DIY rotary DJ mixer

[efinque]

Hello GroupDIY,

and excuse me for being new here.

Anyway, I've been dreaming of making a rotary DJ mixer for about 10 years now. I've searched the web and doodled a bit on paper but I came here asking for some additional advice.

As for some background ; I've been DJing since 2006 and played my first gig in 2007. I've owned and used several mixers over the course of years as well as serviced and repaired them so I have some knowledge on how they work. (edit : basically the design is something along the lines of Bozak/Urei and the boutique stuff like E&S)

Here's the initial sketch of how it would look :

(edit : the width is the standard 19")

Basically, I'd build the preamps from kits as modules with wiring as well as the (optional) isolator and level meter and the headphone amplifier.

For potentiometers I was thinking ALPS 50K stereo log and Alpha 100K dual log for trim controls. (edit : then there are Bourns pots also which are excellent from what I've read) As for the cue it'd be a rotary switch.

I found a 3-way linear phase filter/isolator kit from Analog Metric

I'm a little concerned about the PSU... I've been looking at one that's specs state 30VA, 230V AC / 18V + 18V / 0.833A + 0.833A but I don't know if it packs enough punch to power the whole thing (preamps/headphone amp + isolator)

Another thing is balanced outputs...

What are your thoughts?

-ef

EDIT : as for the isolator it'd be post-master (possibly bypassable)

EDIT 2 : I was also thinking Sifam VU meters but they're pricey albeit very classy looking.

EDIT 3 : Sorry I missed the headphone level control and cue. Updated

EDIT 4 : as some of you might have noticed I'm obsessed with disco stuff.. I've been designing and building a classic dance stack (the "finished" stack is on page 11 but I haven't worked on it since apart from making a driver hole in one of the scoops)

EDIT 5 : Stuff I already have/can make use of : soldering iron, multimeter, a bit of free time... I also have an Arduino Uno board and some components too. Financially speaking it wouldn't be a big deal I guess, back in high school when I had the idea I didn't have any resources to begin with... now there's a slight chance of making one.

EDIT 6 : Sorry for the messy picture but I hope you get the idea.

EDIT 7 : I've also been scheming a studio mixer on Arduino forums.. I guess it'd be better to start from a project of this scale.

EDIT 8 : I'm not an avid DIY-er when it comes to electronics.. I built a radio back in school but that's about it. I did pull a headphone amp from a Presonus FP10 audio interface to my Mackie D4 because I broke a plug inside the connector and couldn't get it out in one piece without damaging the terminal... it works like a charm. I've also been thinking of servicing my Formula Sound PM-90... I asked the company for spares and the costs would've been astronomical, and seeing the mixer was installed at a venue back then I postponed it..

EDIT 9 : I'd guess some would prefer pan controls instead of gain.. which is something I've also considered as it is a rotary. One thing I didn't consider was a cue/pgm switch.

 

[JohnRoberts]

I did some consulting for Bozak back in the day and their 10-2-DL was a classic rotary control DJ mixer.

In 1978 I published a kit DJ mixer (cover of Popular Electronics) DJ mixer with rotary controls... (IMO it didn't suck, but I like all my kids).

JR

 

[efinque]

I did some consulting for Bozak back in the day and their 10-2-DL was a classic rotary control DJ mixer.

In 1978 I published a kit DJ mixer (cover of Popular Electronics) DJ mixer with rotary controls... (IMO it didn't suck, but I like all my kids).

JR

I've looked at the Bozak schematics... they're fully discrete.  Another thing is the placement of the headphone jack (and the master/cue section layout in general)

EDIT : Bozaks have the jack (physically) placed between the program channels and the master whereas in the Urei the cue section is located on the far right. It's not really an issue with an L-connector though.

EDIT 2 : I'm not a fan of FX send/return loops either... they're quite usually found in DJ mixers. However, I've been learning to use one in my live setup but that's another story.

EDIT 3 : Another thing that bothers me is grounding issues... sound quality-wise and I'd also be dealing with 230V here. Booth EQ is a luxury that can't be really justified.

EDIT 4 : I also have an amp module from a pair of old computer active speakers...

 

[JohnRoberts]

IIRC the urie was "inspired" by the much older Bozak....

Nothing about rotary controls really matters all that much, it just was a solid well-engineered design.

JR

 

[efinque]

IIRC the urie was "inspired" by the much older Bozak....

