In general, the term linear, applies to any device or system where the relationship between the input and the output can be described by the equation of a straight line.Suraj143 said:*What’s the meaning of LINEAR devices? Why they call like that?
*What is the meaning of switch mode power supply? What’s the difference between normal power supply and a switch mode power supply?
Thanks
y = mx + b
where m is the slope of the line
and b is the y-intercept
Sorry, but that's completely untrue! - linear power supplies work in that exact same way, using feedback to correct for any variations in the output.TheVictim said:Switching power supplies work in a similar manner to the Phased Locked Loop oscillator in that the actual output is compared with what the output SHOULD be, which generates a reference value known as the Error Voltage. This is feed back into part of the input to correct the problem, until there is no error voltage coming from the comparison circuit. This is why switching power supplies don't vary their voltage and PLL synthesized tuners don't frequency drift.
Have a look:speakerguy79 said:Actually, the true definition of 'linear' in linear systems is this:
f(x) + f(y) = f(x+y)
Which is quite interesting, because contrary to the completely natural seeming assumption that a line equation would obviously be a linear function, y = mx + b is surprisingly not a linear equation, unless b = 0. Technically, a linear system cannot have a DC offset.
This is, of course, if I remember my linear systems and signals course from college oh so many years ago. I only remember this because of a question essentially based on this on a test that killed 90+ % off the class.
Must have been a third rate school with third world TA that slipped that wowser by you. Are you unfamiliar with the concept of translation of the coordiante axes as a linear operator?speakerguy79 said:Actually, the true definition of 'linear' in linear systems is this:
f(x) + f(y) = f(x+y)
Which is quite interesting, because contrary to the completely natural seeming assumption that a line equation would obviously be a linear function, y = mx + b is surprisingly not a linear equation, unless b = 0. Technically, a linear system cannot have a DC offset.
This is, of course, if I remember my linear systems and signals course from college oh so many years ago. I only remember this because of a question essentially based on this on a test that killed 90+ % off the class.
I don't think you are applying the definition right. Look at the Wikipedia definition. That's what I see in all signals books.speakerguy79 said:Sorry, top ten engineering school in the country. Maybe you guys should go through the math before telling someone they're wrong. Example:
f(x) = 2x + 5
f(2) + f(4) = 9 + 13 = 21
f(2+4) = 17
f(2) + f(4) != f(2+4)
Since f(x) + f(y) must equal f(x+y) for a system to be a linear system, f(x) is not linear (as far as linear system theory goes; it's still a straight line on graph paper though).
This is straight out of the B.P. Lathi textbook that most colleges use.
This is a good general conception of linear versus switch-mode. I would only add that the more efficient switchmode power supplies can be made smaller (due to smaller magnetics due to higher frequency of operation) and that can be critical in some size demanding applications.speakerguy79 said:Yeah, I can be a smart ass for sure
But to answer the OP's post:
Linear regulators take a noisy voltage in, step it down usually by 2V or more to your desired voltage, and filter out a lot of the noise. They are inefficient because they dissipate power in direct relationship to how much voltage drop you have across them, and the current through them.
Ex. A regulator with 35V in and 5V out will dissipate a whopping 45W when supplying 1.5Amps.That's 45W dissipated as heat to 7.5W delivered as power This is of course an extreme example and no one would ever use a regulator like this.
The big plusses with linear regs are very very low ripple and noise, and good output accuract (1% easily acheivable). Also really really cheap. Good to use in analog cktry.
Switchers - don't know much about how they are designed, but they are much more efficient than linear regulators with the drawback of high frequency noise introduced onto the power supply lines. Switchers actually increase in efficiency as output power goes up, maxing out at about 80% of their full output power (at anywhere from 75 to 95 percent efficiency). Switchers can be quite a bit more expensive.
I only use switchers where efficiency is a concern or I need to generate multiple and/or positive and negative voltages from a single DC voltage source. Other than that I always use linears.
HI,TheVictim said:I'm waiting for a small package I can plug into a circuit like a 78xx regulator but is actually a tiny switching mode supply. That way if you want 5v and all you have for supply is 12v, you aren't wasting a buttload of power as heat.