Tips to reduce electrical noise in my workspace?

Thread Starter

cspwcspw

Joined Nov 8, 2016
78
Hi - I have a study / workspace with all sorts of electrical things: a laptop, a second screen, a scope, a signal generator, a guitar amp, a sound system, some lighting, a phone charger, and and air conditioner (for the human!) The room is carpeted with whatever came with the house, synthetic fibre of some sort. The "workbench" is an old wooden dining-room table, with countless layers of varnish and furniture polish, I guess. I have too many multi-plugs, but they all chain back to a single wall socket.

I'm looking for tips to reduce excessive electrical noise.

If I hold the scope probe tip with one hand and hold the ground in my other hand, I get pretty dirty 50HzPeak-to-Peak of about 4 volts. If I dump my elbows onto the table, that goes up to about 7V P-P. The usual things like turning off the aircon, screen unplugging things, etc. don't yield an obvious culprit.

Similarly, my DVM can generate a 5V square wave. If I place the DVM on the table (in its rubber case, etc), there is noticeable signal distortion visible on the scope,
20200203_151755.jpg


If I move the DVM onto an adjacent table it cleans up the signal considerably, (not perfectly).
20200203_151707.jpg
So I'm looking for tips: ways to identify the key culprits, and then useful tips for mitigating this, e.g. get an anti-static mat for the carpet first, or cover the table with a mat, or try varnishing the table with a conductive varnish. If I get some mats, should they be insulating or conductive?
 
Last edited:

Audioguru again

Joined Oct 21, 2019
6,671
That is why all audio and video cables are shielded cables and no Mickey Mouse solderless breadboards are used.
Even my oscilloscope probe uses a shielded cable so it picks up no interference.
 

OBW0549

Joined Mar 2, 2015
3,566
My situation is similar and I've found there isn't a whole lot that can be done about it. There are a few things that help, though.

Turning off all CFL and LED room lighting helps. Unplugging all test equipment other than what I'm using at the moment helps a bit, too, as does unplugging anything else with an internal switching regulator like wall warts and USB phone chargers. Sometimes I'll disconnect the two data aquisition laptops and take them out of the room so their power supplies don't add to the din of electrical noise. With a lot of electronic stuff these days, simply turning them off doesn't help much; you have to unplug them completely.

If I'm working on anything that's extremely susceptible to electrical interference and all else fails, I have a shielded enclosure made from a large cookie tin that I can place my work in, with shielded feed-throughs for power and signals in and out.
 

ebeowulf17

Joined Aug 12, 2014
3,307
If I hold the scope probe tip with one hand and hold the ground in my other hand, I get pretty dirty 50HzPeak-to-Peak of about 4 volts.
It's worth pointing out that you're deliberately creating a worst case scenario there. Using your body as an antenna is a great way to pick up stray electrical noise. If there's mains power nearby, grabbing a scope probe with your fingers is a great way to find it!

How bad is the noise when you're not inserting yourself unnecessarily into the circuit? If you scope something that isn't mains derived (any steady voltage in a battery powered device) and take your hands off the uninsulated part of the probes, how much 50Hz noise do you see?

Don't get me wrong - there's nothing wrong with learning how to minimize the noise in your environment, but I'd hate to see you spend a lot of time or money fixing this if it doesn't impact the measurements of anything but your own body!
 

SamR

Joined Mar 19, 2019
5,031
Grounding and Shielding! PWM to LEDs was one problem I found that I thought was being generated by fluorescent lighting which is also noisy. We live in an electronic soup and do the best we can.
 

OBW0549

Joined Mar 2, 2015
3,566
PWM to LEDs was one problem I found that I thought was being generated by fluorescent lighting which is also noisy.
That's a good point, one that I forgot. Anything operating via pulse width modulation is a potential source of interference.

We live in an electronic soup and do the best we can.
Amen to that. The lower part of the EM spectrum is full of noise anymore, largely due to the prolifieration of switching regulators.
 

Thread Starter

cspwcspw

Joined Nov 8, 2016
78
It's worth pointing out that you're deliberately creating a worst case scenario there. Using your body as an antenna is a great way to pick up stray electrical noise. If there's mains power nearby, grabbing a scope probe with your fingers is a great way to find it!

How bad is the noise when you're not inserting yourself unnecessarily into the circuit? If you scope something that isn't mains derived (any steady voltage in a battery powered device) and take your hands off the uninsulated part of the probes, how much 50Hz noise do you see?

Don't get me wrong - there's nothing wrong with learning how to minimize the noise in your environment, but I'd hate to see you spend a lot of time or money fixing this if it doesn't impact the measurements of anything but your own body!
:) Thanks - yeah, I'm forever chasing things that might not be necessary! I'm strictly a DIY hobbyist so there aren't any "specifications" for how electrically clean I need the environment to be. Using the DVM to generate its square wave was my "hands off, don't insert yourself into the circuit" scenario.

But I'd stumbled across the same noise problem a few weeks ago: I was playing with monitoring power usage with a Current Transformer Clamp. I sampled the resulting A/C wave on an Arduino / ESP32 and integrated the positive and negative halves of the wave separately. I thought symmetry should get the two halves to be about equal areas (and they were close), and I reported the "imbalance" as my error. I built it on a breadboard, and then re-did it on a soldered board and added some shielding and filtering, etc. to reduce the noise. But then I also found that moving the processor off my worktable greatly reduced my noise. So I've been carrying around a niggle that my space is overly noisy, and the niggle was provoked back into the foreground again today. (That project is at https://github.com/cspwcspw/Yet-Another-Energy-Monitor if anyone is interested.)
 

