a low cost hearing aid

Discussion in 'The Projects Forum' started by luc, Mar 5, 2011.

  1. luc

    Thread Starter Member

    Mar 4, 2011
    153
    1
    hey guys, i do search for old post and found same topic, but my design is quite difference with it, so i decide to open new post .

    first, i plan to design a hearing aid which contain a low pass filter(butterworth 2nd order) to filter off the frequency that greather than 20khz; a pre-amp to amplify the signal that received from microphone. attachement below are my circuit design. I am using tlc2274 as my op-amp for pre-amp and also filter design.

    For first picture,i search from net about the design of pre-amp and i was apply it as my pre-amp . the value for those component are listed as below:
    * R1 - 1 kOhm resister
    * R2 - 3.3 kOhm resister
    * R3 - 1 kOhm resister
    * R4 - 540 kOhm resister
    * C1 - 0.0054 uF capacitor
    * MIC1 - an electret microphone

    *actually i would like to learn about how to design a pre-amp and calculate the R and C value, is that anyone can teach me? thank!

    For second picture, that is my lowpass filter that use to filter the frequency that above 20khz. I do test this circuit with function generator and oscilloscope, i observed the result shown that when my function generator generated a signal that greather than 8k and above toward 20khz, my output signal's amplitude will be keep on reduced. I assume that the filter is work.

    Now, the question is, after i combine both pre-amp and lowpass filter circuit, when i try to use my handphone to make a call, i found out that my phone wil interference my circuit, which mean that it will have some "pu pu" sound heard from earphone. May i know a handphone will generate how much frequency to interface with electronic device? Or it show that my filter does not work pefectly which mean that it cant filter off the noise that cause by interference of handphone? I do search from net about frequency of phone but i cant find it..

    Lastly , is that i need to improve my filter order to reduce noise? May i know where can i get the infomation for how to design a 3rd order butterworth low pass filter? (cause i only can found the filter design that up to 2nd order)

    Thanks !!
     
  2. wayneh

    Expert

    Sep 9, 2010
    12,128
    3,049
    Welcome to the AAC forum.

    This may have to do with shielding than with your circuit design. Phones emit a lot of electromagnetic noise - if it's near my computer's speakers, I can tell when my Blackberry is receiving e-mail by the sounds I hear in the speakers.

    It may not be practical, but for testing you could try putting your circuit inside an Altoids tin, to see if it stays quiet when it is shielded. If the phone is affecting the mic itself, you might be forced to use a different mic technology or a different phone.
     
  3. luc

    Thread Starter Member

    Mar 4, 2011
    153
    1
    Thanks for answering my question, i will try to shield it later and test it again :). May i know is that the frequency of EM noise that emited by phone is greather than 20kHz? Because i would like to know whether that is possible to filter it or not. Thanks!
     
  4. luc

    Thread Starter Member

    Mar 4, 2011
    153
    1
    i test already,my earphone still interfere by my handphone :-(
    when i try with difference type of phone, i found out that certain phone will interfere the mic when call in or call out, but certain phone would not.
    May i know is that anyway to filter off the EM noise that generate by every type of phone? Because my main point is to filter off the EM noise that generate by every type of phone..

    The attachement below is the pic that i capture from oscilloscope show that how my filter working. 1st pic show that when my input frequency is 20khz, the output amplitude will reduce.. so on for 2nd and 3rd.

    Anyone can help?
    Thank!
     
    Last edited: Mar 6, 2011
  5. DickCappels

    Moderator

    Aug 21, 2008
    2,656
    632
    That noise from the cell phone will be difficult to get rid of; most likely it is the pulsed RF being rectified, or at least handled nonlinearly inside one or both of the opamps. RF filtering on the input and especially the output and battery connection should help some. Ferrite beads on the input and output connections, then good quality RF capacitors to ground, several hundred pf should be plenty, and the capacitors should be on the circuit side of the ferrite beads. The Altiods tin probably didn't help because the RF was getting in through your earphone leads and the battery leads (if the battery was not inside the box).

    Just wondering: What is the purpose of the 20 kHz low pass filter?
     
