Audio frequency of a humidifier

[DaveC426913]

TL;DR

Trying to figure out something - anything - about this annooying noise from my humidifier

Brought out a small humidifier much like this Homedics Ultrasomic Humidifier:

to humidify my bedroom.

It makes a whine that is piercing when the room is very, very quiet, like when you're trying to fall asleep. The whine has number of harmonics, the lower ones simply buzz, which is perfectly fine, like white noise, but the high pitched note is a pure, perfect sine wave - you know, like a synthesizer on default settings.

The lower ones are pretty immune to self-interference, but the purity of the highest note makes it very susceptible to self-interference.

I manage to lie with my right ear in the spot where the frequency is loudest. I assume constructive interference requires either two identical sources or a good reflecting surface somewhere. I don't know why a humidifier would have two sources, and this device does seem to make the make the same noise whether on the floor, on the dresser, or at either end of the room, so I'm not sure it's a reflection. I've movrd it aorund a litte, pointing it this way and that, to see if it alters the location of the nodes, but not much luck so far.

Where do you suppose I'm getting self-interference from?

Anyway, I found a second spot with maximum constructive interference about a foot away from the first (alas, it puts my head almost off the edge of the mattress, so untenable).

I tried to use a(n online) frequency analyzer, but it's too faint to register.

I also tried to match it to a played tone, but its very hard to do. It's somewhere in the ballpark of 2.5-5 kHz.

Two aspects: one scientific, one practical.

Can I deduce its freq from the node distance? Also,can I baffle it somehow?

 

[Baluncore]

Can I deduce its freq from the node distance? Also,can I baffle it somehow?

If we knew the exact model number, we could look up the operator's manual.

The complexity of resonance, in a living space, would make it difficult to identify the path geometry. To identify the nodes, you would need a simpler adjustable organ pipe. Alternatively, you could record the sound with a mobile phone or computer, then take the FFT to see the spectrum. Post the waveform and the spectrum here.

To baffle the sound, you would need to disrupt the resonance of the room. Maybe hang a heavy curtain on a wall, especially near a three corner reflector, such as is made where two walls meet at the ceiling.

You should not be able to hear ultrasonics. Maybe it generates short trains of ultrasonic pulses, and it is that you are hearing. Does it have a switching power supply that is noisy, or which might beat with the ultrasonics?

Maybe, something like the plastic case, is oscillating mechanically. To identify what is resonant, run your hands over the device, twist it, and lift it off the surface A ballast weight of modelling clay might then be used to prevent the oscillation detected.

 

[DaveC426913]

If we knew the exact model number, we could look up the operator's manual.

Homedics UHE-CM65

https://www.manualslib.com/manual/1350286/Homedics-Uhe-Cm65.html?controller=view#manual

The complexity of resonance, in a living space, would make it difficult to identify the path geometry. To identify the nodes, you would need a simpler adjustable organ pipe. Alternatively, you could record the sound with a mobile phone or computer, then take the FFT to see the spectrum. Post the waveform and the spectrum here.

It seems to be too quiet for my mike to pick up. But maybe I will post it and someone can pull the signal out.


Sorry. PF does not accept .wav files or .mp3 files. Oh well.

It's a pity I have to guess what it will accept. I'll try converting it to antoher type, and after I have oen of those Ill see of the uploader recognizes it. There's no other way I know of find out what PF will accept. I'm just rolling the dice here - crapped out twice so far.

Update: Nope. PF does not accept any form of uploaded audio file.

I will post it to my personal webhost and link to it.

To baffle the sound, you would need to disrupt the resonance of the room. Maybe hang a heavy curtain on a wall, especially near a three corner reflector, such as is made where two walls meet at the ceiling.

I would have thought that almost any movement would destroy the consructive interference.

You should not be able to hear ultrasonics.

