I have an application where I need to devise a method to add Three LED indicators to a speaker to determine the Gross amount of power being fed to the speaker. This is an industrial application where they use large PA speakers to literally shake devices.
The problem is that the amps are in one remote location and the control room is in another location. Access to the speakers is limited so the indicators must run purely from the speaker lines. the speakers are viewable by camera remote control.
I need to figure a way to light a Green LED's at 1 watt, and then some division of power there after. something like 200 or 250 watts triggers a yellow LED and then a Red LED at 500 watts which is the program rating of the speakers used. something like that.
But, whatever circuit is devised cant affect the audio signal being fed to the speaker.
I was thinking of a resistor and a Zener that would feed the LED and current limiting resistor. Just duplicate that 3 times with zeners/resistors calculated for the various power levels.
BUT, when the diodes conduct. wont that effect that audio signal? or would the effects be negligible???
Zc
The problem is that the amps are in one remote location and the control room is in another location. Access to the speakers is limited so the indicators must run purely from the speaker lines. the speakers are viewable by camera remote control.
I need to figure a way to light a Green LED's at 1 watt, and then some division of power there after. something like 200 or 250 watts triggers a yellow LED and then a Red LED at 500 watts which is the program rating of the speakers used. something like that.
But, whatever circuit is devised cant affect the audio signal being fed to the speaker.
I was thinking of a resistor and a Zener that would feed the LED and current limiting resistor. Just duplicate that 3 times with zeners/resistors calculated for the various power levels.
BUT, when the diodes conduct. wont that effect that audio signal? or would the effects be negligible???
Zc
Thanks, but the LM chips require power. In this application there is no power available and batteries are impractical. The circuit needs to run entirely from the speaker line.
I thought about using a diode bridge across the speaker line and then take the DC from the bridge and use that to feed the three LED's with current limiting resistors just calculated for the thee voltage levels. But would the bridge effect that speaker signal???
They dont seem to effect the output of a transformer...but what about at higher frequencies??
Zc
I thought about using a diode bridge across the speaker line and then take the DC from the bridge and use that to feed the three LED's with current limiting resistors just calculated for the thee voltage levels. But would the bridge effect that speaker signal???
They dont seem to effect the output of a transformer...but what about at higher frequencies??
Zc
I cant go into much detail about what they are doing. they wont even tell me all of it. But. No other AC or DC power lines are allowed in the room during testing. and the cables that do extend into the room now were installed as part of the structure with extreme measures taken for shielding etc. Lets just say this place is making things you or I cant buy (legally)nor could we ever afford. and they basically have one use and thats not a good one....
Anyway.
I once had a pair of PA system monitor speakers made by community sound that had Green "Signal present" LED's in them. If i still had the dang things i would just take them apart and see what they did.
But if they can put a LED in a speaker without having to resort to battery's or external power. then this should be an easy thing to come up with.
The client has given me a directive. find a way to make a level indicator with at least 3 steps that works without an external power source. speaker line only.
SO, the LM signal chips are great. But for this application they are out of the question. so lets talk about other methods.
Now if community put a signal present LED in there speaker. they had to have done it somehow.
Now if i used a diode bridge on the speaker line and rectified the AC coming in to DC, then the next problem becomes how to trigger the LED at 2.8V for the first step and how to limit the voltage to it so that it doesn't die from gross over voltage???
Anyway.
I once had a pair of PA system monitor speakers made by community sound that had Green "Signal present" LED's in them. If i still had the dang things i would just take them apart and see what they did.
But if they can put a LED in a speaker without having to resort to battery's or external power. then this should be an easy thing to come up with.
The client has given me a directive. find a way to make a level indicator with at least 3 steps that works without an external power source. speaker line only.
SO, the LM signal chips are great. But for this application they are out of the question. so lets talk about other methods.
Now if community put a signal present LED in there speaker. they had to have done it somehow.
Now if i used a diode bridge on the speaker line and rectified the AC coming in to DC, then the next problem becomes how to trigger the LED at 2.8V for the first step and how to limit the voltage to it so that it doesn't die from gross over voltage???
