Mains hum

If it's not twice mains frequency, it's probably not acoustic. The item has a metal box, so light sensitivity is likewise ruled out. That leaves magnetic coupling, i.e. ground loops. Less likely, could be capacitive.

Pairing the input wires, twisting the pairs, would be one approach. Keeping the area between the wires small, and orienting it differently, will change the coupling. Also, find a room with no large current-draw appliances, and no knob-and-tube wide separation of AC hot and neutral wires. In the old physics building, sensitive experiments were done with lights-off in the library, after hours. If there are transformers or motors nearby, unplug 'em (a 12V high intensity light transformer is a BIG magnetic AC source).

Old-school treatment: put a twin-tee filter in the signal path to trap that pesky power frequency. New-school treatment: FFT the signal, and zero out the power frequency: maybe look at the harmonics, too, for other artifacts.

What happens when the leads are properly terminated? In use, they are attached to sensor pads that are stuck near the collar bone (1) and on the left and right side of the belly

Ed

Yandex. Russian. Makes sense.

I see the signal, it is clearly not sinusoidal. So you will need to do more than just trap the fundamental, but also deal with harmonics. Since the fundamental frequency drifts, it won't be possible to use an extremely high Q filter. what frequency range are you interested in? Is it possible to perform post capture processing to remove the background noise?

It looks like your leads are made to attach to the patches placed on a person. I would have thought to use coax wire, but maybe there's no point since the person it's attached to is not able to be shielded. Are your probes connected to a subject when you see this signal? Have you tried attaching them to a subject? It may be that the high impedance input is seeing a high noise signal which would essentially be grounded through the patient if the leads were connected... just a thought. Also look at the signal without the input leads at all.

BTW, why do you obscure the readout data in the scope capture? I count 4.3 divisions per cycle of the waveform. I can't think of a multiplier that would make that a power line frequency anyplace I know of.

Cursitor Doom wrote: =================

** 50/60 Hz electric fields are everywhere inside wired buildings. That is why you shield things. ** But you have a DSO connected and that is supply grounded. That is your problem creating a lot of common mode hum. ** Dynamic mics are low impedance and well shielded. ECGs are very special devices. ..... Phil

Ricky the IDIOT puked:

=====================

** That is the f****ng heart beat you moron !!! The 50/60 Hz hum is the superimposed fine wiggling.

Rest of this retarded wanker's absurd drivel snipped.

.... Phil

Er, yeah, that's why I obscured the time/div., to avoid confusion for those in the US. Well, that was the plan, anyway. The period is 20mS which equals 50Hz. I'm in Yurp and that's the mains frequency here.

I had a similar problem before with another such die-cast box housing a very small signal amp. On that occasion I was able to kill the hum stone dead by clamping the box to an earthed metal sink. But that hasn't worked this time for some reason. :-/

Some good suggestions all in all; cheers.

** ?????

More like 60 bpm.

...... Phil

How does hiding info "avoid confusion". I think we all here are aware that the world is not on 60 Hz.

So this is an actual heart beat with the tiny 50 Hz noise? That doesn't look so bad. I think a notch filter will clean that up very well. That can be implemented easily in the digital domain if you are capturing the signal using an ADC.

On a sunny day (Tue, 05 Apr 2022 19:05:23 +0100) it happened Cursitor Doom snipped-for-privacy@notformail.com wrote in snipped-for-privacy@4ax.com:

One way is perhaps to pick up mains with some wire and then 180 degrees inverse it with some opamp and then subtract that from the output?

This is sometimes done in ham radio to cancel noise from wallwarts in the house etc small antenna indoors to subtract from the real big antenna outside,

If the pickup is by the wires themselves, the details of the arrangement would modify the details of the pickup, so a static counter signal would not be effective.

Much better to just narrow band block that frequency. I'm not sure what impact that might have on the signal. I believe an EKG is analyzed by eye, so that the sort of distortion caused by most filtering would not be noticeable. It's easy enough to digitize the signal and run a few tests.

Distinctly unhelpful since at least one poster mistook the heartbeat waveform rather than the wiggles for the mains hum. I can't imagine why.

One distinct possibility is that the ECG is actually filtering out

*60Hz* mains rather than 50Hz. Most such devices have a notch reject filter for mains with a moderately high Q so that they will give good rejection for the spot mains frequency +/- 0.05 Hz.

Try it out in a field away from local mains and on battery power and see if it is clean then. Some of your problem could be coming from the bench psu.

Most of it I expect comes from the flying leads capturing magnetic flux so plat them together leaving just enough free lead to connect it up.

Any high gain amplifier will amplify unwanted mains hum but in these systems most of it should be common mode and rejected by the front end.

Finger on the input test works well enough for most high gain audio amplifiers if a signal generator is not to hand.

My guess is that the hum is coming from your patient / target.

The capacitance of an average person is about 20 pF to the active phase line and 200 pF to ground. You can assume that the patient has about 10% of the line voltage through a pretty high impedance voltage divider.

When we made ECG equipment in the early 1970's, the solution was to make the input of the amplifier differential and as high impedance as possible, including guard bootstrapping the shield braids of the input cables. The impedance of the connection electrodes are hardly ever identical, and this creates voltage dividers with the amplifier input impedances. The imbalance works directly to convert the common-mode hum into differential input.

Tauno Voipio wrote: =================

** Spot f****ng on. ** Yep, has the OP never handled a live audio line ?

Bluuuurp, bluuup.....

** As does grounding the diff amp metal enclosure. If the whole kaboodle is *floating* - the diff disappears.

FYI:

The OP is a notorious idiot and is "dreaming" since his posted scope image is good.

...... Phil

Although I believe my pacemaker is set to let me get a bit slower before it starts geeing me up...

Yeah, I know what you mean but no cigar for you this time. I was able to clean up the hum by using good old fashioned ferrite beads inside the box at the point where the power lead comes in. Worked like a charm. Sometimes the best solutions are the old ones...

Ingenious, Jan; ingenious. I'd have loved to have tried that idea out if I'd had a bit more time to spare. In the end, ferrite beads did the trick. Very analogue! :-)

The 60 beats is just a number near the average. For people that are in very good health and do a lot of physical activity the heart may beat slower than the 'nornal average' when at rest, others may beat faster if not very active and in good physical shape. I think mine is mnore like

70 some BPM.

From what I see this is a simple decvice and maybe has only 2 leads. The heart monitors I am familiar with has 3 leads where the internal circuits filter out the stray electrical noise picked up by the monitor. Big difference in a device under $ 100 and the professional devices. Then there are the multilead devices.