost eliminated.
4. As the gap is increased in either core, the inductance goes
down but the dc current handling capability goes up. This is
very useful when the core has to handle high dc current such
as in a dc converter used for flashlights or battery chargers.
The drawback is that we have to add turns to get back to a
useable inductance value, but in the long run we end up with
a smaller inductor.
THE BASIC IDEAS BEHIND USING A GAP
We want the core to handle dc current so we insert a gap,
because the gap allows us to use higher dc current. Because doing
this lowers the inductance quite a bit, we must add more turns to
get the inductance back up to something suitable for the application.
The reqired additional turns might mean winding three times the
original number of turns on the core, but the resulting inductor
will be much more useful even with a somewhat higher dc resistance
because it will actually work in an application where it couldnt
before the gap was introduced, and we end up with a smaller inductor
as well which takes up less board space.
NEXT TIME:
A real life example using National Semiconductor's 52kHz Simple Switcher
and a very small toroid core.
Take care,
Al
Hello again,
As the subject title suggests,
I decided to put one of my ideas for building inductors to the test.
Since different kinds of toroid cores are found in the surplus market
at prices almost ten times less than a single 'store-bought' inductor,
wouldnt it be nice to be able to use one in a dc converter configuration?
This is especially true if you already have some cores laying around
not being used for anything.
Unfortunately, dc to dc converters often have to run with relatively
high dc current and small toroid cores with high permeability saturate
quite easily, meaning the toroid cores, as is, wont work. Sure, there's
bound to be a larger toroid out there that will work at the required
dc current but even on the surplus market the larger cores are usually
more money, and, after all, why would we want to use a big clunky
core if we can get away with using a smaller, more compact core?
Power supply engineers have been doing this very thing for decades
(although using toroids is still a relatively new technology when
compared to E cores and the like) by inserting a small air gap
in the magnetic path of the core. The small air gap makes the core
look much bigger than it really is, and so fools it into working
with higher dc current, even though the size doesnt change much.
There's a bit of a problem, however, that must be overcome when
using run of the mill surplus toroid cores. That is, in order to
get a precision gap cut into the tough, brittle material that the
toroid is made of you'll need to purchase a diamond tooth cutting
machine that can cut gaps as small as 0.005 inches into cores made
from materials as hard as ceramic. Im sure the purchase price is
well within anyone's budget, he he. But if you're working with
a low budget like i am, you dont have $50,000 to shell out just
to make one 0.005 inch cut in a toroid core costing 25 cents just
so you can save three dollars from not having to buy a ready made
inductor.
Ok, so we have a problem -- we have a small toroid core that doesnt
have an air gap, and we'd like to add a small air gap so we can use
it with a converter that carries significant dc current.
Machines that will cut a gap into such cores are very expensive, so
we need an alternate strategy to get a small air gap into the core.
Here's where it gets interesting...
If the core is wrapped in a soft white cloth or white paper napkin
and placed into a vise, gentle but firm pressure will cause the
toroid to crack into four approximately equally sized pieces.
The core, when glued back together, exibits all the properties of
a core with a small air gap, because now it has four tiny air
gaps which make the core much more useful in dc applications.
The only problem remaining then is what kind of glue to use.
Since the gap has to be very small in most cases when the core
is small (maybe 0.5 inch OD) Super Glue is the first that comes to
mind. When the glue is applied and the core is reassembled the
thin layer looks like it's around 0.002 to 0.003 inches wide,
which is just about right. Since there will be four such gaps
in a core with four separate pieces, the total gap length in
the core will be about 0.008 to 0.012 inches. This works out
pretty well when the core has to carry dc current. The
down side to using Super Glue is that the useful temperature
range probably goes down quite a bit, although i havent found
any real data on the max temperature of cured Super Glue yet.
