Maker Pro
Maker Pro

How do I calculate the ESR of a capacitor please?

I posted this(https://www.electronicspoint.com/th...psu-without-ic-components-help-please.278584/) thread sometime ago and thanks to you guys, I have made some progress. I am going with designing a buck converter circuit and I found a few helpful resources. I have been using this( http://www.arrl.org/files/file/QEX_Next_Issue/Jan-Feb 2013/Dildine.pdf)material and I have been able to calculate my inductor value. However when it got to the value of my output capacitance, I found this (http://www.ti.com/lit/an/slva477b/slva477b.pdf) material because the formula is clearer.
My problem here is that there is a value of ESR that I need to get, I have no idea how to get that value and I seem to have searches all over. Can you please help?
My parameters are:
Vin= 220v;
Vout= 15v;
I(load) = 1A;
Duty Cycle = 15/20 = 0.0682;
F = 50hz.
Thank you.
 

hevans1944

Hop - AC8NS
You cannot calculate it. It is either measured or specified by the manufacturer. The goal you seek is to minimize the ESR. That may also require that you specify a maximum acceptable value, so you would need to look at specifications for candidate capacitors and make a choice. I doubt you want to pursue research into capacitor design, so just look through on-line catalogs to find anything suitable that has an acceptable ESR at the switching frequency and voltage applicable to your design. Ceramic dielectrics might seem attractive at first because of their relatively small size, but widen your search to other types. Capacitors designed for pulsed energy storage generally have very low ESR but are probably rated for higher voltages than you require. It's worth a look however.

Do you have to actually build this and demonstrate it? Perhaps a note on your schematic that you submit for grading can just mention that Cxx (or whatever) needs to have a maximum ESR of 35 mΩ at 50 Hz. BTW, that is waaay too low a frequency for a buck converter. Your inductor will be huge. 50 kHz is more appropriate. How did you arrive at 50 Hz? The switching frequency has absolutely NOTHING to do with the power-line frequency in your country! Or was that a typographical error? Forgot to put the "k" in front of Hz?

Please post a schematic, with component values and anticipated waveforms, of what you have accomplished so far. Don't just copy the stuff in the QEX article.
 
You cannot calculate it. It is either measured or specified by the manufacturer. The goal you seek is to minimize the ESR. That may also require that you specify a maximum acceptable value, so you would need to look at specifications for candidate capacitors and make a choice. I doubt you want to pursue research into capacitor design, so just look through on-line catalogs to find anything suitable that has an acceptable ESR at the switching frequency and voltage applicable to your design. Ceramic dielectrics might seem attractive at first because of their relatively small size, but widen your search to other types. Capacitors designed for pulsed energy storage generally have very low ESR but are probably rated for higher voltages than you require. It's worth a look however.

Do you have to actually build this and demonstrate it? Perhaps a note on your schematic that you submit for grading can just mention that Cxx (or whatever) needs to have a maximum ESR of 35 mΩ at 50 Hz. BTW, that is waaay too low a frequency for a buck converter. Your inductor will be huge. 50 kHz is more appropriate. How did you arrive at 50 Hz? The switching frequency has absolutely NOTHING to do with the power-line frequency in your country! Or was that a typographical error? Forgot to put the "k" in front of Hz?

Please post a schematic, with component values and anticipated waveforms, of what you have accomplished so far. Don't just copy the stuff in the QEX article.


Oh wow, I had no idea that the switching frequency is different from my line frequency. Can I just assume a value for my switching frequency or is there a formula?
 

hevans1944

Hop - AC8NS
Oh wow, I had no idea that the switching frequency is different from my line frequency. Can I just assume a value for my switching frequency or is there a formula?
I went back and read your original post (April 12, 2016) on this topic. Your OBJECTIVE was to "develop circuits design and construction skills" and the due date is Friday, April 22, 2016. Based on your posts so far, I think you are not going to be able to finish on time.

You have been given several links in this thread to websites that describe the circuit theory for switching power supplies. IMHO it is too late to play "catch up" to learn and build an operating power supply circuit of any kind, even a simple transformer/rectifier/filter capacitor design with zener diode shunt regulation, which would appear to satisfy most of the requirements stated in your paragraph A.i.b. Any way you cut it, it's waaay too late to include a switching regulator with the basic power supply. And, if I read the requirements correctly, you also need a bipolar ± DC voltage output and possibly a variable output, so two switching regulators are required.

Perhaps you will have better luck the next time you take this course...
 
I went back and read your original post (April 12, 2016) on this topic. Your OBJECTIVE was to "develop circuits design and construction skills" and the due date is Friday, April 22, 2016. Based on your posts so far, I think you are not going to be able to finish on time.

You have been given several links in this thread to websites that describe the circuit theory for switching power supplies. IMHO it is too late to play "catch up" to learn and build an operating power supply circuit of any kind, even a simple transformer/rectifier/filter capacitor design with zener diode shunt regulation, which would appear to satisfy most of the requirements stated in your paragraph A.i.b. Any way you cut it, it's waaay too late to include a switching regulator with the basic power supply. And, if I read the requirements correctly, you also need a bipolar ± DC voltage output and possibly a variable output, so two switching regulators are required.

