[linuxworks]

here's a UI question that I'm trying to work out-

I am determined to get a good clean implementation of a motorized volume control on my 'rackmount' headphones amp. its all DIY so I can design the user interface any way I want.

I've been using a standard IR remote and allocating 2 buttons for volume up and down, as you'd expect. it works ok but I'm starting to think that a dual speed up/down style might be better. 2 more buttons that move up and down but run at 2x the speed or something like that.

has anyone done that before? liked it or found it overkill?

when I use a '12v motor' pot and send 5v to it it runs kind of slow - BUT that's a good thing for small steps or 'nudges' when you want to move 'a bit' up or down. otoh, if you are watching a movie and a commercial comes on, you *know* it will be 3-6db louder than the main program (sigh) and so you *want* a really fast 'shuttle' on the volume knob. right?

I'd like to see comments on this topic. is it worth doing a dual speed pot implementation. how would you do it (using standard h-bridge drivers)? would you use 2 levels of constant DC or some kind of variable PWM?

and also, is this really a good use of 2 extra buttons on a remote? I do have 14 buttons I can use but ultimately I plan to copy the IR codes over to my learning remote and that will be the one that will have to send out as fast-up, slow-up, fast-down or slow-down button press; and I'll have to find ways on the universal remote to map these 'extra' 2 buttons, somehow.

is it worth it? I guess that's the first question

 

[error401]

I'd definitely use PWM for this, I don't know about the specific motors in the pots you've got, but many DC motors aren't too happy running at much lower than rated voltage, they make nasty noises, draw too much current etc. It's probably easier to implement though if you want to avoid micros, though it is possible to do PWM without a microcontroller, especially if you only want a couple of duty cycles.

I've personally tried 'acceleration' based schemes and I've never found them comfortable. I always end up fighting the acceleration one way or another - it starts accelerating just before I find the point I want, or I have to sit there holding the button too long waiting for it to speed up. It could just be a matter of tuning the algorithm or using some sort of log-based scheme, but I've never been satisfied with this. I have personally found that I'm okay with fine adjustment and an 'attenuate' button that drops it quickly down 24dB+ and restores back to the original setting afterwards.

Random thought: I wonder how hard it would be to interface a rotary encoder over a standard RC5 (or whatever) IR carrier...

 

[linuxworks]

please see discussion at:

http://www.head-fi.org/forums/f6/nee...45/index2.html

I am proposing a circuit that is based on the regular old h-bridge and uses 2 voltage rails (dc) to get the 2 speeds. do you think it would work?