Noob needs circuit- If speaker is working blue LED is on.

[Captain Colorado]

I need to build 20 units.
each one needs to fit into a single gang box, mounted in a wall.
blue LED will be powered by a 16vac transformer.
typical outputs from 5w amplifier into 16 ohm speaker - volume at 10% 10mV~60mV, volume at 50% 200mV~800mV, volume at 99% 2V~11V.
parts need to be available for production by Friday, 11/12.

Thanks,
CC

 

[blueroomelectronics]

Unusual project, single blue LED just to show the speaker is connected?
What's your budget?

 

[crutschow]

You could use an op amp to amplify the speaker signal, and rectify/filter that to get a DC signal. Use this to trigger a comparator and light the LED.

You also need a rectifier and filter for the 16VAC to get DC for the circuit.

But good luck getting parts for all that by Friday.

 

[Blueteeth]

When you say 'speaker is on', does this mean it is producing constant audio? or that it is connected/routed to a system? I mean, measuring a tiny voltage across the speaker is all well and good, but if your audio system, is giving it silence, then your blue LED won't light. You could add a cap on the comparator input so the LED stays on for a while after being 'tripped', but then how long does it stay on for?

 

[Captain Colorado]

I am monitoring a LogiSon Acoustic Network(logison.com) that produces white noise in the human voice spectrum(125 hz to 10,000 hz). The speakers are always on. The output volume level limits are 35 dbA(10%) to 85 dbA(99%) or off. The typical levels in a tuned system are 50% to 75%, and constant.
I need to know if the speaker wire has been cut or disconnected, in which case the blue LED goes off.

Budget for the project is $100 for the design, and $20-$30 for each sensor. I need at least 16 for this job.

 

[KeepItSimpleStupid]

How about series resistor + clamp (+-0.6V) + Precision rectify (probably level shift) + comparitor + voltage ref

Half-Wave Rectifier

 

[KeepItSimpleStupid]

You might be able to get away with resistor + clamp + compare, but there are still som issues.

1. power (16 VAC may be too high)
2. Power supply noise (switching regulator may not work, 10 mV is pretty small)

Nature of output:
1. If line to voice coil, that creates lots of complications.
2. If there is an amp at every speaker, that may case issues too.
3. If output is bridged, that causes problems too.

 

[canadaelk]

Is this a soundmasking system with problem employees?
A modified circuit as used in automotive lightbulb monitoring could do it (Analogs AD22001 or similar)
Your design budget and timeframe is way under though!
E

 

[Boncuk]

Hi Captain Colorado,

here is a circuit which might suit your needs.

Brief circuit description:

AF enters the amplifier (IC2A) via C1 (AC-coupling). There it is amplified by the factor of 101 (1+R1/R2). So the weakest input voltage of 10mV will be amplified to 1.010V.

The amplified signal is then rectified (D1) and fed into a comparator. C2 and R3 hold the signal for a short time while the input signal might be lost.

The comparator output swings high if the amplified signal is over the reference voltage of 909mV (voltage divider R4/R5)

For a blue LED you might use a 470Ω resistor (instead of 400Ω) for current limiting. 400Ω result in almost 20mA LED forward current.

The PCB design is purely single sided and measures 2.5750X1.9".

Boncuk

 

[Boncuk]

Your design budget and timeframe is way under though!
E

Oops!

Where do you normally purchase electronic components?

Without the transformer the part costs are below US$5.

Boncuk

 

[ronv]

Led

Here is a radio shack version of Boncuk's circuit. You shouldn't loose to much money if you run right down there. :+)

 

[4pyros]

Boncuk; I think the OP needs something in series with the speaker to see if it is still conected. Not something in paralle to see if thare is a signal. Andy

 

[Captain Colorado]

Andy,
you are so right.
I need the blue LED to go off if someone has cut the wire to the speaker.
And, of course, the circuit can't affect the output to the speaker.

Someone at a local electronics store suggested that I need a current transformer.
At that point I knew I needed help from the experts...

the time frame has expanded, but fast help is much appreciated.

CC

 

[Blueteeth]

You sohuld be able to get away with a current sense resistor, although at such low power the resistance would have to be quite high which would affect the speaker - but so would a current transformer. For 10mV p-p in your 16ohm speaker, thats under a miliamp. For a 1ohm resistor, thats under 1mV < - prone to all sorts of noise. 16ohms would give some noise immunity, but a value approaching the impedance of your speaker will abosrb a significant amount of the signal.

You mentioned the speaker is always kicking out some audio, so perhaps a telephone pick-up coil at the back of the speaker would pick up a signal. Amplified, rectified, to a comparator. Whether the LED lights for a 'speaker active' or 'speaker inactive' just depends on how you wire up the comparator. Or use a transistor inverter. I'm guessing a dual opamp, a diode or two, a transistor, and some passives for the detector.

