LM358 single supply band pass filter

Thread Starter

fiatuno

Joined Jan 28, 2011
47
Dear members,
I have a problem designing a non inverting single supply band pass filter, with the use of an LM358. I have worked the calculations for the gain and they seem to be right, Then I worked out the calculations from my school book as attached in one of the attachments and simulated on proteus something like in the below attachment, I can't seem to see what is going wrong with my simulation, since I need an bandpass filter for a 150mV ac signal to be amplified, letting frequencies between 200Hz-20kHz to pass. The outputs can't seem to be what I need as I need around 25.66dB in gain. Any help would be greatly appreciated as I can't seem to find a way out, Thanks in advance.
Regards
Fiat Uno
 

Attachments

Last edited:

Thread Starter

fiatuno

Joined Jan 28, 2011
47
Am I doing the right configuration or is there any other one as I seem to be at a standstill and really need to solve this problem if someone could give me some help. Thanks in advance
 

Audioguru

Joined Dec 20, 2007
11,248
Check your math.

Your input capacitor of 796pF into 100k reduces frequencies below 2010Hz.
Your feedback grounding capacitor of 723nF reduces frequencies below 201Hz.
Your 80pf output capacitor feeding the 100k load reduces frequencies below 20kHz.
The lousy old LM358 reduces frequencies above 1kHz to 20khz depending on the output level.

Since the values of two of your capacitors are too small to pass most of the frequencies you need then you have no gain. Since you are using an extremely old and poor quality opamp then you also have no high frequency gain.
 

Thread Starter

fiatuno

Joined Jan 28, 2011
47
First of all I would like to thank you for your reply, but that one I have already used it as I had to design a bandpass filter using an op amp with dual supply, but now I need to do the same thing with single supply and I opted for the LM358. This configuration this time doesn't seem to work just by substituting the ic and just introducing the bias to the input of the non-inverting input. Thanks anyways.
 

Thread Starter

fiatuno

Joined Jan 28, 2011
47
Check your math.

Your input capacitor of 796pF into 100k reduces frequencies below 2010Hz.
Your feedback grounding capacitor of 723nF reduces frequencies below 201Hz.
Your 80pf output capacitor feeding the 100k load reduces frequencies below 20kHz.
The lousy old LM358 reduces frequencies above 1kHz to 20khz depending on the output level.

Since the values of two of your capacitors are too small to pass most of the frequencies you need then you have no gain. Since you are using an extremely old and poor quality opamp then you also have no high frequency gain.
So what would you suggest as an opamp that I can use?
What is wrong then with my calculations as I double checked the equations I have done and seem to match. Just to confirm if I am on the right track please as shown in the attachment above:
f1 was used with 200Hz R2=200k C1 came to be 796pF
f2 was used with 20kHz RL=100k C2 came to be 80pF
f3 was used with 200Hz R1=1.1k C3 came to be 723nF

Thanks in advance.

Regards,
Fiatuno
 

Audioguru

Joined Dec 20, 2007
11,248
So what would you suggest as an opamp that I can use?
I use and many stereo manufacturers use the TL071 single, TL072 dual and TL074 quad low noise, low distortion and wideband inexpensive opamps.
The lousy old LM358 is very noisy (hissss), has horrible crossover distortion and has troulble above only 2kHz.

What is wrong then with my calculations as I double checked the equations I have done and seem to match. Just to confirm if I am on the right track please as shown in the attachment above:
f1 was used with 200Hz R2=200k C1 came to be 796pF
Two 200k resistors make 100k, not 200k.
1 divided by (2 x pi x 100k x 200Hz)= 8000pf, not 800pF.

f2 was used with 20kHz RL=100k C2 came to be 80pF
Your 80pF capacitor feeding 100k at the output cuts low frequencies, not high frequencies. The lousy old opamp cuts mid frequencies and cuts high frequencies.

f3 was used with 200Hz R1=1.1k C3 came to be 723nF
Correct.
EXCEPT, Now you have two RC networks cutting low frequencies below 200Hz so their total cutoff frequency is 400Hz, not 200Hz.
 

Thread Starter

fiatuno

Joined Jan 28, 2011
47
Have you tried using Texas Instrument's FilterPro application to check your work?
Yes I did try to use it already, but it seems to have a different configuration and doesn't cater for single supply as long as I know. I wish to edit the design attached from the book to make it workable. Thanks anyways KJ6EAD
 

Thread Starter

fiatuno

Joined Jan 28, 2011
47
I use and many stereo manufacturers use the TL071 single, TL072 dual and TL074 quad low noise, low distortion and wideband inexpensive opamps.
The lousy old LM358 is very noisy (hissss), has horrible crossover distortion and has troulble above only 2kHz.


Two 200k resistors make 100k, not 200k.
1 divided by (2 x pi x 100k x 200Hz)= 8000pf, not 800pF.


Your 80pF capacitor feeding 100k at the output cuts low frequencies, not high frequencies. The lousy old opamp cuts mid frequencies and cuts high frequencies.


Correct.
EXCEPT, Now you have two RC networks cutting low frequencies below 200Hz so their total cutoff frequency is 400Hz, not 200Hz.
Hi Audioguru,

First of all thanks for the TL071 as I matched some information from the datasheets and showed that it is a much better solution and also for f1 I have understood your reasoning and I agree.

