thec

Hi. First post to this message board which looks promising! I'll make sure to bookark it, great work!

I'm thinking of building myself a mixer. But as a digital nerd I am (computer programming is my main interest) I'm not too familiar with the analog circuts. I did go a electric education in secondary upper school but can't remember most of it :-)

I'm building a digital mixer... there is a digital potentiometer out there which I'll be using for this purpose. Controlled by the computer (of course).

However, the basic mixing circut is:



Yes, this is a simple volume controller, with a 47k variable resistor (in my case a digital one but makes no difference). So far soo god.

When connecting a bunch of these and try to blend the signal (doh, building a mixer), we get into a problem. Have a look at the following image:



The output may need an amplifier I've been told, to make up for some loss of signal, I'm not sure what is meant by that, but I guess you could tell me, but that's not the real question.

The question of the week: The ? in the post. The blending. How is it done without signal A going in reverse to signal B,C or D? I mean, if A and B potentiometers are set to 50%, wouldn't like 25% of the signal from A go to B's input, and 25% of B blend into A's input?

A diod solution is of course out of the question, so my question is, how do I solve this and DO I solve it or does every output (CD,DVD,soundcard(PC) etc) have some kind of blocking of reverse audio? I wouldn't like the fact that signals are crossing my system all over.

So.. any ideas? Thanks for reading, now reply ;-)
//Albert "thec" Sandberg

mechie

See a posting called "advice needed on designing power supply" from 440uf.
He(she?) links to http://www.ffynnon16.freeserve.co.uk/schem.html (thanks 440uf ) I will pinch it while its still there :!:
This mixer avoids feeding back out of other inputs by using a 'virtual earth point' as the summing node. Effectively all the inputs think they are connected to ground (via all the input resistances), there is therefore no chance of cross-feed. 8)
Other mixer circuits do allow a percentage of one input to feed out of others as you suspected. :shock:

In your second circuit you have a box labelled [?]; the simplest option is to simply connect the pot wipers together so [?] is a terminal. BUT if one pot is set to minimum ? ... the output is shorted to ground. ops:
For this reason your [?] will be a resistor for each pot input, all connected to the common output. Common sense suggests a value equal to the pot's resistance. These resistors prevent the shorting out of the output but add considerably to the losses in the circuit; hence usually requiring an amplifier to boost the signal back up.
Your calculated 25% of one signal feeding back to another input is therefore reduced to something under 10% ???

Hope that helps (wordy clap-trap) :wink:

thec

Thanks a lot for the fast answer! .. The polyesters ( --||-- ) .. should do to stop reverse signals for AC (such as audio), but nevermind, now I see how others have done it, and it becomes quite clear (edited: no, actually not since they're not polarized I noticed later, but I have SOME truth in there... damnit, I lost track ... ).

I can't see the virtual earth point, he's using the same line for all grounds as I can see, but maybe you're reffering to something with the OP amplifier, which I'm not too familiar with (but I could check that up).

Sure, adding resistors to the outputs from each channel sure helps avoiding shorting the output which then makes sence for an amp... thank you for the explanation :-)

Why would it be common sence to use the same size of that resistor as the variable one? .. could you explain that to me please? I don't have any argument against it, would more like to know why :-) Something with impedance?

You gave me a perfectly clear and easy to understand answer, thank you!
//Albert "thec" Sandberg

Btw, the 25% signal thingy was just a value out of the blue, I just wanted you to see what I ment with reverse signals :-)

mechie

OK Thec,
The capacitors in the circuit are to STOP DC, they will let ac signals pass unscathed but by not letting dc through they all ow the input of the operational amplifier (the inverting input in this case) to have a dc bias of around half the supply voltage; so in this design the capacitors will prevent the dc feeding backwards out of the inputs - it could wreck some signal sources, and if the source upsets this bias the amp won't work properly!
The 'virtual earth point' I refer to is the inverting input of the op-amp. (see description 'advice needed on designing power supply').

Resistors... hmm... the choice of resistors to match the pots' resistance is based on attempting to maintain; as far as possible; a constant input impedance (the pot track) and output impedance (the resistor +plus some of the track to ground- in parallel with the remainder of the track to input load). This is a bit contentious.

thec

Yeah, I got the capacitors part ;-)

I didn't realize you inverted ground on the OP, sorry for that, but I'll sure check it up because it looks interesting.

Thanks
//Albert "thec" Sandberg