Use an analog meter
Get a 0 to 5V meter.
Run the amp output through a resistor.
R~= Vrms max the amp puts out * 0.9 / amps needed for full scale meter deflection.
Wattage = 4 X Vrms max^2 / Resistor value
Resistor goes to a 4 diode fullwave bridge (ultra fast diodes are first choice, I've used 1n4148s).
If buying a diode bridge with 4 diodes already in side it. The meter goes from + to - and the resistor goes to one AC input and the amp return goes to the other ac input.
Put a cap across the meter if desired.
10 years ago, Radio Shack would have had all the parts to build this.
Get a 0 to 5V meter.
Run the amp output through a resistor.
R~= Vrms max the amp puts out * 0.9 / amps needed for full scale meter deflection.
Wattage = 4 X Vrms max^2 / Resistor value
Resistor goes to a 4 diode fullwave bridge (ultra fast diodes are first choice, I've used 1n4148s).
If buying a diode bridge with 4 diodes already in side it. The meter goes from + to - and the resistor goes to one AC input and the amp return goes to the other ac input.
Put a cap across the meter if desired.
10 years ago, Radio Shack would have had all the parts to build this.
The people at Community were nice enough to supply me with a schematic of the speakers i used to own that had the signal present LED.
They are just using a Diode string and a resistor. if my math is correct and please correct it if it is wrong. each diode provides a .7v forward drop. so three of them in a string should be 2.1V. with the LED across those three diodes. the LED should start to conduct at 2.1V correct??
But there is the forth Diode that also has a .7v drop. I am assuming it i wired as such to also protect the LED from high reverse voltages. so this really should be a forward voltage drop of 2.8V before the LED conducts correct???
This would make sense as that is 1 watt at 8 ohms.
Now if the diodes are just providing a voltage drop. couldnt i use a reversed Zener rated at whatever voltage i want in place of the lower three diodes??
63vrms at 8 ohms = 500 watts, the max program rating of the speaker.
couldn't i use a Zener as close to 56v as i can find. as the lower diodes with everything else the same? i bet i would have to change the value of the resistor wouldn't i?
Zc
They are just using a Diode string and a resistor. if my math is correct and please correct it if it is wrong. each diode provides a .7v forward drop. so three of them in a string should be 2.1V. with the LED across those three diodes. the LED should start to conduct at 2.1V correct??
But there is the forth Diode that also has a .7v drop. I am assuming it i wired as such to also protect the LED from high reverse voltages. so this really should be a forward voltage drop of 2.8V before the LED conducts correct???
This would make sense as that is 1 watt at 8 ohms.
Now if the diodes are just providing a voltage drop. couldnt i use a reversed Zener rated at whatever voltage i want in place of the lower three diodes??
63vrms at 8 ohms = 500 watts, the max program rating of the speaker.
couldn't i use a Zener as close to 56v as i can find. as the lower diodes with everything else the same? i bet i would have to change the value of the resistor wouldn't i?
Zc
An externally hosted image should be here but it was not working when we last tested it.
acid_k2 said:At 63Vrms, the 220 ohm resistor have to dissipate about 18 W
and:
are you sure that your 0.28A current (@ 63V) flows trough the diodes only, without burning your LED? I'm not.
In my opinion, for the 1W-LED you need a more efficient current regulator...
I did some playing around with the breadboard last night. the circuit i have pictured is from a set of PA speakers i once owned. thats exactly how the factory has there LED wired up.
I duplicated the setup on the breadboard as shown with the exception being i only had a 2 watt 220 ohm resistor on hand. and it works of course. the LED seems to come on around 2-3 volts of amplifier output. and the voltage at the LED does in fact get quite high! with the amp output approaching 35V the LED is seeing close to 3V and that 220 ohm resistor was getting quite hot.
and i did play around with some 12V Zeners i had on hand and i now understand a little bit better of what the problems are.
it looks like i am going to have to try and figure out something more complicated. using some small transistors and zeners as constant current sources to turn on and hold the LED's at certain voltage levels and at a constant current as the voltage continues to rise past the turn on point of the LED.
