Alternatives to LDR for audio signal.

[makaze]

Hello. I am hoping to be able to pan audio across four outputs. from my limited experience with electronics the only method I think may work is to use 4 LDR's to act as gates triggered by timed LED's.

If this is the only way, I don't have the knowledge to build it yet and I'm unsure what chip would turn LED's on or off. Any input and information appreciated.

 

[KrisBlueNZ]

Any kind of VCA (voltage-controlled amplifier) will allow you to vary the amplitude of an audio signal using a control voltage. Do a Google search. A common way to make a VCA is using an OTA, operational transconductance amplifier, such as the LM13700, which contains two of them. Google that as well.

Regarding generating the control voltages for the VCAs, this depends on what you want the signals to do. Saying that you want to "pan audio across four outputs" doesn't explain much. Describe exactly what you want to happen to the signals at the four outputs, how you want to trigger the circuit, and so on.

 

[makaze]

Thanks for the reply Kris.

I have a quadraphonic amplifier, it has 4 inputs (mono channels), each input goes to a separate (within the unit) power amp section which feeds a separate speaker. Each channel has tone and volume potentiometers.

With a speaker in each corner of a room, I want an audio source (for example a tape of footsteps) to be able to pan round the room at a variable speed, from no speed to high speed. I would like to control the speed of the pan from a potentiometer.

I suppose it is an audio version of a light chaser.

 

[davenn]

The obvious and much more compact and reliable alternative to a LED and LDR would be an opto coupler

Dave

 

[KrisBlueNZ]

You can't use an optocoupler for variable audio attenuation.

 

[davenn]

You can't use an optocoupler for variable audio attenuation.

OK
reason being?


Dave

 

[KrisBlueNZ]

You need something that is (a) resistive, not a simple semiconductor, so it doesn't make a huge distorted mess of the audio signal, and (b) smoothly and continuously variable over a fairly wide range of resistance, so the attenuation can be smoothly controlled.

Using an LED and a LDR, the LDR is connected as one half of a voltage divider. For example, the top resistor. The audio is passed through this voltage divider. When the LED is brightly lit, the LDR has a relatively low resistance and most of the signal passes through, so there is little attenuation. When the LED is dimmer, the increased resistance of the LDR causes more attenuation. When the LED is not lit, the LDR has a very high resistance, and almost no signal gets through the voltage divider. At all points in between those extremes, the attenuation varies smoothly, and the resistive nature of the LDR ensures that there is very little distortion.

Nothing like that can be done with an optocoupler; even a linear optocoupler is only used to transfer an analogue quantity across an isolation barrier and can't be used like that.

 

[KrisBlueNZ]

To the OP. As I said, you will need four VCAs (voltage controlled amplifiers). Google up some schematics. I would recommend using operational transconductance amplifiers for these, as it's fairly simple. They would all be driven from the same signal source, and each one would drive a different channel of your quadrophonic amplifier.

Generating the control signal is a bit more difficult I think.

I guess you only want to pan the sound between pairs of speakers - for example, start at front left and pan smoothly to front right, then pan smoothly to rear right, then pan smoothly to rear left, then pan smoothly to front left again and repeat.

In that case, no more than two channels are ever active at the same time, and you could use that fact to simplify the design. If you want the sound to move in a circular pattern, then three channels need to be active simultaneously, and that would complicate the design significantly I think.

With the two channel limit, you just need a circuit that will generate two output voltages (to drive two VCAs), that represent left and right volume values, that change over time to produce the effect of the signal moving smoothly from left to right and back. You then need some logic to switch each of those control voltages between two different VCAs at the appropriate times.

Creating the effect of a signal source moving smoothly from one speaker to another without changing its apparent volume is not as simple as it sounds, because of how the ear percieves volume, especially when the signal is coming from two separate sources. There are several approaches to panning, and I'm not well-versed on the subject. I suggest you Google some keywords like smooth pan audio left to right constant perceived volume. You may be able to find circuits that will do this.

You also need to make sure that the VCA's response to the control signals follows the correct "law". I believe an OTA-based VCA's voltage gain is directly proportional to the control current. So you need to find a pan circuit that generates control voltages (voltages can be converted into currents using op-amps) that are designed to drive an amplifier with linearly controlled voltage gain.

If you find some VCA and pan control circuit designs you like, post the links here and I'll try to suggest how you can connect them together.

