Voltage Controlled auxiliary input attenuator

Thread Starter

h2opolo

Joined Aug 2, 2013
38
I am trying to attenuate the stereo signal in an RCA cable. I would like it to do exactly what this product does:

http://www.amazon.com/gp/product/B0...1-99ac-45e5-8c23-e9e0811a2b2c&pf_rd_i=desktop

However, instead of using a knob, I would like to control it from a 0 - 5 volt signal. 0v translating to a gain of 0, and 5 volts translating to a gain of 1 or something similar to that.

I found this attenuator but I had a hard time understanding the data sheet. http://www.farnell.com/datasheets/1846875.pdf Is this what I am looking for?
 

crutschow

Joined Mar 14, 2008
34,285
That digital attenuator requires a serial digital input to control the attenuation, as from a microprocessor.
Here's a voltage controlled variable gain circuit.
And here's one that can be controlled from two push buttons.
I believe most (if not all) of the variable attenuator chips are controlled with some type of digital signal, not a voltage.
 

AnalogKid

Joined Aug 1, 2013
10,987
Besides Wally's circuit, there are Voltage-Controlled Amplifier (VCA) chips, although most of them prefer bipolar power supplies. If by "gain of 0" you mean off (no signal output), how off is off? Another approach is a simple 2-resistor attenuator, where one of the resistors is a linear or LDR (light-dependent resistor) optocoupler. Basically a cadmium-sulfide LDR and an LED in a light-tight package. Vary the LED current and the output AC changes. Full on is slightly attenuated, full off still lets some signal through, and two cells won't track perfectly throughout their attenuation ranges, but you can't beat it for simplicity.

ak
 

Thread Starter

h2opolo

Joined Aug 2, 2013
38
Thank you for the replies. Crutschow, I wouldn't be opposed to using the linear gain controlled amplifier that you posted but I don't quite understand it.

AnalogKid, I did find one of the VCA chips you are talking about. http://www.ti.com/lit/ds/symlink/vca610.pdf Is there a chip that can create a bipolar power supply for this?

And to answer your question... the gain doesn't have to be completely zero. If it attenuates the signal to about 10% (volume level), that is great. I like the idea of the VCA chip. Would it be hard to implement the bipolar power supply? If this is too difficult, I would also consider using the optocoupler idea. I attempted to design a circuit for it. Would something like this work? (I attached the PDF)
 

Attachments

crutschow

Joined Mar 14, 2008
34,285
The VCA610 is obsolete. It's been replaced by the VCA810.
This device can be used with a signal supply and a mid-supply ground reference voltage (see Section 10, page 31 of the data sheet).
The device is rather expensive however. :(

Here are some other variable gain amps that may work for you also, some of which are significantly cheaper.
 

Thread Starter

h2opolo

Joined Aug 2, 2013
38
Crutschow, I am looking for whatever produces the best sound quality because it will be passed through an amplifier and played out of speakers. I don't mind paying extra if the VCA810 is the best product for what I need. Is there something you recommend based on the specifications of all these products?

Also did you happen to view the optocoupler circuit I attempted to design in my last post? Is that a viable option?
 

crutschow

Joined Mar 14, 2008
34,285
The opto coupler circuit should work, but you need the LEDs to be tightly and equally coupled to the respective LDR or the two attenuations will not track.

Also the LDR likely won't work well in series with the signal. You need to use it in a voltage divider (attenuator) circuit.

I have no particular recommendations for those listed by TI.
Take a look and see which ones seem most appropriate to you.
 

Thread Starter

h2opolo

Joined Aug 2, 2013
38
I made an attempt to try to figure out the VCA810 circuit. I have access to +12V so I am trying to make it work for that. I don't know if I am looking at it correctly, but I am here to learn and any criticism is welcomed. I am also confused at what to use for the negative output to the rca cable. I attached a pdf of what I have so far. Will someone please let me know if I am heading in the right direction?
 

Attachments

Thread Starter

h2opolo

Joined Aug 2, 2013
38
There is one last thing I am still concerned about. The same voltage source that I will power the VCA810 chip with, powers an amplifier that the output of the VCA810 will feed signal to. Will this mean the input to the amp will have a DC offset of 6 volts? I feel like that will be very bad for the amp. Will I need to run the output of the VCA through a highpass filter to eliminate the DC offset? I attached a graph of the voltage signal that I am concerned about.
 

Attachments

crutschow

Joined Mar 14, 2008
34,285
There is one last thing I am still concerned about. The same voltage source that I will power the VCA810 chip with, powers an amplifier that the output of the VCA810 will feed signal to. Will this mean the input to the amp will have a DC offset of 6 volts? I feel like that will be very bad for the amp. Will I need to run the output of the VCA through a highpass filter to eliminate the DC offset? I attached a graph of the voltage signal that I am concerned about.
Yes, you are correct. You may need to capacitively couple the circuit to eliminate the offset.
But if the amp you are feeding operates from the same supply then it may already be capacitively coupled at its input.
If so, you wouldn't have to add another capacitor.
 

AnalogKid

Joined Aug 1, 2013
10,987
THAT makes the very best VCA's and audio compressors, but their parts have...personality. That is, it takes a lot of attention to detail to get them to meet their reputation. Not what I would call a beginner's part.

ak
 

Thread Starter

h2opolo

Joined Aug 2, 2013
38
After searching some more, I decided I would go with something actually made for audio. I found this one. http://www.ti.com/lit/ds/symlink/lm4811.pdf In the data sheet it says: voltage supply 2V-5.5V.. Does this mean I can give 5 volts to the positive rail and 0 volts to ground? I am assuming 12V is too much and I need a regulator to bring it down to 5... Can someone explain to me why this chip doesn't need a negative voltage? Why wouldn't the negative voltage saturate at 0V (gnd)?
 

crutschow

Joined Mar 14, 2008
34,285
After searching some more, I decided I would go with something actually made for audio. I found this one. http://www.ti.com/lit/ds/symlink/lm4811.pdf In the data sheet it says: voltage supply 2V-5.5V.. Does this mean I can give 5 volts to the positive rail and 0 volts to ground? I am assuming 12V is too much and I need a regulator to bring it down to 5... Can someone explain to me why this chip doesn't need a negative voltage? Why wouldn't the negative voltage saturate at 0V (gnd)?
Because the internal circuits are likely biased at ½ the supply voltage and the inputs and outputs are capacitively coupled to block this DC.
 

AnalogKid

Joined Aug 1, 2013
10,987
Yes, 5 V only.
Yes, 12 v will kill it.
Looking at the internal diagram on page 2, the chip generates an internal reference voltage that is midway between the external power rails. Both amplifiers use this internal reference, so while the chip has 0 V and +5 V externally, internally it thinks it has +/-2.5 V. Coupling capacitors on all inputs and outputs justify this with the outside world.

This is not a simple part to control. Instead of a knob to turn, you need two pushbuttons and a clock circuit.

ak
 

Thread Starter

h2opolo

Joined Aug 2, 2013
38
I tried wiring the circuit today. I only wired one channel and left the input and output to the second un-wired. I intentionally grounded the up/down pin and the clock pin. My reasoning was that if they are grounded the gain should not change. In the data sheet, it says that on startup the default gain is 0db. So I figured that the signal coming from the ipod should be the same at the output. This was not the case though. I wired the output to my headphones.. The signal was at full volume on the ipod but the volume level in the headphones was very low. I turned off the power supply and it didn't even make a difference. It's like the chip wasn't even doing anything. I tried using another chip and got the same result. Does my circuit look correct? Can anyone find any issues? Is the impedance of the headphones too low? I attached a pdf.
 

Attachments

Top