  6. luc

    Thread Starter Member

    Mar 4, 2011
    153
    1
    Thanks for answering my question :). the purpose of the 20kHz low pass filter is to filter off the frequency that above 20kHz which is above a normal human hearing range ( from 20Hz-20kHz). Actually before that, i plan to design a lowpass and high pass filter with the range 20Hz-20kHz, but after that i think that normally people may not hear the sound with frequency around 20Hz, so a lowpass filter may more than enough. As i search from net, human voice range is around 500Hz-2kHz and the peak hearing range is about 2-4kHz. I have to make sure that my circuit can function well between 200Hz to 5kHz and can filter off the noise that above 20kHz.
    Due to my title is low cost hearing aid, so i plan to use low cost component to build a hearing aid which can have better function.I also have to think how to minimize the size of my circuit.

    I plan to filter off the noise that cause by cell phone, it is because, when i using a cell phone to make a call or pick up a call, it will have noise that can hear through earphone, even after i pick up, during chating by handphone, the noise still there (Before that i thought noise will be there during call and receive a call only, but after i try to answer the phone, the noise still there:() I try few type of cell phone and i found out that, certain phone wont have that problem. Is that have certain type or phone wont generate EM noise? or is that the EM noise that generated by each friend have difference frequency ?
     
  7. DickCappels

    Moderator

    Aug 21, 2008
    2,656
    632
    Not sure why some phones are worse than others. It might be that the phones you tried are on different bands, and that could affect how well (or badly) the circuit picked up the RF.

    You already have AC coupling between the microphone and the amplifier, that is probably enough of a high pass filter. I would bet that the low pass filter has no effect on the use of the circuit. As you pointed out, people can't hear past 20 kHz anyway. I can only imagine it helping in the case of very loud rodent or bat signals, or maybe a squeaky hinge, well above 20 kHz that could case some distortion to sounds at lower frequencies. In other words, you might consider simplifying your circuit and lowering your battery drain by leaving the filter out.
     
  8. luc

    Thread Starter Member

    Mar 4, 2011
    153
    1
    Actually i think that cell phone will generate a EM noise which have higher frequency (more than 20kHz), so the filter might filter off that noise. That also another main purpose that i make a filter there. I wondering is there really no way to filter off the noise? example laptop speaker wont affect by EM noise when i try it on my laptop, but external speaker that connect from laptop will have this kind of problem.
     
  9. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    The value of your coupling capacitor C1 (0.0054uF) is so small that it reduces frequencies below 7400Hz. But some deaf people need the opposite equalization.
    The very low value of R2 helps C1 cut most low audio frequencies and its value is so low that it reduces the signal from the mic.

    Your output amplifier is completely wrong and needs to be fixed as per my sketch.

    The output opamp has gain of 54dB which is much too high. The TLC2274 opamp cuts frequencies above about 2khz when its gain is so high.

    The voltage gain of the Cmos opamp drops when its load resistance is less than a few thousand ohms. Its output resistance is 150 ohms so you must find a earphone with a very high impedance, or use a power amplifier.
     
  10. Adjuster

    Well-Known Member

    Dec 26, 2010
    2,147
    300
    It is quite difficult to make a sensitive amplifier which will be immune to the effects of a mobile phone at close range. The problem is not just limited to getting the schematic right: physical layout is also important, because of the intense high frequency radiation that causes the interference.

    Any filtering components need to be as close as possible to the device they protect, and serious effort is needed to reduce the area enclosed by any loops of conductor within the filtering - including the DC power lines.

    An overall metal screen joined to the amplifier common may be helpful, preferably containing the microphone, battery and amplifier. The earphone output could come out via small "feed-through" RFI filters grounded to the case. It may be sufficient to make the microphone aperture quite small. If it is big enough to let some interference in, it could be covered by fine wire mesh thoroughly connected to the case (all around, connection at one point may not work). If the microphone must be external, the job becomes more difficult - but a good quality screened lead and RFI filters at the case ought to help.

    If all this sounds rather involved, bear in mind that the advent of mobile phones and other portable wireless devices has caused a serious problem for professional hearing aid manufacturers. It is clearly not an impossible task though: after all, the microphone circuits in mobile phones themselves work with minimal interference, despite their location.
     
  11. luc

    Thread Starter Member

    Mar 4, 2011
    153
    1
    Thanks for reply. I think my sequence of attachment was wrong, the 1st pic( left hand side) should be the pre-amp, and the 2nd pic ( right hand side) should be my lowpass filter.

    Actually my design will be like this : mic with pre-amp>> lowpass filter>> another amplifier to amplify my signal>> earphone.
    But now i just want to test whether my filter work or not, so the 2nd diagram haven't complete yet with the amplifier.The output of filter should be connect with an amplifier then only connect with an earphone.

