Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Is this filter stable? stable sallen key filter

Status
Not open for further replies.

seyyah

Advanced Member level 2
Advanced Member level 2
Joined
Oct 7, 2001
Messages
646
Helped
8
Reputation
16
Reaction score
8
Trophy points
1,298
Activity points
6,234
stable sallen key filter

I have a 2nd order sallen key filter in order to filter a 100 kHz pwm signal to DC. The output should not exceed 5V and the input does not exceed 5V normally. I have some doubts about the circuit because i sometimes get some weird results in my circuit and i suspected from here though i could not be able to prove it. Can you help me? Is this circuit has a drawback, or stability problems? Does adding a diode clamp at the output has a side effect? Is LM358 suitable for this circuit? Thanks.
 

sallen key filter stability

First of all, you should check the filter response with a simulation program.
How did you calculate the values for R and C ?
Your lowpass has a cutoff frequency of app. 5 Hz ! Is this your goal ?
 

checking lm358

LvW said:
First of all, you should check the filter response with a simulation program.
How did you calculate the values for R and C ?
Your lowpass has a cutoff frequency of app. 5 Hz ! Is this your goal ?

Cut-off freq is ok. Which program can characterize this circuit properly? Psspice? A simpel and free one would be better :). What do you say for the output diodes because i realized in some circuits that when they turns on they cannot recover but a zener does.
 

Re: Is this filter stable?

For simulation you can use each of the available programs, of course.
For example PSpice (evaluation for free) or LTSPICE (free from linear.com).

But one question: Why you have mentioned 100 kHz ? Is a specific damping required at this frequency ? As you will see, the Sallen-Key topology will create some problems at this frequency - at least when you use a rather simple opamp like 741 or LM324.

Concerning the diodes: If the input does not exceed 5 volts the ouput also will never grow above 5 volts as the gain of your filter is lower than 1.
 
Re: Is this filter stable?

I'm filtering a constant frequency signal which is at 100 kHz.
 

Re: Is this filter stable?

seyyah said:
I'm filtering a constant frequency signal which is at 100 kHz.

Sorry, but this is no answer.
When you filter a specific signal you have to fulfill some requirements.
Some frequencies have to be suppressed (how many dB?) and some not.
Have you no detailed damping specification? Shall the 100 kHz component be damped or not ?
 

Re: Is this filter stable?

seyyah said:
LvW said:
First of all, you should check the filter response with a simulation program.
How did you calculate the values for R and C ?
Your lowpass has a cutoff frequency of app. 5 Hz ! Is this your goal ?

Cut-off freq is ok. Which program can characterize this circuit properly? Psspice? A simpel and free one would be better :). What do you say for the output diodes because i realized in some circuits that when they turns on they cannot recover but a zener does.
 

Is this filter stable?

The LM358 is too slow to be a 100kHz filter.
You can count its voltage gain on the fingers of one hand. Its slew rate fails above only 5kHz at a low output swing.

An OPA2134 is a dual opamp with a voltage gain of 100 at 100kHz and a good full output slew rate to 220kHz.
 

Re: Is this filter stable?

Ok, i could not be clear i think. This is a pwm signal of 100 kHz. So i do not want to filter some frequencies. Actually i need only the dc component which is at zero frequency. So this filter does the averaging of the input signal. The output should be the dc average of the input filter which is a 100kHz pwm signal. So all frequency components should be filtered. Practically this would be 10Hz and above. Slew rate is not very important in my application as long as it suffers to much since it does not require immediate response. I'm checking lm358 with a signal generator now and i could not see a major problem. I swept frequency from 1kHz to 12 MHz (lm358 has a BW of 1 MHZ). Since the input has constant amplitude and duty cycle, i did not observe a failure. Also i kept frequency constant at 100kHz and swept input amplitude from 0 to 10V and there was not any problem again. I was concerning if LM358 could not response the signal properly or it may latch-up or it may not work around zero etc. But it does not seem to be a issue. But i still won't use it at least for the operating temperature issue. I have opa2735 in my hand. Is this opamp suitable for this task?

Added after 21 minutes:

I realized that at some frequencies e.g. 500Hz, 5kHz, 10kHz the output changes slightly.
 

Is this filter stable?

I have seen tutorials about active filters where they say the response of the opamp must exceed the response of the filter. They show a lowpass filter where the dropping response comes back up at high frequencies when the high frequency response of the opamp is not adequate.

The OPA2735 is extremely noisy and has a frequency response that is not adequate.

The LM358 (and its sister the LM324 quad) is the first low power opamp. So it has a horrible slew rate where it barely makes it to 5kHz at 13V p-p. With a supply of from 10V to 15V its frequency response is to only 300kHz where its gain is 1.
 

Re: Is this filter stable?

Audioguru said:
I have seen tutorials about active filters where they say the response of the opamp must exceed the response of the filter. They show a lowpass filter where the dropping response comes back up at high frequencies when the high frequency response of the opamp is not adequate.
.................
.................
.

For a lowpass with 10, 20 or even 100 Hz cut-off, the opamp response always will "exceed" the filter response. No doubt about it. The 324 will do the job.
However, as I have mentioned already: The "dropping response" (term of audioguru) comes back for higher frequencies NOT because of an unadequate opamp.
The reason is the Sallen-Key-topology - and that´s its major drawback.

Explanation: For rising frequencies the output of the opamp goes down (as wanted for a lowpass), but at the same time a rising portion of the input goes the direct way through the feedback capacitor and produces a voltage at the finite output impedance of the opamp. Therefore, the output voltage increases again for high frequencies. Other filter topologies (dual feedback or negativ Sallen-Key) less sensitive to those parasitic effects.
 

Is this filter stable?

If a faster opamp is used then the Sallen and Key lowpass filter will work fine at higher frequencies.
Why use the oldest and the slowest opamp to filter 100khz?
 

Re: Is this filter stable?

Audioguru said:
If a faster opamp is used then the Sallen and Key lowpass filter will work fine at higher frequencies.
Why use the oldest and the slowest opamp to filter 100khz?

Do you think that my arguments concerning disadvantages of the positive Sallen-Key structure are false ? Do you need some corresponding references ?
 

Is this filter stable?

I have always used non-inverting Sallen and Key filters with fast opamps and they have always worked fine.

When I first saw the horrible performance of the LM358 and LM324 opamps with their low bandwidth and high crossover distortion then I haven't used them since.
 

Re: Is this filter stable?

For your information I have attached a figure (pdf) comparing 4 different Sallen Key structures (simulation with SPICE). My recommendation for you is to simualte by yourself and verify the poor performance of the positive SK structure - independent on the opamp type.
 

Re: Is this filter stable?

Comment to the figure: It was the purpose to show the properties of each topology for higher frequencies (pole at 50 kHz).

For lower pole frequencies the last 2 structures are fine, however the problem at higher frequencies of the positive SK filter still exist.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top