Def not ultrasonic. I can hear at least into the 13k range. This is less than 4k or thereabouts

Maybe it generates short trains of ultrasonic pulses, and it is that you are hearing. Does it have a switching power supply that is noisy, or which might beat with the ultrasonics?

Buh?

Maybe, something like the plastic case, is oscillating mechanically. To identify what is resonant, run your hands over the device, twist it, and lift it off the surface A ballast weight of modelling clay might then be used to prevent the oscillation detected.

Hard to believe. Again, it's not the buzzing. This is a pure sinusoidal waveform.

(I learned a little about waveforms when I was deeply into customizing music on the SID Editor for my C=64.)

 

[DaveC426913]

Listening to it in bed. It might even be less than 2.5kHz. I think it could be doable on a piano.

When I was trying to get a freq reading with an analyzer, there was a peak at 1012Hz. C5 on a piano is somewhere near 1000Hz.

 

[Baluncore]

Does the fan have a brushless DC motor?
Can you change the fan speed?

 

[DaveC426913]

Does the fan have a brushless DC motor?
Can you change the fan speed?

It has a dual with which i can turn up the output. Not sure if that increases the fan speed.

But no, changing the dial does not make the tone go away.

 

[DaveC426913]

https://www.davesbrain.ca/misc/Record-1.wav

 

[Averagesupernova]

Yeah that would drive me nuts. What are you seeing the peak at 1KHz with? Have you downloaded Audacity and looked at the spectrum of what the mic is picking up?

 

[sophiecentaur]

Does it have a switching power supply that is noisy, or which might beat with the ultrasonics?

Sounds like a possibility. A good old practical problem.

But I do have a problem with the principle. If you evaporate 1 gram of water then you need 2.2J of energy. You can provide this with a heater ('steamer') or moving air over a wick (swamp cooler) which will cool the air down OR this ultrasound method. Is the operating theory that the ultrasonic method needs just the 2.2J , which is enough and that other methods (particularly the boiling kettle) use more?

 

[DaveC426913]

Does the fan have a brushless DC motor?
Can you change the fan speed?

I should say it does not sound like a power supply or a fan.
The lower buzzing might be a fan, but that's not the noise that's intrusive. The one that's intrusive is a pure sinusoidal electronic tone.

This seems to be a perfect match, in both frequency and tone:
https://www.szynalski.com/tone-generator/
Set to 1006Hz


If I play the above in the same room with the humidifier, I can actually get a "beat" above - around 1010Hz - and below - around 1000Hz - which strongly suggests to me I've nailed the freq nicely.

 

[DaveC426913]

So, speculation here: if the freq is ~1000Hz, and I find two nodes ~1 foot apart, is that because the speed of sound is ~1000feet/second?

I think it must be a coincidence. If the sound from the device were 2000Hz, instead of 1000Hz, I would expect the nodes to be six inches apart, not two feet. So it must must be coicidence that I it on 1000Hz = 1ft.
i.e:
Wrong:
500Hz > .5ft nodes
1000Hz > 1ft nodes
2000Hz > 2 ft nodes

Right:
500Hz > 2ft nodes
1000Hz > 1ft nodes
2000Hz > .5 ft nodes

 

[Averagesupernova]

Don't be so certain that the 1000 Hz that you hear is actually there. I've mentioned before on PF that the ear is non-linear and can intermodulate two different frequencies to produce the perceived new frequency. I got a fair amount of pushback about this.
-
I did a quick search and got some interesting hits. This is why I am wondering if you used Audacity or a similar program to see what's actually in the spectrum. Is the tone that you hear actually there or is it a product of intermod.

https://www.google.com/search?q=doe...00LjcuMcgHswKACAA&sclient=mobile-gws-wiz-serp

 

[Baluncore]

Knowing the exact frequency will not resolve the problem.
You must get hands on, and experiment.

Maybe, something like the plastic case, is oscillating mechanically. To identify what is resonant, run your hands over the device, twist it, and lift it off the surface.