Data rate of ultrasound in air

Discussion in 'General Electronics Chat' started by Renegade243, Mar 16, 2014.

  1. Renegade243

    Thread Starter New Member

    Feb 5, 2013
    24
    0
    Hi everyone,

    I am doing a project at university which requires me to transmit a voice signal using ultrasound. I was thinking of using an ADC to convert the voice signal and then use on off keying on the ultrasonic transducer (piezoelectric) to transmit the binary data and then convert it back on the other end. I have some uncertainties regarding the maximum data rates which can be achieved using on off keying. Maximum data rate meaning the point at which the receiver can just make out the binary levels, that will not result in binary errors.

    I cannot simply test this as this comes out of our budget and if it doesnt work it would be a waste of money. I have had a look around on the internet but I havnt found evidence that someone has actually done this for a project. I have found papers on the projects, but they do not show consistent results in data rates achieved. I have created some graphs and flow diagrams which may help you understand what I am trying to achieve, feel free to contact me if you would like to see them.

    Also this form of communication is our only option and we cannot use anything else. I dont know a great deal about communication circuits and data rates and have no experience in wireless communications so far.

    Thank you
     
  2. k7elp60

    Senior Member

    Nov 4, 2008
    478
    69
    I have actually done this but a different way. I used the radio superhetrodyne principal. That is I mixed the voice with an ultrasonic frequency fed it to a transducer, and then at the receiving end mixed again and selected the difference frequency. The range is limited by the transducers and the the emitted power.
     
  3. ErnieM

    AAC Fanatic!

    Apr 24, 2011
    7,386
    1,605
    Ultrasound is just sound, albeit at a higher then you can hear frequency. Whatever transducers you get will determine your "carrier" frequency.

    k7 describes one way of doing an AM modulation/demodulation. AM is perhaps the simplest way of sending the signal, it's good old fashioned basic analog stuff.

    Going digital may be harder then using what has worked for over 100 years.
     
  4. Renegade243

    Thread Starter New Member

    Feb 5, 2013
    24
    0
    ErnieM,

    Thank you for your reply, with a piezoelectric ultrasound transducer, do you apply a DC voltage and then get a fixed frequency output similar to buzzer? Using piezoelectric I dont think AM modulation would be possible, since it would be difficult for the receiver to distinguish the slight changes in amplitude.

    Are you thinking of potentially a different device for emitting ultrasound?
    Also we are limited from websites we can buy from Farnell, Rapid, RS etc (only big names), so far on these websites ive only been able to find piezoelectric ultrasonic emitters..
     
  5. studiot

    AAC Fanatic!

    Nov 9, 2007
    5,005
    513
    Ultrasonic controllers were used before infra red and now RF.

    The BBC also had a project whereby they modulated binary onto audio cassettes to distribute software (I will try to remember the name, perhaps someone else can help here).

    There were many amateur projects doing this in the magazines of the time.
    Since this is a college project, get you colllege library on the job of searching out some references. Again I will try also to find some for you.

    edit a quick google into the past

    http://en.wikipedia.org/wiki/Kansas_City_standard

    http://www.instructables.com/id/Storing-files-on-an-audio-cassette/

    both these use techniques useful to you.
     
  6. sirch2

    Well-Known Member

    Jan 21, 2013
    1,008
    351
    You need to drive the piezo trasnducer at, or close to, it's design freqency so you need to apply AC at the design frequency. Most MCUs can easily generate ultrasonic frequencies and you can just use the square wave output to drive the transducer. But to get more power a tuned circuit approach is better.

    The size of the changes in amplitude are up to you. Detecting the AM on top of the received signal is a case of subtracting the carrier and amplifying the result. AM radios seem to be able to do this...
     
  7. Renegade243

    Thread Starter New Member

    Feb 5, 2013
    24
    0
    "Sirch2"

    Thank you for your reply, ah this changes a lot of things.. so the piezo transducer cannot be operated from a DC voltage? You have to generate the AC voltage yourself and then apply it? If this is the case what modulation technique would you recommend?

