The ECG(heart Pulse) circuit problem...

[kingkong83]

i doing a project about the ECG sytem and unable to get the accurate heart pulse from the body... Can anyone tell me what is my circuit problem or my design have something wrong?? I have upload my circuit.. UAF42 is notch filter, INA114 is instrumentation filter.. i supply 9V to each component.. Hope can send u all reply soon... :wink:

 

[kingkong83]

my notch filter is used to remove the 50 Hz frequency.. Maybe the connection on the INA114 have some problem but i cannot solve it.. i give the the gain of INA114 is 1000.. after that, i pass the signal to the low pass filter and to notch amplifer.. and so on.. Any suggestion can u give me.. Urgent pls.. i have get 1 circuit about INA114 from the datasheet but not so understand how about the connection.. i have upload the circuit..
in my circuit, should i add the summing amplifer and inverting ampilfer?? I need pass this ECG signal to ADC..

 

[Optikon]

my notch filter is used to remove the 50 Hz frequency.. Maybe the connection on the INA114 have some problem but i cannot solve it.. i give the the gain of INA114 is 1000.. after that, i pass the signal to the low pass filter and to notch amplifer.. and so on.. Any suggestion can u give me.. Urgent pls.. i have get 1 circuit about INA114 from the datasheet but not so understand how about the connection.. i have upload the circuit..
in my circuit, should i add the summing amplifer and inverting ampilfer?? I need pass this ECG signal to ADC..

Did you wire up the drive circuit like shown in this example? You say you cannot get accurate heart signal. What do you get? Do you only get noise out of INAMP? Do you get a heart signal but it is too small / inaccurate / distorted? You have to tell us what you have measured. If you do not provide the guard drive connections like shown in example, leakage current can kill your signal. Single stage gain of 1000 might not be a good idea for noise purposes. Consider multiple low noise gain stages. You have to be more specific about what is failing in your circuit. I am not familiar with what the heart signal looks like. I presume it is in microvolt level so low noise design will be very important.

 

[kingkong83]

No, i just connect the electrodes to the right and left wrist only.. It is because i didnt understand the connection and i see some web site also didnt connect it. But can u explain the part i point(1 and 2). I dunno how to connect this two part..

i just can get a waveform that look almost like sin waveform only.. The voltage of the heart pulse very small only.. Just few milivolt.. So i amplify it with so larger gain. The heart pulse signal i get not i expect after i amplify, filtering with some filter. After i filtering i still get it a lot of noise.. Thanks for ur helping..

 

[Roff]

You definitely need the feedback drive circuit. It provides bias current and a common mode voltage reference to the INA114. It will also reduce the common mode input voltage (i.e., 50Hz) by a large amount, reducing or eliminating the need for the notch filter.

 

[Optikon]

No, i just connect the electrodes to the right and left wrist only.. It is because i didnt understand the connection and i see some web site also didnt connect it. But can u explain the part i point(1 and 2). I dunno how to connect this two part..

i just can get a waveform that look almost like sin waveform only.. The voltage of the heart pulse very small only.. Just few milivolt.. So i amplify it with so larger gain. The heart pulse signal i get not i expect after i amplify, filtering with some filter. After i filtering i still get it a lot of noise.. Thanks for ur helping..

I think you need the drive circuit as suggested. Where did you see an example that did not do this? That example may be under a different situation. Also, you mention sinusoidal wave, it may be oscillation. The point 1 and 2 as you have drawn point to the shield of the cable.
If the shield is driven at the same potential as the signal, then the shield capacitance is effectively "guarded" out. This will prevent a leakage current path that would otherwise distort your signal coming from a high impedance source.

Connect the drive circuit as shown then see what you get and report back.
Did you have some specific reason why you (cannot) use the drive circuit?

 

[red4925]

You definitely need to use a ground probe/lead on your leg or hip. You also want to ground the shielding of the other two leads to your chest/wrist. 1000 gain for the diff input is too high - listen to Optikan on that. You should be able to see an EKG signal at the output of your instrumentation amp (maybe a few millivolts peak-to-peak). If you're not seeing that due to noise, then check the output of your notch filter; you should see a small EKG signal there. I think it's not working because you're not using the grounding lead. The 1 and 2 refer to the shielding of the cable you're using as your leads; the schematic is showing you how to ground the leads to your leg through the feedback amps.

You may want to try connecting the leads to your chest (just under your pectorals); the signal there may be stronger than at your wrists. And make sure your leads to your chest/wrist are twisted together (if u r using separate cables) to minimize noise - you want identical noise on both so your input amp can cancel the noise.

 

[kingkong83]

i will taking all of your suggestion. Can u tell me more clear about the point 2??? this point is connect to ground or IN-?? Coz i really not sure how about the connection..

 

[Roff]

i will taking all of your suggestion. Can u tell me more clear about the point 2??? this point is connect to ground or IN-?? Coz i really not sure how about the connection..

Your RA and LA wires should be a shielded twisted pair. Don't connect point 2 to ground or IN-. Point 2 is the shield on the cable. Connect the end of the shield nearest the INA114 to the output of the OPA2604 voltage follower, exactly as shown in the schematic.

 

[kingkong83]

Ron H.. You mean that i no need connect the point 2 to any place is it???

 

[Roff]

Ron H.. You mean that i no need connect the point 2 to any place is it???[/quote
I can't imagine how you drew that conclusion from my comments. See the schematic below.

 

[kingkong83]

Ron_H.. You mean i should connect the crocodile click between the shield and the first output(OPA2604)?? But i cannot understand what will effect make from this point?? It is because the shield cannot conduct..

 

[Optikon]

Ron_H.. You mean i should connect the crocodile click between the shield and the first output(OPA2604)?? But i cannot understand what will effect make from this point?? It is because the shield cannot conduct..

yes connect as shown.

Not true. Keep in mind the shild has leakage paths to the signal. If the signal is moving, leakage capacitance will conduct signal current. Also
there is noise on your signal so its guaranteed to be moving. You can make the noise and error situation worse if you do not drive the shield at the same signal potential. The output of the OPA does this. The shield cannot conduct only in a DC sense. But you do not have a DC signal.

 

[Roff]

Ron_H.. You mean i should connect the crocodile click between the shield and the first output(OPA2604)?? But i cannot understand what will effect make from this point?? It is because the shield cannot conduct..

This is difficult to simulate accurately, but trust me (and Optikon), it is very important for minimizing common mode noise such as 50Hz and its harmonics. Don't worry about this connection for simulation - just make sure you do it in hardware.

 

[ENG_MOHAMMED]

HELP ME

PLZ SEND ME ANY LINK TO HELP IN DESIGN ACG CIRCUIT

 

[DaleM]

ECG Circuit

The circuit you are trying to copy uses the return drive to the leg to control the common Voltage on the patient via feedback. The ECG signal is only about 1 mV from the right arm to the left leg(largest directional amplitude). The trick with getting a good ECG signal is eliminating the common Voltage on the leg and arm so the amplifiers do not saturate. The body picks all sort of signals as an antenna.

An alternative to the back driven circuit is to attach three wires to the patient. Positive and negative signal leads and a common mode pick-up. Using a differential amplifier(use three operational amplifiers) you end up with [(Right Arm - Left Arm)-(Left leg - Left Arm)]. The Voltage on the left arm cancels out on the common mode input leaving the differential Right Arm - Left Leg. As the body floats the common signal will be compensated.

You could also go to your local hospital's biomedical department and get a schematic for the front end of an ECG machine. Get one for an older unit that has Radio Shack style parts. The drawing you have is for a higher end back fed patient potential.