again, microphone input high school project

[zgnoh]

Circuit Diagram Here

sorry again,,, I know I've been quite weak at electronics.
About the circuit diagram:
Input from Microphone ->
Retifier/Filter Circuit to convert the input wave to steady direct current of voltage similar to amplitude of wave ->
Variable Resistor ->
Potentiometer to adjust ->
if Input voltage is high enough, activate relay

questions:
1) any bugs? feasible?
2) I wonder if the voltage of microphone, as we speak, is enough to drive a circuit !@???@_@

other questions later
thanks very much >< :-[

[audioguru]

Just one little bug: The microphone needs a preamp.
The output voltage of a microphone is about only 20mV if you shout closely. The voltage requirement of the rectifier bridge is at least 1.3V. A big difference of 65 times. If you talk normally at a normal distance from the mic, the difference is about 260 times. So the preamp must have a voltage gain of about 260.

You don't need a bridge rectifier, a single rectifier diode will be fine.
You probably also don't need the inductor, 2nd capacitor and rheostat. Just connect the potentiometer directly to the output of the rectifier diode and the filter cap.

The "buffer gate" can be an ordinary PNP transistor, with a power supply for it and for the preamp.

[zgnoh]

Just one little bug: The microphone needs a preamp.
The output voltage of a microphone is about only 20mV if you shout closely. The voltage requirement of the rectifier bridge is at least 1.3V. A big difference of 65 times. If you talk normally at a normal distance from the mic, the difference is about 260 times. So the preamp must have a voltage gain of about 260.

You don't need a bridge rectifier, a single rectifier diode will be fine.
You probably also don't need the inductor, 2nd capacitor and rheostat. Just connect the potentiometer directly to the output of the rectifier diode and the filter cap.

The "buffer gate" can be an ordinary PNP transistor, with a power supply for it and for the preamp.

Thank you very much. i've been wanting to know the voltage generated by speaking to a microphone!!!~

I still have a few problems:
1) why do I not have to use the bridge rectifier but a single rectifier diode instead?~

2) Should I use a PreAmp before the rectifier? Would it make the circuit not sensitive enough because there'll be noise in the PreAmp?
moreover, is that any preamp will do the job provided that it amplifies the voltage by 260 times?

3) Actually the circuit will be extended so that different amplitude will activate different switch. Sensitivity is quite important, do you think my circuit will be sensitive enough so that when people are singing songs, the circuit reponds differently to different amplitudes?~ the variation of amplitude is a song is large enough?~

4) Actually i'm not very familiar with Transistors, so is it okay if I just use the buffer gate?~

haha, thanks very much~~~ :-*

[audioguru]

Hi Zgnoh,
1) You only need to rectify the AC audio into smoothed DC. A single diode will do it with a filter capacitor.
2) A single rectifier diode needs at least 130 times the voltage from a microphone to turn-on, so of course the microphone's signal must be amplified by a preamp before being rectified. When you build the preamp, use parts (an audio opamp) designed for being used in a low-level, low-noise preamp and there won't be any noise.
3) Since your circuit needs to detect different sound amplitudes and give separate outputs for them, you should use an LM3915 bar-graph IC which is designed to drive 10 LEDs but can also be used as your switches. Then you don't need transistors or "buffer gates".

What are you trying to do with different levels? The microphone output for a group of people singing at a distance from it will be about the same level as a single person singing or talking fairly close to it. Just listen to a good PA system, its output level is always about the same.
You can design the preamp and LM3915 to have as much sensitivity as you want. The circuit can activate a switch if a pin drops. But it will also activate when a truck drives by outside.

[Kevin Weddle]

There is one problem. You may not be able to get a DC from the signal. A filter capacitor will only capture the DC level if it is there. It is likely that you will end up with a very small ripple but no DC. And as far as I can tell you would be hardpressed to even get that DC from the signal because there is always some rate of change to it.

[audioguru]

Hi Kevin,
Of course you will get DC out of the rectifier and its filter capacitor when it is being fed many volts of signal from the mic preamp. That is what a rectifier is for.

Also, of course the changing level of the signal will also change the level of the DC, but the rectifier can be designed for average-responding and have a big filter cap to smooth any quick level changes. You can even design the rectifier to have adjustable attack and release times if you want.

[Kevin Weddle]

Audioguru, you know you can't get DC from trying to filter a triangle wave.

[audioguru]

Kevin, you won't have a triangle wave if your filter cap is big enough.
It is filtering audio, not rotations of the moon.
You will just have nice smooth DC with hardly any ripple. Hey, the cap barely has any load current (25nA input current of the LM3915) so stores its charge for a long time.

[zgnoh]

Hi Zgnoh,
1) You only need to rectify the AC audio into smoothed DC. A single diode will do it with a filter capacitor.
2) A single rectifier diode needs at least 130 times the voltage from a microphone to turn-on, so of course the microphone's signal must be amplified by a preamp before being rectified. When you build the preamp, use parts (an audio opamp) designed for being used in a low-level, low-noise preamp and there won't be any noise.
3) Since your circuit needs to detect different sound amplitudes and give separate outputs for them, you should use an LM3915 bar-graph IC which is designed to drive 10 LEDs but can also be used as your switches. Then you don't need transistors or "buffer gates".