Back to the drawing board......anyone want to help?
Zc
What they will tell me is that most of the time, the audio is a complex full bandwidth signal. they wont give me specifics but they said its like white noise. they also do some band limited and some specific single frequencies.
They understand how the speakers work and know they cant do full power at the higher frequencies.
What they want. is a Green LED at 1 watt just to show there is signal there.
The next two LED's get more complicated. The speakers are rated for 250 watts continuous Pink Noise. 500 watts program and 1000 watts peak.
What they have asked for, is a Red LED at 500 watts RMS and a Yellow LED at -6dB from 500 watts.
If my math is correct. then 500 watts @ 8 ohms = roughly 63V and -6dB = 125 watts or 31.5V
SO, I need a Green LED at 2.8V then a Yellow LED at 31.5V and a Red at 63V. But these are RMS voltages. i "Think" i need to figure out the peak voltages for the LED to light at the correct RMS voltages? 3.92V for 1 watt 44.5V for -6dB and 89V for 500 watts.
and it needs to be fairly accurate as i see in the calculations that a 1 V difference at 63V can be a change of about 16 watts.
I am figuring i will probably need to add a Small cap across the Red LED to keep it on a little longer so that peaks trigger it and keep it on for about 1 second or so.
Zc
They understand how the speakers work and know they cant do full power at the higher frequencies.
What they want. is a Green LED at 1 watt just to show there is signal there.
The next two LED's get more complicated. The speakers are rated for 250 watts continuous Pink Noise. 500 watts program and 1000 watts peak.
What they have asked for, is a Red LED at 500 watts RMS and a Yellow LED at -6dB from 500 watts.
If my math is correct. then 500 watts @ 8 ohms = roughly 63V and -6dB = 125 watts or 31.5V
SO, I need a Green LED at 2.8V then a Yellow LED at 31.5V and a Red at 63V. But these are RMS voltages. i "Think" i need to figure out the peak voltages for the LED to light at the correct RMS voltages? 3.92V for 1 watt 44.5V for -6dB and 89V for 500 watts.
and it needs to be fairly accurate as i see in the calculations that a 1 V difference at 63V can be a change of about 16 watts.
I am figuring i will probably need to add a Small cap across the Red LED to keep it on a little longer so that peaks trigger it and keep it on for about 1 second or so.
Zc
The problem with using the voltages is that the peak to rms ratio will vary depending on the waveform. For sine waves the ratio is sqrt of 2, but for everything else, the ratio is different. White noise probably has a peak to rms ratio near 1.
How fast does the thing need to respond? If not too fast, you could use the signal to heat a resistance, and convert temperature rise into power. But if you can measure temperature, maybe it's better to measure the voice coil temperature and solve the problem directly.
Probably easiest to stay with voltages. Use a voltage comparator and switch the threshold voltage depending on the waveform in use. The system operator will have to be smart enough to switch the meter/lights when he changes the waveform.
I_F
How fast does the thing need to respond? If not too fast, you could use the signal to heat a resistance, and convert temperature rise into power. But if you can measure temperature, maybe it's better to measure the voice coil temperature and solve the problem directly.
Probably easiest to stay with voltages. Use a voltage comparator and switch the threshold voltage depending on the waveform in use. The system operator will have to be smart enough to switch the meter/lights when he changes the waveform.
I_F
this does not have to be real accurate. just needs to give some sort of visual indicator. the tech will have to use judgment from there.
If the tech sees no lights on. somethings wrong. if he looks and sees the red light light full, somethings either wrong or just way too high...
It doesn't need to be exact. just a gross indication.
Zc
If the tech sees no lights on. somethings wrong. if he looks and sees the red light light full, somethings either wrong or just way too high...
It doesn't need to be exact. just a gross indication.
Zc
Nordic said:
I like this one best:
http://www.redcircuits.com/Page102.htm
will build it and see if it smokes
Zc
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.