150W boost converter

[electroguy98]

Hello everyone, this is my first post on eevblog. I have basic knowledge about electronics and I'm trying to make a 150w boost converter for driving a 100W strobe led. The calculated values(switching frequency=32kHz) for inductor and capacitor are 100uH and 265uF respectively. The problem is that the inductor requires a current rating of 12.5A but i can't find it anywhere so i have to wind it on a plastic core. I will be using 21 AWG(0.8mm diameter) copper wire. Will it be able to handle the current?
Thanks!

[Siwastaja]

Input and output voltage?

Why such as low frequency? Go higher, and your inductance requirement drops, and you'll most likely find a suitable inductor. For 22 or 33 uH, you should be seeing a lot of results for your saturation and thermal current rating.

Winding your own isn't impossible, but you have more things to calculate and estimate than just the wire thickness. Ferrite cores and bobbins are bloody expensive. This site is the most helpful inductor design tool ever: http://pigeonsnest.co.uk/stuff/core-saturation.html (easy to remember and find by googling "core saturation pigeon"). But it's highly unlikely you need to do custom inductors. I just designed a 750W, 22 to 75V 4-phase boost converter and was able to use off-the-shelf inductors, I even had multiple parts to choose from.

If the ratio between input and output voltage is big, say, over 1:3, you may want to consider multiphase converter, easing ripple current ratings and thermals by dividing the heat over larger number of parts.

[strawberry]

iron powder core  T130-26
core will reduce turns count for same inductance

[Arjunan M R]

https://www.powerstream.com/Wire_Size.htm
I don't think 0.8mm wire can handle that much current.

I don't recommend building an inductor yourself(unless you have an LCR meter)

[soldar]

Ferrite cores and bobbins are bloody expensive.

And sometimes difficult to find what you want. That is why I have crates full of computer and other SMPS where I can recycle plenty of transformers and chokes. Sometimmes I find what I want directly, sometimes I can modify, sometimes I put them in series or parallel.

[strawberry]

I don't recommend building an inductor yourself(unless you have an LCR meter)

Not at all.  L=N^2 * AL 
for T106-26  30(turns)^2*93nH(AL)=83700nH

[Arjunan M R]

I don't recommend building an inductor yourself(unless you have an LCR meter)

Not at all.  L=N^2 * AL 
for T106-26  30(turns)^2*93nH(AL)=83700nH

I don't understand anything in that formula except L & N.
Can you explain?

[electroguy98]

Input voltage=12V output voltage = 36V
I'm using atmega328p so I'm keeping the frequency low because the pwm gets distorted at very high frequencies. Anyway i do have an LCR-T4 meter so i want to know if an air cored inductor will work. I'll try to arrange for a toroid core.
Thanks!

[Arjunan M R]

Input voltage=12V output voltage = 36V
I'm using atmega328p so I'm keeping the frequency low because the pwm gets distorted at very high frequencies. Anyway i do have an LCR-T4 meter so i want to know if an air cored inductor will work. I'll try to arrange for a toroid core.
Thanks!

Are you using microcontroller as a switching controller.
If it is that's not the right way to do it.You should use a boost converter IC. You don't get all the features in a boost converter IC in a MCU. Like softstart etc. If you're using a MCU you need also a separate MOSFET for switching. In a boost converter ic it's all integrated. Or if you are going to use a external MOSFET there are switching controllers which can achieve a very high frequency.And they are not very expensive. You should use switching controllers when dealing with very high current that you need.

[electroguy98]

Input voltage=12V output voltage = 36V
I'm using atmega328p so I'm keeping the frequency low because the pwm gets distorted at very high frequencies. Anyway i do have an LCR-T4 meter so i want to know if an air cored inductor will work. I'll try to arrange for a toroid core.
Thanks!

Are you using microcontroller as a switching controller.
If it is that's not the right way to do it.You should use a boost converter IC. You don't get all the features in a boost converter IC in a MCU. Like softstart etc. If you're using a MCU you need also a separate MOSFET for switching. In a boost converter ic it's all integrated. Or if you are going to use a external MOSFET there are switching controllers which can achieve a very high frequency.And they are not very expensive. You should use switching controllers when dealing with very high current that you need.

Microcontroller or boost converter IC, i still have to make an inductor. So will an air cored inductor do the job?

[Arjunan M R]

Input voltage=12V output voltage = 36V
I'm using atmega328p so I'm keeping the frequency low because the pwm gets distorted at very high frequencies. Anyway i do have an LCR-T4 meter so i want to know if an air cored inductor will work. I'll try to arrange for a toroid core.
Thanks!

Are you using microcontroller as a switching controller.
If it is that's not the right way to do it.You should use a boost converter IC. You don't get all the features in a boost converter IC in a MCU. Like softstart etc. If you're using a MCU you need also a separate MOSFET for switching. In a boost converter ic it's all integrated. Or if you are going to use a external MOSFET there are switching controllers which can achieve a very high frequency.And they are not very expensive. You should use switching controllers when dealing with very high current that you need.

Microcontroller or boost converter IC, i still have to make an inductor. So will an air cored inductor do the job?

Sorry, I don't know.

[Arjunan M R]

I have never seen an air cored inductor in a smps or boost converter.
That doesn't mean it won't work.

[madires]

So will an air cored inductor do the job?

No, but a ferrite core with an air gap or an iron powder core (distributed air gap).

[strawberry]

I don't recommend building an inductor yourself(unless you have an LCR meter)

Not at all.  L=N^2 * AL 
for T106-26  30(turns)^2*93nH(AL)=83700nH

I don't understand anything in that formula except L & N.
Can you explain?

L = inductor inductivity . N = coil turns . AL = core inductance value (found in datasheet or estimeted for custom core)

[DannyTheGhost]

There is no air core inductors in SMPS mainly because you need a lot more wire turns to get desirable inductance, which means a lot more energy loss in the wire -> less efficiency

[Siwastaja]

You need a feedback. Actually two feedbacks: the main one is current mode control for the boost converter. For this, you need a current sense resistor, and a current sense amplifier. You may be able to use the AVR's analog comparator; but the ADC is too slow at 10ksamples/s IIRC, unless you go for even slower frequency.

Outer voltage feedback can be slow and AVR ADC is fast enough, since you would need to slow it down anyway due to boost converter dynamics (so called right half plane zero; simply put, it will oscillate unless you filter voltage feedback to be rather slow).

This can be a good learning experience. Maybe you could reduce your power targets to, say, just 10W, then you could more easily go for even slower switcing which could make AVR more suitable. It's far from sane design, but for learning purposes, very good.