EKG front end circuit

[Wimberleytech]

I started this a few years ago but never finished it.  Now that I am dealing with some issues, I decided to get it to the next step.

The circuit is pretty simple, using and AD624 inamp, followed by two second-order Sallen-Key LPFs...and a gain stage.  Turns out the best quality signal is output from the first filter.  You need enough bandwidth to handle the QRS pulse with enough cutoff to knock out 60Hz.

Baseline wander is troublesome when just trying to get a strip holding the inputs with left and right hand...it is doable but not user friendly.  When actual sensor pads are attached to the chest, I get a very nice waveform.  My pads are old, so I rejuvenate them with a salt-water solution.  Fresh ones would likely perform better.

If I decide to take this further, I will probably move to a different lower-voltage inamp operating from a single supply. 

For now, what I have does the job.

[tpowell1830]

I started this a few years ago but never finished it.  Now that I am dealing with some issues, I decided to get it to the next step.

The circuit is pretty simple, using and AD624 inamp, followed by two second-order Sallen-Key LPFs...and a gain stage.  Turns out the best quality signal is output from the first filter.  You need enough bandwidth to handle the QRS pulse with enough cutoff to knock out 60Hz.

Baseline wander is troublesome when just trying to get a strip holding the inputs with left and right hand...it is doable but not user friendly.  When actual sensor pads are attached to the chest, I get a very nice waveform.  My pads are old, so I rejuvenate them with a salt-water solution.  Fresh ones would likely perform better.

If I decide to take this further, I will probably move to a different lower-voltage inamp operating from a single supply. 

For now, what I have does the job.

bradicardia?

[t1d]

Excellent, WT! I have played around, with these type circuits, a bit, myself. What I discovered is that they use the exact same technology as lie detectors. So, I began to include those circuits, in my web searches, too. Additionally, component manufacturers are beginning to offer HIDs, that interface better with individual, portable and DIY units. The major on-line component suppliers are beginning to carry those parts. Adding respiration rate and sweat detection are not very complicated and give the doctor even more/better information.

Great project, on the assembly! Do well. Keep us informed.

[SiliconWizard]

I'd suggest adding some input protection.

As a side note, you could take a look at the AD8232 which packs a lot of features.

[Wimberleytech]

bradicardia?

PVCs

[mtdoc]

Normal sinus rhythm. Rate just over 60 bpm. Not bradycardia.

[Wimberleytech]

I'd suggest adding some input protection.

As a side note, you could take a look at the AD8232 which packs a lot of features.

That is a nice part...but...my assembly skills are no match for that package!!

[SiliconWizard]

It's much like a 0.5mm pitch QFN. Doable with some experience.
There are breakout boards though, if you're interested in playing with this chip and evaluate what it can do.
Sparkfun has one: https://www.sparkfun.com/products/12650

[Wimberleytech]

It's much like a 0.5mm pitch QFN. Doable with some experience.
There are breakout boards though, if you're interested in playing with this chip and evaluate what it can do.
Sparkfun has one: https://www.sparkfun.com/products/12650

I could have saved myself some effort!  But I still had some fun.  I will buy one of these and play with it.  Thanks for bringing it to my attention.

[Wimberleytech]

I have one of these: https://store.alivecor.com/products/kardiamobile
Quite amazing little device.

[David Hess]

You need enough bandwidth to handle the QRS pulse with enough cutoff to knock out 60Hz.

Why not notch out 60Hz in a separate stage to relax the low pass filter requirements?

This might be a good application for inserting an integrator between the output of the instrumentation amplifier and its reference input to create an AC coupled instrumentation amplifier with a high pass cutoff at a low frequency.

[TimNJ]

You need enough bandwidth to handle the QRS pulse with enough cutoff to knock out 60Hz.

Why not notch out 60Hz in a separate stage to relax the low pass filter requirements?

This might be a good application for inserting an integrator between the output of the instrumentation amplifier and its reference input to create an AC coupled instrumentation amplifier with a high pass cutoff at a low frequency.

Indeed. Did an ECG for my senior design project..never fully finished it, but we used a circuit like that.

DSP filtering, namely an adaptive FIR filter, would be a good fit for this type of signal.