Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Power efficiency of amp IC?

Status
Not open for further replies.

antknee

New Member
How is it best to calculate the power efficiency of an amplifier?

I'm thinking one factor is a= quiescent current x supply voltage

The other factor: b= output current x output voltage

So the efficiency would be: b/(a+b)

Does that look right?

Thanks,

Antknee.
 
Last edited:
Is it a linear IC? If that's the case the efficiency is the same as that of a linear regulator because it's dropping the extra input voltage inside itself to create the output signal voltage, and because of that the input and output currents are the same (and so they cancel out):

Efficiency = 1 - [(Vin-Vout)*Iout]/[Vin*Iin] = 1 - (Vin-Vout)/Vin

Iin = Iout so they cancel. If you want to take into account quiescent current, then the formula chances slightly, but a lot of the time you can ignore it because it's insigificant compared to the output current:

Efficiency = 1 - [(Vin-Vout)*Iout]/[Vin*(Iin+Iq)]

The bolded term is the power being wasted as heat because the extra supply voltage dissipated as heat to produce the output voltage. The efficiency is changing all the time because the signal is changing. THis means if you want to calculate efficiency you kind of have to know what your waveform looks like. If efficiency is your concern you shouldn't be using a linear amp. THe goal there is linearity, not efficiency. From this though, you can see that amp runs coolest when the current is low (which makes sense) and the magnitude of the output voltage is high (a bit counterintuitive).
 
Last edited:
I'd say it was a linear amp, although I have just looked up the terminology because I'm not familar with it. I was thinking in particular about the OPA561 hi speed hi current operational amplifier. It is rated at input voltage 7-15V output 12Vpp with up to 1.2A.

The waveform doesn't have to be perfect either square or sine would be fine for input or output. What are non linear amplifiers?

I wasn't expecting a chip to run coolest when the current is low and output voltage high, that is counterintuitive.

Thanks.
 
Last edited:
THere are no non-linear amps (we want amplifiers to have a linear input-output response!). The word linear in this case is referring to the construction of the amp and not the input-output response.

Do you know the difference between a linear regulator and a switching regulator? Same difference between a linear amp and switching amp. A linear amp will be all silicon and output the actual analog signal. It works by burning the excess voltage supply as heat inside itself so the closer the output voltage is to the input voltage, the less voltage it will need to burn and it will run cooler. It will also run cooler with lower current because the power being burned off is also proportional to the output current. Just like linear regulators. They are quite similar. The voltage refernce inside the linear regulator is like the input of a linear amp, and the output is the output.

A switching amp will require output PWM digital waveform of some instead of the true analog signal and probably requires some kind of output filter (inductors and capacitors) to get rid of the square-wave switching parts of the digital waveform leaving behind only the average which is the desired analog waveform. THey are much more efficient but larger and more complex and more possibly difficult to build linearily with a good frequency range.
 
Last edited:
Using your equation with Iout=1.2A, Vs=15V, Vout=12V
So the wasted energy = 3x1.2 = 3.6W

That is interesting. Thanks.
 
Ahah! I have an analogy! It's like if you (the amplifier) needed to carve a giant block (input voltage) into a statue (output voltage). THe complexity of the statue aside, it's going to take more work to carve a small statue than a large one since you have to carve out more material. Just because the final statue is small doesn't mean it took less work to make than the large one. Because of the starting materials, the final size is not representative of the work that went into making it. More work = you sweat more.
 
Last edited:
Do you know the difference between a linear regulator and a switching regulator?

I don't know the difference between the two. I'm still surprised for efficiency I should be looking for high volts and low current, rather than the other way around.
 
I don't know the difference between the two. I'm still surprised for efficiency I should be looking for high volts and low current, rather than the other way around.

You're probably driving some kind of pre-determined load so you don't get to pick the required output power, voltage or current. Even if you did get to the pick the load, the application probably requires a certain amount of power and in that case it doesnt matter how you pick the voltage or current because decreasing one increases the other- the power dissipation in the amp will be the same. All you can do is pick the amp supply voltage and even then you won't do much better than picking one that is just a little bit higher than the maximum output voltage.

And it's still going to generate a lot of heat for low output voltage signals because there's more excess voltage from the supply to be burned off as heat.
 
Last edited:
I'm driving a transformer that is 3.2ohms to 1.2Kohms. I work at around 100KHz and tend to get slew rate or GBP limited with the chips I use to output at most 10Vpp. So I need a large current to up the volts.

As my device is portable it is hard to supply 80V with batteries. I would be happy with 80Vpp, it is just that the transformer I'm using offers more than I actually need.
 
Last edited:
High power linear amps levels require heatsinks (have you ever looked on the back of stage speakers? Those giant heatsinks? It's because the speakers were built for linearity with power consumption as a only a secondary goal). If you want lighter and more efficient, switching is the way to go, but I do believe you tend to sacrifice a bit in frequency response and linearity and have an increase in parts count and complexity.
 
Last edited:
I have in the past used audio amps and had to put quite big heatsinks on them. As I don't need a perfect signal I will investigate switching.

I wondered when I would have cause to find out why there were different classes of amplifier. Thank you for your comments.
 
Last edited:
A switching amp is also known as a Class-D amplifier. Class A is the linear amplifier. Each is on the extreme end of trade off scale between efficiency and bandwidth/linearity. THe classes B,C, and AB are somewhere in between (but are more similar to the class A than the class D. These are terms that tend to be used by the audio crowd though so it might take some work to find one that works at your frequencies.

The class E and F amplfiiers are the switching type and more similar to the class D-amplifiers but are apparently rarer (I'm just grabbing this stuff off Wiki now). While the B,C, and AB amplifiers have traded signal integrity for efficiency relative to the class A amp, the class E and F might offer both signal integrity and efficiency improvements on the class D.

https://en.wikipedia.org/wiki/Electronic_amplifier#Power_amplifier_classes

Class G and H is interesting. It's like a linear amplifier (either A, AB, or B) being used with efficient switching regulators on the rails that adjust the rail to be close to the output voltage to maintain linearity and bandwidth but increase efficiency. Sounds like a lot of work.
 
Last edited:
I'm currently on wiki. It will take a while to digest. I have never looked for class when buying chips, do datasheets note that? I will read up on Class D. Thanks.
 
Last edited:
Some datasheets might note it explicity as a class. It'd be easy to tell if it was one of the switching types (D,E,F), one of the linear types (A,B, AB, C), or one of the hybrid types (G and H). They'll probably tell you if it's switching or hybrid right off the bat. They might not say if it's linear though since it might be so common as to be a given if nothing else is said. The actual class might not be specified. But you can find out or at least narrow it down by looking at how the amp works.

You can go about it by picking the performance and efficiency you need whch indirectly refers to the class, or picking the class which indirectly refers to the performance and efficiency. So class is not always explicity given.

Just note that switching types are more complicated to design for becayse more work is left to the user as it requies external components. That's usually the case with switching anythign vs linear anything.
 
Last edited:
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top