    My concern regarding amplitude modulation is; could the receiver be sensitive enough to pick up the slight changes in amplitude, I mean i want to achieve decent voice quality. since this is going to be transmitted through the air (at least a few metres) surely there must be some factors which will limit the quality that can be transmitted. Frequency modulation could be a possibility. I think I will need to consider the most efficient modulation technique that will take full advantage of the behaviour of ultrasound.

    This is the ultrasound transducer I was looking at:
    http://uk.farnell.com/multicomp/mcust10p40b07ro/transmitter-40khz-10mm-plastic/dp/2362667

    I can get any other frequency of ultrasonic transducer if need be..

    Also we have an extremely stingy £20 pound budget.. and if I want to carry out testing the components required comes out of this budget... also we cant use any form of microcontroller or IC that will carry out the entire operation of the system.
     
    Last edited: Mar 16, 2014
  8. ian field

    Distinguished Member

    Oct 27, 2012
    4,413
    782
    Before IR handsets turned up, most TV makers used FM on their US remotes - usually as simple as a resistor chain tone generator to modulate the 40kHz and a FM discriminator or ratio detector in the TV front panel to decode the tones into voltage levels.
     
  9. studiot

    AAC Fanatic!

    Nov 9, 2007
    5,005
    513
    Since my last post was so well received I will have another go.

    Have you looked at the specs on the pdf?

    What do you think about the units suitability for FM modulation of high quality speech.

    That is what characteristics would you look for?
     
  10. Renegade243

    Thread Starter New Member

    Feb 5, 2013
    24
    0
    "StudioT"

    Many apologies, I did have a look, lets just say I am a bit confused at this point. I cant really make much sense of the pdf, the information I need is not there, such as how to operate the transducer.

    I am not quite sure what you mean by:
    "What do you think about the units suitability for FM modulation of high quality speech"
     
  11. studiot

    AAC Fanatic!

    Nov 9, 2007
    5,005
    513
    The ultrasonic transducer is nominally 40khz.

    But the pdf includes a response curve.

    If you are going to modulate the transducer what bandwidth of audio are you thinking of for 'good quality' and will this unit achieve that?

    The pdf has a second curve that shows a highly directional output.
    Is this what you require.

    Considerations like these are part of doing a college project, and a good writeup of them will get you good or extra marks.

    But they should be your consideration - although you should feel free to run them past us here as you are doing.

    By the way why was this question not in the project or homework section?
     
  12. sirch2

    Well-Known Member

    Jan 21, 2013
    1,008
    351
    Google will get you plenty of AM circuits. I have measured 20 to 40V across one of these transducers with a well tuned circuit on the transmit side. Modulate this by up to 50% and you will be getting mV changes on the receive end, unamplified.

    The thing about using FM is that the voice frequency is a significant proportion of the carrier and so may affect the transmitted power.
     
  13. Renegade243

    Thread Starter New Member

    Feb 5, 2013
    24
    0
    "StudioT"

    I am looking to use a frequency range 200hz - 3.2khz which is the bandwidth for a telephone conversation. Thank you very much for your input, I have looked around more and consulted my tutor and AM is a very strong candidate. I am in the process of designing a block diagram of the circuit and will post it in here when it is completed.
    This question was not in the homework section because I thought I was just gonna ask a single question and get an answer instead I found that I was mistaken which caused a redrafting of all my ideas.
     
  14. Renegade243

    Thread Starter New Member

    Feb 5, 2013
    24
    0
    "Sirch2"

    Thank you, yes I am looking up AM circuits and techniques, I dont think FM will be a great idea actually because the relationship between sound pressure and frequency is not linear.
     
  15. studiot

    AAC Fanatic!

    Nov 9, 2007
    5,005
    513
    You will need your thinking cap for AM with that transducer.