Nothing about rotary controls really matters all that much, it just was a solid well-engineered design.

JR

I read the monoaural Bozaks were initially designed to be used for film production.

EDIT : I have nothing *against* per channel EQs... I just rarely even use them while playing anymore because the Mackie has a pair of filters. The Formula Sound has +-12dB EQ (in fact I think they were the first to implement per channel EQs) and I use it to cut the bass, nothing else. I think the EQ kind of kills the idea of a rotary mixer (I'd use it for playing vinyl & cd at home and maybe also as a master isolator for the dance stack)

EDIT 2 : I'm sure there's something I didn't notice... yet another thing is that most mixers have an output module of some sort. I was thinking of running it from the isolator. One option would be to use tubes to warm things up a little bit (I've looked at the 12AX7) and as a bonus they'd light up the Sifams quite nicely. I'm not sure whether there's space for all the aforementioned stuff like the booth EQ, Sifam metering etc but I have a rackmount Omnitronic DB-100 led meter which could be used in conjunction with the mixer, eliminating the need for metering.

EDIT 3 : ..but due to the rotary heritage I'd rather make it solid-state. There's a company called Bozure that does (or used to do) discrete rotary kits

 

[PRR]

> In 1978 I published a kit DJ mixer

http://www.americanradiohistory.com/Archive-Poptronics/70s/1978/Poptronics-1978-09.pdf
Cover and page 61 (page 53 of PDF).

 

[efinque]

> In 1978 I published a kit DJ mixer

http://www.americanradiohistory.com/Archive-Poptronics/70s/1978/Poptronics-1978-09.pdf
Cover and page 61 (page 53 of PDF).

Fascinating... I remember reading somewhere that DIY guides such as the one above were very popular back in the day (for guitars etc) (edit : another thing I noticed they mention the NE5534 opamp; I didn't know it dates back to the 70's..)

But I guess it's a dying scene.  I've sourced some of the materials and made (rough) initial estimates on the costs (in eur) :

-3RU 19" rack case ~21€
-3 band linear phase filter kit ~70€
-2x RIAA preamp kits ~15€/pc
-3x line preamp kits ~15€/pc
-1x microphone preamp kit ~10€
-1x headphone amp kit ~15€
-8x ALPS 50k stereo log pots ~15€/pc
-7x Alpha 100k dual log pots ~3,5€/pc
-1x rotary switch 6-pos (for example SR17A1620F69N) ~3,3€
-1x transformer 2x15-24VDC ~22€
-14x RCA female panel mount connectors ~1€/pc
-3x 6,3mm panel mount stereo jack ~3€/pc
-2x XLR male panel mount ~3€/pc
-1x XLR female panel mount ~3,5€
-1x 2x10-15 LED meter kit ~20€ (optional)
-1x power switch ~2,3€
-1x power connector 3 pin with fuse/breaker box ~3,5€

(edit : that's ~500€ in total... tempting, seeing a mixer like this costs anywhere near 1,5-2k and upwards. With Sowter output transformers it'd be close to 1k, the Jensens are a bit cheaper)

The calculations are missing the knob heads and other stuff like the internal wiring and veroboard for tidying up the panel mounts.

EDIT : minor fixes.. I also think one would benefit from putting a fuse somewhere in there..

EDIT 2 : Another thing is gain staging and dynamic range/headroom.. as for "tuning it"; I don't have access to an oscilloscope.

EDIT 3 : What about transformers? I've looked at Jensen and Sowter units but from what I gathered they're used in analog mixing consoles.. I guess the 19" chassis would house a pair nicely.

EDIT 4 : I think one could get away with 50k pots for trim... I can't really tell. I'd also steer clear from building a PSU from scratch although it would be beneficial due to the different voltage needs here and there. As for the tech specs I'm aiming for 20..30Hz-20kHz frequency response at a minimum... I don't know about output values; the Bozaks were known for being LOUD which is why they found their way in the broadcast industry and eventually in discotheques.

EDIT 5 : Here's a quick block diagram mockup :

(edit : revised)

I left the PFL metering open since there are a few ways to do it... maybe a 2-pos switch between master and the PFL bus. I've also been thinking about an unbalanced REC out.. I'd probably omit the mic PFL and put the master channel there instead. As for the isolator I'm unsure but I accidently put it before the master. The reason I want to use trim controls is because most DJs including me are used to matching levels on their program channels and playing with the volume controls at 10. I guess some would argue that the taper wear is more noticeable in the extreme ends due to the mechanics involved, go figure.

