Go Down

Topic: Is it safe to connect Arduino to human body? (Read 6842 times) previous topic - next topic

Garnagar

Jul 15, 2015, 05:48 pm Last Edit: Jul 15, 2015, 08:15 pm by Garnagar
Hi

I want to create some project which uses Arduino to collect and display ECG. It uses two electrodes which are attached to body (on chest) and signal si transmitted through op-amps to Arduino ADC-pin. Is it safe for both human and Arduino board to be connected this way? Should I make some safety pracautions and if so, which?

Thanx for answer. (Sorry, if this is wrong forum but i'm not sure where to put this.)

Garnagar

far_1

Hmm. Not sure. The arduino does pack enough current to cause damage or even death, but voltages are far too low for the human skin. Even when wet and good conduction voltages below 10V are not particular dangerous, since you usually notice the pain and react. At 5V, i would guess it's not that of a concern.
You might have however some HV voltage quirk on the lines (as i did, but that's a different story).

Personally if i'd probably wouldn't worry too much, since i've experienced worse things. But proceed at your own risk.  There's always a chance of something going wrong, especially if your circuit is connected to a HV source in any shape or form.

Cidos

If you stay below 20mA for a longer time its not critical.


Regards
Ben

DVDdoug

#3
Jul 15, 2015, 07:25 pm Last Edit: Jul 15, 2015, 07:36 pm by DVDdoug
Yes, it's safe.    The Arduino runs off 5V.   Your op-amps may run off slightly-higher voltage.

Anything less than 50V is generally considered safe.   (The safety regulations for medical applications may be more strict than household/industrial standards.)         

I'm touching 5-12V circuits every day at work.   You'll notice that normal batteries or wall-wart power supplies/chargers are not insulated to prevent you from touching the contacts.    There's no danger with these low voltages.

It's current  that can kill you, but current depends on voltage and  resistance.   The high resistance of the human body means you can't get dangerous current without high voltage.    Even if you're soaking wet (lower resistance) you are not going to feel 5V.    (You can feel the voltage from a 9V battery if you touch it to your tongue. :D )

Quote
to collect and display ECG. It uses two electrodes which are attached to body (on chest) and signal si transmitted through op-amps to Arduino ACD-pin.
If everything is working properly, you are not applying voltage to the body, you are reading voltage naturally generated by the body.    But of course, you have to allow for something going wrong.

polymorph

Yes, and no.

As others have pointed out, the Arduino runs from 5V.

However, if you are not using batteries, then leakage from the powerline becomes an issue. An ungrounded power supply may be floating at 1/2 of the AC line voltage. And your circuits specifically strive to make a low resistance connection to your body.

I've been shocked before on dry skin by a floating switch mode power supply that had a missing ground pin on the power plug. I was touching actual ground, and case ground, and got a bit of a tickle with dry skin. High impedance voltage measurement was about 60Vac (120Vac powerline), and short circuit current was around 25mA.

I'd run it from a battery.
Steve Greenfield AE7HD
Drawing Schematics: tinyurl.com/23mo9pf - tinyurl.com/o97ysyx - https://tinyurl.com/Technote8
Multitasking: forum.arduino.cc/index.php?topic=223286.0
gammon.com.au/blink - gammon.com.au/serial - gammon.com.au/interrupts

Garnagar

But I need connection to PC becase I'm sending there data. Is there any way I can send data and not power Arduino from PC?

far_1

Yes, and no.

As others have pointed out, the Arduino runs from 5V.

However, if you are not using batteries, then leakage from the powerline becomes an issue. An ungrounded power supply may be floating at 1/2 of the AC line voltage. And your circuits specifically strive to make a low resistance connection to your body.

I've been shocked before on dry skin by a floating switch mode power supply that had a missing ground pin on the power plug. I was touching actual ground, and case ground, and got a bit of a tickle with dry skin. High impedance voltage measurement was about 60Vac (120Vac powerline), and short circuit current was around 25mA.

I'd run it from a battery.
I had almost the same issue. Something was wrong with the PSU and i kept getting zapped with 70 or so volts. Went right trough normally conditioned skin.

But I need connection to PC becase I'm sending there data. Is there any way I can send data and not power Arduino from PC?
Power the arduino trough battery and just use the TX pin (with a ressistor for added protection).


But thing is...if your equipment is in good working condition there probably wouldn't be any problem even if PSU powered. Normally things get bad when you're messing with currents and voltages you have poor control over. During normal use there's a small chance things would get lethally wrong.
In the usuall scenario if you mess something up you're probably going to wreck either the USB port or the power supply, which might end up zapping you as i've described above.

Of course there are ways to have maximum safety, but often they are overkill. To be hones i have no idea how exactly is your wiring, what are the inputs, outputs, etc. I'm just guessing right now.

Archibald

#7
Jul 15, 2015, 09:49 pm Last Edit: Jul 15, 2015, 09:51 pm by Archibald
Is it safe for both human and Arduino board to be connected this way? Should I make some safety pracautions and if so, which?
You should not be relying on advice from forum members who probably have no in-depth professional knowledge of electrical safety (that includes me!).

With electrodes connected close to the heart, you obviously need to be especially careful about safety. In this situation, don't even rely on the advice of a qualified electrician.

DVDdoug may assure you that it will be safe, but see the requirements for 'Separated or Safety Extra-low Voltage' in this Wikipedia article:
https://en.wikipedia.org/wiki/Extra-low_voltage.
With electrodes close to the heart, commercial ECG equipment may adopt an even higher standard of safety.

far_1

Well yes. Having electrodes close to vital organs, operated by insufficiently skilled people does require a certain degree of bravery...or carelessness.

However people work constantly within these levels and rarely there are lethal accidents. Also, considering the electrodes are attached to the skin surface...the safety margin widens.