Perhaps you will have better luck the next time you take this course...

The due date has been shifted till further notice and I am NOT going to take this course a second time. One way or the other, I will find the help I need. Most of my classmates also have zero experience like me and are paying some off someone to do all the work for them. I chose not to. I decided I wanted to at least design and simulate the circuit and then pay someone off to do the construction(because marks are awarded for soldering neatness and I have never done that in my life). All I am asking for is some guidance but with discouraging comments like this, man is it even worth trying to learn.
 

hevans1944

Hop - AC8NS
All I am asking for is some guidance but with discouraging comments like this, man is it even worth trying to learn.
Well, it is what it is. In this forum we don't discourage anyone from learning, because none of us here was born with an innate understanding of electronics and things electrical. Most of us acquired what we do know by the application of time and effort over long periods of time. But only you can decide what is worth trying to learn. We are here to help, but learning is up to you.

As for "guidance" please show us what you have done so far. We can't offer any guidance unless we know what path you are on. To date, it appears to me that you may have found one path in the QEX article, but I see no evidence that you understand that design. Example: your question about ESR (equivalent series resistance) specification for capacitors. Example: your question about assuming a switching frequency and asking if there is a formula.

Engineering is NOT about collecting a portfolio of formulas, although many of us start out doing just that. Well, I did, but I was prepubescent at the time and didn't know any better. Formulas are a crutch, a substitute for real knowledge and understanding. Every "formula" you use in engineering (any field of engineering) you should be able to derive (at least once in your lifetime) from first principles. Engineering is applied science, not a collection of recipes and formulas. We do often refer to "cookbooks" of circuits for inspiration, but no good engineers that I have ever met copy circuits verbatim. They take what they find, or have encountered in the past, and adapt it to their specific task at hand.

There were plenty of suggestions on the other thread, but I have yet to see you post anything resembling your own block diagram, your own schematic with parts specified, or your own calculations. You did specify your parameters but without much specificity. Also, the image you posted specified bi-polar output instead of a single output. Has something changed? Is it even necessary that your design be a switching mode power supply? In other words, is that a "given" or is it an option? Would a power supply with a linear regulator be an acceptable design? Does the output have to be adjustable? If so, over what range? How much ripple can you tolerate in the output? How fast must the power supply respond to changes in load? Should that response be critically damped, over damped, or allowed some oscillation before settling in to a final stable output? How much can the output be allowed to change as a function of load? How much can it change for variations in the input voltage? Over what temperature range must it operate, and to what specifications? Yada, yada, yada. Point is, there are many factors to consider, some more important than others, depending on how the PSU will be used.

Okay, so now you have some "breathing room" to finish your design. I wasn't aware that it is "standard practice" to pay off someone to do the work to secure a grade. Although the practice is not unknown here in the USA, I consider it to be dishonest. Good for you that you have decided not to do that with your design effort. The "reward" may not be a higher "grade" but you will have the satisfaction of actually learning something.

As for hardware build and test, this would appear to be heavily weighted against you and any of your classmates with zero experience, compared to someone who is, for example, an electronics hobbyist and familiar with the various assembly techniques. That's too bad. As you may have noticed, the world is full of unfairness and challenges to moral integrity. Please don't let it discourage you.

Prototyping experience is valuable, even essential, if you are just learning the basic principles of an unfamiliar circuit. I hope you are able to "visit" the Electronics Laboratory as often as possible to breadboard sections of your power supply, test it, take measurements with meters and oscilloscopes, keep a notebook of the things you did, and note what did and what didn't work. Perhaps you think you can learn everything you need to know by reading the appropriate texts, but IMO you would be wrong. You need hands-on experience too and a well-kept journal.

Most working engineers now "out source" circuit board manufacturing to companies that specialize in that, including not just PCB (printed circuit board) layout and manufacturing, but also "board stuffing" with vendors providing pick-and-place automation of component selection, placement on the board, and re-flow soldering in an oven. Doing all this by yourself would definitely be character building, perhaps even worthwhile toward a career in engineering. But if you can afford the service, and your instructor does not prohibit it, why not pay someone to provide a "professional" looking board? Or learn to do it yourself and perhaps save a little money. It all depends on how much time and money you have, and how much of that you want to spend to become a well-qualified hobbyist as well as an engineer.

It is my personal opinion that really good electrical or electronic engineers are also intimately familiar with "bench work" and enjoy it immensely. Unfortunately, management at large corporations does not agree. All they are interested in is that paper ticket granted by an accredited school that says you are an engineer. After hiring you they will train you fit their idea of what an engineer does. They hire technicians (not engineers) to read color-codes and solder parts. I have worked with so-called engineers who literally didn't know which end of a hot soldering iron to pick up and hold in their hand. When I once interviewed to be hired by one of these mega-corporations, I was told that all the unionized technicians would go on strike if I were to pick up a soldering iron (by either end). I turned down their job offer.