Of course, this all depends on 'where' you suspect the speakers wires will be cut. If its anywhere from its source to the 'speaker/LED detector' then sending a constant small high frequency signal from the source (too high for the speaker to output) would give a nice clear indication of a connection. Your circuit could then look for this and keep the LED off if its there.

Perhaps a block diagram would help us!

 

[4pyros]

Can you tell us whare the sound source, the speaker, and the LED are to be? Andy

 

[Boncuk]

Andy,
you are so right.
I need the blue LED to go off if someone has cut the wire to the speaker.
And, of course, the circuit can't affect the output to the speaker.

Someone at a local electronics store suggested that I need a current transformer.
At that point I knew I needed help from the experts...

the time frame has expanded, but fast help is much appreciated.

CC

Hi Captain Colorado,

I guess sensing an audio signal doesn't necessarily require series connection of a sensing resistor or current transformer.

The main thing is: If the speaker cable is cut or broken the speaker won't sound anymore. The amplifier won't get any useful signal either.

All you have to take care of is a sabotage proof connection between speaker and my suggested circuit.

With the speaker in the same enclosure as the sensing circuit that problem will almost be covered.

Another problem would be if besides the broken speaker cable somebody gets the idea to connect a small button type battery directly to pin3 of IC2A. That would make the blue LED to illuminate until the battery is depleted to a level <8mV.

Additionally a circuit with a current sensor and amplifier (without reducing speaker volume) would most likely bust your budget.

Boncuk

 

[Captain Colorado]

here are the parameters of my problem.
I am selling a sound masking system to a govt contractor. We are trying to make a SCIF room(Security Campartmented Information Facility) that is audibly secure. There are specifications for the walls to make them sound resistant but not sound proof. To lower the cost of sound proofing we add sound masking(random sound in the human voice spectrum).

Inside the room will be the sound masking system consisting of multiple 5w amplifiers, hung in the ceiling above the ceiling tile and next to the walls. Each amp can feed 2 speakers, the speakers will be outside the room.

I like the idea of a high frequency output to the speaker and listening for it on the other end. Both the HF source and the blue LED will be inside the SCIF.

If my cost has to go up then so be it, better to buy a BMW that works than a Yugo that doesn't.

 

[Boncuk]

The problem is the extremely low AF voltage which has to be made visible using an LED.

10mV equivalent DC voltage on a 16Ω load generate a current of 625µA.

The second problem is the fact that the sensing circuit is not supposed to reduce loudspeaker volume.

I doubt you'll be able to find a current transformer for that kind of extremely low power.

Here is another approach using a current sensor ACS712 from Allegro micro. It puts out 66 to 185mV per Ampere, which is pretty high for your application, but it should work with high amplification after the sensor circuit.

I used the sensor with the foregoing design with some modifications.

The sensor is connected in series with the loudspeaker. It has an internal resistance of 1.2mΩ which should not affect loudspeaker performance. The output is connected to an amplifier with an amplification ratio of 1,000 : 1.

The remaining circuit is unchanged with the exception of the LED current limiting resistor (now 180Ω) since the supply voltage had to be reduced to suit the ACS712 for a VCC of +5V.

Here are the altered schematic and PCB design. PCB dimensions are: 2.3815X1.56875".

If an amplification factor of 1,000 : 1 won't suffice you could use an instrumentation amplifier with a factor of 10,000 : 1. This of course will increase the cost for the circuit considerably.

Boncuk

 

[Captain Colorado]

Boncok,
thanks for all your help.
I should say that the value of 10mV will probably never be used.
the settings for the system when tuned to the sound masking specification for the job will almost always(98% of the time) be 100mV or higher.
I expect that normal settings will produce at least 200mV 80% of the time.
Can we add or change the value of a cap to hold 200mV for 1-2 seconds?
will the top end of 11V cause any problems for the circuit?
CC

P.S. - the transformer will be external to the gang boxes and one transformer will supply power for all the circuits. No need to place the transformer on the board.

 

[Boncuk]

Hi Captain Colorado,

the ACS712 is designed to measure AC and DC currents of ±5A. 100mV at its input won't do any harm to the chip since it can also measure mains currents.

Additionally the minimum isolation voltage is 2.1KV(RMS) between pins 1 to 4 and 5 to 8.

With the speakers almost supplied with maximum power the chip should generate a usable output voltage.

C3 (4.7µF) together with R4 (1MΩ) should hold the signal for one to two seconds. If the time period is too short you might use 2.2 or 4.7MΩ for R4.

This design is purely SMD and also purely single sided.

The supply voltage should be well filtered and stabilized for +5V.

PBC dimensions are: 1.39X0.87"

If you wish to use this design please PM me your email address for the Eagle files and a BOM.

Boncuk