Hi everyone

I am in the process of building a MPPT (based on Buck topology).The setup has the following parameters.

PV panel: 24v 550W
Battery: 12v Lead Acid 150 AH
Vmp of PV: 43v
Imp of PV: 13A
PWM frequency: 36KHz
Controller: Teensy 4.0

I have a confusion regarding the Inductor.According to my calculation on buck calculation.xlsx,the Inductor value should be 22uH.I have chosen two inductors!
This One having high saturation current(near it's max rated current).
The second one here has much more lower saturation current than it's max rated current!
I've read the defition of saturation current but unable to understand which one to choose.Please help !

 

I looked at the large inductor and it will work. Maybe it will run 6C temperature rise.
I would not use the 2915H.

Look at the 2918H.
If you sent 20A through it DC there will be a 20C rise in temperature.
I don't see an easy to know the AC losses but you are running slow. On the other hand you are pushing hard into saturation which will cause heat!
The inductance starts to drop at 12A 10%, 13A 20%, 14A 30%. If you have fast cycle by cycle current limit, like in a current mode PWM, you need to shut down the MOSFET at 14A peak.
Just doing the numbers in my head I think your peak current will be 25 to 30A. (do the math and check me) I think by 20A the inductance is approaching zero.
My conclusion is that you need a 2x more current inductor.

Also I think the inductance is low.

 

The inductance is way low for 36KHz. I would guess it should be at least 1000 times that. Did you get the units wrong?

 

Thanks a lot for your replies guys ! Another thing i forgot to mention that this is a synchronous buck topology.

the datasheet for the 22uH shows a saturation nearly at 35A.Please take a look.


I also followed this link.This guy designed a working MPPT controller.As per his calculations,the Inductor value for my setup came 22uH.I am attaching his excel as well.

But I am flexible,I can increase the PWM frequency & the inductor value as well.Please guide if any changes to be made.Please provide some calculations if any available.....

 

In the steady state, the inductor will have roughly 10 to 11V across it.

dI/dt = V / L = 10 / 22e-6 = 455,000 A/sec

At 36 KHz 50 % cycle is 1.4e-5 sec

So the current will change by
4.5e5 * 1.3e-5 or 6.75A .

That’s what I call ripple!

 

In the steady state, the inductor will have roughly 10 to 11V across it.

dI/dt = V / L = 10 / 22e-6 = 455,000 A/sec

At 36 KHz 50 % cycle is 1.4e-5 sec

So the current will change by
4.5e5 * 1.3e-5 or 6.75A .

That’s what I call ripple!

Thanks! Will this inductor suffice in this design then ? You can also suggest me some other general purpose high current off the shelf Inductor from Mouser for my setup.I will change my PWM frequency in my setup according to that.....
The reason I am choosing Mouser cause i live in a remote village in India & i can only get parts from a Mouser Redistributor here & custom made inductor is nearly impossible here due to limited component supply.....

 

Here is a simulation with your 22u inductor (blue), and ones 10X(green) and 100X (red) larger. Note that the ripple (about 7A) is about what I predicted (6A) with my simplified calculation.

 

Here is a simulation with your 22u inductor (blue), and ones 10X(green) and 100X (red) larger. Note that the ripple (about 7A) is about what I predicted (6A) with my simplified calculation.

Thanks for the work. My guess of 25A pk current was close. lol
So a 220uH will have 16.5A pk.

CoilCraft AGP4233-223ME, AGP4233-333ME or -473ME These are options. The 33uH or 47uH will reduce the ripple.

One option to reduce the size of the inductor needed is to run faster. 36khz and 220uH will have the same ripple as 360khz and 22uH. Example; run at 100khz, (reduce the ripple to 1/3) and increase the inductor to 47uH (ripple to 1/2) = total of 1/6 the p-p ripple. Now you can use a 16 or 17A inductor (saturation of 10% at 17A).

 

No, do not go by the actual currents! These depend on the duty cycle and characteristics of the battery which are not modeled properly.

The purpose of my simulation was just to show the ripple current with a current somewhere in the ballpark.

Note that my circuit is also not a synchronous converter, though the waveforms would be similar.

 

Ok,so far,I can understand that i need to make the following changes

1) Increase the PWM frequency.100KHz is a suitable value (I can do that in Teensy)
2) Increase the Inductor value to 47uH ( I can also do that)

CoilCraft AGP4233-223ME, AGP4233-333ME or -473ME These are options. The 33uH or 47uH will reduce the ripple.

I took a look at the 47uH from coilcraft. AGP4233-473ME from coilcraft has 18.6A saturation current & also a bit costlier.

I found this one from Wurth which a high saturation rating of 31A & comparatively cheap.Can you please take a look & check if it can be used ?

 

Ok,so far,I can understand that i need to make the following changes

1) Increase the PWM frequency.100KHz is a suitable value (I can do that in Teensy)
2) Increase the Inductor value to 47uH ( I can also do that)

That will get the ripple current down to a little more than 1A. Is that acceptable? Why such low switching frequencies? Typical controller chips use 500 KHz or more.

 

I found this one from Wurth which a high saturation rating of 31A & comparatively cheap

Looks good.
What is a Teensy?

 

Typical controller chips use 500 KHz or more.

Actually,with general purpose Microcontroller like Arduino/Teensy,it is quite a bit difficult to achieve that high PWM frequency.Although,I went through it again & look what i Found! With Teensy 4.0 which is my controller,i can achieve much high pwm frequency as compared to arduino(but higher the frequency,it will lose the precision of bits).Look at this chart.I can go 145KHz with 10bit precision! Though,i can achieve 585KHz PWM as per the spec sheet,but that would lose the precision of 10 bits!(Only 8 bits precision with 585KHz).

What is a Teensy?

Sir,Teensy is a programmable microcontroller like Arduino but much more capable & fast!

Another good news is,i found some power toroids online which can be delivered to my location,i will order them as well & try to build the inductor myself rather than using off-the-shelf! I am attaching the Toroids link below,you may check.....

Toroid1
Toroid2

I will update with my design.....