    About the pre-amp capacitor C1 and R2,do u mean that i need to change capacitor and resistor with larger value? may i know how much value is suitable to use for it cause if it will reduces frequencies below 7400Hz, then there is a problem for me since i would like to maintain the frequencies below 20kHz. Can you teach me how to calculate it?

    I do test my circuit, the oscilloscope show that when my input is around 6kHz, my output amplitude starting to reduced as show in attachments, is that a normal case for butterworth 2nd order lowpass filter?

    Sorry for my poor understanding and lack of knowledge,may i know the attachment that sketch by you is refer to pre-amp or output amplifier? And how to get to know about the output op-amp have a gain of 54dB?

    Thanks!!
     
  12. luc

    Thread Starter Member

    Mar 4, 2011
    153
    1
    Thanks for reply. Now i haven't solder all my component in pcb yet, i sitll using breadboard to test the circuit. I heard from some senior said that aluminium foil can be a shielding to my circuit, so i put my circuit inside a small box, then wrap it with aluminium foil which connected to ground. i just leave my mic and earphone outside the box and then trying to test my circuit with mobile phone, but i found out that it is still intefere by mobile phone. I think this should be a problem when bearer using mobile phone, because that sound is quite annoying. Is that mean that no any way to solve it unless the mobile phone manufacturer improving their phone to generate lesser distortion?
     
  13. Adjuster

    Well-Known Member

    Dec 26, 2010
    2,147
    300
    The mobile phone manufacturer cannot help you much, unless the powers and frequencies used by the whole mobile phone system are radically changed. That is unlikely any time in the near future.

    The mobile phone has to work as a radio transmitter in order to work at all. What is giving you interference is the transmitter output which is necessary to carry information away from the phone.

    If you had understood what I had said earlier, you would see that it is possible for a microphone amplifier to work despite being near a phone. The phone itself has a microphone, but does not suffer from much interference.

    Screening against these frequencies is simply rather difficult. You might try getting advice from someone who has experience in UHF / low microwave design - the problems are as much in the area of physical layout as in the schematic circuit.
     
  14. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    It is obvious which is the lowpass filter and which is the preamp.
    You don't need a lowpass filter and the gain on your preamp is too high at (540k/1k) +1= 541. A gain of 1000 is 60dB and half (a gain of 500) is 6dB less so your preamp has a gain of about 54dB. You might need a gain of only 100 (40dB).

    The formula for a coupling capacitor is "1 over 2 pi RC". See my attachement. It makes a highpass filter. For speech, a male voice produces sounds down to about 100Hz. The resistor value should be 10 times or more the value of the source resistance. So if you use a 33k resistor to ground at the opamp then the capacitor value should be 0.048uf. A 47nF film capacitor should be used.

    You don't have a Butterworth filter because your resistor values are not the same. I did not calculate the cutoff frequency of your odd filter.
    You do not need a 20kHz lowpass filter because your preamp cannot go higher when it has high gain. The microphone also probably cannot go higher.
     
  15. luc

    Thread Starter Member

    Mar 4, 2011
    153
    1
    The lowpass filter that i designed was a butterworth filter and my resistor values are also same.I using the formula for butterworth 2nd order to calculate it's value from a cutoff friendquency.It should be correct if i follow the format?

    Is that anyway to design my pre-amp up to 20kHz?

    Thanks!
     
  16. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    I see the same resistor values. If C1 is 2nf and not 2.2nF then it will be a good Second-order Sallen and Key Butterworth 20kHz lowpass filter. Use two 1nf capacitors in parallel to make 2nF. This filter will reduce 8kHz a little because 20khz will have an amplitude of 0.707 times. 40kHz will be 0.25 times.

    The datasheet for every opamp has a graph that shows the high frequency cutoff at various amounts of gain. The higher is the gain then the lower is the cutoff frequency.
    If you need very high gain then simply share the gain between two opamps for a much higher cutoff frequency.

    How will you measure people's hearing to equalize response for the hearing aid?
    Where is the adjustable equalizer circuit that all hearing aids have?
     
  17. luc

    Thread Starter Member

    Mar 4, 2011
    153
    1
    Thanks for reply. I will try it:).

    I plan to connect the output amplifier with a potentiometer to adjust the volume, so i think that bearer can adjust the sound according to their comfortable range.

