3
\$\begingroup\$
  1. Should a high frequency filter be implemented on the differential signals input like in the following Olimex EMG/EKG amplifier, or can it be implemented after the Vinput+ and Vinput- have passed through a unity amplifier and single ended output Vout? what would the HF filter look like if implemented after it?

  2. Is the high frequency filter the same as RFI filter? Do both filter EMI from air and from line? Then how do EMI/RFI filter from air and line differ?

From the SHIELD-EKG-EMG schematic:

enter image description here

\$\endgroup\$
4
  • \$\begingroup\$ Maybe the data sheet of the so-called "instrumental" amplifier will illuminate. BTW it's called an instrumentation amplifier in that it is used for processing signals from instrumentation devices (and not amplifying guitar or keyboard solos). I'd also be concerned about what sources the signal V_REF. \$\endgroup\$
    – Andy aka
    Feb 25 at 11:49
  • \$\begingroup\$ this is datasheet of the olimex but my question is not about the olimex. its just an example..i was just asking if it is better to implement the high frequency or rfi filter at differential inputs or after any amplifier olimex.com/Products/Duino/Shields/SHIELD-EKG-EMG/resources/… \$\endgroup\$
    – Jtl
    Feb 25 at 13:50
  • \$\begingroup\$ It's not as clear cut as "is it better" (one way or the other). It depends on the signals (both differential and common mode) and how much you want to spend on CM capacitors. \$\endgroup\$
    – Andy aka
    Feb 25 at 14:08
  • \$\begingroup\$ I want to see examples where the rfi filter is implemented after the amplifier. i was thinking and read others say it should always be implemented right after the input \$\endgroup\$
    – Jtl
    Feb 25 at 14:19

2 Answers 2

3
\$\begingroup\$

Should a high frequency filter be implemented on the differential signals input like in the following Olimex EMG/EKG amplifier

Sometimes, it's a really bad idea to implement CM filter capacitors as shown here (C2 and C3): -

enter image description here

The reason is because they have a value tolerance and, if that tolerance means that one is 5% high and, the other is 5% low, you will get CM signals (normally ignored when the capacitors are unfitted) being converted to differential signals. You can't do anything about differential signals; they are what you want to amplify but now, with the C2 and C3 capacitors, any significant CM signal is going to muck-up your wanted differential signal.

So, I don't favour them unless the likely CM signals are spectrally miles-away from the wanted spectrum of the differential signal you want to process. Quite often this is the case i.e. you are wanting to process an audio like signal and the CM interference is above (say) 1 MHz. This would be likely called an RFI filter and, if that is what is intended with C2 and C3 then, they are in the right place.

can it be implemented after the Vinput+ and Vinput- have passed through a unity amplifier and single ended output Vout? what would the HF filter look like if implemented after it?

Just a regular RC low-pass filter is good enough but, quite often higher order filters are used like the Sallen-Key filters shown in your diagram.

\$\endgroup\$
14
  • \$\begingroup\$ I thought high frequency filter is the same as RFI filter. they are not the same? what is the difference? \$\endgroup\$
    – Jtl
    Feb 25 at 14:50
  • 1
    \$\begingroup\$ @Jtl if I interpret what you ask correctly, C1 is a differential filter and C2/C3 form a common-mode (CM) filter and, both can be regarded as RFI filters. A HF filter is very likely an RFI filter also. However, neither are very well defined definitions. \$\endgroup\$
    – Andy aka
    Feb 25 at 15:13
  • \$\begingroup\$ So maybe the rule is not to implement HF/RFI filters right after amplifiers with any gain because the HF/RFI can be amplified too and it's not good for the signal. So the only time filters can be used after the amplifier if it is a unity (gain 1) buffer amplifier only? \$\endgroup\$
    – Jtl
    Feb 25 at 15:15
  • 1
    \$\begingroup\$ @Jtl a unity (or low gain) front-end amplifier is a good choice. Anything with significant gain could hit output saturation at the intermediate stages within an instrumentation amplifier. When that happens it is usually unresolvable without lowering front-end gain. \$\endgroup\$
    – Andy aka
    Feb 25 at 15:19
  • \$\begingroup\$ If common mode provokes nonlinear behavior in the amp, you'll want to filter it out. \$\endgroup\$
    – John Doty
    Feb 25 at 15:19
3
\$\begingroup\$

The purpose of the input filter in a low frequency measurement device like this is to keep EMI from driving the input amplifier into nonlinear behavior. So, you can't move it to the output. This circuit has filters on the output of the differential stage. The "regulated gain" stage is configured as a low-pass filter, and it feeds a "Butterworth filter" stage.

\$\endgroup\$
2
  • \$\begingroup\$ If the amplifier is unity with gain of 1. can it contribute to the non linear behavior? what will be the advantage or disadvantage if using HF/RFI filter on the single ended output Vout of the unity amplifier versus directly filtering it in the differential input before the common mode signals were removed by the unity amplifier? \$\endgroup\$
    – Jtl
    Feb 25 at 15:46
  • 1
    \$\begingroup\$ @jtl A unity gain amplifier may behave nonlinearly to input signals outside its bandwidth. A filter on its output will not affect this behavior. \$\endgroup\$
    – John Doty
    Feb 25 at 20:38

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.