# I can't find the problem. 2 stages amplifier with op amps.

Status
Not open for further replies.

#### Hellen Grig

##### New Member
I have the following circuit. I have some problems and I can't identify the reason.
1.Why I have such a distorsioned output voltage?
2.Why Vmax is aprox is 25V ,but Vmin is 29V?

#### Attachments

• 92.9 KB Views: 25
• 70.5 KB Views: 24

#### Ylli

##### Active Member
I don't know what you are using for sim software, but it looks like your sampling rate is way too low. Also, if you are looking for 27 v RMS output, that translates to 27 * √2 = 38.2 volts peak. That will require *at least* +/- 40 volt rails.

#### gophert

##### Well-Known Member
I have the following circuit. I have some problems and I can't identify the reason.
1.Why I have such a distorsioned output voltage?
2.Why Vmax is aprox is 25V ,but Vmin is 29V?
Your first input stage has an RC low pass filter that allows everything under 3500Hz to pass and you are trying to amplify by 60x

Then the second stage has an input filter cutting off everything below 7500Hz and you are trying to amplify with a gain of 90+.

So you've effectively insured nearly no signal will be left in your output.

Finally, you are feeding it a 125Hz signal.

And, to top it all off, you are concerned with the clipping voltage?

#### Hellen Grig

##### New Member
Your first input stage has an RC low pass filter that allows everything under 3500Hz to pass and you are trying to amplify by 60x

Then the second stage has an input filter cutting off everything below 7500Hz and you are trying to amplify with a gain of 90+.

So you've effectively insured nearly no signal will be left in your output.

Finally, you are feeding it a 125Hz signal.

And, to top it all off, you are concerned with the clipping voltage?

At imput stage I have RC high pass filter that allows everything below 350Hz and at the second stage a lowpass filter the allows everything under aprox 7300Hz, or am I wrong?

#### Hellen Grig

##### New Member
I don't know what you are using for sim software, but it looks like your sampling rate is way too low. Also, if you are looking for 27 v RMS output, that translates to 27 * √2 = 38.2 volts peak. That will require *at least* +/- 40 volt rails.
I supplied the second amplifier with +/-40 and the problem remains. I have an offset and I don't know how to fix it.

Last edited:

#### ronsimpson

##### Well-Known Member
I have an offset and I don't know hoe to fix it.
1) I changed the RC on the input of the LM741 to make it more like an audio amplifier.
2) The LM741 is not perfect. The data sheet: "input offset error voltage" is typically 1mV, worse case 5mV at room temp and could be as bad as 6mV. What that means is that the input is not perfect and there typically 1mV of error on the input pins. (could be as bad as 5mV) With your gain of 62 then the 1mV=62mV and 5mV=400mV of error at the output pin. (SPICE might not show this problem) Then you amplify again and the error is BIG. So I put a RC high pass filter between the amplifiers to remove the DC error.
3) The OPA501 also has in input error. Did not look it up but you will have the same problem here. So I added a 10uF cap to the bottom of R3. This causes the gain to decrease below 20hz. By DC the gain will be one. So the input offset error will be amplified by 1 not 93.

These fixes will help the offset.
--edited--
R6, C2 should be reversed.

#### Hellen Grig

##### New Member
1) I changed the RC on the input of the LM741 to make it more like an audio amplifier.
2) The LM741 is not perfect. The data sheet: "input offset error voltage" is typically 1mV, worse case 5mV at room temp and could be as bad as 6mV. What that means is that the input is not perfect and there typically 1mV of error on the input pins. (could be as bad as 5mV) With your gain of 62 then the 1mV=62mV and 5mV=400mV of error at the output pin. (SPICE might not show this problem) Then you amplify again and the error is BIG. So I put a RC high pass filter between the amplifiers to remove the DC error.
3) The OPA501 also has in input error. Did not look it up but you will have the same problem here. So I added a 10uF cap to the bottom of R3. This causes the gain to decrease below 20hz. By DC the gain will be one. So the input offset error will be amplified by 1 not 93.
View attachment 118182
These fixes will help the offset.
--edited--
R6, C2 should be reversed.
Those filters I need for a specified BandWidth. I just added the 10uF capacitor and the offset now is just 15mV. I think it's ok.
Thank you a lot!
One more thing. When I make Transient Analysis Vmax=27V, but when I make ACSweep Vmax=27.5V.
Why this difference?

#### gophert

##### Well-Known Member
Those filters I need for a specified BandWidth. I just added the 10uF capacitor and the offset now is just 15mV. I think it's ok.
Thank you a lot!
One more thing. When I make Transient Analysis Vmax=27V, but when I make ACSweep Vmax=27.5V.
Why this difference?

Let the transient analysis run longer and the DC component will settle out. Depending how many data points you have, speed of computer, frequency of signal, it may take a while to settle. This answer, of course, depends on the simulation tool you are using.

#### audioguru

##### Well-Known Member
All the numbers from Gophert are wrong:
1) The input has a 356Hz highpass filter. It is not a 3500Hz lowpass filter.
2) The Input to the second stage has a 7270Hz lowpass filter. It is not a 7500Hz highpass filter.
3) The input signal is not 125Hz since 1kHz is clearly shown on the schematic and 'scope timebase.

You are using a 53 years old "general purpose" 741 opamp that was never designed for audio and its datasheet does not mention how much distortion it has.
The second opamp also is not an audio amplifier since it has pretty bad distortion on a graph when it has a lot more negative feedback (much lower gain) that you have.

On the other website or other post you were told to use an audio opamp for the preamp with an audio power amp. Then the distortion will be extremely low.

I wonder if your signal generator produces some of the distortion. The waveform looks like an antique ICL8038 function generator:

#### Attachments

• 103.1 KB Views: 6

#### Hellen Grig

##### New Member
Let the transient analysis run longer and the DC component will settle out. Depending how many data points you have, speed of computer, frequency of signal, it may take a while to settle. This answer, of course, depends on the simulation tool you are using.
I use PSpice . The Active BandPass Filter affects my amplification . And I don't know why.

#### Hellen Grig

##### New Member
All the numbers from Gophert are wrong:
1) The input has a 356Hz highpass filter. It is not a 3500Hz lowpass filter.
2) The Input to the second stage has a 7270Hz lowpass filter. It is not a 7500Hz highpass filter.
3) The input signal is not 125Hz since 1kHz is clearly shown on the schematic and 'scope timebase.

You are using a 53 years old "general purpose" 741 opamp that was never designed for audio and its datasheet does not mention how much distortion it has.
The second opamp also is not an audio amplifier since it has pretty bad distortion on a graph when it has a lot more negative feedback (much lower gain) that you have.

On the other website or other post you were told to use an audio opamp for the preamp with an audio power amp. Then the distortion will be extremely low.

I wonder if your signal generator produces some of the distortion. The waveform looks like an antique ICL8038 function generator:
I changed the sampling rate and now I don't have distorsion.

Status
Not open for further replies.