Instructables

Operational Amplifier not working?

Hello,


I'm trying to use the KA324 IC (which consists of four opamps) for amplifying an audio signal.

I've tried almost every schematic from this website: http://sound.westhost.com/dwopa.htm
and also from lots of other websites, but none of them work.

If the output volume is Really low, the sound is pretty good, but very quiet.
When I turn up the volume, the sound is really bad, I almost can't recognise the song I'm playing.
The output volume is also not that really high. If I keep increasing the input voltage, the output stops increasing after a while. It's not really 'loud' at that moment too.

I've connected the IC to 12V DC (between the two middle pins). The + input goes to the ground connection, the - input goes to my POT voltage divider of the input voltage. The output pin goes straight to the speaker.
I also added a 100kOhm resistor between output and - input. Without that resistor, I hear nothing at all but I think that's normal.

I've also tried the three other op-amps from that IC but no luck.

Is this opamp just not ment for audio amplification? or am I doing something wrong?


Greetings,

Electorials

AC
Electorials (author)  steveastrouk2 years ago
so then I'll have to add a DC voltage to it so that the op amp can also amplify the 'negative' voltages right?
No, you just bias ground to be at Vcc/2

This STILL won't give you any power,

Steve
Electorials (author)  steveastrouk2 years ago
I know I won't get any power from it.

So it would work if I add a 2x 10k voltage divider (from Vcc to GND), and connect the input's ground connection to the centre?
It will certainly going to help, though your voltage divider values should be << your feedback resistor values.

Steve
Electorials (author)  steveastrouk2 years ago
ok
Get it working like this, and then you can use an opamp to MAKE a ground.

Steve
Electorials (author)  steveastrouk2 years ago
What will the opamp's output voltage be in this case: (case 1)
GND = 0V
Vcc = 12V
Vin+ - Vin- = 0V
will it be 6V then? or 0V?

and in this case: (case 2)
GND = 0V
Vcc = 12V
Vin+ - Vin- = << 0V
will the output then be just smaller than 6v or will it be 0?

and in this case: (case 3)
GND = 0V
Vcc = 12V
Vin+ - Vin- = >> 0V
output just larger than 6V or just larger than 0V?
If both inputs are connected to your reference ground, the output will be at the same potential as the reference ground, which is 0V relative to that, or Vcc/2 relative to the - terminal on your battery.
Electorials (author)  steveastrouk2 years ago
Ok, I've just tested this and this is what I get:

Vcc = 17V DC
GND = 0V
'fake' GND = 8.50V (no matter what load I add, this stays constant, so that's already working very well, by using opamp number 1)

I connected both inputs of opamp number 2 to the 'fake' GND which is at 8.50V
Vcc is 17V for opamp 2 and GND is connected to the real ground.

Results:
output = 15.73V (output voltage of opamp 2 compared to the real GND)
output = 7.22V (output voltage of opamp 2 compared to the real GND)

It doesn't seem to work like you say :/
IMG903.jpg
Put some feedback resistors in !!! At the moment its got close to infinite gain.
Electorials (author)  steveastrouk2 years ago
Ok, I did that (see picture) and it works.

Is it normal that the first resistor always needs to be larger (100k) than the second resistor (10k)?

and I have another problem now: the output is inverted.

If I connect the input to Vcc I get -8.5V
If I connect the input to fake GND I get 0V (several millivolts)
If I connect the input to GND (the real GND) I get +8.5V

I want it to be the opposite.
I just tried switching the + and - from the second amp but then it doesn't work anymore.

What should I do?
IMG904.jpg
Electorials (author)  Electorials2 years ago
oh no ....
I did something wrong here, sorry
it's fixed now (looked up non inverting opamp, so I added another resistor)

but I don't get 0V anymore, it's 1.5V when I connect the + to the fake GND
You are running before you can walk. The input bias current, and input offset voltage of an LM324 class amplifier is quite high. Reduce the value of your feedback resistors, but keep them in the same ratio
Electorials (author)  steveastrouk2 years ago
It's working!