EDIT 6 : And no, it won't go to 11. I'll probably make the headphone section mono because I'm too lazy find a 3-pole 6-pos switch, and I'll probably be using it with an old school stick headphone anyway. What about a stereo/mono switch in the master?

EDIT 7 : I don't even consider myself a hobbyist... more of an enthusiast? I don't know if this is good as a first "serious" DIY project... if this was my first mixer ever I wouldn't build one unless I really had to... rotaries are an exquisite in the mixer realm these days and a very niche market due to the fact that they're for a certain style of music and mixing.

EDIT 8 : Another version I was thinking would've consisted of a frame and 19" 1-2RU fader modules attached to it vertically...

EDIT 9 : I miscalculated the filter kit price.. it's fixed

 

[efinque]

I'm also having second thoughts about the isolator... most DJs would want resonance for an enhanced effect and it might possibly be harmful to speakers.

EDIT : there also could be phasing issues seeing that the outputs in the kit are separate. I remember reading somewhere that in some filter designs when you attenuate/boost the frequency band starts to drift...

EDIT 2 : Then there's Yaron and his custom, mastering grade mixer but I'm not looking for that much EQ (and I don't think he's giving away schematics anyway).. I know a bit about the Calrec's and stuff but it's a bit of an overkill imo.

EDIT 3 : One solution would be to ape the Allen & Heath S6/V6 design and put a HPF on each channel... that would eliminate the need for a master isolator, possibly dwarfing it down to 2RU.

 

[efinque]

Regarding transformers; I have no experience with them... I've looked at Sowter, Jensen and Lundahl transformers for output.

Another thing that bothers me is the use of kits for basically everything and then trying to get it working when there are a number of different gain stages in the signal path.. it could inversely affect serviceability too.

EDIT : I have this theory that while it looks good on paper and fairly easy to make (and knowing myself a bit) there will be a lot of problems and troubleshooting along the way..

EDIT 2 : I was thinking a pair of Lundahl LL1527 for output transformers.. any thoughts?

EDIT 3 : There was a thread from like 10 years ago (here or diyaudio.com) where someone was designing a rotary but I can't seem to find it anymore..

EDIT 4 : I can't say I'm good with the soldering iron either.. there's a chance that I'll end up with a bunch of components,  modules, transformers and a rack case unused and that kind of worries me..

EDIT 5 : One reason I like DIY is because it's like practical math... I wasn't really good at physics, chemistry or maths in school but I did spend an awful lot of time in the workshop or designing a mixer such as this one on paper at class when I was ought to do something else.. I guess it's better than causing trouble in the neighbourhood at that age (I've been looking at some old notes from high school but I didn't find the schematics)

EDIT 6 : I've built a dj console and another one for a venue.. there are pics of both there, the one I have in my apartment is further down the thread. As you can see, the 2nd rack is empty and could use something to fill it. The "panel" is actually a casing from an Akai tape deck because it seemed to fit in there nicely.

EDIT 7 : As for analyzing the modules before attaching them would feeding, say, a 100Hz sinewave or white noise at an amplitude of 1 and then recording it back to an audio editor spectrum analyzer be a substitute for an oscilloscope to get an idea of the frequency response and other stuff? (because other than that it gives an idea of the equipment it was recorded with but I'd guess it'd prove useful in checking the RIAA preamps)

 

[KrIVIUM2323]

Hi,
The rotary mixer you use as a reference was designed by Igor and the thread is probably burried in here.
The eq on this mixing desk are very probably his stripped down version of sontec eq (the one with only 4 bands).
From what i ve read on the link you posted, phono preamp are Urei cards.

About the topology of your mixer i don t understand what is the use of isolator? To kill bands before the main out? I don t see the possible use of this kind of thing on the main but every dj has his own way of doing.

What i know is that resonance control should be avoided if you implement a filter (except if you plan to use a constant amplitude design) or at least being limited and stop it before auto oscillation.

No crossfader?
What are you planning to use for volume control? Vca, audio potentiometers?

