which amplifier to use

[P.Copper]

Hi all

I've been designing lowpass filters using opamp (LM324N) and my problem is this: most of the time the output signal is out of phase with the input, i've even tried different methods but non works. I've come to the conclusion that LM324N is the problem. What other opamps are best for designing filters? Thanks

 

[FvM]

More likely you are observing the designed phase characteristic of your circuit.

 

[P.Copper]

More likely you are observing the designed phase characteristic of your circuit.

Fvm
what do you mean?

 

[crutschow]

Filters intrinsically have a phase-shift that varies with the frequency around their corner frequency. It's not a problem with the op amp unless you are operating at a frequency that's near the op amp GBW limit.

Why is this phase shift of concern in your application?

 

[P.Copper]

Filters intrinsically have a phase-shift that varies with the frequency around their corner frequency. It's not a problem with the op amp unless you are operating at a frequency that's near the op amp GBW limit.

Why is this phase shift of concern in your application?

Thanks crutschow,

the thing is my i designed my cutoff to be 650Hz but i can only get cutoff of 500Hz. My first thought was the shift is impacting the cutoff somehow

 

[FvM]

Why don't you show the circuit and filter design? LM324 shouldn't have problems to implement a 650 Hz low-pass.

 

[P.Copper]

Why don't you show the circuit and filter design? LM324 shouldn't have problems to implement a 650 Hz low-pass.

here's my design and curcuit

 

[FvM]

It's a correctly designed 650 Hz butterworth low-pass.

 

[P.Copper]

the thing is when i build the circuit on a breadboard i get the shift.

 

[FvM]

The circuit is robust enough to work on a breadboard. I presume you know that the gain drop at cut-off frequency is 3 dB (gain 0.71), second order filter phase shift is -90 degree.

 

[P.Copper]

Thanks FvM
Yes that i know.... now,Generally speaking will the shift have negative impact on my circuit? or should i deal with it now so i don't encounter problems in the long run

 

[Audioguru]

The datasheet for the very old LM324 quad opamp shows that it has slew rate problems at high levels above only 2kHz or less.
You should use a modern audio opamp that works well up to 100kHz or higher.

 

[PA4TIM]

I have read in a book about opamps that it is best to take an opamp with about 10x the bandwidth you need for a filter. So I think audioguru is right. I made a 1 kHz filter using two TL-084 opamps.
But I think your problem can be the breadboard, that adds capacitance . And have you checked the component values with a good LCR meter (many have a 5% tolerance and most multimeters are way of in measuring C. (just in case, but I asume allready tried other values)

 

[crutschow]

Below is a simulation of your circuit. Its -3dB point is 650Hz, as you wanted. So if your actual circuit is significantly different from that, I suspect some of your component values are off. The phase shift is 90° at the corner, which is normal.

 

[FvM]

The latest contributions are riding a dead horse. The OP already stated in post #11 that his low-pass filter has a characteristic expectable by design. As crutschow showed, LM324 will work well for a 650 Hz low-pass, already mentioned previously. Power bandwidth limitations do matter for many LM324 applications, but not this one.

 

[PA4TIM]

The voltage gain A of a LM324 drops from 110dB at DC to 55dB at 1 kHz typical, so it will work but it would not be my first choise. But telling I'm riding the dead horse is a bit over the top. I also wrote:

But I think your problem can be the breadboard, that adds capacitance . And have you checked the component values with a good LCR meter (many have a 5% tolerance and most multimeters are way of in measuring C. (just in case, but I asume allready tried other values)

And that could be the real problem here.

 

[FvM]

The voltage gain A of a LM324 drops from 110dB at DC to 55dB at 1 kHz typical, so it will work but it would not be my first choise. But telling I'm riding the dead horse is a bit over the top.

Obviously not the best choice (for any performance application). But that wasn't the original problem of this thread.

The OP worried about a phase shift which turned out as normal operation of a low-pass (according to the most recent statement). If you don't like "dead horse" say non-existing problem.