Thank you very much for helping me with this, and sorry for al the stupid questions.

When I have a problem, I instantly ask it here, and a second later I fixed it myself already :/ I should learn to stop doing that ^^
I'll have a thought about the problems I have myself first in the future :)

This is the schematic of the pretty good working amp:
IMG905.jpg
READ UP !

You have made an inverting amplifier. For many purposes - like AC work, it doesn't matter.

Look at how to wire a non-inverting amplifier.

Steve
Electorials (author)  steveastrouk2 years ago
Oh, that's what they are used for!
I'm just new in opamps, and I didn't see what their function was. I think I understand now. I'll try it :)
Get some books - like the "Art of Electronics" by Horowitz and Hill

Steve
Electorials (author)  steveastrouk2 years ago
I prefer internet :/ but thanks
"Internet" does not have all the knowledge you need. You need to read books too.

Steve
Electorials (author)  steveastrouk2 years ago
That's true
Electorials (author)  Electorials2 years ago
with 'output voltage' I mean the voltage between 'output pin' and GND (0V)
gmoon2 years ago

More than likely the KA324 isn't designed to drive a speaker load--it has a different (higher) output impedance. And low current handling...

I took a quick look at the datasheet, and in the specs it gives the two values (under Voltage Swing) for load resistance--2K and 10K.  You're using a speaker as a load--what's that, approx 8 ohms? Pretty far from the load.

In short, a general opamp isn't a power amplifier. This would be an appropriate chip for a preamp, or a mixer or something similar. Something early in the signal chain. But it's not the chip to drive a speaker load.
Electorials (author)  gmoon2 years ago
Ok, thanks

I was a little confused because I took apart PC speakers, and I gathered an IC from it which had two opamps in it (one for each channel) and it worked so great!
The output was really loud.
Then I accidently made a wrong connection and I broke it.

After that, I started searching for another IC with opamps and I found this one.
I was wondering why the old one worked perfect and this one didn't.

So where do I find those power amps suitable for audio? only in audio applications like the PC speakers? or can I also find them somewhere else?

Probably wasn't an opamp you found, but an audio amp IC.

There are lots of dedicated amp ICs: LM386, LM389, TDA2003, TDA2005, TDA7294, etc. Several ICs in the TDAxxxx line--some are decently high wattage.

(Lots of online experimenters refer to LM386 amps as "op amps," but they are not).
Electorials (author)  gmoon2 years ago
It 'was' a TEA2025B :)

I accidently connected a capacitor between Vcc and the output instead of between output and that bulk thing.
Hey, we've all "bricked" a few ICs in our time...

There you go--that's really what we've all been saying ;-). It's an audio amp; one that only requires a few extra discrete components. And it's specifically designed to drive a speaker.

You shouldn't need an extra op amp preamp, unless you're using it for a low-power, high impedance signal like a guitar pickup. But if it worked with your setup before, then you already know that.

Incidentally, a bipolar (split rail) power supply isn't necessary with many op amp circuits, unless they are drawing higher current, or they are old op amps--or sometimes if they are "instrument amps" and require a very stable voltage/reference.

Instead, most op amp audio preamps (and audio effects) just use voltage divider to create a "signal ground" reference--i.e., they place two resistors (maybe 10K or 20K) between the pos and neg power supply. The voltage in the center is midway--for example, for 9V power supply the ground reference would then be 4.5V. Then ALL the ground references within the op amp circuit would use that 4.5V as the ground.

Of course, the V+ and GND pins of the op amp would still be connected to the full voltage. And one needs to be careful with a divider "ground reference" if the power supply is being shared with multiple devices in the signal chain...
Electorials (author)  gmoon2 years ago
Thanks! that explains a lot.
won't the 4.5V 'fake' GND voltage drop drastically when a load (a 10ohm speaker) is added to the circuit?