 

[efinque]

Hi,
The rotary mixer you use as a reference was designed by Igor and the thread is probably burried in here.
The eq on this mixing desk are very probably his stripped down version of sontec eq (the one with only 4 bands).
From what i ve read on the link you posted, phono preamp are Urei cards.

The Traxx mixer uses Bozak phono cards from what I've read... anyway, I was thinking either kit or making my own. They shouldn't be overly complex to design, it's just that I don't have the necessary equipment to make PCBs.

About the topology of your mixer i don t understand what is the use of isolator? To kill bands before the main out? I don t see the possible use of this kind of thing on the main but every dj has his own way of doing.

What i know is that resonance control should be avoided if you implement a filter (except if you plan to use a constant amplitude design) or at least being limited and stop it before auto oscillation.

The isolator is for an effect, not so much for transitioning. Some (mostly house) DJs use a separate unit in the master out, there are a few 1RU designs like the Bozak ISO-X and the E&S XFX 3004 to name a few.  I don't know how the Analog Metric kit will do but seeing the commercial isolators are in the 500-1k price range disassembling and putting one in my own design would be expensive and not exactly "sound" DIY ethic. (edit : the filter kit doesn't allow for boost, ie it's from -100 to 0dB... I could use a little boost there, something like +6dB max but I don't know if the filter becomes unstable)

No crossfader?
What are you planning to use for volume control? Vca, audio potentiometers?

Crossfaders in mixers are mostly considered something hip hop DJs use for cutting.. I've never really used one.

The volume controls would be ALPS RK27 50k/100k stereo logs.

EDIT : I don't know if it would be safe to output 3-way directly from the isolator (as it's a crossover) because of the apparent risk of damaging the speakers so I'd sum the signal to stereo. On the other hand I remember reading that the Paradise Garage had a crossover that allowed the DJs to boost the tweeter arrays..

EDIT 2 : And I don't think any engineer in his right mind would let me connect a crossed over 3-way output mixer to a system.

 

[KrIVIUM2323]

The "linear phase filter" could not boost, they are filters not eq. By the way i doubt them to be linear phase filter. Linear phase filter are only achievable through use of digital technology (and this is neither cheap (my dolby Lake processor cost me an arm and a leg!) and neither realtime processing, it needs a delay to proceed, the lower in freq the longuer). I suppose the filters from the kit are some bessel kind of filtering as in analog they are the least harmfull for phase.

I agree crossfader are most used for cut but not only, some perform constant power summing and are usefull for smooth transition too. Anyway all use of that functions depend of your kind of mixing style and genre you are playing.
As a D&B mixer i don t use isolator or other filters, crossfader more often.
Back to your isolator output, yes no one is going to let you determine the crossover freq for a sound system! This is usually fixed and not something an artist will have access to.

 

[efinque]

The "linear phase filter" could not boost, they are filters not eq. By the way i doubt them to be linear phase filter. Linear phase filter are only achievable through use of digital technology (and this is neither cheap (my dolby Lake processor cost me an arm and a leg!) and neither realtime processing, it needs a delay to proceed, the lower in freq the longuer). I suppose the filters from the kit are some bessel kind of filtering as in analog they are the least harmfull for phase.

The kit states they're either Butterworth, Linkwitz-Riley or Bessel.

I'd probably go for the Linkwitz-Riley.

EDIT : and I've mixed dnb too.. but I still rarely use a crossfader.

 

[KrIVIUM2323]

I find it weird to call linear phase filter something that use bw, lr or bessel.
Only fir can do linear phase, but who cares after all if you can use it the way you want.

You mix d&b and don t use crossfade? I m not involved in the scene as i was 15/20 years ago but you must sometime feel lonely in your practice and your style of mix!
From what i've heard most dj's in the genre don t try to mix tracks for long time now, more a collection of cut every minute or less... a la Andy C but not as talented as the guy. Now that you don t have to synchronise vinyl by ear it is even worth, technology is not always something good...
Not my style anyway, i m definately old school about that... i m old guy now!
Even great labels disapeared! Rip Renegade Hardware...

Back to your mixer idea, you talked about tube. Stanton had a mixer with tube preamp years ago, i used it but i thought  it wasn t adapted for d&b, not clean enough in the low end and too soft sounding for my taste, but i m pretty sure it used some starved current kind of circuit, caricature of tube sound in my opinion.
Maybe it could be of benefits for the phono preamps, but i doubt as a final stage. And it will need high voltage psu and filament psu, maybe to complex to implement.

