Choosing optimal voltage and current ripple for boost converter

Thread Starter

zazas321

Joined Nov 29, 2015
437
Hello. Could someone give me advice on how to choose optimal values for ripple voltage and current. Based on the given requirements of the boost converter I am not able to determine what ripples should I use. The only given specifications of the system is :

Vin=20V
Pin=20W

Vout=60V

I need to design and select optimal values of L and C for a boost converter. Any assumptions are allowed!!.

I have derived Inductance and Capacitance formulas, but in order to calculate the values, I am missing current and voltage ripple.

L= (Vin*D) / I(ripple)*F(sw)

C= (Iout * Ton) / Vripple
 

BobTPH

Joined Jun 5, 2013
2,255
There is not an optimum ripple. The larger inductors and capacitors you use, the smaller the ripple will be. What you have to do is choose a level of ripple that is acceptable, then it will give you the sizes of the parts you need.

The acceptable ripple depends on what you are using the output for. If it is for an incandescent light, for example the ripple could be very high, even 50%. For sensitive electronics you probably want it under 1%.

Bob
 

Thread Starter

zazas321

Joined Nov 29, 2015
437
I am using the boost converters for a hybrid power system (wind/solar). I have calculated the Inductance and capacitance based on this article - http://www.ti.com/lit/an/slva372c/slva372c.pdf.

I have assumed the ripple current to be 30% and calculated Voltage ripple to be 1.75%.

Calculated values for L=1.3mH
and value for C=2uF.
I have chosen frequency to be 100kHz.

I am trying to simulate the design and I am not quite sure what should I do with the load and input resistances? Do I need to use them?

The results I get are clearly wrong on the simulation:
Voltage input and output are equal and current=0
upload_2019-1-12_16-25-58.png
 

MisterBill2

Joined Jan 23, 2018
5,188
With a wind generation system you should already have AC power at a changing frequency, so a simple transformer could provide the desired voltage boost. Solar power is a different issue completely.
But my question is why 60 volts?? And why is ripple an issue? Also, how much power is needed?
In addition, simulations produce no useful power, tghat only comes from actual hardware.
To get a good education about switchmode voltage boosters check out the TI website, as well as a number of others that some of the folks here can suggest. You will discover that starting with a simulation is not the best place to start.
 

kubeek

Joined Sep 20, 2005
5,681
What are you doing with the output? Charging a battery? IIRC an AGM battery called for less than 1% current ripple when charging to achieve long service life.
 

MisterBill2

Joined Jan 23, 2018
5,188
The hardware does not exist. This is a theoretical problem in an incomplete simulation. A small portion of a switch mode supply. Another thread by the same poster asks questions that can't be answered.
 

BobTPH

Joined Jun 5, 2013
2,255
Here is a simulation with the inductor and capacitor you calculated:
upload_2019-1-12_16-22-7.png

I get a voltage and current ripple of about 1.7%. Your inductor is way larger than needed, I get essentially the same results with a 100uH inductor.

Edit: Raising the capacitor can give you much lower ripple if needed.
Bob
 

MisterBill2

Joined Jan 23, 2018
5,188
Once again, the partial power supply is running OPEN LOOP, which does not bear any relationship to how an actual power supply operates. In an actual supply, the feedback plays a big part in eliminating ripple and variations.
 

kubeek

Joined Sep 20, 2005
5,681
Once again, the partial power supply is running OPEN LOOP, which does not bear any relationship to how an actual power supply operates. In an actual supply, the feedback plays a big part in eliminating ripple and variations.
Are you sure? The ripple comes from the basic opertion of the buck converter, and I don´t think any form of feedback can change that. Load regulation, overshoot on load change etc. definitely, but it can´t get below the basic ripple.
 

Thread Starter

zazas321

Joined Nov 29, 2015
437
As misterbill mentioned, there is no hardware. I just have to simulate and come up with a concept. I need two different boost converters with two different load ratios. I am trying to do it one by one. First I need to figure out how to design a simple boost converter.
 

Thread Starter

zazas321

Joined Nov 29, 2015
437
With a wind generation system you should already have AC power at a changing frequency, so a simple transformer could provide the desired voltage boost. Solar power is a different issue completely.
But my question is why 60 volts?? And why is ripple an issue? Also, how much power is needed?
In addition, simulations produce no useful power, tghat only comes from actual hardware.
To get a good education about switchmode voltage boosters check out the TI website, as well as a number of others that some of the folks here can suggest. You will discover that starting with a simulation is not the best place to start.
There is no difference if its 60 or 10 for me. I am just trying to understand the boost converters and how to design it. If I can design it with one Voltage I should be able to apply the knowledge to design it for the other one.
 

Thread Starter

zazas321

Joined Nov 29, 2015
437
You appear to have no load in your simulation.

Bob
Yes I didint have a load on my simulation cause I was not sure what load should I use. It is not specified in the specifications what kind of load I should use. The main aim of the project is to design two boost converters to control two power supplies with different load ratios. I do not seem to get the same response even if I add 180 ohm resistance as you did.
 

Thread Starter

zazas321

Joined Nov 29, 2015
437
What are you doing with the output? Charging a battery? IIRC an AGM battery called for less than 1% current ripple when charging to achieve long service life.

It is one of the options. I can make an assumption that it is if it does make it any easier?
 

kubeek

Joined Sep 20, 2005
5,681
Sorry for me making a confusion, the ripple current in the inductor will not have significant impact on the charge ripple current of the battery, if the output capacitance (and impedance) is good enough to keep the voltage ripple low at the output.

Anyway, you either have to make a guesstimate, assumption, or pick a number by throwing dice, but without a number for the allowed ripple you won´t get anywhere, even for the back of the envelope design. I think that somewhere between 1-10% is reasonable, but still it depends on what the load is and how nice power it deserves or requires to have. After that, and after choosing your operating frequency you will arrive at what size inductor and capacitor you need and how much will they cost, and then you can do a second round of thinking to see if you can tweak some of the parameters to suit your preference better if not all was hunky dory.
 

crutschow

Joined Mar 14, 2008
24,329
You show a bipolar transistor with no base resistor to limit the current.
That definitely will now work.
And why is the drive voltage 100V? :eek:
 

Thread Starter

zazas321

Joined Nov 29, 2015
437
Should I use mosfet or igbt? Ah i was jist testing different values to see if it does make difference. It should be the input voltage right?
 
Top