Ultrasound, IoT and wireless

[Ilari Sivil]

https://www.wired.com/story/ultrasonic-signals-wild-west-of-wireless-tech/

I've been thinking about ultrasound a lot lately. Thanks to information on the forum, I've found out that some companders on analog wireless devices may choke out the signal you actually want when dealing with ultrasound due to HF boosts in the system. The noise from an induction stove or an ultrasound motion detector are known examples. Apparently ultrasound is becoming more commonly used in IoT-devices for wake-up signals and inter-device communication in household stuff, so I'm quite curious about the future as well as the present state of things.

How many ultrasound related problems actually come up these days? What types of common sources are there, and how could they be dealt with?

[codyman]

Someone encountered a problem with something like this last year.

 

[Constantin]

I don’t if I‘ve had a problem Ultrasound yet, but I did infrasound. This would engage the limiter on my recorder and got cured by by engaging the hpf on the mic. 

Ultrasound could behave similarly 

[Ilari Sivil]

8 hours ago, Constantin said:

I don’t if I‘ve had a problem Ultrasound yet, but I did infrasound. This would engage the limiter on my recorder and got cured by by engaging the hpf on the mic. 

Ultrasound could behave similarly 

I've had infrasound from handling noise compromise the dynamics of a recording. Probably everyone has. Engaging the HPF helps, as you mentioned. I'm not sure if using an LPF to get rid of ultrasound would be a good idea, though, since it might bite into the hearable top end around 16-20kHz if if the level of the unwanted signal is high enough, unless the filter slope is steeper. Doesn't having a steeper filter slope increase the risk of self-oscillation, though? That's my experience with synthesizers at least, and on cheaper recorders I've seen some pretty obvious lines on the spectrogram at the Nyquist-frequency from the steep antialiasing filter. With modern tech, an automatically setting notch could be possible, perhaps? It would probably be a lot simpler to just skip the compander and have digital transmission.

[It's All Waves]

I've experienced something funny which I believe is the result of ultrasound feeding my old G3 transmitters. It will sound like an irregular low frequency modulation of the audio that's coming through the microphone, and the Tx meter will be constantly peaking at almost any reasonable gain setting, but the receiver's looks relatively normal in size, and the space is relatively quiet. The sound is clearly not right. I assume it's an artifact of the dynamics processing involved in analog transmission. I originally thought this was a result of acoustic modulation by subsonic frequencies from newer A/C systems, since the modulation was more on the scale of a beat, than a pitch, but I could not hear it with the naked ear. I finally came up with the right Google search phrase, and found someone experiencing similar problems that stopped after disabling an ultrasonic light switch (auto light switch that uses ultrasound to detect a presence). I see this at almost every brand new elementary/high school in New Orleans.

 

The last time I dealt with this, I had a lectro SRc, SMQV, and LT running at the same time, and only the Sennheiser G3 was affected. Maybe it's related to the G3's failure of the key test? It would seem the G3 transmitters do not like ultrasonic information! 

 

To sum it up, it seems that G3's are susceptible to ultrasonic interference, where Lectro hybrid systems are not. Have yet to try Wisy or Zax in this environment. If you ever have a strange Tx AF meter reading, and "choppy" sound, and are running a G3, it's very likely that you'll need to disable some ultrasonic detectors in your vicinity!

[borjam]

I really doubt that ultrasound can affect the RF performance. Ultrasounds can create nasty nonlinear effects on a microphone however, that's an entirely different matter of course. Several years ago a friend was developing an audio reproduction system based on strong ultrasounds modulated in amplitude. When the ultrasound beam hit something (a wall or maybe your back) the ultrasound signal was demodulated. The effect was amazing, if it hit you in the back you heard someone speaking to you right behind yourself. He told me that some museums were considering similar systems so that you can have, for example, voice explanations for different elements of the exhibit. A room could have several of those audio recordings played at once with little interference between them and you would hear it when moving to a certain spot.

 

If it's a microphone weakness there's no way to correct it after the fact, because those non linear effects that, for example, demodulated the voice recordings, actually produce sound in frequencies in our hearing range.

 

So, back to RF. Might it be related to some microwave based motion detectors? I always thought those were mostly 10 GHz, but a friend called for help once, trying to set up a WiFi network without success and, voila, when I turned on the spectrum analyzer I saw a dirty, continuous signal within the 2.8 GHz band. It rendered most channels unusable. Something similar can happen with some crappy A/V transmission systems which, to add insult to injury, use wasteful analog modulation schemes. 

 

An example attached, a crappy A/V transmitter (note carrier and two side bands) saw on a WiPry spectrum analyzer for iPhone.

 

 

[Constantin]

1 hour ago, borjam said:

I really doubt that ultrasound can affect the RF performance. 

 

I don’t think It’s All Waves was talking about rf performance. I am think he was talking about audio issues. 

And with your theory on how ultrasound affects microphones I think you may be on to something. 

It‘s also possible that the ultrasound is upsetting the compander circuit

 

another possibility is that it affected the pilot tone, which is also an ultrasound tone. It may get interrupted so to speak, and thereby the receiver opens and closes the gate and that results in a choppy sound. 

Just a theory...