Having high headroom and symmetrical output is great in my opinion. As you don t plan eq headroom is not as critical than when you have eq with +12db boost in low end!
About out level it will depend what you ll drive: if your next stage is a pro converter it make sense to have same standard (+21/24dbu), if you attack directly power amp you ll need much lower.
Maybe plan different out circuit cards ?
Transformers are great, soundwise it ll depend on what you are after and how you implement it.
One last thought, you want to invest lot of money in nice pot but have you tried them already?
I ask because i had surprise with some pot in use: log are theorically better but i ve found some to have a feel not really adapted in use: not enough volume along the traveling range then too much variation on small rotation in the end of scale.
This can be annoying, and sometime a linear pot with a paralell resistor give a more usefull span in the end ( more control on the last half travel of pot).

 

[efinque]

Hmm.. food for thought there.

As for tubes there's the aforementioned A&H V6 which is a tube mixer. I guess the problem with tubes is that they have to be serviced more often and need to be "warmed up" before use. They're also more delicate design-wise.

I don't necessarily *need* a filter. It's just that including one in the build adds up the fun factor of DIY.

As for not getting it working; like I said my piano teacher I'll have the rest of my life to figure it out.

EDIT : as for the output if it was like really ridiculously loud and clean (which I highly doubt I'll achieve) wouldn't the problem be solved by putting a pad switch in the output?

 

[KrIVIUM2323]

Hi,
I didn t know that A&H had a tube mixer, i ll check that. Talking about A&H one of my friend have a 'recent' one (4/5years old) that i found to be well thought about gain staging: it have huge headroom and can be operated at pro level standart without breaking a sweat! And it does include everything you need for connecting to the outside world: unbalanced -10dbv, balanced +4dbu, you name it, it is there...
For me the only drawback are all the gadget(filters, aux send, etc,etc,... as i already said i m old school and i don t see the point about more tweaking than level and eq but that is me!) and the choice of eq points (frequency) which is not adapted to my genre (or my liking i should say).

I don t agree with you about drawback of tubes. Tubes are not likely to die from electrostatic discharge, and are overall strong electronic components (just think about what a guitar amp have to withstand during touring of a rock band...).
I don t think they are more delicate to design with, just different. And about warming, well... every piece of electronic gear have to warm up until it reach it's equilibrium point.
The real drawbacks are the multiple and high voltage you have to deal with and yes you must change them from time to time, which implies to choose a type you could still find in some years ago... And the build may need more space than smd components!

Anyway before making a choice of technology maybe you should define more precisely parameters about input, outputs, operating levels and all that kind of things... as you already done with ergonomic.

If it was my project i would focus on a pristine signal path with tons of headroom, a multiple output master section able to drive anything you ll present to it, a clean pfl/monitor/booth section, a psu able to feed everything cleanly and  the feel of controls adapted to your liking and leave all the gadget for future expansion.
For example you could forget the isolator for now but include some inserts on the master section for future build to come.

As there is multiple different circuits having a modular approach with the build could save on headache with debugging and servicing.
And this will help you when you ll have to try and prototype things (as it will happen).

Edit: i checked the v6 manual, it seems a nice mixer! I could live with this one if it have an external crossfader. Anyway, check the technical spec i think they are spot on.
Well, from what i seen i think you should inspire yours from this one! It have everything you could dream of!
Just forget the tube parts (think about an insert post preamp and you could add a tube warmth processor -whatever it means, most likely a distortion generator...- later).
As you dont want eq you could live with a lesser max level in chanel than 28dbu., probably around 20dbu this should be well enough for a digital source or riaa preamp.

 

[efinque]

I don t agree with you about drawback of tubes. Tubes are not likely to die from electrostatic discharge, and are overall strong electronic components (just think about what a guitar amp have to withstand during touring of a rock band...).
I don t think they are more delicate to design with, just different. And about warming, well... every piece of electronic gear have to warm up until it reach it's equilibrium point.
The real drawbacks are the multiple and high voltage you have to deal with and yes you must change them from time to time, which implies to choose a type you could still find in some years ago... And the build may need more space than smd components!

One thing with tubes I believe is that they're also very sensitive in terms of breaking easily if the unit is pushed around a lot.

Another is that they pose a serious threat.. I've heard stories of people getting electric shocks from guitar amps because of bad grounding.

