Building a noise detecting circuit

[Soilwork]

Hey,
I never used to like electronics, but after doing a unit last semester it has really interested me.
I've gone out and bought simple circuit components and have built a whole heap of simple circuits just because I wanted to learn more of how things behave.
Anyway I still have the notes from that unit I did a while ago and found a pretty tough situation and wanted to at least be able to draw the circuit before getting the components (if I can ).
I am pretty useless at electronics, but it's fun nonetheless and I would like to learn more.

This is what I'm wanting to build:

A circuit that will switch on a tape recorder when the input sound level is above a certain threshold and record for a preset amount of time (adjustable in the range 1 to 10 minutes). Include a filter to stop the recorder being triggered by 100 Hz wind noise. And you use batteries for this.

So anyway I was thinking that you'd need a Microphone, maybe 12 V battery supply, A high-pass active filter, timing device, combinations of resistors and capacitors and this isn't a necessity but it'd be pretty cool to have an LED to let you know that it's recording or not.

I know I'm probably missing some components that I'd need so I was hoping to find out what components and what exactly those components are doing exactly.

I'm probably in way over my head with this one, but I'm determined to build it to satisfy myself that I'm not a complete retard haha :)

 

[Soilwork]

Can someone tell me if this is along the right lines or not?
I've attached a drawing of the circuit I've contructed for the above situation so far, but I don't know if all or any of the parts are correct really.
Really appreciate any feedback.
Thanks

 

[dlgoff]

This is what I'm wanting to build:

A circuit that will switch on a tape recorder when the input sound level is above a certain threshold and record for a preset amount of time (adjustable in the range 1 to 10 minutes). Include a filter to stop the recorder being triggered by 100 Hz wind noise. And you use batteries for this.

Well it really good that you are interested enough to experiment. For your 1 to 10 minute timer, you should consider the 555 timer integrated circuit. Look at the Monostable circuits here: http://www.doctronics.co.uk/555.htm
Regards
Don

 

[willib]

help .. my file associations are messed up ..
whenever i try to look at a .doc file the graphics never come out...
i'm using XP ..
thanks in advance..

 

[Soilwork]

yeah I have the 555 timer in my circuit and I'm using a variable resistor to alter the length it allows the pulse through for.
But like I said I think I'm missing some other vital components and I'm not sure if I've connected everything correctly.
Thanks for the link by the way :)

 

[Soilwork]

help .. my file associations are messed up ..
whenever i try to look at a .doc file the graphics never come out...
i'm using XP ..
thanks in advance..

Sorry I don't know much about this, but I've uplpoaded the image here so you can look at it this way.

http://nucite.net/isWext.php/4426/60351.jpg

 

[dlgoff]

Sorry I don't know much about this, but I've uplpoaded the image here so you can look at it this way.

http://nucite.net/isWext.php/4426/60351.jpg

This doesn't work for me.
I haven't been able to see the circuit.

 

[Soilwork]

sorry about that
Try right clicking on it and saving it to disk that works for me so it might work.
so the word attachment isn't working at all??

 

[willib]

thanks Soilwork, that works for me...

 

[dlgoff]

Your first op-amp will give you amplification but not high pass filtering.

 

[Soilwork]

Ahh wait I need a capacitor there before R1 for it to act as a high pass filter don't I??

 

[dlgoff]

Ahh wait I need a capacitor there before R1 for it to act as a high pass filter don't I??

You can use a passive RC circuit or an active low-pass filter with an op-amp. Check out the last circuit on page 15 of the .PDF data sheet.
http://www1.jaycar.com.au/images_uploaded/LM2900.PDF
Regards
Don

 

[Soilwork]

That high-pass filter is a lot different to the one that was given in the notes I have from that unit I did.
Thanks for that though :)

 

[dlgoff]

Using a quad op-amp (4 amps in a chip) make it ideal for your application. I like the LM324 quad amp. If you want a LED to see when it's on, just put it in series with a resistor on the 555 output.

Regards

 

[Soilwork]

what are you saying I would use the quad-amp for??
Sorry for asking so many questions, but I'm just trying to understand this stuff :(

 

[dlgoff]

what are you saying I would use the quad-amp for??
Sorry for asking so many questions, but I'm just trying to understand this stuff :(

Well I would probably use one amp as a mic pre-amp coupled to a op-amp high pass filter. Okay maybe just two amp for your application.

