Scotty, I need more power!
by Darren Ashby

The meaning of 'buck,' the correct spelling of buss, (or is it bus?) and one way to 'boost' your supply, are all answers found in the follow up to my last article on switching power supplies. I thank the many readers that provided feedback, ⁱ it was valuable and informative. We will again tackle the topic of switchers in a manner that drills the basic operation of the device so that you can understand the myriad supplyⁱⁱ of information available to the designer.ⁱⁱⁱ

Buss, correct spelling or just good marketing?

Reader Gary Lynch puts in his "$2.0E-02 worth," pointing out that whenever two EE's get together you can start a debate over the spelling of the word bus. I used buss, which is apparently longer than the word should really be. After perusing a few on-line dictionaries, I have to agree with Gary. While you aren't going to get lynched ^iv for using the extra 's' there is definitely more support for the word bus when it comes to electricity. So unless you are referring to a 'kiss' or a fuse company, you're better off using bus. As I had no idea the term meant 'kiss.' I suspect that I must have succumbed to the intense marketing of the fuse company to come up with that spelling.

Bucking Bronco

In the first installment, I admitted that I had no idea where the term 'buck' converter came from. Many readers went immediately to their favorite dictionary^v and pointed out that buck means to resist stubbornly and obstinately, or balk. Being raised on a farm, I experienced firsthand the joy of a horse 'stubbornly resisting' my riding him. As this resulted primarily in a sore pelvic bone, I had a difficult time associating it with a switching regulator. While most readers thought that 'buck' also served as a nice contrast to the opposite 'boost' regulator, I was no closer to the history of the term. That is until I heard from reader Eric Stromberg. He pointed out that there is a class of transformer known as a buck/boost transformer...

The nature of this transformer is in how you wire the primary and secondary coils, either adding/boosting the input voltage or resisting/bucking the input voltage. After Paul Rako of National Semiconductor passed on a comment from Bob Pease that this transformer has been in use over 100 years, I figured that had to be it. The term must have made sense enough to apply it to switchers when they were developed. I know it won't hold up in court, but one email and hearsay of Bob Pease^vi is enough for me to settle the matter. If you have any more ideas though, feel free to pass them along.

I especially enjoyed John Kua's comment when he said, "I can't say for sure, but I think it comes from the 'bucking' characteristic of the converter in that as the voltage drops, it closes the switch and kicks, or bucks the voltage up again. Up, down, up, down, like a bucking bronco." Thanks, John, while I don't think that is why they call it a 'buck' converter, you provided a great description to remember how it works.

Getting a little boost

First a correction,^vii Eric also astutely pointed out that I had indicated that a transformer changes power levels in my previous article. This is not the case at all. I should have said voltage levels. Just a slip of the keyboard on my part, I apologize. So if my statement misled you, please know that a transformer changes voltage levels. Power levels remain about the same, minus some loss due to heating, of course. Conversely, a switching supply does the same.

The first glimmer of understanding I had concerning the switchers was the 'boost,' not the 'buck' topology. Introducing the 'buck' supply first was chosen because when I came to a better understanding of these supplies, I felt that the 'buck' converter is a little simpler to understand and also provided the framework to understand the 'boost' topology. The motor braking system we were designing is very much a boost architecture where the motor supplies the power if you look at it in a certain way. It might even explain the red bomb disaster, but I'm not entirely convinced of that yet. Without further ado, lets get right to the 'boost' converter. Here is the basic form in all its glory:

The first thing you should notice is that all the main components are the same. You still have a diode, a switch, a source, cap and a load. They are simply arranged differently. So taking a look at the circuit (keeping in mind current flow^viii) let's figure out what will happen. First close the switch. With the switch closed current flows through the inductor back to the source. The maximum current would be infinite if the source and inductor had no resistance. But they do not, so the current limits out a level determined by this resistance. Also, the amount of time it takes to get to this level is a combination of this resistance and the inductor value. (It is basically a LR^ix circuit at this point in time. I didn't put the R in all by its lonesome, but believe me it's there.) This is important to note because any time you spend with the switch closed that is significantly longer than L*R, you are just wasting energy.

The point here is to get the current flowing through the inductor, and then open the switch. Now what happens? You have an inductor sitting there with a bunch of current. It would like to keep outputting that current, but can't send it home the easy way anymore. Luckily though, there is this path through the diode out to the load where we would really like to have current go. The diode figures it's O.K. for current to flow that way. So now we have power to the load. But wait a minute, this is a boost converter right? How did the voltage get boosted?

The easiest way for me to realize this is to take into account the back EMF that is created across the inductor when the switch opens. Remember how the voltage developed across the inductor (after the switch opens) is opposite to what you have across it when you are pushing current through it? The end of the inductor tied to the diode wants to be higher in voltage than the end tied to the source. The only way for that to be the case is to apply more voltage to the load. Hence the voltage is boosted.

Closing the loop in this design is done similarly to the buck converter. Feed back from the voltage applied to the load is used to control the switch. Thus the process starts all over.

Some more final thoughts

One down side of the boost converter above is that a significant amount of current doesn't ever get applied to the load. So if you are looking for higher efficiencies, this may not be the circuit for you.

As mentioned above, my last article sparked a conversation with National on some specific requirements I had to design to. They have been very helpful over the past couple of weeks, and I recommend their support. I even spent some more time with the web-bench tool discovering what it can do. The worst limitation it has is holding you to National parts. But I can't slight them for that. I'd do the same.^x

----------------------------

ⁱ Even though it further proves my own lack of knowledge. I guess I should study more, but right now I'd rather go snowmobiling.

ⁱⁱ Very subtle pun intended

ⁱⁱⁱ As I am by no means an expert in this field, I will refer the reader to more in-depth links for the formulas and equations involved. I only hope with this article to create a basic knowledge for the engineer who is getting into switching supplies for the first time and feels a bit lost by it all. It is also my hope that the reader will develop an intuitive understanding for the operation of these supplies so he/she will know where to look when the design stops working the way it is supposed to.

^iv Pun definitely intended. I hope you can take a joke Gary. It just sorta fell out onto the keyboard as I wrote and I couldn't bear to delete it.

^v I really did spend several bleary-eyed hours looking for the history of the word 'buck' but not once did I look in a dictionary. All I can say is du'oh!

^vi Being an analog nut myself, Bob Pease is a hero of mine. If you haven't followed my articles back to the beginning, you may not have realized the tongue-n-cheek humor intended by the comment made at the end of my last article on this subject.

^vii I won't hide the correction here in the footnotes. I figure I made the mistake in the main text, I ought to correct it there too.

^viii The fundamental point we made about inductors last time was to think in terms of current. After all, an inductor wants to keep current constant right? Whenever I am feeling lost working on a supply like this, it helps me to remember that the inductor wants to keep current constant. Based on that, you can begin to predict what will happen in the rest of the circuit.

^ix Should it be "an LR circuit" or "a LR circuit?" When I look at LR I think inductor/resistor, so "an" seems right to me. However, my grammar checker insists that it is "a." Sigh, without the wonders of modern technology, I'd never make it as a writer. Either way if you need to brush up on basic components, I recommend this link, mine of course! :)

^x This footnote has no particular importance other than to point out that I may have gone a bit overboard with the footnotes this time. So if you like it or think it's too much, please let me know!

Product Engineering Archive