 

[davenn]

thanks Kris

and congrats on hitting the 2000 posts

Dave

 

[makaze]

Kris googled (amongst many others) ota panning and this first result is a PDF from a site called geofex. I have built the few circuits I have built from other stuff on his site.

Is the circuit at the top of the pdf any use?

It shows down the page a circuit for panning to two outputs that he states is similar to a boss panning pedal, which is a respected panning pedal.

I am happy with the simpler two channel panning (as opposed to a circular three channel, which as you commented, would be more difficult).

 

[KrisBlueNZ]

Thanks Dave! I hadn't noticed!

makaze, I don't see any links in your post. Anyway I can't look into it now. It will be a day or so before I check in again.

 

[makaze]

Yea there is no link, the google search just brings up a pdf link which I can't copy. No rush.

 

[KrisBlueNZ]

OK. The "Application 3: Pan Pedal" design would work, but I don't think the panning will be very smooth. In other words the sound will appear to move at an uneven speed, and will appear to change in volume as well. See if you can find a circuit that's designed to pan smoothly, and with constant volume.

You can generate the control signal for it using a very common design: a two-op-amp integrator/schmitt-trigger oscillator. Add a third op-amp to invert one signal, and you're nearly there. Some kind of DC shifting and voltage-to-current conversion will be needed between the oscillator and the OTAs. Two flip-flops clocked from the square wave output of the oscillator would be used to switch the control currents between the OTAs so that two control currents (nominally Left volume and Right volume) can feed four OTAs for panning across four corners.

So see if you can find a better design for OTA-based panning. It will require some research, I think, and I won't have time to do that for a few days.

 

[CDRIVE]

VCA is using an OTA, operational transconductance amplifier, such as the LM13700, which contains two of them. Google that as well.

Too bad I wasn't familiar with this amp a week or two ago when I was answering an AM Demodulator topic. To simulate a good AM signal requires more than two generators wired in series or parallel. I ended up creating a macro that uses an LDR on the (-) input to Gnd pin of an OpAmp. The modulation gen was fed into an LED with quiescent forward bias. It did create a perfect modulation envelope though.

thanks Kris

and congrats on hitting the 2000 posts

Dave

Hey, I didn't get one of those! Maybe it got lost in all the happy birthdays that happened about the same time. Actually, this is rather weird. What's the odds of 2000 peddled miles and 2000 posts occurring simultaneously?

Chris

 

[KrisBlueNZ]

Yes an LDR is a good solution for amplitude modulation or variable gain. They have a fairly wide resistance range too. They have a pretty slow response though, so I don't think they'd be suitable for modulating a carrier with an audio signal. You would lose any high frequencies in the audio. After demodulation, your music would sound like "oontss, oontss, oontss" but without the "tss"

Belated congratulations on hitting 2000 posts Also, what do you mean that you created a "macro"? Do you mean a building block or circuit module?

 

[CDRIVE]

Belated congratulations on hitting 2000 posts Also, what do you mean that you created a "macro"? Do you mean a building block or circuit module?

Yeah, Tina lets be build a complete sub-circuit, save it as a macro and represent it by my choice of component symbol, as long as the pin-out agrees with the symbol. 500Hz to 1KHz sim'd very well.

Thanks for the 2K point.

Chis

 

[CDRIVE]

Belated congratulations on hitting 2000 posts Also, what do you mean that you created a "macro"? Do you mean a building block or circuit module?

Tina lets me build a sub-circuit, represent it with a symbol of my choice, as long as the pin-out agrees and save it as a macro file. My modulation frequency was 500Hz to 1KHz and it worked out very well.

Thanks for the 2K heads up!

Chris

Edit: This double posted because Marilynne distracted me. That's my story and I'm sticking to it!

 

[KrisBlueNZ]

makaze, I'm afraid I'm going to have to drop out of this project.

I haven't been able to find, or design, a suitable control circuit for the variable-gain amplifiers.

Smooth, constant-volume panning requires half-cycles of sinewave to control the gains of the amplifiers. Sinewaves can be generated from triangle waves using a couple of methods (see National Semiconductor application note AN-263) but panning requires a quadrature signal, i.e. sine and cosine.

There are quadrature oscillator circuits around, but they don't lend themselves to wide variation in frequency, nor to being stopped completely.

I hope you'll be able to find someone who can design a circuit for you on an electronics forum somewhere.

Regards

KrisBlueNZ - Kris in Wellington, New Zealand