Here's the http://www.national.com/ds/LM/LM124.pdf
You can use these application circuits to make all kinds of analog devices. I once made a polyphonic music synthesizer.

Regards

 

[Averagesupernova]

You can't just run an analog signal into the 555. You need to break down into block diagrams and then make a simple schematic for each one. Then combine, then simplify where you can. You need a sound 'detector' block. You have a good start with a mic and amplifier stage. You want to add some high and low pass filtering to make the detector immune to things you don't want to trip the detector. These could be considered blocks within the blocks Then you need the timer block. I assume you want to make the recorder go as long as there is noise and for an adjustable amount of time after the noise stops. Or do you want a noise to trigger it and record for a preset amount of time and then stop regardless of whether the noise is still there or not? These are things that need to be decided before we start drawing schematics. Then there is the output block. A pretty simple part hardly worth mentioning. Basically a switch.

 

[dlgoff]

"You can't just run an analog signal into the 555."

Why not?

"Then there is the output block."

His looks good to me.

 

[Omegatron]

"You can't just run an analog signal into the 555."

Why not?

If you look at the http://www.fairchildsemi.com/ds/NE/NE555.pdf , the trigger pin of the 555 is a comparator checking if the trigger voltage goes above a certain percentage of the 555's power supply. So actually, if the analog signal is scaled correctly, and you want the 555 to trigger whenever the analog signal goes above a certain value, then you can connect the output of an op-amp directly to the 555. :-)

"Then there is the output block."

His looks good to me.

The way he has it drawn, the transistor is connecting ground to ground. It needs to be connected between the power supply and the recorder instead. I would bet that just powering up the recorder wouldn't do what you want, anyway, since there would be a significant delay while it started up. You'd probably want to turn the tape motor itself on and off.

 

[Omegatron]

(By the way, is there a built-in method for drawing diagrams, like there is for TeX math?

I've used klunky for web diagrams in the past, and made my own version (it is public domain) with more symbols and such. It would be really convenient if something like this was built right into a forum like this, though. Save the hassle of screenshots and saving and uploading and so on.)

 

[Averagesupernova]

555 timer:

When pin 2 goes BELOW 1/3 of supply voltage the timing cycle starts when the unit is configured in monostable mode. When in this mode, the timer will time out and wait for the next pulse to start the timing cycle again. He has no negative power supply drawn. This means that we will have half wave rectified audio on pin 2 of the 555. Some op amps are not able to go 'rail to rail' so the voltage might not ever actually get down to zero volts. When there is no sound, the output of the 555 will be at its lowest possible value without figuring an offset. Most likely this will be below 1/3 supply voltage. He has a good start, but needs to work on a few things. He needs a little bit more processing on the front end before the 555. And for that matter, I'm not sure I would even use a 555. It depends on his answer to my original question of how the timer is supposed to function.

I wouldn't say his output stage is wrong, but I wouldn't necessarily say it is right either. The way he has it drawn, the box he calls 'recorder' would be the remote jack. I haven't played with many of them, but then ones I have messed with have the tip of the connector positive and the sleeve negative. When connected together the recorder runs. When that connection is broken it stops. This configuration would work for him. But, I would use a small reed relay or some sort of solid state switch that is bi-directional.

 

[Omegatron]

555 timer:

When pin 2 goes BELOW 1/3 of supply voltage the timing cycle starts when the unit is configured in monostable mode. When in this mode, the timer will time out and wait for the next pulse to start the timing cycle again. He has no negative power supply drawn. This means that we will have half wave rectified audio on pin 2 of the 555. Some op amps are not able to go 'rail to rail' so the voltage might not ever actually get down to zero volts. When there is no sound, the output of the 555 will be at its lowest possible value without figuring an offset. Most likely this will be below 1/3 supply voltage. He has a good start, but needs to work on a few things. He needs a little bit more processing on the front end before the 555. And for that matter, I'm not sure I would even use a 555. It depends on his answer to my original question of how the timer is supposed to function.

oops. yeah i guess i was oversimplifying.

