Digitally controlled analog attenuator

[Wavebourn]

My friend is working on development of remote controlled step attenuator. It is a thingy (either kit or complete box with input and output sockets) that has a ladder-type digitally controlled attenuator. Resistors in the ladder are switched by relays, in 2 dB steps.
Currently, he has prototypes with both rotary switch and pot controls. Pot control works as a A/D converted that senses input DC adjusting attenuation digitally.
One PCB may be used either for balanced IN controlling (4 balanced ins / 1 out), or for 2 stereo channels (4 stereo ins / 1 stereo out).


The question is, what interface is better for such a box to control remotely? I am thinking either of 9-wire control (5 for attenuator, and 4 for input selector relays)

Also, since any resistance is possible, which ones are better to choose?

Pics of what he currently has are attached.

Anatoliy

[SpreadSpectrum]

To control the attenuator remotely? An IR receiver, of course.

Well, you probably mean on the amplifier panel. Why not use the rotary switches with two-bit encoding, you just sense direction and count pulses? Why add so many wires? My vote: two bits for attenuator, two bits for input selector. Or if you want to put input at absolute positions on the panel (otherwise you would need a display to tell which is selected) you could go three-bit grey code for up to 8 inputs.

Is he planning on doing a display to tell what is going on or using single-turn switches/pots?

[Wavebourn]

Currently he supplies options with potentiometer control and 11 - position rotary switch for volume, and 4 position rotary switch for channel selection. A 25 - position rotary switch may be better, but rotary switches with encoding, of course, is the best choose!
Are 5 - bit encoders available? It would allow 2 dB control all over the range!

But the initial question was about remote control interface. For example, 8 - 16 - or 64 inputs on a stage digitally controlled from console... What do you think?

Edit: it is an early morning there in Israel, but here are files he sent me ye http://wavebourn.com/maat/MAT_Assembly.pdf sterday:

http://wavebourn.com/maat/Mattrix_At...plications.pdf

[SpreadSpectrum]

If absolute position is not important, the two-bit quadrature output rotary encoders would be best in my opinion. You could get them without detents for very smooth operation. You could have as many positions as you want, it would just function as a multi-turn pot just like modern commercial receivers/amps. In fact, I'm sure that's what they use.

You would need a display to know what it is set at, though. Could be as simple as a strip of LEDs.

[SpreadSpectrum]

Ahhhh.... I thought you meant remote control like from the couch to the TV.

[SpreadSpectrum]

Wavebourn,

I don't know much about being a sound guy at a concert, but if I were doing this myself all of the information over long distances would be handled by a single TX and RX lines between two microprocessors. If you are going long distances this could easily be handled by a low-cost fiber line. That would be easy to set up and cheap to replace if broken.

[Wavebourn]

Thank you!

There is already thingy on the market called Digital Snake, a local guy Barani from Networksound.com makes them. My idea was, to use his digital snake in my powered mixer, and add remote control to switch inputs (TRS/XLR, mic/instr, signal polarity, phantom power, HPF) and set gain remotely using the same digital snake that works over twisted pair.
Currently I have a box on stage with 8 XLR ins, 8 TRS ins, switches and pors, tube preamps with input and output transformers, and this thingy connects via 25 pin socket to the console.

[Pano]

The Oregon Triode Society has a nice attenuator project going with relays. I believe it can be remote controlled. I'll look into it.

[wa2ise]

This may not really be applicable to the task at hand, but he might consider using dual control pentodes to vary the volume of the audio signal. Changing the DC bias on the G3 will change the amount of signal the plate will get (G2 gets what the plate doesn't get). This DC bias could be a low impedance (to avoid noise pickup over a long control cable), or DC varied by a digital volume control chip (here the chip doesn't touch the audio signal)..

More at 6BE6 volume control

[Wavebourn]

It is totally different: either to use non-linear devices to control volume, or to use linear stepping attenuator. Stepping attenuator allows to get minimal possible distortions, and better possible tracking between channels. The lower is tempco of resistors, the more precise they are, the better is the result. However, complexity is the price we pay for it, but using modern technology it is possible to get it relatively cheap.