Some of those "cookie cutter" engineers I have met eventually learned at least the rudiments of real-world design, which most definitely does include constructing prototypes. Others learned to work the system and became engineering managers. Many, I suspect, left engineering for other careers. If you aren't having fun in your chosen career, it's time to choose another.

So, please don't allow me or anyone else to discourage you from learning.

Hop
 
Well, it is what it is. In this forum we don't discourage anyone from learning, because none of us here was born with an innate understanding of electronics and things electrical. Most of us acquired what we do know by the application of time and effort over long periods of time. But only you can decide what is worth trying to learn. We are here to help, but learning is up to you.

As for "guidance" please show us what you have done so far. We can't offer any guidance unless we know what path you are on. To date, it appears to me that you may have found one path in the QEX article, but I see no evidence that you understand that design. Example: your question about ESR (equivalent series resistance) specification for capacitors. Example: your question about assuming a switching frequency and asking if there is a formula.

Engineering is NOT about collecting a portfolio of formulas, although many of us start out doing just that. Well, I did, but I was prepubescent at the time and didn't know any better. Formulas are a crutch, a substitute for real knowledge and understanding. Every "formula" you use in engineering (any field of engineering) you should be able to derive (at least once in your lifetime) from first principles. Engineering is applied science, not a collection of recipes and formulas. We do often refer to "cookbooks" of circuits for inspiration, but no good engineers that I have ever met copy circuits verbatim. They take what they find, or have encountered in the past, and adapt it to their specific task at hand.

There were plenty of suggestions on the other thread, but I have yet to see you post anything resembling your own block diagram, your own schematic with parts specified, or your own calculations. You did specify your parameters but without much specificity. Also, the image you posted specified bi-polar output instead of a single output. Has something changed? Is it even necessary that your design be a switching mode power supply? In other words, is that a "given" or is it an option? Would a power supply with a linear regulator be an acceptable design? Does the output have to be adjustable? If so, over what range? How much ripple can you tolerate in the output? How fast must the power supply respond to changes in load? Should that response be critically damped, over damped, or allowed some oscillation before settling in to a final stable output? How much can the output be allowed to change as a function of load? How much can it change for variations in the input voltage? Over what temperature range must it operate, and to what specifications? Yada, yada, yada. Point is, there are many factors to consider, some more important than others, depending on how the PSU will be used.

Hop

Hey there, thank you for this reply. The thing is , right now I am going to have to start the little I have done all over because i got advised not to use a buck converter to convert a 220v ,50Hz source to a 15v, 1A one so right now I am at a dead end. The question says "stabilized, regulated, low-ripple, constant and variable output,switched mode" so I would say it is pretty necessary to be switched mode. These were the only specifications given for the circuit so I really cannot give a response to some of your other questions. I think I have a choice there. I don't expect the PSU to be used for anything other than for my grades. My problem now is even where to start working on this project. I have been told to use an isolating power supply like a flyback converter. Problem is , I have found many flyback converter tutorials online but all of them go way over my head. Basically I have been looking for a material that does some explanation and then shows how to calculate the component values. I am not even sure if a flyback converter is the best way to go so can I ask what type of circuit you would recommend that I design?
 

hevans1944

Hop - AC8NS
Actually it says "b. Power Supplies: - stabilized, regulated, low-ripple, constant and variable output (voltage, current), switched mode."

This is probably important enough to go back and ask your instructor very specifically whether a switched mode power supply is a requirement or whether a linear regulator is acceptable. What do you suppose the requirement for "constant and variable output" means? Are those two conflicting requirements? Or does it mean that the output has to be variable but constant after the output voltage is variably selected? Does the power supply have to have both a constant and variable output voltage AND a constant and variable current? Since the statement refers to "Power Supplies" in the plural, does that mean you are allowed to select from more than one design approach?

Another requirement from Area of Focus 2) says " - carry out projects on Power Supply Units: Dual Outputs: ±12V, 2A, ±15 V, 1A, ±9V, 2A" Does this mean your power supply has to provide all these output voltages and currents? Or are you free to pick one? Note that some students are asked to design uni-polar power supplies with various voltage and current specifications, so I must presume your power supply must be bi-polar.

Who is advising you not to use a buck converter? What reason did they provide? Not saying they are wrong, but maybe they thought you intended to provide power directly from the power line without using a step-down transformer, which is dangerous for a novice to attempt. Do you have a dialog with your instructor?

Many commercial switching supplies operate directly from the mains power, first rectifying the power line voltage and then using that rectified DC to drive a high-frequency oscillator and transformer that provides galvanic isolation from the power line as well as a low-voltage DC output. The advantage of this approach is a very wide range of input voltage and frequency tolerance so their PSUs can sell to a worldwide market without needing switches or jumpers to select a particular line voltage.

It would be extremely ambitious of you to try to design, much less build, a flyback converter without recourse to integrated circuits. It is not a trivial design problem and generally requires special transformers that you are unlikely to obtain. I can point you at this Google result, but from your comments it is likely most of the links will be "over your head".
 
Wow..... I am shocked that they would have you designing a SMPS at third year! Perhaps this tool can be of some use: Texas Instrument
Is Unilag that difficult? Or is the teacher not doing a very good job at educating?
 
Top