The question I address is somewhat academic. Power Electronics textbooks usually introduce the analysis of DC/DC converters considering the load to be a resistor.
But in the field of applications, a resistor is not the predominant element as an energy recipient: we can think, for example, of a photovoltaic generator or a wind turbine, which can be used to charge a battery system or feed a DC engine. In this type of applications, the load shows a behavior that can be far from being purely resistive. What implications would it have to modify the traditional teaching approach, considering other types of load, in terms of duty cycle calculation or component selection, for example?

 

But in the field of applications, a resistor is not the predominant element as an energy recipient: we can think, for example, of a photovoltaic generator or a wind turbine,

Those aren't loads, they are power sources.
So which are you talking about?

 

The question I address is somewhat academic. Power Electronics textbooks usually introduce the analysis of DC/DC converters considering the load to be a resistor.
But in the field of applications, a resistor is not the predominant element as an energy recipient: we can think, for example, of a photovoltaic generator or a wind turbine, which can be used to charge a battery system or feed a DC engine. In this type of applications, the load shows a behavior that can be far from being purely resistive. What implications would it have to modify the traditional teaching approach, considering other types of load, in terms of duty cycle calculation or component selection, for example?

If you are really talking inductive or capacitive loads then the main concern is stability. This means careful analysis and testing.
Boost converters are the most troublesome because they have an inverse pulse width vs output voltage as compared to say a buck converter.

 

Many loads are mostly resistive and most inverters can supply them adequately. In communications systems it used to be that a transistor inverter would power a capacitance-type load thru a diode rectifier bridge .
With today's switching power supplies the situation is a bit different, but what is normal is that a supply is sold with voltage and current ratings and if the load requirements do not exceed those rating then all is well. For many applications, including all of them in my industrial equipment systems career, purchasing a power supply based on the power requirements has been entirely adequate. Most standard power supply designs are designed to be load agnostic, A supplier unable to provide such a supply would rapidly be moved to the "Previous supplier" list.
My point being that adequate power supply design is still load agnostic.

 

Those aren't loads, they are power sources.
So which are you talking about?

Think of a resistor as a load in the sense that Vout/Iout is constant, not as a power load. This is an assumption when deriving the relationship between the output and input voltages (or output and input currents) in most introductory Power Electronics manuals. The voltages relationship equals to the duty cycle in a buck converter, for example, provided that the converter operates in continuous mode.
And this is the assumption on which most of the manuals are based. How can you deal, for instance, with a a R-L load? In this case Vout and Iout are not as directly related as employing a resistor load.

 

The difference between a brave man and a coward is a coward thinks twice before jumping in the cage with a lion. The brave man doesn’t know what a lion is. He just thinks he does.
Charles Bukowski.

Think of a resistor as a load in the sense that Vout/Iout is constant, not as a power load. This is an assumption when deriving the relationship between the output and input voltages (or output and input currents) in most introductory Power Electronics manuals. The voltages relationship equals to the duty cycle in a buck converter, for example, provided that the converter operates in continuous mode.
And this is the assumption on which most of the manuals are based. How can you deal, for instance, with a a R-L load? In this case Vout and Iout are not as directly related as employing a resistor load.

 

Think of a resistor as a load in the sense that Vout/Iout is constant, not as a power load. This is an assumption when deriving the relationship between the output and input voltages (or output and input currents) in most introductory Power Electronics manuals. The voltages relationship equals to the duty cycle in a buck converter, for example, provided that the converter operates in continuous mode.
And this is the assumption on which most of the manuals are based. How can you deal, for instance, with a a R-L load? In this case Vout and Iout are not as directly related as employing a resistor load.

If you are talking about a DC output then aside from stability, the output voltage should still be closely related to the duty cycle although the current may not be.
For a resistive load, the current will be Vout/RL of course, and for a resistor in series with an inductor the current will be Vout/R if the ESR is low.

 

DC/DC converters often put plenty of filtering on the output. This is usually sufficient to mitigate any issues with excessively reactive loads. If not, then the design must be re-considered.

Keep in mind that most textbooks use ideal cases as illustrations (think Spherical Chicken Joke). This is fine for an introductory course and it's been that way since schools started teaching Electrical Engineering. You're not likely to change that any time soon.

Most engineers learn real life situations on the job AND, most integrated circuit manufacturers supply plenty of real-world examples for the use of their products as well as applications engineers to guide the way.

 

DC/DC converters often put plenty of filtering on the output. This is usually sufficient to mitigate any issues with excessively reactive loads. If not, then the design must be re-considered.

Keep in mind that most textbooks use ideal cases as illustrations (think Spherical Chicken Joke). This is fine for an introductory course and it's been that way since schools started teaching Electrical Engineering. You're not likely to change that any time soon.

Most engineers learn real life situations on the job AND, most integrated circuit manufacturers supply plenty of real-world examples for the use of their products as well as applications engineers to guide the way.

In addition, most power supplies ARE rather "load agnostic", that is to say that the output filtering does keep reactive loads from affecting the way the supply delivers the required power. Perhaps a special case might be a treadmill motor supply, although even there the motor load does not have a large effect on the supply operation.
So I am wondering about what sort of load would make what sort of power supply misbehave??

My guess is that it would be a poorly designed supply for which the main design target had been minimum cost.

 

I've had more problems with modern DC/DC converter overload protection than direct instability with reactive loads. Many of them designed for steady loads have hiccup overload protection that for a typical motor start load or a heating load with low element cold resistance causes the supply to shutdown the output completely at some current limit, recover, and repeat the hiccup (some will fault until power is cycled) instead of going into a predictable current limiting stage with output voltage reduction. It's very important to specify and chose the type of protection needed for the application.

https://forum.digikey.com/t/what-is-hiccup-mode-hiccup-protection/4212
What is Hiccup Mode / Hiccup Protection?

https://www.microsemi.com/document-portal/doc_view/131948-an-8-hiccup-mode-current-limiting

Hiccup-mode is a method of operation in a power supply whose purpose is to protect the power supply from being damaged during an over-current fault condition. It also enables the power supply to restart when the fault is removed. There are other ways of protecting the power supply when it is over-loaded, such as the maximum current limiting or current foldback methods. One of the problems resulting from over current is that excessive heat may be generated in power devices, especially MOSFET’s and Schottky diodes and the temperature of those devices may exceed their specified limits. A protection mechanism has to be used to prevent those power devices from being damaged.

https://www.nisshinbo-microdevices.co.jp/en/faq/081.html

 

In addition, most power supplies ARE rather "load agnostic", that is to say that the output filtering does keep reactive loads from affecting the way the supply delivers the required power. Perhaps a special case might be a treadmill motor supply, although even there the motor load does not have a large effect on the supply operation.
So I am wondering about what sort of load would make what sort of power supply misbehave??

My guess is that it would be a poorly designed supply for which the main design target had been minimum cost.

Hi,

A boost converter. These can be nasty because of the way the control works. The duty cycle actually gets smaller for a higher output voltage or current.
There are procedures for checking this in theory but it gets a little involved.

 

Posts #10 and #11 demonstrate why some applications require careful selections of the power supply type and ratings. Not every type of power supply design is suitable for every variety of application, and certainly some power supplies handle momentary overloads much better than some other varieties.

This discussion shows why it is in many cases that power supply design selection should be done by an EE that understands the application, rather than by a purchasing agent or an accountant, or an MBA style of supervisor.
It also shows that an accurate understanding of the anticipated power supply loading is important for selecting the power supply capabilities.