    You need a bandwidth of at least twice the modulating frequency (ie 6.4khz).

    The response of the transducer is down more than 20dB at these points.

    Long Wave AM radio has a few stations down below 100khz that function well enough, but their response curves are sensibly flat over the required bandwidth.

    I suggest you look into AGC circuits as well.
     
  16. Renegade243

    Thread Starter New Member

    Feb 5, 2013
    24
    0
    "StudioT"

    I think a Wien bridge oscillator could be used in accordance with an AGC circuit, where the AGC directly controls the gain of the wien bridge oscillator. & the AGC is dependant on the amplitude of the modulating signal. Would this not be a form of amplitude modulation?

    & Then i could just have an amplifier to ready the signal so that it can be applied to the ultrasonic emitter.

    Thanks
     
  17. studiot

    AAC Fanatic!

    Nov 9, 2007
    5,005
    513
    I don't know where you are with your studies but you haven't appreciated my point. Perhaps I didn't make myself clear.

    A wein bridge will oscillate correctly when the gain is exactly 3. The purpose of the AGC in wein bridge circuits is to stabilise the gain to this value.

    You can easily form suitable a cowan modulator or ring modulator at these frequencies.

    Let as take a simple example of modulating a single pure 1khz tone onto your 40 khz carrier.

    This will generate a complex signal comprising three frequencies, 39khz, 40 khz and 41 khz.

    If you do not know this about AM signals you should study it up.

    Particularly as the carrier at 40khz contains no information whatsoever and is often suppressed since it consumes at least 2/3 of the power.

    Either (or both) sideband contains the full modulating signal information.

    Now look again at the pdf.

    If the output of your transducer is 1000v units at 40khz, what is its output at 39 or 41?

    This will be the size of your wanted signal.

    When used as a receiver this tranducer looks even worse.

    Your block diagram should have two blocks feeding into the modulator, one for the carrier oscillator (which could be a wein bridge) and one for the microphone and its amplifier.

    The output from the modulator is then fed in to a (automatic) gain controlled amplifier. The purpose of the gain control is to compensate for the loss of output of the transducer as the frequency of the modulating signal increases.

    If you do not do this the output at 3khz audio will be unacceptably less than at 500hz audio.

    Edit

    I also suggest you read up on 'hf bias in magnetic tape recording', google is your friend,
    For various reasons the audio signal is modulated onto a carrier in the frequency range 25khz - 100khz .


    Incidentally why have you rejected digital?
    If you finally go analog then your reasons will add extra marks to your writeup, but think it over first.

    Ian field has (and I have) already indicated about old fashioned ultrasonic controllers using tones.
    You do not switch them on and off, you modulate tones onto the carrier. Two tones say 100hz and 300hz will be enough to give you binary. This would keep your ultrasonic transmission system within its peak sensitivity range since you would only be transmitting 40khz ± 300hz.
    If you had say 5 cycles of each per data pulse and counted cycles then all your system would have to do would be to distinguish between 100 and 300 hz.
    You should be able to fit digitised 3khz comfortably into this.
     
    Last edited: Mar 17, 2014
  18. wayneh

    Expert

    Sep 9, 2010
    12,093
    3,033
    Don't know about the BBC, but way back when, the Apple II computer could output Basic programs as audio, so you could store programs you had written onto cassettes (or anything that could record audio) and load them back into the computer by playing the tape. This could all take place with no disk drive attached. Disk drives and operating systems weren't much to speak of at that time, so I suppose that's why they did it.

    BTW, I've got a couple of those old Apple IIe I'd be glad to sell. ;)
     
  19. bertus

    Administrator

    Apr 5, 2008
    15,645
    2,344
  20. studiot

    AAC Fanatic!

    Nov 9, 2007
    5,005
    513
    Yessss

    Thanks a bunch Bertus

    :D
     
Loading...