Is my assumption correct?

[D.M.]

I am designing a basic two-lead EKG acquisition circuit to measure the cardiac signal for one of my classes.

We will have Burdick Cardiosens Ultra II “tab” electrodes going into the Burr Brown INA 121 instrmentation op amp.

The cardiac signal is known to be riding on some 150-300 mV DC offset.


MY assumption: <----- Can someone check me please?


Since the input to the INA121 will be a differential input two electrodes placed on both wrists, the CMRR will play a major role in reducing the DC offset and allow the cardiac (AC signal) to pass through and be amplified with little DC offset on the output side.

Is that a correct assumption?

prompt responses are always welcome

 

[ke5frf]

If the DC offset is present at equal amplitudes at both input terminals, yes common mode rejection will null them out, leaving only the cardiac pulses to be amplified. I would imagine, however, that the DC voltage you are trying to null will be present at different potentials, even if slight, all over the body, would it not? And if so, would there be variations in this potential that might be out of sync throughout the body at any time interval? ie, for simplicity, a probe on one wrist might read 175 mV DC at time x and on the other 215 mV...and at time y the first wrist might be 187 mV and the other perhaps 202 mV?

If this is the case, common mode rejection will still play a role, but there will be baseline noise in your amplified signal. If the variations are not equal or predictable there will be a challenge in filtering this noise out.

Just some things to consider. You might have to investigate DC filtering at the output of your amplifier with a pass capacitor for your cardiac pulses and perhaps this might be an application where a common mode choke coil at your amplifier inputs will reduce noise.

 

[D.M.]

thank you, that was very helpful. I appreciate it.

 

[ke5frf]

You are welcome. Let us know how your amplifier performs and what filtering techniques you try in case of noise.

 

[D.M.]

Well i am kind of bummed to admit, but due to the semester's end drawing near i dont think i will have time to implement any other type of filtering. The project is due tuesday.

I am mainly trying to fill my report with concise and well thought analysis.
The acquisition of the EKG signal was stressed but clarity of it was not, as this is only
an under-graduate class.

My lab partner and I successfully measured my cardiac signal, but it was noisy. You can clearly see the QRS wave, but the PQ and ST wave are buried in noise. The signal seems to also be riding on a 60 Hz carrier wave. My lab partner did the Labview which had a 60Hz comb filter and a 0.05 Hz High pass filter but we are not sure how effective it was.
We were only given design criteria and not much else so it was up to us to research and design this. considering i am only junior level i am fairly impressed with my work.

in case youre interested, its not much but here is my cardiac signal as measured by my circuit.

**broken link removed**

 

[ke5frf]

Hmmm,

Well, if you JUST so happen to have a RadioShack nearby, you may still be able to pick up a 1:1 audio isolation transformer, cheap and small. But instead of connecting it like you would in a power supply (i.e. Hot/neutral to primary, secondary to DC rectifier)...You would turn the transformer sideways, so that one of your ekg sensors is connected to one terminal of the primary, and the other sensor is connected to one terminal of the secondary. Likewise, the opposite secondary and primary terminals go to your amplifier inputs. Basically you have two parallel inductors with a common core by turning a transformer sideways like this.

You might be able to null the 60 Hz noise as well as the unwanted DC noise. It wouldn't be difficult to implement and I'm certain you can get the component even at radioshack for not much money. I know what you mean, but why not make the circuit better if you can? Mind you, I'm not 100% certain it will do the trick but I believe it will.

 

[D.M.]

Thank you for the advice, i may try something like that time willing. I will include your recommendation somehow into my report as ways to improve my project.

Thanks again.

 

[Mikebits]

Looking at the image on the right: is the slow rising waveform a result of 60Hz? If so, then the pulses that your are seeing are most likely not your QRS complex, if they were you would be having tachycardia and in need of a crash cart. Typically 1 beat per sec in healthy person. Your waveform would imply multiple beats in a fraction of a second. It is hard to read the time scale on your picture, but I think you are just looking at outside noise.

 

[Diver300]

My interpretation of the right-hand trace is completely different from that of Mikebits, but as neither of us know the time scale, we could be completely wrong.

I think that the spikes are heartbeat, at about 1 - 2 Hz. In that case the slope is slow changes in offset voltage caused by chemical changes or static charge or something. The point is that you are not interested in such slow changes and your high pass filter at 0.05 Hz is far too low. At 0.05 Hz you filter is letting through stuff that takes 10 - 15 heartbeats to change from minimum to maximum. You want to set it to 1 Hz or so.

I know that sounds wrong. You are trying to look at something happening maybe once per second, so filtering out anything at 1 Hz would seem to remove the signal. The fact is that the spikes that you see are very short compared to 1 second, and all of the information within the spike, its shape and size, come from higher frequency information.

(You can think of it like this. Human hearing goes down to about 30 Hz, but we can hear a clock ticking at 1Hz because the click itself if a mix of much higher frequencies)

 

[D.M.]

Great feedback guys. Thank you, these thoughts have eluded me and will add to the content of my report.

I appreciate it very much thank you.

 

[audioguru]

A simple ECG circuit has a 3rd wire that feeds an out-of-phase common mode signal to the patient's leg to cancel the DC and low frequency AC common mode signals.

I recently had many ECG tests done where 10 electrodes were used.

 

[wavestorm1986]

D.M.
i am working on the same project also but have to send the ECG signal using bluetooth module and acquire the signal using labVIEW so, do u mind to attached the labVIEW file ???
now i am testing the cct using oscilloscope hope i can try to test it using labVIEW using the NI-USB 6008

hope you help me with my task

any ideas will be welcome ..............

 

[bryan1]

I recently had many ECG tests done where 10 electrodes were used.

Ah don't tell us AG is becoming or become an ol' git.....

 

[audioguru]

Ah don't tell us AG is becoming or become an ol' git.....

I was 63 years old when I had a heart attack last year. They fixed two blocked arteries that feed my heart and gave me 32 cardio rehab excercise classes. I bought a bicycle. I run. Each heartbeat is stronger and slower now.
The arteries were fixed so quickly that my heart suffered no damage.

I am fine now and feel like I am only 29 or 39 years old again.