I wouldn't say his output stage is wrong, but I wouldn't necessarily say it is right either. The way he has it drawn, the box he calls 'recorder' would be the remote jack. I haven't played with many of them, but then ones I have messed with have the tip of the connector positive and the sleeve negative. When connected together the recorder runs. When that connection is broken it stops. This configuration would work for him. But, I would use a small reed relay or some sort of solid state switch that is bi-directional.

yeah that makes sense now. would you use a relay because you're expecting large currents in the remote jack?

 

[Soilwork]

well what I'm wanting is for the noise above 100 Hz to trigger the timer so that it allows a pulse through for a preset amount of time hich will trigger the recorder. It really doesn't need to go for ten mintues I guess.

 

[Averagesupernova]

Hmmm. Not real descriptive. Do you want continuous noise to cause the recorder to continue to run? Or do you want the recorder to run for X time and then shut off until the noise stops and then starts for a new cycle?

 

[Soilwork]

Yeah so if you've set the timer for 10 minutes say.
Then once the noise starts it all off the recorder will record for ten minutes and then I guess regardless of whether the noise is still going the recorder will stop.

 

[Averagesupernova]

OK. So then if the noise stops for a couple of seconds and starts again the whole cycle starts over? Or do you want it to have to stop for more than a couple of seconds? You want some type of reset or literally a one shot with a manual reset?

 

[Soilwork]

Yeah I think one shot monostable with manual reset would just be fine.

 

[Averagesupernova]

I'll think about this for a while.

 

[Soilwork]

So is this what I'd have then?

power supply connected to mic which is connected to a pre-amp, which is in turn connected to a high pass filter (is a passive filter alright??).
I don't particularly know what else would go between the high pass filter and the timer though because I don't quite get why you can't run an analog signal into the 555 timer.
And do I leave everything after the timer as is?
Well actually I can probably remove the wire connecting the tape recorder to ground because there is only one input there.

 

[Soilwork]

I'll think about this for a while.

Thanks for this guys I really appreciate you helping me understand this stuff.

 

[dlgoff]

power supply connected to mic which is connected to a pre-amp, which is in turn connected to a high pass filter (is a passive filter alright??).

Yes

I don't particularly know what else would go between the high pass filter and the timer though because I don't quite get why you can't run an analog signal into the 555 timer.

The trigger (pin 2) of the 555 is the negative input of an analog op-amp used as a comparator. When you high-pass filter ouput is grater than 1/3 of your supply voltage, the 555 will trigger.

And do I leave everything after the timer as is?

Build it on a bread board. Remember if you use a quad op-amp, you have a couple of extras for say an inverter or driver.

Well actually I can probably remove the wire connecting the tape recorder to ground because there is only one input there.

Look at the applications data sheet for the op-amps I linked you to. Check out using an op-amp and a transistor as a current sink.

Regards

 

[Omegatron]

So is this what I'd have then?
I don't particularly know what else would go between the high pass filter and the timer though because I don't quite get why you can't run an analog signal into the 555 timer.

First, it depends if you're running your op-amp from "split supplies" or from a single-sided supply. For instance, an op-amp run from +15 V and -15 V that outputs 0 V with no signal would be a split supply. An op-amp that is driven by 0 V and +9 V and outputs +4.5 V would be a single-supply. Split supplies are a little easier to build, but require two batteries. Single supply requires voltage dividers to create the 4.5 V reference.

Second, the input to the 555 trigger is powered from 0 V to the positive rail (+9 V in the above example), and wants the input signal to drop below a certain percentage of it's rails.

"As you can see, the trigger input is held HIGH by the 10 kW pull up resistor and is pulsed LOW when the trigger switch is pressed. The circuit is triggered by a falling edge, that is, by a sudden transition from HIGH to LOW."

 

[Soilwork]

thanks guys :)
And yeah I will use a split supply with two batteries.

 

[Averagesupernova]

I wouldn't use a split supply. I'll try to explain why later today when I have more time.

 

[Soilwork]

haha ok I guess I won't use a split supply then :)

 

[Averagesupernova]

LOL Hey it's your deal not ours. But gimme an hour or two and I'll post some design tips for you. The decision is still yours.