    I have some question regrading to the pre-amp that you sketch,May i know the bias voltage is similar to the attachment diagram? Is that i need to add a series of resistor as show in the diagram for bias voltage? May i know what is the function for bias voltage?

    Previously i have design a high pass filter, but i found out that,when i setting my input with an offset of 2V, but my output does not have offset and maintain the signal in the reference ground.I think that is because of the coupling of capacitor in highpass filter. Is that something relate to bias voltage?

    Beside that, the calculation of capacitor with R3 is that using the same formula as 1/2piRC? May i know what is the missing filter that connected with R1? Is that mean the input of signal should pass through a filter before entering an op-amp?

    Sorry for asking so many question, thanks!!
     
  18. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    A volume control is necessary.

    The transistor circuit you show has its bias voltage made with two resistors forming a voltage divider. I did the same thing with the opamp. The bias voltage causes the output of the trasnsistor and opamp to be at about half the supply voltage so it can swing equally up and down the make the AC signal.

    I don't know what you did.

    Each coupling capacitor attenuates low frequencies so usually the capacitor in series with R3 has a higher value than the calculation. I added the capacitor to my corrected circuit.

    I added one smoothing filter.

    The smoothing filter smooths the DC voltage to the microphone and to the bias resistors. The supply current for the opamp fluctuates with the signal and causes the supply voltage to jump up and down a little. the fluctuations will be amplified and will cause problems if there is no smoothing filter.
    The smoothing filter does not filter the audio.
     
  19. luc

    Thread Starter Member

    Mar 4, 2011
    153
    1
    Really sorry for my lack of knowledge,for this part "The bias voltage causes the output of the trasnsistor and opamp to be at about half the supply voltage so it can swing equally up and down the make the AC signal" , do you mean that it will make my circuit become +-5v instead of dc? That is because i am using single supply ,so i setting offset to my input, would it cause the distortion to my output like the graph show below when i am supplying +5v to my circuit?( this graph i take from my previous high pass filter which cause by ac coupling capacitor that reduce offset)


    May i know how you calculate out the value for each component in this pre-amp? cause i would like to learn about the design of an pre-amp.


    Just want to know that ,if i am not added one smoothing filter but just directly connect the pin of bias voltage with Vcc, would the circuit work ? Actually now i am building this pre-amp but it seem cant function without smoothing filter ( when i connect the ouput with earphone,it doesn't have any sound receive from mic)

    In the diagram that you sketch,there have a capacitor connected to Vcc and ground, may i know that capacitor value is same as the capacitor in the filter?


    Do you have any source that teaching how to design an pre-amp?Or can you teach me some step on how to design a pre-amp? cause i really want get to know and learn it.

    Really thanks for your help!
     
  20. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    If your circuit is powered from +-5V then 0V is at "half the supply voltage". The output can swing equally up and down to make the undistorted AC signal.

    You show a 'scope waveform that is rectified (extremely distorted) because your circuit is biased at 0V with a single polarity supply then it cannot produce the missing negative part of the waveform. It should be biased at half the supply voltage for the output to swing equally up and down.

    You don't have a highpass filter circuit. Your lowpass filter was biased wrong at 0V.

    You need to learn about the basics of electronics. Not from me.

    No, its output will be saturated at Vcc instead of at half the supply voltage. The output will probably be a rectified signal that is just the lower half of the waveform.

    I don't know why you do not use the smoothing filter. The circuit should be built the way I showed it.
    Since you have a single supply voltage then the output DC voltage is at half the supply voltage. An earphone is driven with AC, not DC so you need to add a series output capacitor to pass the AC but block the DC. The capacitor value is calculated with the same "one over 2 pi RC" where the impedance of the earphone is the R. But an opamp doesn't have enough power to drive an earphone so it will be very faint and will be extremely distorted if you try to play it louder. If the mic can hear the earphone then the sound will go around and around making acoustic feedback howling.

    Every electronic circuit needs a supply bypass capacitor. If an audio circuit is powered from an AC to DC power supply then the power supply has a pretty big bypass capacitor so the capacitor on the circuit board can be from 0.1uF to 10uF. it stops the circuit from oscillating. If the supply is a little battery then the capacitor on the circuit board should be from 47uf to 1000uF.

    I learned electronics many years ago in university. There are many good tutorials on the internet. I recently downloaded a huge good book "The Art of Electronics" but it cannot be posted here because they want you to buy it. Maybe you can find somewhere to download it from.
     
Loading...