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.

Simple Audio Amplifier

Status
Not open for further replies.

Frosty_47

New Member
Dear Komrades,

I wanted to build a simple audio amplifier for a very long time. Today I took a first shot at it, and was successful. The output is distortion free on the oscilloscope display from 100Hz to several MHz range. The audio is very clear coming out of 8 Ohm 100watt Pioneer speaker. My input was taken from the back of a computer sound card. Please comment and share your designs.

**broken link removed**

Beethoven symphony No.9 and Rostropovich Prelude from Bach's Cello Suite No. 1 sounded best!
 
Last edited:
What are the FETs?
 
What are the FETs?

P-Channel E-MOSFET = IRF9520

N-Channel E-MOSFET = IRF510

They survived some brutality as I did not heat sink them for the first dozen test runs. The Fets melted through my breadboard, yet they still work perfectly. I attached them both to CPU heat sinks with fan so they stay cool now.

However, I will be getting a matched pair, that way the Drain to Source resistance will be matched so there will be no need for capacitor C1.
 
Last edited:
Congratulations! I think you might be able to s**t can the output capacitor, and just connect the speaker directly.
 
The Mosfets are turned off at idle so there must be massive crossover distortion.
The circuit does not have a method to adjust the idle current that will be very different for each set of mosfets.
The negative feedback for the opamp does not include the Mosfets so their general distortion must be pretty high. If the Mosfets are included in the negative feedback of the opamp then C1 will not be needed.

The Mosfets are source-followers so if the opamp also uses the plus and minus 12V supply then its output is only plus and minus 11V. Then the max output from the Mosfets will be only about plus and minus 6V. Then the max power to an 8 ohm speaker is only 2.3W.

Why use an expensive high voltage and high current power opamp?
 
I second all of AudioGuru's comments.

I applaud the fact that you want to build your own audio amplifier. However, I would suggest that you do a google search on Mosfet Audio Amps. Countless individuals have taken a shot at these amplifiers, and you can learn a lot from prior art, and incorporate the techniques into your own design.

Another recommendation: if you do not have an oscilloscope and a sinewave generator, please beg, borrow, lease or purchase one. It is extremely informative and educational to follow the waveforms in a circuit.

Additionally, once that you've finalized your circuit, again beg or borrow for a distortion analyzer. You'll find again how subtle circuit and/or component changes have an impact on the results.

There is a magazine devoted to do-it-yourself audiophiles, that I strongly recommend if you are interested in the field:
audioXpress
 
Last edited:
Power the opamp from a much higher voltage than the output stage, in this case I would go for +/-18V; obviously check that he op-amp can handle this. If you're using a transformer then you could use a voltage doubler and possible voltage regulator to power the op-amp.

Include the MOSFETs in the feedback loop to minimise distortion.
 
Thank you all for your suggestions!

I have blown IRF510 today, and I am happy that I did because when I swapped it with IRF520 and adjusted VDD to +/ - 30V the output was sounding much better. I think I had the source and drain inappropriately hooked up so maybe that is why it blew. Regarding the feedback, I should be able to take the feedback directly from the output, hopefully this will reduce distortion (not that I noticed any on the oscilloscope or by sound of ear).

Oh and I absolutely love OPA-452. Its best Op-amp I ever used. None the less, I got 4 of them for free from Texas Instruments.
 
Last edited:
Feedback Directly from Output

I just tested my circuit with the feedback taken directly across the load and it sounded horrible! A lot of distortion could be heard. My original configuration sounds clear so I will leave the feedback as it was before...
 
That's wierd, it's just about the opposite of what I would expect.
 
Weird, very weird. Something is not right here.

Could you post the actual circuit that sounded horrible? How about voltage readings? Waveforms?
 
If you connected the feedback to the speaker side of the cap, then you lost your operating point. If so, try to connect to the other lead, or else just remove the cap.
 
That's wierd, it's just about the opposite of what I would expect.

Yes it is weird indeed. However, the above posted circuit is distortion free (at least non visible on oscilloscope display).
 
If you connected the feedback to the speaker side of the cap, then you lost your operating point. If so, try to connect to the other lead, or else just remove the cap.

What do you mean by "other lead" ?

I am afraid to remove the cap because the drain 2 source resistances are very different. This will cause a significant shift of the waveform.

I will look for some quality differential pair mosfets tomorrow, perhaps that way I can get rid of the cap.
 
Last edited:
Weird, very weird. Something is not right here.

Could you post the actual circuit that sounded horrible? How about voltage readings? Waveforms?


The one below had audio out with lots of distortion. The last one had no out at all....

**broken link removed**

**broken link removed**
 
The Mosfets are turned off at idle so there must be massive crossover distortion.
The circuit does not have a method to adjust the idle current that will be very different for each set of mosfets.
The negative feedback for the opamp does not include the Mosfets so their general distortion must be pretty high. If the Mosfets are included in the negative feedback of the opamp then C1 will not be needed.

The Mosfets are source-followers so if the opamp also uses the plus and minus 12V supply then its output is only plus and minus 11V. Then the max output from the Mosfets will be only about plus and minus 6V. Then the max power to an 8 ohm speaker is only 2.3W.

Why use an expensive high voltage and high current power opamp?


How can I adjust idle current?

Wouldn't the waveform still be shifted if C1 is removed due to differences in drain to source resistances ?

I received OPA-452 for free when I ordered samples. It comes in handy now that I use +/- 30V supply.
 
I am afraid to remove the cap because the drain 2 source resistances are very different. This will cause a significant shift of the waveform.
Very wise. Since you are still experimenting it is advisable to leave the cap in place. Without the cap all you need is one wiring mistake and the speaker is toast!
 
Additionally, once that you've finalized your circuit, again beg or borrow for a distortion analyzer. You'll find again how subtle circuit and/or component changes have an impact on the results.

There is a magazine devoted to do-it-yourself audiophiles, that I strongly recommend if you are interested in the field:
audioXpress

Thank you for your suggestive feedback. It is difficult to find distortion analyzer however, I have some quality oscilloscopes in school that have the "MATH" function. So I can clearly see the difference in input/output. I did not see any significant difference when I tested my circuit under CH1 - CH2 Math function.

CH1 being input...
 
Last edited:
Very wise. Since you are still experimenting it is advisable to leave the cap in place. Without the cap all you need is one wiring mistake and the speaker is toast!

Yeah especially that I blew one FET already. However it was fun to see 20+ people duck under the table in the lab when it happened :)
 
Last edited:
Mosfets have a very wide range of drain to source resistances. Unless you measure yours they might be well-matched or be either way mis-matched. The negative feedback should take care of the difference.

You will not be able to buy a "matched pair' of complementary Mosfets. Amplifier manufacturers buy thousands and match them themselves.
Usually the circuit adjusts the bias and feedback on the Mosfets to match them.

Your second circuit with the feedback connected properly to the Mosfets has the Mosfets idling at a very high current that might overheat them.
 
Last edited:
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top