Speaking strictly personally...if it was my project, tested on myself...i'd probably go forward without that much hesitation. Especially if it's more of a one time only kind of thing. If it would be tested on others in variable circumstance...then it would be very wise to give it a second thought.

If you're really paranoid then i guess running the system on low voltage batteries and saving the data on an SD module for later analysis is a...doable way.

polymorph

Small chance, but high risk. I'd use an SD card, or use optoisolation. With EEG pads and gel, you can possibly get a very low resistance akin to penetrating the skin.
Steve Greenfield AE7HD
Drawing Schematics: tinyurl.com/23mo9pf - tinyurl.com/o97ysyx - https://tinyurl.com/Technote8
Multitasking: forum.arduino.cc/index.php?topic=223286.0
gammon.com.au/blink - gammon.com.au/serial - gammon.com.au/interrupts

Paul__B

OK, the situation is that ECG measurements are performed with electrode pads whose very purpose is to have an excellent connection to the skin and thus, the internal body structures.

The very design of an ECG (electrocardiograph instrument) is that it measures such signals and applies no voltage at all to the electrodes - at any time.  So your task if you propose to build one is to ensure the design fulfils this requirement.

There are two aspects to this.

One is the requirement that the operating voltages of the measuring amplifier cannot be applied to the electrodes.  This is generally ensured by using high impedance buffer op-amp circuitry and high-value series resistors in each lead, such as one megohm.  There are special components made for such protection which behave somewhat like the polyfuse in the Arduino, but "open" much faster and at microamp levels rather than milliamps.  In general, a design which uses 1 megohm series resistors feeding the "+" inputs of op-amps wired as non-inverting buffers, will perform this task.

The second aspect is to ensure complete isolation from any other power source, including a mains power supply or a USB interface.  This is generally achieved using warranted DC-DC isolation converters.  Obviously that is precisely what is contained in my own USB interfaced ECG which the manufacturer has of course, been required to have approved by the relevant certifying authorities (which is part of the justification for a price tag such as this, similar to the one I use).

A simple approximation to such isolation is to use some form of wireless data transmission from a battery-powered device completely enclosed in a fully-insulated case.

raschemmel

#11
Jul 16, 2015, 04:26 pm Last Edit: Jul 16, 2015, 06:00 pm by raschemmel
Quote
You should not be relying on advice from forum members who probably have no in-depth professional knowledge of electrical safety (that includes me!).

With electrodes connected close to the heart, you obviously need to be especially careful about safety. In this situation, don't even rely on the advice of a qualified electrician.
  
I worked as a technician in a factory where they made hospital bedside patient monitors. (EKG+RESPIRATION MONITOR +PULSE MONITOR)

The op amp circuits required for all the filtering (active filters) and signal conditioning
were all dual supply circuits +/- 15V (if I remember correctly . That was 1984)

The dual supply voltages were generated by inductive coupling  charge  pump circuits that used magnetic coupling to create an electrical isolation barrier. On the
input side was all the circuitry that ran off the ac wall power. On the patient side (all the ekg electrode monitoring circuitry) was 100% electrically isolated from the the AC input side by the magnetically coupled charge pump power supply circuits.
(not sure if that is the correct name for that technology but it is the same thing used for wireless cell phone battery chargers, more or less) The signal conditioning op amp circuits ran off +/- 15Vdc that was totally isolated from the input side. It was all FDA approved. FYI, the signal conditioning circuitry was extensive and complex. It is not the kind of circuitry that could be DIY'd, if you get my drift (unless you're a BSEET with 20 yrs experience). You can DIY something but the signal you end up with won't be a perfect duplication of the heart signal like on the hospital monitors. The acquired conditioned heart signal output was isolated by opto isolators on the op amp inputs of the circuitry that supplied the video signal to the monitor.
Arduino UNOs, Pro-Minis, ATMega328, ATtiny85, LCDs, MCP4162, keypads,<br />DS18B20s,74c922,nRF24L01, RS232, SD card, RC fixed wing, quadcopter

MarkT

Definitely electrically isolate if any connection possible to the mains in any form.  This might be
as simple as only powering from an isolated (no charger) laptop.

The issue with EEG signals is interference and the high impedance signal source. For ECG
they use special coax-in-coax to allow a bootstrapped middle coax conductor to cancel out
the cable capacitance from loading the signals. (As featured in "The Art of Electronics").  And
I presume its all differential sensing.
[ I will NOT respond to personal messages, I WILL delete them, use the forum please ]

Paul__B

The dual supply voltages were generated by inductive coupling  charge  pump circuits that used magnetic coupling to create an electrical isolation barrier. On the input side was all the circuitry that ran off the ac wall power. On the patient side (all the ekg electrode monitoring circuitry) was 100% electrically isolated from the the AC input side by the magnetically coupled charge pump power supply circuits.
That's another way of saying an Isolated DC/DC Converter as I described.  These have of course, for some thirty-odd years been available as a fully-specified module.  Like any other switchmode converter, they use high frequencies (tens of kHz to possibly MHz) to permit the use of small transformers (such as pot cores) with substantial insulation barriers between windings

raschemmel

#14
Jul 17, 2015, 12:55 am Last Edit: Jul 17, 2015, 09:00 pm by raschemmel
Of course FDA approval is contingent on many factors, not the least of which is the isolated supply. The high price tag is probably due more to the test & documentation than to the cost of the technology (components) itself.

Not sure if this is any help:
ECG Processing

ECG Processing hardware
Arduino UNOs, Pro-Minis, ATMega328, ATtiny85, LCDs, MCP4162, keypads,<br />DS18B20s,74c922,nRF24L01, RS232, SD card, RC fixed wing, quadcopter

Go Up