But for f2 since I used the 80pF capacitor which is feeding the 100k at the output, why is it cutting low frequencies if I need it to cut higher frequencies which are greater than 20kHz. Do I need to exchange the position between the capacitor and the resistor? I have tried it already and the output is more like a square wave which has max amplitude at 3.5 and min amplitude at 2.5.

For the f3 why is it that I have 400Hz, which parts are doing such thing.

Thanks for your reply and for your help,
 

Jony130

Joined Feb 17, 2009
5,487
All your filter in you diagram are high pass filters.
So you need to add low pass filter capacitor (C3).


For example
R1 = R2 = 200KΩ then

C1 = 0.16/( R1||R2 * F) = 0.16/ ( 100K * 200Hz) = 10nF

Av = 1 + R4/R3

So for the gain around 26dB

R4 = 20 * R3 = 20 * 10K = 200KΩ

R3 = 10K and R4 = 200K;

C2 = 0.16/( R3 * F ) = 0.16/ ( 10K * 20Hz) = 1μF

C3 = 0.16/( R4 * F) = 0.16/ ( 200K * 20KHz ) = 47pF

R5 = 100K

I assume Rload _min - 1K

C4 = 0.16/ ( 20Hz * 1K) = 10uF
 

Attachments

Audioguru

Joined Dec 20, 2007
11,248
First of all thanks for the TL071 as I matched some information from the datasheets and showed that it is a much better solution and also for f1 I have understood your reasoning and I agree.
Not only does the TL07x have a higher frequency response than the lousy old LM358 but its slew rate allows full output to 100kHz. The very low slew rate of the LM358 allows half of full output to only 5kHz and full output to only about 2kHz.

But for f2 since I used the 80pF capacitor which is feeding the 100k at the output, why is it cutting low frequencies if I need it to cut higher frequencies which are greater than 20kHz. Do I need to exchange the position between the capacitor and the resistor?
It is a highpass filter (it passes high frequencies and cuts low frequencies).
Instead you need a lowpass filter (to pass low frequencies and to cut the high frequencies). You cannot simply swap the 80pF capacitor with the 100k resistor because then the 100k resistor will attenuate everything if there is a load.

Post #11 shows a lowpass filter but it does not cut the high frequencies completely.

For the f3 why is it that I have 400Hz, which parts are doing such thing?
You have an input highpass filter and another highpass filter in the negative feedback to ground. Each has a cutoff frequency of 200Hz but they add together and produce a total cutoff frequency of 400Hz.
 

Thread Starter

fiatuno

Joined Jan 28, 2011
47
Hi Jony130,

I have tried it with the values you have given to me, but it seems that it doesn't work as I desire, maybe I am doing something wrong. I have attached a simulation with a graph to show what outputs I have been having and the configuration made, maybe someone can see my error please as I can't find a way out for some reason. Thanks in advance,
Regards,
Fiat UNO
 

Attachments

Audioguru

Joined Dec 20, 2007
11,248
Most opamps use darlington output transistors that cause a voltage loss of about 1.5V each when there is no load. Then if the opamp works from a supply as low as 5V, its maximum output is 2V peak-to-peak which your simulation is showing.
Your input is 150mV peak and the gain of the opamp is 21 so the output of the opamp is trying to go to +5.65V (above the power supply voltage) and down to -0.65V (below ground). Obviously the opamp cannot do it so the output is clipping severely.
The output of the opamp can swing up to only +3.5V and swing down to +1.5V.
Turn down the input signal level to avoid clipping.

The TL07x opamps have a minimum spec'd supply that is 7V so they do not work properly with only 5V. Your simulation software probably doesn't know it.
 

Attachments

Thread Starter

fiatuno

Joined Jan 28, 2011
47
First of all I would like to thank you a lot for your help, all of you, as with your help I managed to reach the desired results. But is there any other op amp which works at 5V, without having to use the 12V please, as I need to make it workable from 5V and I tried various op amps such as the opa237, opa1013 and they do not give me what I need.
 

Audioguru

Joined Dec 20, 2007
11,248
An OPA134 works with a supply as low as 5.0V. It has extremely low distortion and noise and a full output bandwidth of 1.5MHz when the supply is 5.0V.
 

Thread Starter

fiatuno

Joined Jan 28, 2011
47
But why am I having half the gain, I think that this happens because of the supply, because I simulated your suggestion and it still gave me half the gain when, using 5V as supply to the op amp. This single supply op amp gives a very hard time :(
 

Audioguru

Joined Dec 20, 2007
11,248
But why am I having half the gain, I think that this happens because of the supply, because I simulated your suggestion and it still gave me half the gain when, using 5V as supply to the op amp. This single supply op amp gives a very hard time :(
If you use an OPA134 opamp, then the gain is 21 at 2050Hz. Because the filters are extremely simple they cut frequencies above and below.
214Hz is cut -3dB and 20.5kHz is cut -3dB.

With a 5V supply the maximum output is 2V peak-to-peak so the maximum input level is 47.6mV peak.
 

Attachments

Top