Need help with ECG sgnal, still distorted

[integralx2]

I have built an ECG system. What I am not sure is why the signal is still distorted. Now i think its the wiring. I am using just basic breadboard wiring that came with breadboard. Maybe its not shield enough ? What i am struggling still to understand is why do we connect an electrode to the lower leg? What does it provide ? The instrumental amplifier i am using is an AD620. I have included the signal i get where it says oscilloscope connection. PLEASE ANY HELP WOULD BE GREAT.


THe amplifier on the most left side is a TL071 then going to the right its the AD620 then another TL071 then last amplifier is a LM339. The system is correctly setup, and this design does work! (just mine does not).

 

[rfsystem]

I see that the instrumation amplifier deliver a, I guess, buffered signal to the guard drive amplifier which use the mean avlue of the buffered inputs. But what I am missing is the ground connection to the body at all. I assume that the input common mode voltage of the instrumentation amp is limited. So if there is no ground reference at all I guess that the diff inputs are hanging somewhere at the edge.

A good design feature would be to indicate the out-of-range in a digital way to the user.

 

[karesz]

Hi RFsystem,
I think you will be in short in trouble with "patient safety", pls refer to the ADI script...
For me is only a cable shielding is missed (but it is maybe not a reason of distorted signals), but not a GND contacting from patient self.
K.

 

[rfsystem]

But how to avoid that the mean level of your patient goes outside the inputs of you amp?

 

[karesz]

Maybe this stuff gives some ideas...
www.nottingham.ac.uk/nursing/practice/resources/cardiology/function/placement_of_leads.php
https://en.wikipedia.org/wiki/Electrocardiogram
Open ECG Project:
www.open-ecg-project.org/tiki-read_article.php?articleId=1
K.

 

[rfsystem]

I think there is no other way than isolate your measuring system. The prefered way is to have the isolation in the digital interface. Probably you have to use special transformers to supply the "patient"-side. Like have a equipment earth connected shield foil between primary and secondary windings. That is cheaper if you only supply the analog instrumentation amplifiers and the ADCs.

I you want to get legal stickers there is certain some rules which you can use for conceptional design.

 

[keith1200rs]

I am still suspicious of the common mode leg drive - I mentioned it before in your other thread. It appears to be inverting to me and I don't think it should. The Analogue Devices AD620 data sheet, where your diagram comes from, gives a contrary connection for a common mode shield driver, which is what I think the leg connection is. It is also the way you will see it drawn in many other data sheets such as the INA121.

Keith.

 

[Audioguru]

The common mode drive to the leg is inverting so that it can cancel most of the common mode signal.

 

[rfsystem]

Sorry I overlook the inverting amplifier configuration.

I found a more detail here:

https://www.analog.com/library/analogDialogue/archives/37-11/ecg.html

The high negative gain drive the overall common mode to the ground symbol connection of the driving opamp.

But the issue with the floating ground remain. If the common mode rejection amplifier drive some current, what is the return patch for this driving current.

Furtheron ADI clearly state the requirement of isolation.

 

[keith1200rs]

I was mixing it up with a common mode shield drive - looks similar but no inversion. The link you posted makes more sense than the one on the AD620 datasheet because it has the closed loop leg drive on the patient side of the isolation whereas the AD620 datasheet puts it on the circuitry side.

Keith.

 

[karesz]

Hi,
I think AD620 datasheet makes the same to suggest as here(fig.41),than he copied surly the same picture in the topic, but the "Common mode cable shield driver" on fig. 45 is really noninverting...
Regards!
K.