summing resistors on groups?

[daArry]

Hey

Just wondering - i'm having some groups on the mixer i'm planning (still!) would i place summing resistors on each chan like with the main L/R bus?

that correct?

thanks!

 

[PRR]

A mixer is a mixer. It doesn't know or care if it is mixing groups or main output.

However if this is a passive mixer, with the switch there, then as you put more or fewer inputs on one group, the mix-loss and level will change. To keep loss constant, unused mix resistors should be grounded (actually, grounded through a resistor similar to the pan-pot impedance). That won't be a standard switch.

You are running into the reason all non-simple mixers end up Active Mixing.

 

[daArry]

Hey PRR

Yes, this is an active line mixer...

However if this is a passive mixer, with the switch there, then as you put more or fewer inputs on one group, the mix-loss and level will change. To keep loss constant, unused mix resistors should be grounded (actually, grounded through a resistor similar to the pan-pot impedance). That won't be a standard switch.

Still good to know! So, as I have the group selecter switch on the ckt above it's okay? I guess I was wondering if the values i'd use would be the same as that on the L/R bus...

Soz to seem mundane with the q's sometimes but frankly I dun give a shite - cos it's worth looking like a tit when u end up with the next level gear places like this enable self tought dabblers like i to create...

cheers

 

[NewYorkDave]

Is that Cockney rhyming slang? Translation, anybody? :wink:

 

[daArry]

where'd i lose ya dave?

since you've stepped in tho - workin out the resistor values for the panpot, and from what i've seen there seems a big dif in selected values amongst a few designs...in your buttamxr.gif and genpurposemxr.pdf its 24K with a 10k lin pot. In the mixer project at the ESP site theyve got a 3K6 with a 10k lin pot. And in a fossel doc (8chsum_1.pdf) there's a 15K used...

What's the apple me ol china?

:thumb:

 

[NewYorkDave]

Orban, Robert. Notes on panpots. Journal of AES, Volume 19 Number 11 pp. 935-939; December 1971

It's a tradeoff between input impedance and signal loss. Orban's paper gives the formulas to allow you to pick your tradeoffs and optimize your values for a correct constant-power panning law. I'm afraid I can't post the paper because it's copyrighted by the AES and, as I'm a member, I don't want to annoy them. :wink:

 

[daArry]

well, i'll just wing it then...

just tried a few values, from 3k6 to 27k...guess u settled for 24k as it's as decent as it can get with a single 10k pot. I have noticed hpwever that when panned hard left or right theres still a slight low distorted signal still remaining on the opposing chan - this most likely down to crappy pots?

 

[NewYorkDave]

You'll never get 100% attenuation in the opposite channel with this arrangement. Using the higher-valued series resistors, as I did, does get you closer. I wouldn't judge the pots too harshly because ALL pots have some end-of-track resistance.

Reconsider your grounding arrangements carefully and you can probably improve it further. Poor ground bussing can degrade your isolation by many decibels.

 

[daArry]

Reconsider your grounding arrangements carefully and you can probably improve it further. Poor ground bussing can degrade your isolation by many decibels.

Well i've just been testing with a single channel module on the bench - psu's, xlrs, veras, cups etc everywhere and there's no audible hum, hiss or noise... i guess once I've done a proper psu (just using a mickey mouse one atm to test) and sorted out a chassise layout etc all should be good (easy said)...still, here's a PDF of the trackside and overlay of the module thingy I done, the upper pcb flips over (so copper side up) the one underneath and the 2 and joined together with 30mm spacers - these will be stackable...

http://notapplicable.co.uk/audiolab/sumix/SumMIXER-INA.pdf

I ran this into a JFP for makeup and it sounded fierce :thumb:

 

[daArry]

no gurus round here see a prob with the above then? I'm gonna etch, drill n stuff 16 of these fellas n/w!

 

[NewYorkDave]

You're assuming we've all had a chance to study it. :wink:

I know I haven't.

 

[daArry]

new cam landed today so was able to snap a cpl pics of the test module i made up:

tidied a few bits up since this version, but ready to roll on with mo chans soon as....landed a 6U chassis n all

 

[soundtrack2life]

What happened to the links of the PCB & Pictures. They are not working.
Joe