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.

Register Log in

Active Filter butterworth 8 order

Status
Not open for further replies.
TXRX

TXRX

Full Member level 2
Joined
Mar 7, 2012
Messages
143
Helped
2
Reputation
4
Reaction score
5
Trophy points
1,298
Activity points
2,252
Hi,

I built an Active Filter Butterworth 8 order.

I receive at the output noise of 5mVptp.

Does anyone know why i receive a very high noise at the output ?

Thanks !
 

Have you measured the noise on the input and on your power lines ?
 

TXRX said:
Hi,
I built an Active Filter Butterworth 8 order.
I receive at the output noise of 5mVptp.
Does anyone know why i receive a very high noise at the output ?
Thanks !
Click to expand...

Depends (1) on the selected topology (cascade or direct design?) and (2) on the gain of each stage.
Thus, without any additional information no answer is possible.
 

You should also say what kind of filter it is (lowpass, highpass, bandpass, bandstop).
At approximately what frequency is the output noise? Can you give the circuit structure and
maybe a picture of the measured noise?

v_c
 
Last edited:



Hi Guys,
I attached the Schema.

I measure the Vptp noise at the next point and receive the next Results:

There is no signal at the Input.

The noise at the REF_Volt_FLT is 1mVptp (this is the limitation of the scope).
The noise at the +5VDC_RX ,PIN 4 is 1mVptp (this is the limitation of the scope).
The noise at the Output of U20A ,PIN 1 is 2.5mVptp.
The noise at the Output of U20B ,PIN 7 is 2.8mVptp.
The noise at the Output of U20C ,PIN 8 is 6.6mVptp.
The noise at the Output of U20D ,PIN 14 is 8.8mVptp.

You can see that the noise does not create at the Voltage and not at the REF_Volt_FLT.

Do you have explanation to this phenomenon?

Doron
 

This looks like a 6th order bandpass filter (three, 2nd order Delyiannis-Friend circuits).
Are the two coupling capacitors (C109, C116) necessary? I cannot read their values -- it looks like 470 xF where x is unreadable.
It actually looks like 470uF, but I am not sure ... if this is an electrolytic, then at high frequencies (including 50kHz -- the center frequency of your bandpass filter),
I am guessing the capacitor will be inductive.

Also, you did not mention the frequency at which the noise is being observed.

v_c
 
Last edited:

Hi,

The C109 & C116 are 470nF.

The frequency of the noise is around 50KHz.
 

v_c said:
This looks like a 6th order bandpass filter (three, 2nd order Delyiannis-Friend circuits)............
Click to expand...

Not to important - however, I think (a) it is a 4-stage bandpass of 8th order and (b) it is the classical MFB topology (the Deliyannis modification consists of an additional resistive positive feedback path).
 

The frequency of the noise is around 50KHz.
Click to expand...
It shouldn't be surprizing to see noise mainly in the pass band of a band pass filter, isn't it?

Your filter implementation involves a high Q of each biquad, which also results in a high noise gain. Without calculating the details, I assume that the noise level can be explained solely by OP and resistor noise. Can you please tell the full filter specification (including bandwidth and gain)?
 

The total Gain design to 0dB but i measure G= -2.5dB.
The Center frequency is 50KHz and the BW is around of 5KHz.

How can i receive noise gain if the there no Gain ?

---------- Post added at 11:42 ---------- Previous post was at 11:23 ----------

Hi,
I attached the Noise that i receive at Output of U20D ,PIN 14.

 

The total Gain design to 0dB but i measure G= -2.5dB
Click to expand...
Can be assumed to be a result of component value tolerances
How can i receive noise gain if the there no Gain?
Click to expand...
Review your analog design text book. Noise gain is another word for the inverse feedback factor of an OP circuit. A noise voltage feed to the OP input (e.g. the OP's input noise) appears at the output amplified by the noise gain. It' e.g. about 240 for the first filter stage.

A SPICE noise analysis can show you all relevant noise sources and it's individual contribution to total output noise.
 
  • Like
Reactions: LvW

    LvW

    LvW

    Points: 2
    Helpful Answer Positive Rating
If i decrease the Resistors by 10 and increase the capacitors by 10 - I will receive a lower noise ?
 

In principle yes, because resistor noise is reduced by a factor 3. In your circuit, OP noise voltage is the dominant noise source (about 12 nV/√Hz OP input referred), so the only options are OPs with lower noise voltage and preferably a filter topology with lower noise gain.
 

Hi FVM,

Is this noise phenomenon relate to "oscillations" ?

Can you reccomend on other OPA with GPB > 200MHz ?
 

I assumed that you checked your circuit for possible oscillations. The questions cant be answered from your measurement without knowing the measurement conditions.

I basically wanted to mention, that your filter topology involves a high noise gain. I estimate a noise of about 1.5 mVpp from the OP data for the first stage. This about half the measured value, so additional noise sources seems to be present, but it's at least the order of magnitude.

To reduce the OP noise considerably, you need to refer to bipolar OPs.
 

H FVM,
How you calculate the output noise ? I attached the data sheet - I attached the input voltage from the data sheet.

 

Input referred noise * effective bandwidth^0.5 * noise gain * 5(Vpp/Vrms) = 12 nV * sqrt(10e3) * 240 * 5
 
  • Like
Reactions: LvW

    LvW

    LvW

    Points: 2
    Helpful Answer Positive Rating
Why you take Gain of 240 ? the Gain for the first OPA is around 1 ?
What is the meaning of this number 5(Vpp/Vrms) ?
 

I explained about noise gain before. You can analyze it by inserting a voltage source in series with the positive or negative OP input terminal.

Vpp/Vrms is 2.8 for sine voltage. For noise, you need to supplement an empirical factor. Due to the statistical nature of noise, there's no strict upper bound. Factor 5 is an estimation that applies in many cases. For an exact noise measurement, average the power (V²) over a sufficient time interval.

---------- Post added at 14:33 ---------- Previous post was at 14:29 ----------

See below noise gain of the first stage:

 
  • Like
Reactions: LvW

    LvW

    LvW

    Points: 2
    Helpful Answer Positive Rating
Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top