 

[Averagesupernova]

Ok. Here is what I would do. Go with a single supply because going with a dual supply complicates things unnecessarily. Also, if you run the 555 on the same dual supply it is no longer a ‘dual supply’ so to speak. Running the 555 on the dual supply means that we need to ensure that the voltage to pin 2 of the 555 will not go below the 555s negative supply. Ground is really just a reference and we can move references wherever we like to most of the time.

Make an input stage using an active high pass or bandpass filter with one or two of the op-amps on the quad op-amp package discussed earlier. Of course you will have to bias up the op-amp to run at half the supply voltage. When choosing an op-amp, make sure to get one who’s output will go ‘rail to rail’. What I mean by this is that the output should be able to go all the way to the either supply voltage. Some op-amps will only go to within a volt or even more of the supply voltage.

Use a capacitor to couple into the next stage. Configure an op-amp as an inverting amplifier biased at about 8.5 volts. This is assuming you are using a 9 volt power supply. Gain of this stage can be adjustable using a pot for feedback resistor going from the output to the inverting input. What you will have out of this stage is rectified audio pulses going negative from 8.5 volts. Can you see where this is going?

Now do me a favor and draw me a schematic of what I described and post it. I guess maybe I’m lazy or something but I think it is good practice for you to make a schematic based on a text description. That’s what I was told at one time anyway. I promise I won’t leave you hanging and I will help you finish this.

 

[Omegatron]

Ok. Here is what I would do. Go with a single supply because going with a dual supply complicates things unnecessarily.

I always had trouble with op-amp stability when I did stuff like this, which is pretty difficult to diagnose when you are just learning and don't have access to a scope, etc.

Also, running the op-amp from +-9 V and running the 555 from +-9 V would allow the trigger point to be -6 V, correct? So you would only need to scale and rectify the signal (so it goes from 0 downwards) and when it got loud enough it would dip below -6 and trigger.

 

[Averagesupernova]

Remember what I said about moving references around? That's all you have done. There will be no connection to 'ground' on the 555 at all with a split supply and the op-amp will have connections to ground as bias points on input pins. Original poster please take my advice on this one. I've done this sort of thing with op-amps many times.

 

[Soilwork]

OK this is a rough schematic of what I think you've said, but I'm most certainly wrong.
You didn't really say anything about the transistor section so I don't really know if that's right and have left it as it was pretty much.
Thanks again for your help.

 

[Soilwork]

Oh and how does the capacitance coupling along with the inverting amplifier result in a rectified signal?

 

[dlgoff]

OK this is a rough schematic of what I think you've said, but I'm most certainly wrong.
You didn't really say anything about the transistor section so I don't really know if that's right and have left it as it was pretty much.
Thanks again for your help.

Can you attach a .jpg file please?

 

[Omegatron]

Oh and how does the capacitance coupling along with the inverting amplifier result in a rectified signal?

It won't. You need diodes. Something like this:

http://www.play-hookey.com/analog/full-wave_rectifier.html

Though just a regular bridge rectifier would probably be fine for your purposes...

In fact, not rectifying the signal is probably fine for your purposes.

 

[Averagesupernova]

It won't. You need diodes. Something like this:

http://www.play-hookey.com/analog/full-wave_rectifier.html

Though just a regular bridge rectifier would probably be fine for your purposes...

In fact, not rectifying the signal is probably fine for your purposes.

Nope. No diodes. Please don't confuse him. My workday is not yet done but when I have time in a couple of hours I will explain. He does not have it drawn quite right.

 

[Averagesupernova]

Ok. I notice several things about your schematic. Some are important, some are not. You did not draw an active filter on the first stage. That is fine, we can add it later. Let’s just get the main functions working first. R3 needs to be omitted and the non-inverting input (+) on the second op-amp needs to be tied to about 8.5 volts instead of ground. This can be accomplished with a voltage divider. I assume you know how to make a voltage divider out of 2 resistors tied between the 9 volt supply and ground. If not, just ask. When I say ground, I mean the negative side of the 9 volt supply. The negative side of the mic is also tied to what I call ground. Also, tie the non-inverting input of the first op-amp to a 4.5 volt reference. Use a voltage divider to get it.