What are you trying to do with different levels? The microphone output for a group of people singing at a distance from it will be about the same level as a single person singing or talking fairly close to it. Just listen to a good PA system, its output level is always about the same.
You can design the preamp and LM3915 to have as much sensitivity as you want. The circuit can activate a switch if a pin drops. But it will also activate when a truck drives by outside.

Thanks vy much again. Perhaps it'll take me some time to digest :-*

I'm doing a project trying to control the switches so that there'll be some kinda musical fountain with a music input from a CD. A microphone version is for demonstration. when people sing with the microphone,,, switches open and close.
The CD version need a variable resistor I suppose because input from CD is fed by user where the user can adjust the volume.
This is a high school project and I'm trying to achieve the task above using the simplest way (difficult ways i'm not capable of that,,,:p)
That's why I chose to use amplitude instead of frequency though i'm afraid amplitude won't show much variation.><

About the problem being discussed by you guys!~! I know the amplitude of a song is always varying... that's why I'm quite worried><

[zgnoh]

Hi Kevin,
Of course you will get DC out of the rectifier and its filter capacitor when it is being fed many volts of signal from the mic preamp. That is what a rectifier is for.

Also, of course the changing level of the signal will also change the level of the DC, but the rectifier can be designed for average-responding and have a big filter cap to smooth any quick level changes. You can even design the rectifier to have adjustable attack and release times if you want.

then do I have to use a specially designed rectifier?

moreover, as I mentioned in the last reply, u have a brief idea of my project.
but will the rectifier give a very smooth DC even the amplitude is varying?
thanks vy much.

bed time~

[Kevin Weddle]

What I am saying is that you will have trickles of DC charge at all levels with a microphone signal. You can't expect the peaks to give you much DC. Conversely, when you charge a capacitor with mains you get a nice DC charging. It is also imperative that it completely charges to the voltage in the time allowed. The peak of the waveform has a time to it. There must be a short RC time constant. And of course it is not because you are using a large filter capacitor. So I really don't not think that the signal contains the stuff to develop a good DC level. You are however creating a small ripple on the filter capacitor, but no DC.

In fact I hooked the circuit up and realized this fact.

[audioguru]

Hi Zgnoh,
It would be cool to see your water fountain squirt by the various levels of music. With most pop music it would squirt to the beat of the music!
1) The LM3915 can have a switch to select a single output called dot mode when it lights LEDs, or bar mode when it lights all LEDs below the one that it is indicating. The outputs of the LM3915 can drive electrical water valves or relays to drive them.
2) You can put a volume control on it to adjust the range of input.
3) With CD input you reduce the gain of the preamp but the circuit still needs a rectifier circuit and filter cap.
4) A very simple rectifier circuit is shown on the LM3915's datasheet that works much better than just a single diode.

[audioguru]

Kevin,
The charging time of the rectifier's filter cap depends on the cap's value and impedance of the source.
If you use a 0.1uF filter cap, the milli-ohms output impedance of the preamp will charge it through the diode to the peak of the audio immediately. Then you could use a 1M load resistor to fully discharge the cap in 0.5s, or a 10M resistor in 5 seconds.
Mixer consoles and compressor/limiter circuits use this theory all the time.

The circuit you tried it on probably didn't have a very low output impedance opamp circuit feeding your rectifier diode, and also probably didn't have a very high impedance load like the input of an LM3915.

[Kevin Weddle]

Hello, Audioguru

I managed to experiement with the triangle wave. The results where untinteresting, the peak to peak was low and so was the DC. You see every cycle charges the capacitor at the point where there is zero rate of change. Otherwise you simply have a diode capacitor filter. And as you can see that if the voltage fall, it's not the diode that holds the voltage from falling across the capacitor, its the DC that was built up across the capacitor. This why you have some ripple to begin with, it's because the voltage goes up and goes down regardless of the diode and even without a load.

You see a capacitor will not hold much DC when you have a triangle applied because there is always a rate of change. Just apply the triangle to the diode and capacitor and you will see the ripple.

[audioguru]

Kevin,
Your cap should have charged to the peak voltage of your triangle wave, or any other AC, minus a diode voltage drop. Since you say the DC voltage was low, then you also need a preamp as described in my very 1st reply to this topic.
With a load resistance and capacitor value time constant of 0.1 seconds (1M and 0.1uF), there will be insignificant ripple for signal frequencies above about 70Hz. At about 70Hz, the cap's voltage will drop 20% between cycles, which won't be detected by the 3dB steps of the LM3915.

[Kevin Weddle]

I'll be honest, this is an AC triangle applied to a diode and capacitor and the voltage out is 1.5 VDC and the input is 2.5 volts peak. This is more DC than I expected. You have to realize that when the voltage falls, it also falls at the capacitor. It's the DC supply from the capacitor that allows the diode to reverse bias.