For anyone disliking this drawback, i've found a glue called
"PC11" which is like a thick epoxy which wont drip. This glue
works up to about 90 deg C when cured. The only drawback to
using this glue is that the gap may come out somewhat larger
than with the Super Glue, because more material will be inside
the gap once the core is glued back together. Also, this
glue takes at least 12 hours to cure, meaning turnaround time
for constructing a core will be much longer than with Super
Glue, which takes about 1 minute or so to harden.
Ok, so now that we have a method that allows us to gap small
toroid cores, what exactly are the effects on the properties of
the resulting inductor? Do we really get something more useful?
Some tabular data is in order here, in order to compare properties.
The most important properties are:
1. The resulting inductance
2. The max dc handling capability
so these properties are shown here for cores made from two types
of materials: one made from high perm (mu) material and the other from
much lower perm material.
The format for the tables below is:
The header "mu= " shows the permeability for that core, and
the first entry is the gap length in inches (stepped every 0.002 inches),
the second entry is the inductance in Henries,
the last entry is the maximum dc current handling capability at
Bdc=2500 gausses.
For example, for the material with mu=8030 and a gap of 0.002 inches
the inductance is 234uH and the max dc current is 0.418 amps.
mu=8030
0.000 0.003267 0.030
0.002 0.000234 0.418
0.004 0.000121 0.805
0.006 0.000082 1.193
0.008 0.000062 1.581
0.010 0.000050 1.968
0.012 0.000041 2.356
0.014 0.000036 2.744
0.016 0.000031 3.131
0.018 0.000028 3.519
0.020 0.000025 3.907
mu=830
0.000 0.000338 0.289
0.002 0.000144 0.676
0.004 0.000092 1.064
0.006 0.000067 1.452
0.008 0.000053 1.839
0.010 0.000044 2.227
0.012 0.000037 2.615
0.014 0.000032 3.002
0.016 0.000029 3.390
0.018 0.000026 3.778
0.020 0.000023 4.165
From simple inspection of the two tables above, we can see some
very interesting properties about cores with and without gaps...
1. Notice how the high mu material (8030) with no gap can only
handle 30 milliamps dc current? Gee, not very useful.
As the gap goes up, so does the dc current handling ability.
This is basically because the gap makes the core look like it has
a longer magnetic path, and a longer magnetic path means the
core can handle higher dc currents.
2. For two cores of equal size, one with high perm and one with
low perm, the high perm core has higher inductance. This can be
found by looking at the two cores with 0.000 gap length (no gaps).
3. As the gap is increased in both cores, the inductance becomes
very much the same even though the perm of each core is very different.
Looking at the inductance for the two cores with gaps of 0.020 inches
one has L=25uH and the other has L=23uH. Thus, the inductance is
almost the same for both cores. This means the gap begins to take
over control of the inductance level, not the permability of the
core! This is a very useful property indeed, because surplus cores
from the same batch might have perm's that vary as much as 2:1 even
though they are binned the same. Without a gap, this could change
the inductance by a factor of 2 in either direction. With a gap
this variation is almost eliminated.
4. As the gap is increased in either core, the inductance goes
down but the dc current handling capability goes up. This is
very useful when the core has to handle high dc current such
as in a dc converter used for flashlights or battery chargers.
The drawback is that we have to add turns to get back to a
useable inductance value, but in the long run we end up with
a smaller inductor.
THE BASIC IDEAS BEHIND USING A GAP
We want the core to handle dc current so we insert a gap,
because the gap allows us to use higher dc current. Because doing
this lowers the inductance quite a bit, we must add more turns to
get the inductance back up to something suitable for the application.
The reqired additional turns might mean winding three times the
original number of turns on the core, but the resulting inductor
will be much more useful even with a somewhat higher dc resistance
because it will actually work in an application where it couldnt
before the gap was introduced, and we end up with a smaller inductor
as well which takes up less board space.
NEXT TIME:
A real life example using National Semiconductor's 52kHz Simple Switcher
and a very small toroid core.
Take care,
Al
