0.5mV amplify to 3V-5V

Thread Starter

reshin

Joined Oct 6, 2011
6
The design is amplify the heart beat signal which is 0.5mV to 3V-5V. My idea is to used op amps to amplify the the voltage. Do anyone have any suggestion for my idea?

Thanks
 

thatoneguy

Joined Feb 19, 2009
6,359
You only need a gain of 6-10. What will the output go to?

A rail to rail low noise op amp would be a big part of your solution, but a bit more info is needed, especially if this is more than a hobby device, anything having to do with human health circuits have a ton of regulations.
 

tom66

Joined May 9, 2009
2,595
0.5V (half a volt) or 0.5mV (half of a thousandth of a volt)?

Amplifying 0.5V is no big deal - you could do it with a jellybean op-amp like the LM324 or TL071. Amplifying 0.5mV on the other hand will be tricky, although not impossible. You'll probably need an auto-zeroing low noise op-amp. And if you want any kind of bandwidth (more than 1kHz or so), you'll need a high speed one. These are expensive.
 
Last edited:

Audioguru

Joined Dec 20, 2007
11,248
First we must know if your idea is for an audio electronic stethoscope or if it is an electrical ECG device.

Opamps are used for an audio electronic stethoscope and instrumentation amplifier ICs are used for ECG.
 

Adjuster

Joined Dec 26, 2010
2,148
0.5V (half a volt) or 0.5mV (half of a thousandth of a volt)?

Amplifying 0.5V is no big deal - you could do it with a jellybean op-amp like the LM324 or TL071. Amplifying 0.5mV on the other hand will be tricky, although not impossible. You'll probably need an auto-zeroing low noise op-amp. And if you want any kind of bandwidth (more than 1kHz or so), you'll need a high speed one. These are expensive.
Auto zeroing, chopping or other extreme measures aimed at ultra-low DC offset may not be required here.
Even if this is a true ECG (US: EKG) signal, a perfect recording of the DC baseline may not be required.

One would not expect to see much DC difference between the electrode potentials in the first place, and even if there were some, e.g. due to differences in their contact potentials, it might be considered as an artifact in the recording. Even some lower frequency AC variations (at lower than the heart rate) can be regarded as interfering "baseline wander". Analogue or (increasingly) digital filtering is often used to eliminate these LF components. This American Heart Association article appears to recommend an LF cutoff of 0.05Hz. http://circ.ahajournals.org/content/115/10/1306.full

Here is a piece on wavelet filtering - it's all too modern for me, but may interest the OP?
http://www.waset.org/journals/waset/v3/v3-4.pdf
 

SgtWookie

Joined Jul 17, 2007
22,230
If it is for an EKG, have a look at the datasheets for the INA118 and INA128 instrumentation amplifiers. There are applications for EKG monitors in both datasheets on page 11, including how to drive the shielding for the wiring (which really should be twisted pair to minimize noise pickup) and the guard potential drive for the subject.

Datasheet:
http://www.ti.com/lit/ds/symlink/ina128.pdf
 

Audioguru

Joined Dec 20, 2007
11,248
Many medical words about the heart include "cardio", not kardio.
Then why do Europeans call ECG things EKG?

There are some very old ECG circuits done by school kids using old 741 or LM324 opamps on the internet. I doubt if any of them worked properly.
 

SgtWookie

Joined Jul 17, 2007
22,230
Many medical words about the heart include "cardio", not kardio.
Then why do Europeans call ECG things EKG?
Because the physician who did the initial development work on it was Dr. Emil du Bois-Reymond, who was German, where it is spelled "Elektrokardiogramm". I can only surmise that the abbreviation EKG is used in the English language as a nod to the German physician who originated electrophysiology. [eta] It's more likely because the first machines were made by a German company in Munich; see below.

http://en.wikipedia.org/wiki/Emil_du_Bois-Reymond

The first person to actually record an electrocardiogram of the human heart somewhere around 1887-1888 was Dr. Augustus Desire Waller:
http://profiles.incredible-people.com/augustus-desire-waller/

The first commercial EKG device was developed by Dr. Willem Einthoven, a Dutch gentleman:
http://en.wikipedia.org/wiki/Willem_Einthoven
Einthoven's electrocardiograph was initially manufactured in Germany by Edelmann and Sons of Munich. He later went with the Cambridge Scientific Instrument Company, Ltd., of London.

The first EKG machine introduced to the United States was an Edelmann String Electrocardiograph brought by Alfred Cohn in 1909.

There is other information on this page:
http://rmccrory.tripod.com/historyekg.html
but there are no resources cited, and you will receive at least one pop-up, and likely one pop-under advertisement.
A more complete history, with at least some of the information sources referenced:
http://www.ecglibrary.com/ecghist.html

There are some very old ECG circuits done by school kids using old 741 or LM324 opamps on the internet. I doubt if any of them worked properly.
Yes, the 741 opamp circuits would leave a lot to be desired. The 324 opamp circuits might work somewhat, but the CMRR (common mode rejection ratio) would be terrible due to resistor tolerances.

For those who are new to instrumentation amplifiers, I recommend reading at least the first few pages of Intersil's Application Note #1298:
http://www.intersil.com/data/an/an1298.pdf

You don't have to go through all the math if you really don't want to; just read the text and look at the schematics. Try to get through at least pages 2 thru 12.
 
Last edited:

Audioguru

Joined Dec 20, 2007
11,248
Sometimes the common-mode signal is very high. The DC offset voltage is amplified a lot and can also be high.
Many ECG circuits feed negative feedback along the shield to the patient's body to cancel DC offset and common-mode AC and DC.
 

SgtWookie

Joined Jul 17, 2007
22,230
Sometimes the common-mode signal is very high. The DC offset voltage is amplified a lot and can also be high.
Yes, exactly. For the sake of our new hobbyists, that's why a high CMRR is important; you want to reject as much of the signal that's the same on both inputs as possible; and this is very difficult to do with "home-made" instrumentation amplifiers as documented in the Application Note that I referenced earlier.

Many ECG circuits feed negative feedback along the shield to the patient's body to cancel DC offset and common-mode AC and DC.
This is just what is shown in the example application schematics in the INA118/INA128 datasheets on the bottom of page 11.

It will help new readers to go through some of those pages I referenced; as Dr. Einthoven discovered that placement of the electrodes is important.
 

Adjuster

Joined Dec 26, 2010
2,148
Ah, did not realise DC offset was not a problem. In which case an auto-zeroing op-amp is not necessary.
DC offset can be a problem, as others have pointed out. It's just that since we don't need a zero-frequency response, the effect of DC offset can normally be suppressed by feedback or other filtering methods.
 

Audioguru

Joined Dec 20, 2007
11,248
Hi audioguru,
It is for ECG.
Your idea to use old opamps was tried 40 years ago. I don't think anybody was able to make a circuit that worked.
Today an excellent ECG circuit is shown in the datasheet of most instrumentation amplifier ICs.

The opamps and resistors in an instrumentation amplifier IC are matched.
 

JMW

Joined Nov 21, 2011
137
I don't know what your project is, but medical equipment is pretty tough. This site has a Tos for automotive, medical is far more restrictive. Usually for medical components you need a waiver from the manufacturer, if they even sell to the medical community.
Just my 2 cents.

Regards,

Jim W

Really like this site.
 
Top