The power amplifier would be an additional stage, after the op amp preamp. It wouldn't utilize the divider voltage reference as ground. The preamp stage is strictly "small signal," so little current is necessary to shape the waveform.

If the main power supply "sags," then the divider as ground might be a problem. But that's usually only a characteristic of cheap, poorly designed amplifiers w/ underrated power supplies (or of old tube amps ;-).

Interstage coupling would take care of any bias issues between the stages. If you're using the same 9V power supply for multiple stages/devices, you just need to be sure everything is capacitive coupled.

But if the TEA2025B worked for you before, you shouldn't need the op amp preamp at all. Just find a similar "general purpose" audio amp IC.

I was only making a comment about bipolar power supplies and op amps in general...
Electorials (author)  gmoon2 years ago
Yes, If I get one like the TEA2025B again, I'll just use that one, but now I'm trying to figure out some stuff about regular opamps (see conversation above) :)
The GND from the voltage divider is made, and works great already ^^

Now I'm trying to figure out how to use a second opamp from inside that same IC, that can use the fake GND from the first opamp, to create positive and negative voltages as output.
Cool, I'm glad it's progressing. Looks like you're getting a "crash course" in op amps. There's lots to learn, but you're well on the way.

No problem chaining op amps together. If you're using them in "inverting amplifier" configuration, each stage inverts the previous one. So even numbers of inverting stages result in the same phase as the original signal.

Normally you would want to capacitively couple the stages together, but direct coupling can work too. If you "direct couple" two op amps, you'll need a resistor between the two--without it the gain is "infinite." I've used 4.7K with good effect (but it depends on the op amp and the application).

I'm always more comfortable with capacitive coupling 'cause they form filters (and the theory is easy to understand), but I've used direct for at least one op amp project and it worked fine. If you have a strong input signal there would be less headroom with direct, however.

RE: inverted signals--unless you're mixing signals together, having an inverted output isn't a real big deal for audio.

(Also--RE: your "op amp/transistor" amp above--you'll need to block DC offset somewhere in that circuit after the op amp with a cap...)
Electorials (author)  gmoon2 years ago
Ok thanks ;)
Sure, no prob. Using a capacitor on the output of any of your amps is probably a must (look at the datasheet for the LM386 to see what I mean).

Without a cap, the offset DC bias voltage becomes "standing DC" on the speaker, which increases current draw and ruins the audio. Capacitors pass AC and block DC...
"Audio" is not the same as "power amp". The old NE5534 op-amp is a superb AUDIO signal amplfier, but it is NOT a power amplfier.
Since steve and g have pointed out that this is NOT a power amplifier - rather it would be better suited for a preamp (that's to bring the signal up to a usable level FOR the power amp) - I would also like to point out that it sounds as if you have connected the power backwards. Positive should go to pin 4 (VCC, middle on the left as you are looking down on the top) while negative should go to pin 11 (GND, middle on the right as you are looking down on the top. What purpose does the voltage divider perform? 12V is within it's range.

For further help, you should post a pic or actual schematic of how it is currently hooked up. Just draw the opamp as a block (with pin 1 marker) and the pins coming out and show how things are physically hooked together. This would tell us a lot about your build.

Qa
Here's the PDF of the datasheet. It does not need a split supply, though it can use one. However, It has been pointed out that it is a small signal device, not a power device. Follow the advice given elsewhere in these answers and look for a stereo audio power amp IC that uses 12V. Do you still have the original? If so, what are the numbers on it?

Qa
Electorials (author)  Quercus austrina2 years ago
Yes, I read the pdf but just for the connections and maximum voltage rating.

The old IC was a TEA2025B

Well, you are in luck as a quick search shows that Hobby Engineering has the original TEA2025B in stock for a paltry $2.00 US. I'm not sure what that is in Euros at the moment. That would be the best thing to do if you are still trying to repair the original board.