Audioguru again

Joined Oct 21, 2019
6,671
I got a visit from the RF cops when I was a teenager in my first apartment. It did not have air conditioning so I made a fan with a motor I found. The motor had sparking brushes and the RF cops said it ruined radio and TV for hundreds of people. The first people who complained bought a new fan for me.
 

ebeowulf17

Joined Aug 12, 2014
3,307
:) Thanks - yeah, I'm forever chasing things that might not be necessary! I'm strictly a DIY hobbyist so there aren't any "specifications" for how electrically clean I need the environment to be. Using the DVM to generate its square wave was my "hands off, don't insert yourself into the circuit" scenario.

But I'd stumbled across the same noise problem a few weeks ago: I was playing with monitoring power usage with a Current Transformer Clamp. I sampled the resulting A/C wave on an Arduino / ESP32 and integrated the positive and negative halves of the wave separately. I thought symmetry should get the two halves to be about equal areas (and they were close), and I reported the "imbalance" as my error. I built it on a breadboard, and then re-did it on a soldered board and added some shielding and filtering, etc. to reduce the noise. But then I also found that moving the processor off my worktable greatly reduced my noise. So I've been carrying around a niggle that my space is overly noisy, and the niggle was provoked back into the foreground again today. (That project is at https://github.com/cspwcspw/Yet-Another-Energy-Monitor if anyone is interested.)
Ah, well if it's noticeably affecting readings when you're testing your actual circuits, then it's certainly worth further investigation.

A few of your examples have mentioned on vs. off the table. It's odd to me that a wooden table itself would be trouble, but it sounds like it is. You might find your easiest improvements will involve isolating things from the table, as opposed to eliminating all noise in the area.

We ran into trouble when some of our machines were installed on metal counters which worked as great antennas for injecting outside noise into our machines, causing glitchy displays, etc. Future designs will need better shielding, insulation, decoupling, etc... but the short term fix for machines in the field was simply adding tiny little rubber feet. I don't know if it was the electrical insulation, the extra 1/4" of spacing, or both, but the little feet solved nearly all problems.

I'm not saying that little rubber feet will fix everything, but it *might* be possible to decouple your table noise from circuits and test instruments with something similarly simple.
 

SamR

Joined Mar 19, 2019
5,031
LOL I bought a resistor sparkplug for my neighbor and even volunteered to install it for him. Saturday morning he would mow his grass and DESTROY the HAM bands while I was working Europe.
 

danadak

Joined Mar 10, 2018
4,057
Sounds like a good product for mass production would be some form
of addressable power strip. In my lab, unplugging not an option, bad knees.
Plus where I shove all this wiring. Also handle 10/100 cables, they are
nasty as well.

My estimate in lab there is 1 long power strip, maybe 12 outlets, 3 short 6
outlets/strip. And not enough. Idea is to eliminate all the antennas and
remove power to off circuits in instrument, just got a spectrum analyzer
and power switch is configured as heartbeat LED power switch. Thats
when my profession is not always displaying intelligence.

Regards, Dana.
 
Last edited:

djsfantasi

Joined Apr 11, 2010
9,156
Sounds like a good product for mass production would be some form
of addressable power strip. In my lab, unplugging not an option, bad knees.
Plus where I shove all this wiring. Also handle 10/100 cables, they are
nasty as well.

Regards, Dana.
Addressable Power Strips are made. They are used in remote computer hosting facilities for lights out computing. I had several installed in my hosting racks, so if a single piece of equipment needed rebooting, I could do it from home on my laptop without having to travel into the city.
 

Thread Starter

cspwcspw

Joined Nov 8, 2016
78
Ah, well if it's noticeably affecting readings when you're testing your actual circuits, then it's certainly worth further investigation.

A few of your examples have mentioned on vs. off the table. It's odd to me that a wooden table itself would be trouble, but it sounds like it is. You might find your easiest improvements will involve isolating things from the table, as opposed to eliminating all noise in the area.

We ran into trouble when some of our machines were installed on metal counters which worked as great antennas for injecting outside noise into our machines, causing glitchy displays, etc. Future designs will need better shielding, insulation, decoupling, etc... but the short term fix for machines in the field was simply adding tiny little rubber feet. I don't know if it was the electrical insulation, the extra 1/4" of spacing, or both, but the little feet solved nearly all problems.

I'm not saying that little rubber feet will fix everything, but it *might* be possible to decouple your table noise from circuits and test instruments with something similarly simple.
Thanks. My experiment is compromised, in that all the noisy stuff sits on the table too. So moving stuff off the table is also increasing the distance from laptop, screen, phone chargers, etc. So is it removing the table, or increasing the distance, that improves my result? I don't know yet.

At my workplace building we had separate circuits for computer equipment and "all the rest". Perhaps getting two separate supplies into my room and putting all the switched-mode noisy stuff on one (in a steel box or cage or some metal skirting if I have to) might help. But from the comments on this thread it seems that EMF radiation, rather than "dirt on the mains" might be a more common culprit.
 
Top