Anyway before making a choice of technology maybe you should define more precisely parameters about input, outputs, operating levels and all that kind of things... as you already done with ergonomic.

I have no idea about output impedances and THD... I guess there's a chance that it could sound ok but I'd think no high end output transformer can save a bad preamp.

If it was my project i would focus on a pristine signal path with tons of headroom, a multiple output master section able to drive anything you ll present to it, a clean pfl/monitor/booth section, a psu able to feed everything cleanly and  the feel of controls adapted to your liking and leave all the gadget for future expansion.
For example you could forget the isolator for now but include some inserts on the master section for future build to come.

I've already considered the possibility of send&return loop with some sort of level control.. I guess it wouldn't be that hard to make, a couple of pots and panel output connectors. Since the unit isn't really "powering" anything one could pull as many outputs from it as needed. However, the front and back panels will be crowded, especially if I choose to attach the transformers or something back there. (edit : the Formula Sound PM-90 has an external PSU.. a thing that I've also considered but I'd rather take the risk and put it in the case)

I can't really make it deeper than 300ish (in mm) due to the fact that most consoles/racks can't house a unit larger than that (one has to remember that the connector&power cables and stuff also add up to the overall depth) (edit : the case I ordered is actually somewhere around 200mm)

As there is multiple different circuits having a modular approach with the build could save on headache with debugging and servicing.
And this will help you when you ll have to try and prototype things (as it will happen).

Edit: i checked the v6 manual, it seems a nice mixer! I could live with this one if it have an external crossfader. Anyway, check the technical spec i think they are spot on.

I think the V6 had an external crossfader module available..

EDIT : I've decided to scrap the linear phase filter and substitute it with a 3 band tone control instead (and pretend it's a filter, which it basically is) due to the fact that the filter network has 26 opamps in total. Another solution would be a simple 2 band Baxandall EQ on the master which would allow me to house it in a 2RU chassis.

 

[KrIVIUM2323]

About output z (impedance) just check the v6 manual page 36. This is listed. Thd will depend of circuit used and headroom.
I repeat, from what i ve read the tech spec of the v6 is a nice target (read profressional standard).

You seems to want absolutely an output transformer... ok but why? No need to focus on one costly component if you don t know why... put differently there is other possible answers that can fullfill requirement with lower cost.

That said you are right, a very good component can t change a signal path, this is a chain and the weakest link determine the overall quality of the chain.

I talked about insert not aux. Do you know difference between this two terms? If no, just do a search and think about it: no needs to have complex circuitry when you can use a simple one.

About transformer and the need of them (or not) try to read white paper on Jensen site. And take a look at line drivers and receiver from 'that' (1606 and 1646), designed by Jensen.

I don t want to seems to be patronising you, but if you choose to design something it is better to make educated guess.

 

[efinque]

I talked about insert not aux. Do you know difference between this two terms? If no, just do a search and think about it: no needs to have complex circuitry when you can use a simple one.

The old rotaries had jumpers on the send & return which means they were made a bit differently than the ones found in todays mixers.

EDIT : as to using transformers they add up nicely to the overall weight, and weight is power, so overweight is overpower, right?

EDIT 2 : On a serious note if the back of the mixer is a lot heavier than the front then it puts additional stress to the rack ears if the case is flimsy and also it's much more harder to screw it in place.

EDIT 3 : and if I don't get it working, at least I can use it as a weight for fishing net.

EDIT 4 : Then there's this age old debate of American vs British sound..  the PM-90 is very transparent. Of course I have nothing to compare it to but I've read the Bozaks had a nice low end.

 

[KrIVIUM2323]

I don t know specifically of the mixer you are talking about or their functionnality but an aux and an insert are different things, this is independant from the level at which they work.
An aux is a parallel feed with an indepandant return which is mixed back to a bus. It usually have volume controls (individual sends, master sends then a fader to mix in the receiving buss).
An insert is a series feed, without volume controls as the reference levels is predetermined (pro or semipro line level or anything that please you) and it just need a on/off switch to be activated.

In your overall layout your isolator is in serie between your mix bus and your main output card, no need for an aux.
Same thing in your chanel if you want to include some processing in individual chanels.

An aux is interesting for external effects (mainly time related like reverbs, delay, chorus,... you name it) can sometimes be  used as more output or to make parralel treatment but i doubt you ll use them like this.