Raspberry Shake - Washing machine spectrogram collection

[Mauri Niininen]

Creating this topic to collect washing machine signatures. 

Here is spectrogram of  Bosch Axxis+ on cotton cycle.  

[John Beale]

Here's a LG WM2301HW on normal/cotton cycle, it runs too long to fit in one spectrogram. Notice in the second plot around  23:41 UTC between 18 and 25 Hz, I see a faint wiggly line plus what looks an awful lot like a folded-over (aliased, or inverted) copy of that same signal. Maybe there isn't much low-pass anti-aliasing filtering done before the ADC?

[d3noob]

Fisher and Paykel model GW712.

Rinse and then Spin cycle

[stickhog]

Here's a Kenmore Model 417.47102700, rinse then final spin cycle. Notice the additional line at around 22Hz, and how it momentarily increases as the final spin speed increases at 20:40, but then suddenly drops off and settles in just below 20Hz when the spin is stabilized at 15Hz. 

[John Beale]

Any signal ramp that goes up to 25 Hz and makes a sudden turn down at the same angle it went up, is almost certainly a frequency alias or "foldover" effect due to the Shake's 50 Hz sampling rate. Any input frequency above 25 Hz will appear in the output as f(out) = 25 - f(in). See also: "Nyquist sampling rate"

[Skip]

That line around 22 Hz looks like the third harmonic of the strong signal between 5 and 10 Hz. Zoom in on the segment between 20:30 and 20:32, display it in "Spectra View", and see if you can identify both peaks. I bet they're 1:3 in frequency. I wonder if this is being generated by the washing machine, or if it's from distortion in the analog amplifier.

From another conversation here, I suspect the peak just below 20 Hz is an artifact. It seems pretty constant through the whole spectrogram.

I agree with John about the alias. 25 Hz (half the 50 Hz sampling frequency), called the "Nyquist frequency", is also called the "folding frequency"; these displays make it obvious why. As a side note, it will "fold" again when it gets to 50 Hz (signals at 49 Hz and 51 Hz will both appear as 1 Hz - an "alias", 48 and 52 as 2, etc.), and again for every multiple of the Nyquist frequency. See https://en.wikipedia.org/wiki/Nyquist_frequency

Skip

[stickhog]

Yup, it's a 3:1. 7.645Hz / 23.025Hz. Good call.

I've heard about Nyquist frequencies briefly once before, in the context of sampling rates of wheel speed transducers for an antiskid brake system on a large airplane. Neat to see its applicability in another domain. Gotta go read up on that again to remind myself of the details.

Thanks for the comments. This is a really interesting topic. Appreciate everyone sharing their expertise.

Howard

[Skip]

This isn't a washing machine, but I'll be darned if I know what it is. From this morning at about 6:45 AM Local. We got up a little after 7. I haven't noticed anything like it in the weeks before, or since.

Skip - R3AA7

[John Stuart]

Skip, I have had similar ramp-up, then ramp-down vibrations here too, but much longer in duration, starting about 4:36PM local (see attached).

I'm pretty sure its from one of my neighbors, possibly a variable speed heat pump, or spin-dry cycle in a clothes washer.

The hunt /game is afoot !

John Stuart

[John Beale]

I have seen similar ramp up, then steady-run traces like that, in early morning when I would not expect neighbors to be doing laundry, and nobody close by has a pool. Not sure if it's related, but there is some kind of small motor in my wife's Toyota Matrix that I have heard run long after the car has been turned off. The sound comes from the rear of the car, maybe related to the fuel system; probably like this: http://www.subaruoutback.org/forums/104-gen-4-2010-2014/52843-fuel-pump-running-car-off.html

[Skip]

Well, I do have a 2015 Subaru Outback. Coincidentally, I'm also an active member of that forum, too!

It's a pretty strong signal on the helicorder, though, and the car's in a detached garage maybe 100' from the Shake in the basement. Plus, that's an awful slow run up and "spin down" for something like a fuel pump, but it's still something to consider.

Thanks for the tip!

Skip - R3AA7

[John Stuart]

Last night Google led me to "Variable Frequency Drive (VFD) Pumps" that are used on solar water heater panels to prevent water hammer.  Slow ramp-up to fill panels, faster ramp-down to drain panels(?).

My neighbor has a system on a hillside about 200 ft below my RShake's buried sensor. His solar panels are at ground level, mounted on concrete piers, so pumping vibrations could easily couple into ground and come further up the hill to my RShake.

I'll let you know if I can confirm, , ,

John Stuart

Lafayette CA

[chris...@aol.com]

Hi Skip, 

    The signal takes ~30 seconds to run up to ~10 Hz and it is only on for ~1 m 30 sec with a rapid 

fall off at ~6.45 am local. One thing that I thought of is a garage door motor, or similar. Is this possible ? What sort of heating system do you have ? What time does the sun rise with you now ? I was wondering about solar heating systems - but I would expect them to be on most of the day time - or at least intermittently. The 10 Hz bothers me - 4 pole motors are common, but six pole ? Swimming pool solar heaters ? It would need to be quite a powerful local motor for you to detect it. 

    Regards, 

    Chris Chapman

From: Skip <sandfwh...@gmail.com>
To: RaspberryShake <raspber...@googlegroups.com>
Sent: Mon, Jan 2, 2017 10:14 pm
Subject: [Raspberry Shake Community Forum] Re: Raspberry Shake - Washing machine spectrogram collection

This isn't a washing machine, but I'll be darned if I know what it is. From this morning at about 6:45 AM Local. We got up a little after 7. I haven't noticed anything like it in the weeks before, or since.

Skip - R3AA7

On Sunday, December 11, 2016 at 8:19:01 PM UTC-6, Mauri Niininen wrote:

Creating this topic to collect washing machine signatures. 

Here is spectrogram of  Bosch Axxis+ on cotton cycle.  

--
Some useful links:
 
Manual: http://manual.raspberryshake.org/
Shop: https://shop.raspberryshake.org/
Website: http://raspberryshake.org/
Do It YourSelf Page: http://raspberryshake.org/do-it-yourself

[Skip]

Thanks for the comments, Chris. You noticed the same characteristics as I did. The peak freq. is near 10.138 Hz, according to the spectrum display, which corresponds to just over 608 RPM if it's rotating machinery. Garage door motors don't run that long (maybe, what, 10 seconds or so at a time) and don't take so long to start up or stop. There are no significant solar installations within a block of me that I know of. Our neighbor does have a very large whole-house emergency generator (like 80 KW or so(!) - it's huge) that starts up every Monday at noon and runs for 15 minutes or so, but those are 60 Hz and, again, start up and stop fairly quickly. I haven't noticed it on the shake, either. On the other hand, 60 Hz would get aliased to 10 Hz on these systems, so maybe that's something to consider. I'll need to find the frequency-response chart they published and see if there's still a lot of sensitivity at 60 Hz... I recall that there was, and was surprised there was no anti-alias filter in the system whatsoever.

We have two forced-air gas furnaces, but those are constantly cycling on and off and (surprisingly, since they are maybe 15 feet from the shake) don't make a noticeable signature on the helicorder, unlike this thing. We also have a heat pump for the finished part of the attic, but it hasn't been on for a month or more. It certainly doesn't take tens of seconds to start up, either. Maybe there was a service truck of some kind out on the street, but if so, it wasn't there for long.

Still stumped!

Skip - AM.R3AA7