EveryCircuit - Class A amp.

8 years ago

Looks good. A few changes I would recommend would be to change the left 1uF capacitor (the next to the sine wave signal source) to 47uF and the left 100nF capacitor (the one hooked up to the OP-amps output) to 22uF

jason9

8 years ago

Actually, it's 22nF, not 22uF.

jason9

8 years ago

Actually, even better, instead of changing the 100nF capacitor change the 1mF capacitor on the output to 4.7mF to improve the frequency response on the low end.

jason9

8 years ago

Something like this: http://everycircuit.com/circuit/4590284219613184

selman

8 years ago

High continues power dissipation through the MOSFETs

pajaka

8 years ago

Ofc, it is A class

hurz

8 years ago

@jason9, have you sorted all your capacitor recommandations? One word about the output coupling one. To support 20Hz at a -3dB corner at 8Ohm the formula for the cap is 1/(20Hz x 2PI x 8Ohm)=995uF. So 4.7mF is good for about 4.23Hz and quite over engineered!

jason9

8 years ago

Yes, @hurz, but there was a quirk of high gain for the lower frequency end and the bigger capacitor is to but more stress on the amp at lower frequencies to help reduce that quirk.

hurz

8 years ago

Quirk? You mean you started tuning from backend to fortend to find out the backend values are wrong and start all over at tuning the frontend ones? Turning in circles is a Quirk? Just take some math and formulas and make your calculation on given values and facts. No quirk will happen.

jason9

8 years ago

What I mean is that if you look at the bode plot of this amp on the wire before the output capacitor then you will see a higher gain at the low frequency end than at the middle or higher frequency end. This would cause the amplifier to clip even if the 1mF capacitor made a matching decrease in gain at the lower end of the frequency spectrum. I simply messed around with some values and it seemed to work. That's all.

jason9

8 years ago

Also, I cannot find the source of this gain increase at the lower end. No matter what capacitor I increase to extend its effect to the middle and higher frequencies, the mysterious gain is still there.

jason9

8 years ago

Never mind. It turns out that the mysterious gain occurs because the amp is not being loaded as heavily at the lower frequencies because the 1mF capacitor is blocking the signal and because of the nature of the amplifier that causes it to increase in gain at that frequency causing it to clip.

jason9

8 years ago

Look at the bode plot of that wire and play with the capacitor values to see for yourself.

hurz

8 years ago

You just look at the bode plot and play with the caps. Right this is exactly what Im afraid of. Dont play, take formulas and do the math.

2ctiby

8 years ago

It's no good telling him to do the Maths Hurz, since you haven't a clue on how to set up a Mosfet for amplification.... you rant about 1/2pi f r etc...but why did you suggest that, when the Kn or beta etc had not been set, nor Q configured properly.

hurz

8 years ago

Wow, a troll in action. Open a new thread and let the world learn from your trollish knowledge.

jason9

8 years ago

Here are some things I think you should learn: 1, I don't do math, and 2, math doesn't always get it right or something is being excluded that causes the math to fail. In this case, the way the amplifier works is that the top MOSFET is a constant current source while the bottom MOSFET acts as a voltage controlled current source. Since there's no feedback, this makes it so that the amplifier outputs a certain amount current depending on the input. At low frequencies, the output capacitor starts to act as a significant impedance, and a fixed current combined with a higher impedance means higher voltage. This causes clipping. The only way to get rid of this is to either add a negative feedback, possibly requiring a total redesign, or to make the capacitor bigger. I chose to make the capacitor bigger.

hurz

8 years ago

@2cent, when can we expect some clever calculation you present for an output coupling capacitor 8Ohm loudspeaker and 20Hz lower corner frequency???

