0
\$\begingroup\$

I want to design a very low frequency (<1Hz) high pass filter. I was thinking of using a 2 pole sallen-key opamp design with a pair of r's and c's. Is there anything special I need to consider when choosing component values/types. It looks as though the caps will need to be in the 100's of µF range.

Thanks

\$\endgroup\$
2
  • \$\begingroup\$ You can make the caps bigger or you can make the resistors bigger \$\endgroup\$
    – endolith
    Commented Feb 10, 2011 at 16:53
  • 2
    \$\begingroup\$ Not sure where the output of your filter leads, but could you use a digital filter? \$\endgroup\$
    – Nick T
    Commented Feb 10, 2011 at 18:21

3 Answers 3

3
\$\begingroup\$

It depends on how accurately you are expecting to control your cutoff frequency. A few point come to mind ...

  • Tolerance

High value electrolytic capacitors have wide tolerances, indeed cheap ones can be as wide as +100%/-50%. You won't get much better than ±10% and stability could still be an issue. Solid electrolyte (aluminium/tantalum) have better stability but will be much more expensive.

  • Leakage

Electrolytic capacitors will have a finite leakage current which will produce dc offsets given that your resistor values will also be high.

  • Polarization

Make sure that your circuit biasing keeps capacitors correcty polarized.

  • Charge/Discharge

High value capacitors will have to charge & discharge somehow if there is a non-zero dc bias (ie single rail). This will cause turn-on 'thumps' as the circuit settles-down which may take a many (tens of) seconds. At turn-off, the capacitors may discharge into the op-amp causing damage although given that your resistor values will also be large, this is less likely to be a problem.

The lowest frequency filter I have ever built was a 5Hz ±20% two-stage S&K (4 pole) maximally flat design which worked perfectly well.

You might also want to look at a Gyrator circuit to simulate a high value inductor.

\$\endgroup\$
0
\$\begingroup\$

I would vote for 10-100uF ceramic capacitor with large resistor + temperature controlled chamber OR ceramic cap + programmable resistor to compensate for temperature changes.

This should be more stable over time, but less stable over temperature. Also, ceramic caps should have way less leakage.

\$\endgroup\$
0
\$\begingroup\$

It's very difficult (or even impossible) to do such filter with a simple 2-pole analog filter, depending on your requirements (fp, fs, As). Download and install FilterPro (from Texas Instruments) and test some configurations. That's a very good and simple program. You enter the requirements of your filter, click on a button and then you get the circuit and the values for resistors and capacitors automatically.

Probably, with a digital filter you can achieve much better results. Take a look at IIR filters. What's the application of your high pass filter? Maybe DC removal? If so, google for IIR+DC removal, you'll get plenty of results.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.