[LarryF]

1 hour ago, borjam said:

I really doubt that ultrasound can affect the RF performance. Ultrasounds can create nasty nonlinear effects on a microphone however, that's an entirely different matter of course. [snip]

Ultrasound (ultrasonics) is well known to create overload in the audio compander section of the audio processing system of some wireless. Particularly those that have a great deal of pre-emphasis wrapped around multiple sections of the audio circuits. This well known (but occasionally surprising) artifact of some popular wireless can be prevented by not using pre-emphasis (or only a moderate amount) in the audio and by sharp roll off of frequencies above 20 kHz.

Best Regards,

Larry Fisher

 

[borjam]

Affecting some pre-emphasis circuitry in the audio domain would be possible of course. Sorry, when I read about it affected "wireless" I was confused, I didn't consider the analog noise reduction systems. I stand corrected

[It's All Waves]

On 6/28/2018 at 1:33 AM, borjam said:

Affecting some pre-emphasis circuitry in the audio domain would be possible of course. Sorry, when I read about it affected "wireless" I was confused, I didn't consider the analog noise reduction systems. I stand corrected

 

borjam, thanks for your reply and story about that awesome ultrasound tech! I feel that sound is so underrated in its usefulness, compared to EM, so I love hearing about awesome sound-powered tools!

 

I wasn't totally clear when explaining my situation. When I mentioned the problem with the dynamics processing, I was talking about audio-only processing, but that which seems to be necessary to prepare for a healthy analog transmission, i.e. companding, HF rolloffs, etc... Aaaand that's about as deep as my knowledge goes with RF transmission. 😁

 

Actually, the first time I experienced the G3 problem, I thought it was the mic, and called Countryman, sent them a recording, and a spectrograph, and they never responded, ha! The last time I experienced this, I had Cos-11's on all my Tx's, so that rules out the microphone, as far as I can tell. 

 

Regarding the point about microwaves, I didn't realize motion detectors use microwaves! Aye! But that is very good to know for many reasons, ha! I do think it was Tx-related, though, which in my limited understanding, would suggest that it was something coming through the microphone. The bit about seeing the Tx AF meter go crazy, but the Rx not, also tells me it was at the Tx. But I've rather recently become aware of the fact that Tx's can be hindered by other transmissions. So I don't really know!

 

On 6/27/2018 at 4:59 PM, Constantin said:

 

I don’t think It’s All Waves was talking about rf performance. I am think he was talking about audio issues. 

And with your theory on how ultrasound affects microphones I think you may be on to something. 

It‘s also possible that the ultrasound is upsetting the compander circuit

 

another possibility is that it affected the pilot tone, which is also an ultrasound tone. It may get interrupted so to speak, and thereby the receiver opens and closes the gate and that results in a choppy sound. 

Just a theory...

 

Thanks for clarifying for me! I hadn't considered the pilot tone bit–that's interesting... 

 

On 6/27/2018 at 5:06 PM, LarryF said:

Ultrasound (ultrasonics) is well known to create overload in the audio compander section of the audio processing system of some wireless. Particularly those that have a great deal of pre-emphasis wrapped around multiple sections of the audio circuits. This well known (but occasionally surprising) artifact of some popular wireless can be prevented by not using pre-emphasis (or only a moderate amount) in the audio and by sharp roll off of frequencies above 20 kHz.

Best Regards,

Larry Fisher

 

 

Larry, I feel like you nailed it! Mostly because of your background.😉 I imagine this explains the "key test" failure in the G3's as well. Have also had problems with super sibilant talent. Have had to set the gain lower than I'd like, to avoid it.

 

Do you think the bit about interference with the pilot tone could have potentially played a roll? I don't recall the audio straight dropping to silence, as I would imagine a loss of pilot tone would incur. It reminded me of modulating something on a synth with a random step generator.

 

 

[borjam]

16 minutes ago, It's All Waves said:

I wasn't totally clear when explaining my situation. When I mentioned the problem with the dynamics processing, I was talking about audio-only processing, but that which seems to be necessary to prepare for a healthy analog transmission, i.e. companding, HF rolloffs, etc... Aaaand that's about as deep as my knowledge goes with RF transmission. 😁

 

I see  Note that microphonics exist! Especially old valve equipment could transmit voice if you shouted at it. I have an old Leader grid dip meter here that doubles as an AM transmitter if you shout loud enough! But with modern SMD electronics that's almost impossible, don't worry!

 

16 minutes ago, It's All Waves said:

Regarding the point about microwaves, I didn't realize motion detectors use microwaves! Aye! But that is very good to know for many reasons, ha! I do think it was Tx-related, though, which in my limited understanding, would suggest that it was something coming through the microphone. The bit about seeing the Tx AF meter go crazy, but the Rx not, also tells me it was at the Tx. But I've rather recently become aware of the fact that Tx's can be hindered by other transmissions. So I don't really know!

 

Some of them use microwaves although I doubt they can affect your equipment unless you are in the same frequency. So, careful with equipment on the WiFI/ISM 2.8 GHz band. Although I believe that 2.8 GHz detectors aren't common nowadays for obvious reason. At home I have a 10 GHz detector.