EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: atwoz on April 25, 2014, 04:31:22 pm
-
Hello everyone. I want to know what your opinion is about an idea I have on a method to measure (an approximation actually) the volume level of an Audio Amplifier without actually hacking the amplifier itself.
We need a way of detecting if the volume is modified (via the volume knob) and hopefully know a % of change with respect to the original setting. I can hack the input and output cables but not the amplifier itself and the amplifier model is unknown. I was thinking that if I read the input and output signals with an ADC and if I integrate the signal during a defined time and divide on by the other, I get a ratio. Do you guys think that this ratio could be used to know if the volume level changed and if it did, by what percentage?
I believe there are many other factors that I have to consider, like equalization, an who knows what else. I'm in the process of building a test circuit but it will take some time, I would love to hear what other people think about this before I even spend time building it.
Thanks!
-
Use plenty of low pass filtering on the output in order to handle class D amplifiers.
One trick that would be independent of the incoming signal content would be to inject a low level ultrasonic tone (something like 20kHz at 1% or less full scale) and specifically look for and measure that tone.
-
so, what you want is to know the current atten level of some black box amp via measurement?
is it really worth it? does not sound like a good solution. is this for a display, so you can see the direct dB level or something?
-
Mike: I like your idea, could be a backup plan if the first one doesn't work.
linux-works: I want to be able to know if someone modifies the volume of the amplifier, by monitoring remotely any changes and alerting the user. It is a business opportunity so its worth it, if I can manage to find a way.
Thanks for your comments!
-
is there any more direct way of doing what you want? not sure what kind of business could be made from guessing a black box's atten level, but hey .... ;)
if I had to do something, I'd just not use a black box amp and supply my own atten. (ob disc: I designed a few vol controls, relays and solid state as 2 engine choices).
guessing some atten level while its playing material is just not going to work. it sounds very hacky and sloppy, to be honest. maybe if you could give more info on what the real issue is. so far, it still does not make sense why anyone would want to do what you propose.
-
As long as there's some reasonable signal, it should work fine. I wrote software to control my HP audio analyzer and do basically the same thing for every test- measure the output, compare to input and calculate the gain. You just need a way to measure the levels over the same time period- the ratio is the gain.
-
but WHY?
you'd have to capture the input value and the output value, both quickly, and do the division. you'd also have to sample and average things and even still, if its a live signal and not some static tone, I doubt it would be very accurate. maybe accuracy is not needed, though, but I still can't understand why this would be a useful thing to do.
there are so many vol controls out there that DO allow you to directly set and get their values, just seems odd to try to get an indirect value from some black box.
you can use any amp you want, just leave its volume fixed and front-end it with a true programmable one. its really not hard or expensive and I bet its cheaper to just have your own vol control than to use 2 a/d stages, a cpu and all that hackery.
-
My idea after about 3 seconds of thought;
How about scaling the output down to the nominal input level and use a difference amplifier. You'll need to get the phasing correct so they subtract, not add.
In theory the difference should be minimal if the gain (of the amplifier from in to out) has not changed.
-
see, that won't work reliably, either. phasing can change even as freq changes. I would not count on a phase linear (lol) black box. that would assume too much.
so many reasons why this won't be a great idea. for simple test signals, you could get a close value but for moving data, like music, I really doubt you'll get this to work very well thru 'any unknown' preamp or amp box.
-
If both signals are sufficiently strong, you can simply rectify them and adc the dc signals. Otherwise, google precision diode.
-
Trivial with a bidirectional voltmeter, adjust R2 to "setpoint". Upgrade to precision rectifiers, and exact values left as an exercise for the op.
-
Well maybe I'm old school, but I would never use the term of output volume. You can of course measure output voltage as most audio amps are voltage amps using negative feedback. But if you actually mean to measure output power (in watts) without having the amp attached to a load, then no, I can't think of a way you can measure the output power of a amplifier without it be attached to an output load.
So help this old fart, what do you actually mean by measuring 'output volume' ?
Or perhaps you mean to measure the voltage gain of a unknown amp?