Now, to explain why this circuit does what I say. When the non-inverting input on the second stage is tied to 8.5 volts the output will also be centered around 8.5 volts. Since the cap is on the input the DC gain of the stage is one. When you couple an AC signal in through the cap the output should swing positive and negative by an amount determined by the feedback resistors. However, it cannot swing farther up than the 9 volt supply so it will mostly swing down. And obviously it cannot swing lower than ground. When there is no AC signal coming in then the output will be sitting at about 8.5 volts. This is just what we want. With adjusting the gain we can set the circuit for just the amount of noise we want to trigger the timer.

You also drew the output of the second op-amp going directly into the 555. We are going to change this. Insert a diode in between the op-amp output and pin 2 of the 555 with the cathode towards the 555. Tie a resistor from pin 2 of the 555 to ground.

Next we are going to make the manual reset circuit. We’ll make a sort of Schmitt trigger circuit out of one of the op-amps left on the quad package as a type of latch. Tie a resistor from the non-inverting input to the output. Tie another resistor from the non-inverting input to ground. This is positive feedback that makes a Schmitt trigger always go to either power supply rail. This is just what we want. Tie the inverting input to a 4.5 volt reference made out another voltage divider. Tie the output of this op-amp to pin 2 of the 555 with a diode inserted in series. Cathode towards the 555. Insert a diode with the cathode on non-inverting input on the op-amp and the anode on the output of the 555 timer Also insert a resistor in series with this diode. Now install a momentary switch from the non-inverting input to ground. This is the reset button. Also tie a .47 microfarad capacitor in parallel with the switch.

Draw the schematic, ask any questions and I’ll be glad to help you.

 

[Soilwork]

Thanks so much for all the help.
I've got exams on in a week and I haven't been keeping on top of my work so I'm a bit worried about it all at the moment.
Anyway I will post my schematic of what you suggested in the next two weeks if that's ok.
That's when I will be on holiday and I will have plenty of time to practice making circuits.
I just want to make sure that no one minds in particular if I post quite a lot of questions on this board about how things operate and all??

Thanks again guys (especially averagesupernova)
Tom

 

[Soilwork]

Right I've finished my exams now so I'm going to start working on those suggestions you gave me. The schematic should be up within the next few days :)

 

[Soilwork]

Tie the output of this op-amp to pin 2 of the 555 with a diode inserted in series. Cathode towards the 555. Insert a diode with the cathode on non-inverting input on the op-amp and the anode on the output of the 555 timer. Also insert a resistor in series with this diode. Now install a momentary switch from the non-inverting input to ground. This is the reset button. Also tie a .47 microfarad capacitor in parallel with the switch.

Draw the schematic, ask any questions and I’ll be glad to help you.

OK I have a couple of questions here :):

1. Why isn't that first op-amp considered a high-pass active filter? I got it from those lecture notes I had from the subject I did last semester. The input is coming from the mic and going through the capacitor, resistor and into the inverting input. The non-inverting is connected to ground. But also when you look at the complex impedance of the capactor it is 1/jwC (w is omega). So obviously the impedance is really large at low frequencies, but low at high frequencies. I know that everything needs to be changed, but my question is, if you were just looking at that one op-amp would that be a high-pass active filter?

2. I've done the corrections individually for the op-amps (1 & 2) but I haven't actually connected them up because I have sort of lost what you were saying. It says connect op-amp 2 to pin 2 of the timer, but then when you are referring to the "new" op-amp (schmitt trigger) with positive feedback that I need to add, you say that I connect that to pin 2. I can't cascade op-amp 2 and the "new" op-amp, because the inverting input has its own supply and the non-inverting is effectively connected to ground. So which one is connected to pin 2?

3. Sorry I don't really get what the bold statement in the quote above is saying?

4. The capacitor in parallel with the 'manual' switch is so that the switch is bypassed when the circuit is active right?

Those are all the questions I have for now :)
Thanks again

 

[Averagesupernova]

OK Soilwork. I don't have time at the moment to address your issues but just wanted to let you I've been watching this thread and haven't forgotten about you. I hope to get all of your specific questions answered in the next 12 to 24 hours.

 

[Soilwork]

yeah man that's fine, whenever you have the time.
I'm grateful for any time spent helping me :)
Take as much time as you need/want to :)