1

I was studing a linear's datasheet, a step down DC to DC switching regulator.

More specific: the LTC3615.

I am a little bit confused with ripple curent which involved to inductor value calculation. In the datasheet, PAGE 25, "Design Example" a ripple current considered 1A.

And here is my problem. In my research on google i found plenty information about this and all of them refer the ripple current, for properly application, a value of 10 - 30% of maximum DC Load current on the Output. And in the datasheet, Linear choosing that value 1A, for 3A maximum output. 1/3 of maximum current.

Is there some information that i haven't notice?

MrBit
  • 1,973
  • 4
  • 27
  • 54
  • Ripple current can be much bigger than load current so please provide links to what you read. – Andy aka Sep 19 '15 at 23:13
  • http://electronics.stackexchange.com/questions/49280/how-to-choose-a-inductor-for-a-buck-regulator-circuit/50597#50597 – MrBit Sep 19 '15 at 23:36
  • http://www.cooperindustries.com/content/dam/public/bussmann/Electronics/Resources/App%20Notes%20and%20Tech%20Info/Coiltronics/BUS_Elx_App_Notes_4031_Switching_Regulators.pdf – MrBit Sep 19 '15 at 23:39
  • http://powerelectronics.com/content/guide-selecting-inductors-switching-regulators – MrBit Sep 19 '15 at 23:40
  • OK so what specifically don't you understand? – Andy aka Sep 20 '15 at 08:59
  • In Linear's example, ripple current considered 1A0. For the one case: 1/3 of maximum current! And for the other (second output) case: 100% of maximum current. And Not below of 30%. – MrBit Sep 20 '15 at 09:57
  • You possibly need to understand the difference between continuous mode and discontinuous mode of operation? – Andy aka Sep 20 '15 at 10:40
  • Yes, i understand that. Amount of inductor current should not be so small as it drops to zero and regulator enter in discontinuous mode. Also should not be so bigger (in continuous mode) to make capacitor damage. But i don't understand from where's the choosed value of 1A and amount of "10%-30%" – MrBit Sep 20 '15 at 11:27
  • 2
    I think the value of the ripple current is one of the initial design choices ! So saying that the ripple current should typically be 10 - 30% of the DC load current can only be true for a limited set of cases. For example when the load current does not very a lot (for example: 0 A (off) or 1 A). But when the load current does vary a lot (for example, 0 A, 1mA, 100 mA or 1 A) and for all these currents you want a decent efficiency then a different choice for the ripply current (under full load) might be more advantageous. – Bimpelrekkie Sep 20 '15 at 11:32
  • In PAGE 25 of datasheet, the output one is 1A and output two is 3A. Ripple current considered both of them as 1A. 1/3 of maximum current for output two and 100% of maximum current of output one. But why is the same both of them two outputs? – MrBit Sep 20 '15 at 11:32
  • 1
    I think that for the 1A output you can choose a 1A ripple current as the maximum current will always be below the maximum current the switch can handle. In te same way 3A DC + 1A ripple might just be OK but 3A DC + 3 A ripple will be too high for the switch so the ripple must be reduced to 1A. – Bimpelrekkie Sep 20 '15 at 11:40
  • @FakeMoustache, Foy my application that the load current may have very "strong" and instantaneously spikes up to the top line I should choose bigger ripple current, right? – MrBit Sep 20 '15 at 11:42
  • I think so yes, a higher ripple current gives the converter more "freedom" to handle the dynamics at the output. So if your load is very dynamic then yes, choose a higher ripple current. – Bimpelrekkie Sep 20 '15 at 11:44
  • @FakeMoustache, "the maximum current the switch can handle". That case i haven't think before! – MrBit Sep 20 '15 at 11:45
  • There are MANY things to think of when working on switched mode converters ;-) – Bimpelrekkie Sep 20 '15 at 11:45
  • But i think 3A max current and 3A ripple can handle by IC. Last page, page 32. Single output, 6A! :) But those amounts aren't for dual outputs! – MrBit Sep 20 '15 at 11:49
  • How can i determine the ripple current for my own application? I can't find the right value! As i understand from all of them, as low it's that value so much better. – MrBit Sep 20 '15 at 15:54
  • It is good to have lower ripple current by higher inductance? How low ripple current can be? – MrBit Sep 20 '15 at 19:40
  • I can't understand the magnitude of "ripple current". I desire to select a ripple current 0.9A but a fixed value lead me to choosing bigger value and get as result lower ripple current – MrBit Sep 20 '15 at 20:11

1 Answers1

1

The term "ideal ripple current" is valid. I have used the term "porching factor" to define this. Marty Brown defines a ripple factor in his book which is correct.

The ripple factor is important because if it's too small the orthodox current mode chip won't have a proper ramp to sense and you will get instability and/or subharmonics, and you may have to fight with slope compensation to get the thing to behave.

There is good credence to the fact that if the slope compensation was large and ripple factor or porching factor was very small then you are almost back to voltage mode anyway. So, too little ripple is BAD in current mode.

Now, if ripple is big then it is true that peak currents are higher than "necessary" for a given output current.

Dave Tweed
  • 168,369
  • 17
  • 228
  • 393
Autistic
  • 14,235
  • 2
  • 27
  • 65