Qa

Electorials (author)  Quercus austrina2 years ago
but I don't want to fix the speakers. They worked fine, but I didn't use them anymore. I only wanted the amplifier IC from it.
But that would allow you to use the original board which already had all the engineering done and worked. If you want to do something else with it, then by all means, do so. I was under the impression you wanted to use the original board.

Qa
Electorials (author)  Quercus austrina2 years ago
hmm yes but I only needed the IC from it :/
to make my own amplifier.
Yes,. absolutely right. Electrorials: Don't TALK circuits, draw them.

steve
Electorials (author)  steveastrouk2 years ago
I added it below.
Electorials (author)  Quercus austrina2 years ago
I'll post a picture of it this evening.

And I didn't cunnect the power backwards. I must say that the position of Vcc and Gnd is weird because mostly Vcc is on the top left (right in the datasheet) and GND on the bottom right(left centre)
An op-amp is NOT suitable for "power" amplification, and it sounds like you are operating it "single ended", when it really needs either split supplies, or special input biasing arrangements.
Electorials (author)  steveastrouk2 years ago
Ok, I also tried split supplies but that didn't work. Thanks for the reply :)
Then as I suspected you are trying to use an opamp as a power amp.: It won't work. It won't work as a signal amplifier without split rails either
Electorials (author)  steveastrouk2 years ago
Ok. I also tried connecting the output to a transistor but the sound was still terrible.
Don't keep talking about a circuit, post a diagram of what you have. An amplifier isn't built by slapping in a transistor without any design.

Steve
Electorials (author)  steveastrouk2 years ago
Could you also please explain me this?

What type of voltage is coming out a 3.5mm audio plug,
referenced by the thickest metal piece of the connection?
- variable DC voltages? (only positive voltages)
- variable AC voltages? (so also negative)
Electorials (author)  steveastrouk2 years ago
I also tried multiple resistor values for the 100k one and the 1k one.

I can get sound from the opamp if I connect the speaker from output to ground (so without the transistor) but that's very quiet. With the transistor, the sound is louder but really crappy.
IMG902.jpg
sshuggi2 years ago
By the sounds of it, your opamp is only operating in the positive... I've connected the IC to 12V DC (between the two middle pins) The opamp needs something like a +6V and a -6V in the power terminals as apposed to +12V and 0. That way, i can actually amplify sounds when the signal swings into the negative. And it does with audio.

Another reason it might be sounding bad is you're clipping the signal. Say your gain(amplification) is x15, and you have a waveform input that reaches+/-1V. With your current configuration, it will try and amplify the signal to 15V, but all it has to work with is 12V on the positive side. So, the signal is amplified normally up to 12V and then flat lines from there until the input goes back down. With any negative voltage signal, all it can do is output 0V... This is possibly what is giving you poor sound, especially when you try and turn it up.

Like everyone else is saying, this configuration only gives as much power as whatever is powering the opamp. If that's a battery, don't expect much more out of it that mini speakers.

Something to think about: Before you go hooking everything up to a massive power supply, check the ratings on you parts. The opamp has a limit to the current that it can output and the speaker has a max wattage.
P=I2R   The standard minispeaker has R≈10 Ohm and P≈0.5 W. Plugging this into the equation, the max current is something like 220 mA. There's probably other things to consider, but just know the limits of your parts.
Electorials (author)  sshuggi2 years ago
Oh yes, aboit the clipping, that's possible, but it doesn't start clipping at the Vcc voltage (12v) but a lot earlier (5v or something like that). It's probablybecause the output current for the 10ohm resistor is too high. However the IC doesn't even get warm, but I think there must be an internal current limiter.

You also say I need positive and negative supply voltage. It sounds logical, but with my previous opamp (audio intended opamp) I also used +12v and 0v and the sound was perfect. Tye IC was a TEA2025B. In it's original divice it also didn't have positive and negative voltage i think because it was just a regular 6V transformer.