Power Inductor Optimization...
 

This is not directly HAM releated,
but Ham radio is used in the project.

I'm building a Solar powered RC plane.

In that plane I have 6 different solar cell planes.
For each plane I'm building a maximum power point tracker.
The heavest component of the MPP is the inductor.

I need the lighetest most efficient inductor possible.

Specifications

30uH min, 50uH prefered.
6 amps DC current max.
7 amps peak current max.
50 to 150Khz switching freq.

(The current ripple has to be low in releation to the DC current to
keep the solar cell in the "Sweet" spot.

Inductor can not be bigger than 100mm in diameter and 25 mm thick.

I'm thinking I want a Aluminium wire air core toriod, but I'm not
sure.

The weight trade off:

At the point of marginal flight. (About 3 amps in the inductor)
Weight is at an absolute premium. I can trade one watt of inductor
loss (at 3 amps) for 190 gm of mass.

Assuming 12 Ga wire in a cylindrical core 4" in diameter:
12 Ga two layers of 10 turns each gives about 45uH.
251 inches of wire for a wt of: 57gm
and a DC loss of: 1/2 W at 3 A and 2 W at 6 A

I'm a bit lost here .... any suggestions on how to proceede?
Should I consider using a toroid core?
Adds weight, and switching losses, but reduces the total feet of wire
and thus the weight.

It would be a lot easier of solar cells were not low voltage high
current devices... alas they are.



Paul (Kl7JG/6)

Consider using a faster switching chip -- this reduces the amount of
inductance you need to store energy.

Don't try to minimize the ripple -- optimize it. This will also allow you
to use smaller parts. Also, carefully select the ESR and inductance of the
filter cap for the application -- at a high frequency the filter cap will
also look like an inductor. All trace lengths add inductance = impedance =
loss of energy.

You can get 1 MHz chips from TI, National, Linear Tech etc. I myself
haven't used a chip above 100 kHz, so would suggest that you peruse the
manufacturer's data sheets for an appropriate PCB layout.

Jack

wrote in message
...
This is not directly HAM releated,
but Ham radio is used in the project.

I'm building a Solar powered RC plane.

In that plane I have 6 different solar cell planes.
For each plane I'm building a maximum power point tracker.
The heavest component of the MPP is the inductor.

I need the lighetest most efficient inductor possible.

Specifications

30uH min, 50uH prefered.
6 amps DC current max.
7 amps peak current max.
50 to 150Khz switching freq.

(The current ripple has to be low in releation to the DC current to
keep the solar cell in the "Sweet" spot.

Inductor can not be bigger than 100mm in diameter and 25 mm thick.

I'm thinking I want a Aluminium wire air core toriod, but I'm not
sure.

The weight trade off:

At the point of marginal flight. (About 3 amps in the inductor)
Weight is at an absolute premium. I can trade one watt of inductor
loss (at 3 amps) for 190 gm of mass.

Assuming 12 Ga wire in a cylindrical core 4" in diameter:
12 Ga two layers of 10 turns each gives about 45uH.
251 inches of wire for a wt of: 57gm
and a DC loss of: 1/2 W at 3 A and 2 W at 6 A

I'm a bit lost here .... any suggestions on how to proceede?
Should I consider using a toroid core?
Adds weight, and switching losses, but reduces the total feet of wire
and thus the weight.

It would be a lot easier of solar cells were not low voltage high
current devices... alas they are.



Paul (Kl7JG/6)

Consider using a faster switching chip -- this reduces the amount of
inductance you need to store energy.

Don't try to minimize the ripple -- optimize it. This will also allow you
to use smaller parts. Also, carefully select the ESR and inductance of the
filter cap for the application -- at a high frequency the filter cap will
also look like an inductor. All trace lengths add inductance = impedance =
loss of energy.

You can get 1 MHz chips from TI, National, Linear Tech etc. I myself
haven't used a chip above 100 kHz, so would suggest that you peruse the
manufacturer's data sheets for an appropriate PCB layout.

Jack

wrote in message
...
This is not directly HAM releated,
but Ham radio is used in the project.

I'm building a Solar powered RC plane.

In that plane I have 6 different solar cell planes.
For each plane I'm building a maximum power point tracker.
The heavest component of the MPP is the inductor.

I need the lighetest most efficient inductor possible.

Specifications

30uH min, 50uH prefered.
6 amps DC current max.
7 amps peak current max.
50 to 150Khz switching freq.

(The current ripple has to be low in releation to the DC current to
keep the solar cell in the "Sweet" spot.

Inductor can not be bigger than 100mm in diameter and 25 mm thick.

I'm thinking I want a Aluminium wire air core toriod, but I'm not
sure.

The weight trade off:

At the point of marginal flight. (About 3 amps in the inductor)
Weight is at an absolute premium. I can trade one watt of inductor
loss (at 3 amps) for 190 gm of mass.

Assuming 12 Ga wire in a cylindrical core 4" in diameter:
12 Ga two layers of 10 turns each gives about 45uH.
251 inches of wire for a wt of: 57gm
and a DC loss of: 1/2 W at 3 A and 2 W at 6 A

I'm a bit lost here .... any suggestions on how to proceede?
Should I consider using a toroid core?
Adds weight, and switching losses, but reduces the total feet of wire
and thus the weight.

It would be a lot easier of solar cells were not low voltage high
current devices... alas they are.



Paul (Kl7JG/6)

Consider using a faster switching chip -- this reduces the amount of
inductance you need to store energy.

There are multiple losses here.
MY nominal array voltage is about 7V my converter is boosting to
10 to 20 V (Depends on the amount of solar power availible and the
loading conditions on the prop/motor)

So if I run a normal boost converter with a good Shotkey diode I get a
drop of 0.4V or 4% (at 10V) of my total power goes into the diode
drop. So I am using syncronous rectification.
I can't find a highside fet driver that will efficently run much
faster than 100Khz.

I've thought about (mis) using one of the low voltage syncronous
buck converters and running it backwards, but the the chips are
generally too smart about startup and the error conditions and current
sense choices are wrong.



Paul (Kl7JG/6)

Consider using a faster switching chip -- this reduces the amount of
inductance you need to store energy.

There are multiple losses here.
MY nominal array voltage is about 7V my converter is boosting to
10 to 20 V (Depends on the amount of solar power availible and the
loading conditions on the prop/motor)

So if I run a normal boost converter with a good Shotkey diode I get a
drop of 0.4V or 4% (at 10V) of my total power goes into the diode
drop. So I am using syncronous rectification.
I can't find a highside fet driver that will efficently run much
faster than 100Khz.

I've thought about (mis) using one of the low voltage syncronous
buck converters and running it backwards, but the the chips are
generally too smart about startup and the error conditions and current
sense choices are wrong.



Paul (Kl7JG/6)