0
\$\begingroup\$

First of all, sorry, I missed the links / images where I have found the solutions mentioned below (I hope, if you are familiar with this topic, you know them):

I have to design an ECG / EEG analog amplifier with a gain around 1000. Sigma-Delta is the preferred AD converter architecture. As I search trough the net, I see some implementation, where:

  • the input is AC coupled trough RC high pass filter
  • the instrumental amplifier's reference pin is driven by an integrator (servo feedback)
  • after that an active high pass filter is realized

The design tread off that I cannot decide / I don't understand (we know, every components add noise to the signal present on the output):

  • as the input should be AC coupled due to electrode offset cancellation, why should I apply servo drive and HP filter?
  • some designs throw away the AC coupling, but requires lower instrumental amplifier gain, that reduce CMRR
  • some another designs left the servo drive and connect the instrumental amplifier's reference to a predefined potential

Can somebody help me to decide about these solutions / constraints?

Regards

Edit: This is my concept design: enter image description here

User69795 said:

The problem with the high pass filter in this application is due to the very slow response time of the filter ... When the input is overdriven, ... the filter takes a really long time to settle, so the signal output is not visible to the doctors, so they don’t know if the patient is still in fibrillation or is dead or is OK.

Overcoming this response time problem while still providing the correct high pass filtering leads to the different ways of implementing the high pass function that you have noticed. A good solution calls for clever design.

As I understand you correctly, the C15 R13 U4A HP filter should be removed due to the large settling time? And what about the AC coupled input and the common to differential mode transformation caused by the component mismatch?

\$\endgroup\$
  • \$\begingroup\$ Do you need to design the amplifier itself or just make an ECG/EEG device? If the latter, then there are off the shelf chips for measuring biopotentials like ADS1194. \$\endgroup\$ – filo Sep 21 at 19:55
  • \$\begingroup\$ If "As I understand you ..." is a response to a reply or comment you should say so. As it is peop;e rading t do not know what you are referring to. It can be useful to co[y in a small part of what you are responding to and prefix it with ">" symbol which marks the block. I have added a short example to your question. Remove or alter as desired. \$\endgroup\$ – Russell McMahon Sep 22 at 11:34
  • \$\begingroup\$ The goal is to develop an ECG like device (due to proprietary reasons let's say "ECG like") where I have to measure the signal with at least 16-20 bit resolution. If I use ADS1194, ADS1299 or similar AFE, the signal in interest will be represented on 5-10 bits. So I have to design an amplifier to its output signal span the input range of the AD converter. \$\endgroup\$ – moravas Sep 23 at 5:56
0
\$\begingroup\$

The problem with the high pass filter in this application is due to the very slow response time of the filter (because the filter is at such a low frequency). When the input is overdriven, ( for example, when the patient is defibrillated), the filter takes a really long time to settle, so the signal output is not visible to the doctors, so they don’t know if the patient is still in fibrillation or is dead or is ok.

Overcoming this response time problem while still providing the correct high pass filtering leads to the different ways of implementing the high pass function that you have noticed. A good solution calls for clever design.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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