Help please?: some leads on how to design this electronic volume control?

Anyone able to help? The plan is an audio volume control using 40 relay-operated L-pad attenuators per stereo channel. Control is passed in succession to the relay array elements from an IR remote control circuit or the amplifier touch control screen; there is no mechanical control as such - no "knob". Why do this? Pure resistor networks sound sonically far superior in a high-end stereo than even the best motorised Alps pot, and a universe better than a solid state pot. The relay and resistor networks are not a problem, nor are the remote control and touch screen circuits; we know how to do these. But the successive control of the relays is a problem. Microprocessor? Or a circuit like this ---http://www.technologystudent.com/elec1/count1.htm --- using a 555 timer and 4017B ? The need and the trouble is: the circuit needs to move up/ down through digits 0-39 providing successive outputs. Each of these switches on a separate relay, thus giving an incremental or decremental attenuation according to the volume "up" or "down" direction. Around a 20 second period is needed to scan the array of relays (i.e going from max. to min. or vice versa), and the control needs to store the "position" when switched off, so the selected relay stays on. And when the control system is switched on again - either to adjust the volume up or down - it needs to "know" which the currently selected position is to pick up from. Hope that's explained OK. Anyone know how best to design the control circuit? Just some rough guidelines and/ or a pointer to the best ICs or m/processor would be fantastic. Thanks.

First - why 40 relays? - if you can accept parallel or series combinations then surely a binary selection is better - needing only 6 relays for 40 (or up to 64) settings.

Secondly - is this really a 'quieter' system - the associated contact noise, relay spikes (back emf - suppressed or not) floating around, and large geometry (more aerials to pick up ground loops etc.) - surely make this susceptible

Finally - anything which has to 'know' (your words) inpllies a microprocessor - it will certainly give the flexibility you need - provided the clock noise is acceptable! (just clock it real sloooooow).

Hi Jeff, and thanks for your reply. 1st: we've done the subjective tests using a few "positions" each with its own L-pad and relay, and the results are well worth pursuing. Contact noise is not really a worry in practice - it has certainly proved, with the relays we've selected, much better than conventional pot wiper distortion by some margin, even a very good quality pot like the Alps. My partner is darned clever and has made the circuit thus far extremely quiet and as long as the control circuit can follow suit, I'm not too worried. Also, the unit is to be in a screened case with hard-wired OFC leads to numerous points, minimising the degrading effects of long copper tracks. Careful attention has already been paid to capacitance and hum/ noise, and this has paid off. Finally, series networks most likely will increase noise, and the topology has in any case to use 2-pole relays, one per position, for safety reasons: each series resistor (actually, there are only 4 values), goes through one contact to the grounding resistor (39 of those, max. vol. being no resistor - ok, no relay needed for that), then back through the other contact to the output rail. Thus if either relay contact fails, the signal never gets to the amplifier without its grounding resistor; thus a relay failure cannot suddenly cause the amplifier to go to full volume which could prove expensive when it comes to loudspeaker drive units. However, I take your point that we are probably talking microprocessor control. Just as long as it can be switched off when not being "asked" to move the volume up or down. That would minimise noise when the user has selected the desired setting. But that's exactly where we lack any expertise: neither of us knows where to begin with microprocessors in achieving the control we want. So back to the initial question!!! Any guidance there gratefully received. Thanks again.

Quentin, before you invent your own IR interface you might see what quasi-standards are out there.

http://www.innotech.com/primer1.pdf

http://en.wikipedia.org/wiki/Remote_control

If you chose an existing "standard" you can use the dedicated ICs designed for it as well as an inexpensive off-the-shelf IR unit.

You should be able to round up a consultant to design the digital interface for your project.

Thanks Quentin, all your coments taken on board and I bow to you and your partner's knowledge of the analog signal path and knowing what you want. I also echo thet whilst there will be contacts moving and processor activity whilst changing volume, any digital circuitry can be idle whlie waiting for an input.

I agree that there are dedicated IR receiver chip sets, but there may be issues regarding 'wake up' if you want them to remain unclocked. In my opinion using a low power fully CMOS microcontroller (like the PIC 18F series, or the Cypress PSoC parts - don't use a baby part unless you're counting pennies - they'll just be harder to program the way you want) would be a more secure (you are in control of what you are getting) route.

I know that the Vishay series IR receiver series devices could be used to trigger a level based response from a micro - which could be static (no clocks) otherwise - although my own experience of detecting IR does offer a warning. Whilst the IR receivers contain filters for 'out of band' IR there is often some noise in band - which would cause the controller to switch on to receive (and then reject of course) the unwanted interference. In other words, whilst using the telly remote in the same room won't activate or change your volume control - it will wake up your micro - so it might not be 'parked' in a quiet mode all the time.

At the end of the day - it depends how much importance is attached to a low noise detector - in which case a non-standard 'home made' sloooooow protocol will have many benefits.

Cheers,

Jeff

i know little (as my previous posts show ) but a lack of knowledge allows me to think outside the norm LOL
but to drive the relays, something like a few lm3194 led bar driver ics in series in dot mode, instead of driving an LED, a transistor to control the relay.
the input is voltage controlled via a simple digital pot voltage divider circuit with up and down control done via a 4093 ic.like below
http://electroschematics.com/wp-content/uploads/2008/12/digital-potentiometer-schematic.jpg
quite how you connect to IR is beyond my knowledge, but should be quite easy to achieve

OK, all I can offer are seeds of ideas 'cuz I don't have time to work out all of the logic, but here goes:

The basic premise is to use CMOS and a backup battery or super-cap to hold the state when powered down.

The first diagram shows using serial to parallel shift registers to move a bit with each button press. Only one direction is shown -- you would have to work out the logic to make it go both ways and to make it stop at one end or the other.

The second diagram shows using synchronous counters (because ripple counters would cause noisy glitches) and 1-of-16 decoders. Again, only one direction shown and stopping at each end not worked out.

But, hopefully this primes the creative juices

Hi David, and thanks for that. We already have our own IR system running, but it may be we need to revisit it. It was fine for running an Alps motorised pot up and down, but may need to be designed again from scratch for this more complex way of doing things. Will take on board your comments if this proves to be the case.

Hi Sarah. Definitely worth looking at this - will discuss the approach and let you know in a while how it goes. Very much along our lines of thought, but your comment has the added advantage of giving us some IC numbers to look at. Thank you.

Another possible way forward - thanks. I'm very grateful to all who have responded to this query and will let you all know the final route we take.

Other possibilities:

* 74AC299
* CD40104B