Low-pass RC filter design questions

Thread Starter

giovannelucas_

Joined Jan 5, 2023
38
Newbie here.

Hello, I´m currently working on a project that requires a 60Hz low-pass RC filter. I selected a 1K resistor and a 3uF capacitor as in the schematic.
Are these appropiate?

Is my schematic appropiate? I saw that the load must be connected to ground.
What would your advice be to someone starting out in electronics? What are the best resources to learn?

I´d appreciate any advice on changes to the design to make it better.

Thanks!Captura de tela 2023-01-05 173119.png
 

Papabravo

Joined Feb 24, 2006
21,305
R-C filter should connect to the input and feed the opamp configured as a buffer. You have it backwards. Also lose the 741, since it is obsolete and has very poor performance characteristics compared to numerous successors. Your corner frequency is:

\( f_c\;=\;\cfrac{1}{2\pi RC}\;=\;\cfrac{1}{2\pi(1k\Omega)(3\mu F)}\;=\;53.05\text{ Hz} \)

The response will be 3 dB down at that frequency. Is that what you intended?
If it was me, I'd put the corner just above 60 Hz.
 

crutschow

Joined Mar 14, 2008
34,837
It might be better to use the op amp in a active filter configuration, which can generate a 2 or 3 pole response from a faster rolloff after the passband to better suppress signals outside the passband.
Here's an article on doing a 3-pole active filter with a rolloff three times faster than your 1-pole filter..
 

Thread Starter

giovannelucas_

Joined Jan 5, 2023
38
Let me see if I understood Papabravo, I changed the position and connected it to the power input. Should it be like that or connected directly to the OP AMP?
I´ll consider changing the IC, is the LM324N a good choice?Captura de tela 2023-01-05 180157.png

As for the frequency I wanted 60Hz, for that I´d need a 2.65uF so I rounded it to 3uF. Should I use 2.5uF?
 

Papabravo

Joined Feb 24, 2006
21,305
Not quite.
  1. R1 || R4 forms the bias network on the (+) input which should not be connected to GROUND.
  2. the 741 will not operate properly in single supply mode
  3. You want the output of the RC network to be connected to the inverting input
I'm not sure you have a viable connection for the opamp. You should start with a unity gain buffer which requires no resistors.
 

Ian0

Joined Aug 7, 2020
10,269
I think you are intending to amplify signals from leg and chest (looks like it might be an electrocardiogram).
All that you have achieved in post #6 is filtering the signal from the battery, and I suspect that is not what you intended.
I also think you need to measure the difference between the voltages at those three points, so you need a differential amplifier.
If you place the filter on the output as in post #1, whatever you connect to the output becomes part of the filter and changes its cutoff frequency, so I would suggest (agreeing with @Papabravo) a buffer (an amplifier with a gain of 1) after the filter, so that whatever you connect to the output doesn't change the filter.
I'm also guessing that you are amplifying a rather tiny signal, so you might need a better op-amp that a 741 or an LM324, otherwise you'll never find the signal amongst the noise. I think you have a high source impedance, so a JFET op-amp such as a TL072 would be an improvement.
 

Thread Starter

giovannelucas_

Joined Jan 5, 2023
38
Hi @Ian0 and @Papabravo.

So I changed the LM741 for the TL072 and used the TL072 in both the filters. I have two inputs that must be filtered therefore I need to filters, correct?

Can I round the resistors and capacitors values for the filter?

Is the circuit correct? I used the output of the filters to as the input of my OP AMP.
 

Papabravo

Joined Feb 24, 2006
21,305
Hi @Ian0 and @Papabravo.

So I changed the LM741 for the TL072 and used the TL072 in both the filters. I have two inputs that must be filtered therefore I need to filters, correct?

Can I round the resistors and capacitors values for the filter?

Is the circuit correct? I used the output of the filters to as the input of my OP AMP.
The website allows you to select the resistor tolerance. I select the E96 series which corresponds to 1% values.
 

MisterBill2

Joined Jan 23, 2018
19,398
Welcome to AAC. Quite an interesting thread here. One thing about "ground". It is better to use circuit common, because then the amount connected to it is much more limited. Also, common is an adequately defined connection, meaning that only what you connect to it is connected. The whole world is connected to "ground". The images of the circuit look reasonable, I did not go thru with the numbers, though. And at last the "ground" on the output signal line is gone. So now the circuit can possibly function. The results should be close to what is desired.
 

Thread Starter

giovannelucas_

Joined Jan 5, 2023
38
Welcome to AAC. Quite an interesting thread here. One thing about "ground". It is better to use circuit common, because then the amount connected to it is much more limited. Also, common is an adequately defined connection, meaning that only what you connect to it is connected. The whole world is connected to "ground". The images of the circuit look reasonable, I did not go thru with the numbers, though. And at last the "ground" on the output signal line is gone. So now the circuit can possibly function. The results should be close to what is desired.
Thanks @MisterBill2, I´ll order the parts now and try to assemble it and make it work.
 

dcbingaman

Joined Jun 30, 2021
1,065
Please keep in mind that capacitor tolerances are not that great, typically +/-5%. So the cutoff will change accordingly. You can get higher accuracy but they cost more.
 
Top