Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Square wave audio out from the hearing device!

Status
Not open for further replies.

Willen

Well-Known Member
Some were saying that the circuit was sometime sold as a hearing device. From where I got the circuit, he have a very old device too, which is fully based on the circuit. That sounds good. Today I did simulation but I got totally square wave output. Simulation has totally same components.

How much distortion we can neglect in such situation? Please give a run the attached fine on LTspice.

I am trying to do experiment by making such amazing circuit myself. So If you simply can correct then it would be great to me.

Regards
 

Attachments

  • Hearing device Square wave out.zip
    209.9 KB · Views: 172
>You have TOO MUCH GAIN.
>V1=0.5mV lower the input level to 1/10
>Cut open from V1 to R8 and a 10k resistor. This will reduce the gain. I don't know what kind of mic but I think it has resistance.
>The bias seems a little off so I changed R1 to 100 ohms. Not so important.
 
Hi,

Oh yes! May be I used 5mV for Mic which might be very low in real device. The circuit can has very high gain then why the manufacturer didn't reduce the number of transistors?

I am little confusing around Q1, what it is doing here? Isn't it shorted?
 
Q1 acts like a diode. more or less. It is making a 0.547 volt supply. All your transistors have a temperature stability problem. This bias supply changes with temperature to make the amp stable with temp.

Much of you simulation problems are that you are using a 5mV low impedance signal source while a mic is more like a 0.5mV high impedance signal.
 

Attachments

  • Draft4.asc
    4.6 KB · Views: 189
Hi,
Oh wow! How tricky the Q1 bias! Are there any web pages (links) to learn about such bias circuit? I like the idea very much and want to know how it works.
 
You are using a signal generator with an output impedance of ZERO ohms. It is shorting the negative feedback provided by R9 so of course the voltage gain is too high.
I added a 3.3k resistor (the output impedance of an electret mic) between the signal generator and R5/C1 then with an input of 1.5mV peak, the output from Q4 looks pretty good.

If R8 was a 50k volume control then why did you change it to only 2k ohms? A volume control should be shown as two resistors in series making a voltage divider.
Please show how R8 and R11 are connected in the original circuit.

If the 3.3k ohms is added to the input and the volume control is turned down then the output looks pretty good.
 
Edited file by ronsimpson and suggested modification by audioguru both has nice output.

Hi AG,

I am little confused about 3.3k series resistance on signal generator. The device has a single condensor Mic and other all same parts. I think 3.3K resistance is created itself or...should I have to use real 3.3k resistor?

I also used 10K pot for Q4 volume controller. Lets see-
 

Attachments

  • Draft3.asc
    4.9 KB · Views: 187
Transistors Q2, Q3 and Q4 are an inverting amplifier with their voltage gain set by the ratio of R9 divided by the internal resistance of the signal source. The resistance of your signal generator was ZERO ohms so the voltage gain of the amplifier was much too high.

The internal resistance of the Jfet inside an electret mic is about 3.3k ohms when it has a DC voltage of a few volts. I just noticed that your total supply is only 1.5V then the voltage on the mic is very low so its impedance is probably higher than 3.3k ohms.

Did you notice the even-harmonics distortion at the output? The top of the sinewave is squashed and the bottom is stretched.
 
Oh if Mic has internal impedance then it has better output. I never heard harmonics as a audio so cannot find harmonics. But if input is higher than necessary then I can hear audio with distorted noises.

I just can simulate the circuit and can see harmonics on waveform.

Another- If an amplifier has total capacity of 200mV out peak to peak max but if we just use it to get 150mV p-p then in this situation remaining 50mV is 'headroom'? Or what and how can we get headroom?
 
I never heard harmonics as a audio so cannot find harmonics. But if input is higher than necessary then I can hear audio with distorted noises.
A single transistor with no negative feedback has high even-harmonics distortion because the base-emitter voltage swing must be exponential to cause a linear collector output voltage swing.
Even-harmonics are musical (exactly one octave higher) then some people cannot hear them.

Another- If an amplifier has total capacity of 200mV out peak to peak max but if we just use it to get 150mV p-p then in this situation remaining 50mV is 'headroom'?
Yes, it is headroom. Radio stations use a compressor/limiter circuit to decrease sound levels that are too loud so that the average level is louder and less headroom is needed. The compressor function causes the gain to pump up and down which is audible.

Or what and how can we get headroom?
Reduce the input level or increase the power supply voltage.
 

Attachments

  • sim transistor distortion1.PNG
    sim transistor distortion1.PNG
    34.2 KB · Views: 186
Oh wow my simple assumption about 'headroom' is totally correct! :)

2nd simulated circuit has very nice swing 1V to 9.5V p-p, wow!
But how you found exactly 40% distortion on LTspice? Is there any tool or just your assumption by seeing its wave?
 
But how you found exactly 40% distortion on LTspice? Is there any tool or just your assumption by seeing its wave?
I guessed the horrible distortion was 40%. LTspice can show you exactly how much distortion and the level of each harmonic but I did not look at them.
 
Hi again,

Today I made a simplified version of hearing aid (fixed above). I removed some optional parts and used parts which I have (few near value resistors). But got tiny (EXTREMELY low) level of audio which is almost nothing. I made simulation (uploaded here too) and it shows it's doing almost nothing. Can you point out the faults?
 

Attachments

  • simplified failure version.asc
    4 KB · Views: 149
hi W,
One fault is the Base/Emitter voltage of Q3.
E
 
You don´t have enough voltage at base of Q3 to open it, which makes Q4 fully on. I have no idea why is this stage done in such manner, but I guess a normal three stage AC coupled amplifier would work better.
 
The output transistor is class-A that wastes a lot of battery power all the time even when there is no sound, use a class-AB complementary output stage instead.
You should NEVER put DC through an earphone transducer which might burn it out or cause its membrane to be pushed to one side, use a coupling capacitor to feed it AC.
You do not have a condenser mic that needs a 48V supply, instead you have an electret mic that has the 48V built into its electret material and has a Jfet to reduce its output impedance.
The preamp transistors are biased wrong so they do not work. You do not need three transistors to make a simple mic preamp, two transistors biased correctly are fine like this:
 

Attachments

  • wrongly biased preamp.png
    wrongly biased preamp.png
    12.9 KB · Views: 134
  • mic preamp.png
    mic preamp.png
    22.9 KB · Views: 143
Is that how you simulate an electret mic as a voltage source with R bias+gain?
 
An electret mic has a common-source Jfet at its output that uses about 0.5mA of current when it has 3V Vds, or less current with a lower voltage. Its output resistance is about 3.3k ohms which is shown in series with the signal generator. With a 1.3V supply, the drain resistor (that you call R bias) for the Jfet should be about 4.7k or 5.6k as shown. I do not know if the Jfet in the mic has gain because most electret mics have almost the same output level which is abouyt 10mV RMS when speaking in a normal conversation level at a distance of 10cm, which is about the same level as a dynamic mic.
 
I noticed that the low frequency response of my schematic was poor so I increased the value of the emitter bypass capacitor without increasing the value of the supply filter capacitor then it began a motorboat low frequency oscillation.
 
Thanks AG, using your simplified two transistor based preamp (above) I got pretty nice amplitude of swing. Now I am having problem in output transistor. In the base of it, I got so nice sine, but output distorts badly. I changed collector, emitter and feedback resistor many times but got failed. As you suggested I do not want here to use AB amplifier here because it's just an experiment and I want as simple as possible. Here's SIM-
 

Attachments

  • Problem through output.asc
    4 KB · Views: 138
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top