[audioguru]

Kevin, yeah.

[Kevin Weddle]

But there is a problem here. Since my generator keeps clipping at 2.5 volts peak I will surely see there is a difference of a diode drop. I can't even generate a good amplitude with my gnerator. So how can you say for certain that I am wrong. I know for a fact there is ripple at the capacitor without a load. Not much though there is good filtering.

I want you to try something. Connect a diode and resistor to a variable voltage source. When you drop the voltage, there is no clamping action, just dropping voltage. A capacitor however will act as an opposing power supply, maintaining the voltage when the signal drops. However, I know what you are thinking in that the voltage applied to the capacitor is all the energy it is going to store. But the voltage it takes it to is not what you think. The voltage it takes it to is the DC plus the peak of the ripple, regardless of the size of the input.

[audioguru]

Kevin,
Without a load (but maybe your meter or 'scope) the cap won't be discharged by anything so won't have any ripple.
Maybe your rectifier diode is too slow for your frequency. 1N400x diodes are very slow, try a fast 1N4148 diode.

[Kevin Weddle]

Read the statement before your reply.

[audioguru]

Kevin,
Ripple doesn't come from the input. It is the result of the cap discharging a bit into the load between charging pulses from the diode.

[zgnoh]

Hi Zgnoh,
It would be cool to see your water fountain squirt by the various levels of music. With most pop music it would squirt to the beat of the music!
1) The LM3915 can have a switch to select a single output called dot mode when it lights LEDs, or bar mode when it lights all LEDs below the one that it is indicating. The outputs of the LM3915 can drive electrical water valves or relays to drive them.
2) You can put a volume control on it to adjust the range of input.
3) With CD input you reduce the gain of the preamp but the circuit still needs a rectifier circuit and filter cap.
4) A very simple rectifier circuit is shown on the LM3915's datasheet that works much better than just a single diode.

thanks~
where can I find the LM3915 datasheet?

I'll have to do the written part of the project, I'll be using the diagram below to briefly illustrate the flow of the circuit:
Click here, circuit
Then in practice, I'll use the IC, do some calculations on the values of capacitance, resistors.

Do you think it's okay?
moreover, what do you think I can put in the other side of the electromagnetic relay?
I propose using a big water tank and some tubes. What I have to do is just control the opening and closing of some switches or valves.
If I couldn't figure out how to set a valve circuit, I'll have to end up showing LEDs, which is much less fun than seeing a musical water fountain:(

thanks

[audioguru]

Hi Zgnoh,
At the bottom of this page is a link "WWW Search" to the Google search engine. You can enter part numbers, circuit names or anything. I entered LM3915 and it linked to this datasheet: http://www.national.com/ds/LM/LM3915.pdf
You will see that it has 10 outputs arranged in steps of amplitude of the input. The 1st output is activated with a low-level input. The 2nd activates with more input amplitude and so on. It needs only 2 resistors to be calculated, for its maximum input voltage range and the brightness of its LEDs.

On your drawing the left triangle is the LM3815. You show one output driving a relay. You can draw more outputs on the LM3915 driving more relays up to ten.
Each relay can activate an electrical water valve that are used for timed sprinklers for example.
For your demonstration use LEDs instead of relays.

[zgnoh]

Hi Zgnoh,
At the bottom of this page is a link "WWW Search" to the Google search engine. You can enter part numbers, circuit names or anything. I entered LM3915 and it linked to this datasheet: http://www.national.com/ds/LM/LM3915.pdf
You will see that it has 10 outputs arranged in steps of amplitude of the input. The 1st output is activated with a low-level input. The 2nd activates with more input amplitude and so on. It needs only 2 resistors to be calculated, for its maximum input voltage range and the brightness of its LEDs.

On your drawing the left triangle is the LM3815. You show one output driving a relay. You can draw more outputs on the LM3915 driving more relays up to ten.
Each relay can activate an electrical water valve that are used for timed sprinklers for example.
For your demonstration use LEDs instead of relays.

thanks,
I have some problems drafting the circuit regarding LM3915
just silly problems.

http://www.national.com/ds/LM/LM3915.pdf
according to page ten of this factsheet FIGURE 1, the amplifier and rectifier.
the input,,, only one input...>!?!?!?!..<<??!?
how can I connect a microphone to it... a microphone,,, two wires... only one place for input...... and one for output,,,
one output to SIG5 is okay,,, but only one input... how to draw?

also, if i want to voltage gain to be variable, which, R1 or R2, R3, R4 should I change to variable resistor?

I know it's not the page's error,,, just my poor knowledge about circuit...
thanks.

[audioguru]

Hi Zgnoh,
The LM3915 isn't connected to the microphone, it is fed from the volume control like the left triangle in your previous diagram. The volume control is R2 of the "Half-wave peak detector" circuit in figure 1 in the datasheet. The half-wave peak detector circuit is DC-coupled from the output of a preamp circuit you still need. The preamp needs a gain of a few hundred and its input is the microphone.

It ain't simple but you're getting there.