-
I thought the original request was clear enough:
We need a way of detecting if the volume is modified (via the volume knob) and hopefully know a % of change with respect to the original setting
-
Reading through the replies it seems others also had problems trying to figure out what the OP was really trying to measure, and if he was even reading right thing for whatever the goal was. Maybe just a sound level survey: http://www.microdaq.com/reed/sound-level-meter-datalogger.php?gclid=CMSDh-iQ_b0CFU4R7Aod5T4A5Q (http://www.microdaq.com/reed/sound-level-meter-datalogger.php?gclid=CMSDh-iQ_b0CFU4R7Aod5T4A5Q)
-
More about the reason behind my question:
We have a big customer, that has over 1000 units installed that have an a comercial amplifier (could be any of 4 possible types they commonly use). They play audio and other sounds but they have a problem, people change the volume setting manually on the amplifier (don't ask why, they do and it can't be stopped, think of it like there are ghosts or sneaky monkeys). The amplifiers are connected to several speakers, number, size and brand are also random.
I need a way to detect if this volume setting changes, and it's not possible to use some kind of sound level sensor since the actually sounds being played are also random, the physical space is always different and changing constantly and there are other source of noise, so it would be hard to get a setpoint.
It has to be monitored constantly, if the volume setting is changed, the user needs to be alerted immediately (in less than a minute). I am able to tap off (put some kind of terminal with my own circuitry) in the cables that go to and from the amplifier.
This all my sound dumb to some people but there is actually a big need for what I'm describing. I know there are some real pros in these forums that could provide awesome ideas and guide me the right way.
Thanks again!
-
So do what I did for some heat sealers with user settable controls, and disconnect them and provide an internal fixed resistor divider instead. That way you have a fixed sensitivity and the playing with the knob has absolutely no influence on the output. That solved the problem of the sealers being turned up to full power and then burning the insulating PTFE sheeting, and burning a hole through the plastic being sealed.
Others with a wide range of adjustment I added a set of padding resistors so that the control range was limited to a narrow window. That way you still have a very small adjustment range for the control and the output can still be fine tuned but it cannot be set too soft or too loud.
-
Oh, I forgot to add. We cannot modify in any way the amplifier (there are reasons). That's why I was planning to tap both the inputs and outputs since we can modify the cables that go in and out.
-
My thought...
Terms: Amplitude vs Volume.
Connections AC coupled to avoid any environmental issues / offsets.
Sample input and Output as discussed. Use micrpo tpo averae over a period of time - e.g. seconds, and compare the levels with a margin of error - say 5%.
Then fire your event to notify the user.
If it's stereo - you could also monitor phase in the same way - to either auto-switch, or also notify user.
-
p.s. You could also lpook for clipping or excessive level on the input.
I've done all these in monitoring video streams (for audio content/quality)
-
They play audio and other sounds but they have a problem, people change the volume setting manually on the amplifier (don't ask why, they do and it can't be stopped, think of it like there are ghosts or sneaky monkeys). The amplifiers are connected to several speakers, number, size and brand are also random.
If the volume has no need to be adjusted, then locking racks have been the standard solution:
http://www.sweetwater.com/images/items/750/SECL4-large.jpg (http://www.sweetwater.com/images/items/750/SECL4-large.jpg)
If they do need to have control, but not beyond a certain point you'll need to setup some user adjustable preamps or limiters ahead of the amp, and still lock down the final gain on the amplifier (with above solution).
This all my sound dumb to some people but there is actually a big need for what I'm describing. I know there are some real pros in these forums that could provide awesome ideas and guide me the right way.
There really isnt a need for something the way you describe it, when you keep coming back with more details about the requirements you only annoy the people who tried to lend a bit of helpful and free advice.
It is a business opportunity so its worth it, if I can manage to find a way.
Time to employ someone who knows what they are doing.
-
Depending on the type of amp, what about monitoring its current draw from the mains? Very simple and completely non intrusive.
I do 'get' what you are trying to achieve and why BTW
-
probably unlikely, but if its class-A amp, then the current draw is the same no matter if signal or no signal. current will be constant on class A.
-
Agreed. That's why I said depending on the amp.
-
Measuring the amplifier's power consumption is one of the more err... "interesting" ideas thrown out there.
Is this ultimately for a system to keep the sound level within bounds to avoid noise abatement issues?
-
a bandpass peak detector on both input and output in the 1Khz band where phase and gain are likely to be most stable?
sample both at the same time with an analog switch, then use a window comparator with the input sides gain adjusted to tell if its out of range,