2ctiby

8 years ago

I didn't say that I would do that Hurz....I said that you have no idea about how to start with a Mosfet amplifier...When are we going to see you do that instead of being a fraud pretending that you can?

hurz

8 years ago

Nice try

2ctiby

8 years ago

LOL http://everycircuit.com/circuit/6551211735252992

jason9

8 years ago

@hurz, as long as your still here, it must mean that you don't yet understand what I'm trying to say. Look at the bode plot if the wire just before the main 1mF capacitor (after you increase the input decoupling capacitor from 1uF to 47uF). There is some exceptionally high gain at the lower end which causes clipping. How do you explain that?

hurz

8 years ago

So idiots, jason recommand 4.7mF as coupling output capacitor to drive a 8Ohm loudspeaker! If you guy just step back and see whats needed and whats over-engineered then we might continue on this topic.

hurz

8 years ago

Here we go and nothing else I said, idiots http://everycircuit.com/circuit/5841475268247552

jason9

8 years ago

Are you actively trying to ignore me? Why don't you actually try what I say?

jason9

8 years ago

I understand what you're trying to say, but do you understand what I'm trying to say?

hurz

8 years ago

Why? Because LOGIC! There is only one valid solution if we specify 20Hz as low corner frequnecy for this amplifier "1mF" ! Below and above is NOT.

jason9

8 years ago

Yes, that works for some types of sources, but not others.

jason9

8 years ago

Here: http://everycircuit.com/circuit/5103165491118080 Orange is voltage source, green is -3dB output as you say should happen with a normal amp, and the blue is the output of the feedback-less class-A amp. As you can see it is clipping because it acts as a current source and that makes it so that the output after the capacitor is always the same regardless of frequency but before the capacitor there is a +3dB causing clipping.

hurz

8 years ago

This is a transconductance amp. Anyway, the 20Hz corner frequency is equal in both cases, but you try to undergo it with this transcoductance concept. Do not listen to the fire raiser @2cent, he is long time gone and just here for his propaganda, somehow pissing around seems to satisfy him.

hurz

8 years ago

http://everycircuit.com/circuit/6088109134708736

jason9

8 years ago

The only thing you changed in my circuit was the gain. Why would you change the gain to make it not clip? The person with the amp would just increase the input to compensate making clipping a problem again. Also, what you said about 20Hz corner frequency is equal in both cases is correct, but in a way they are not because with the amp it causes a voltage rise on the amps side causing clipping because 3dB is actually quite significant, so I have to make the 20Hz corner frequency at a much lower dB by increasing the capacitor to make it less significant. Is a "transconductance" amp the type of amp I'm talking about which outputs a certain current rather than a certain voltage? Also, I'm not listening to @2ctiby, our ideas are simply similar.

hurz

8 years ago

I adapted your 125mA/V to make both amps comparable. I decreased it by -3dB better said 125/1.414=88.4mA/V and set this way both amps at 20Hz corner frequency to equal output power! You transonductance amp is a nice idea, but you get the higher impedance only for the price of higher supply voltage by equal resistive output. And this cause clipping, right. But the original design gives us only single supply 40 while you can spend to be fair +-20V in your amp. If you need more it does cause clipping. Or you change the lower corner frequency to 4Hz by increase the cap. Its your decission. However, Its quite common to specify amps in a range of 20-20kHz.

jason9

8 years ago

Like I said, the decrease in amp gain will make whoever's using the amp go simply increase the input making clipping a problem again. Also, the +-20V vs. +40V doesn't make a difference, the idea is the same. Another way to think of my increasing the output capacitor is not to increase the frequency response to 4Hz on the low end but to increase the quality of the 20Hz lower corner frequency at 3dB to something much lower than 3dB. It's still 20Hz lower corner frequency, just a higher quality one.

jason9

8 years ago

Approximately 0.2dB.

hurz

8 years ago

+-20V dual and 40V single is fair. And if you start clipping you must increase the rootvoltage to e.g. +-30 and/or 60V single to get some transconductance behaviour. But do not in increase the ouput cap and move the lower corner frequency out of spec 20Hz. This is how I would do It.

jason9

8 years ago

Yes, but whoever uses it would just increase the input to take advantage of the higher supply voltage making it a problem again. 20Hz lower corner is simply by your standards of 3dB. This amp, because if that problem, requires a lower corner frequency at the standards of 0.2dB, which at 20Hz and with load of 8ohms makes a 4.7mF capacitor.

gasboss775

6 years ago

Which MOSFETs do you recommend to use in a real circuit?