Sourcing parts for pan circuit in a passive summing mixer

[Ego Tripper]

Hey guys,
Here's my situation. I recently moved into a new space with a couple of guys from my day job. Between the three of us (mostly them), we have a pretty respectable selection of outboard gear of which I would like to take advantage while mixing. Upon entering the arrangement, I had assumed they had a good way to do this already, but I guess I was wrong. The one engineer is currently mixing in-the-box, but would like to be taking advantage of his sweet outboard (hard to do in Pro Tools w/o HD's hardware delay compensation), and the other engineer tracks & mixes on his Mackie d8b / HDR24/96 combo, taking advantage of the outboard, but having to send his signal through multiple passes of sub-standard conversion with the d8b's insert points (not to mention the original A/D pass during tracking).

Not only would I prefer to avoid the d8b's converters when we have a pair of Black Lion modded Delta 1010s on the Pro Tools rig, but I'd really rather not deal with the headaches of transferring my existing Pro Tools sessions to the HDR26/96 (lots of non-contiguous audio files and FTP sucks).

So I figured I'd be mixing out of Pro Tools via the BLA-modded 1010s, into the various pieces of outboard, and summing through the "Sumthing" passive summing box that's there. The problem, of course, is that the Sumthing was not designed to do this, maxing out at 16 channels and lacking a way to pan signals (without taking up two inputs, limiting me to 14 channels or less and not even possible when the signal in question is going though any outboard after the DAW).

So the logical solution would be to build a passive summing box tailored for this scenario, which would need a way to pan signals. I found some 4P3T shorting switches, but this would just give me Left, Center, & Right. I often find myself panning things to the in-between positions (like toms and even guitars, to give a not-so-exaggerated spread at times), so Left-Center-Right would be pretty limiting.

The ideal part would be a center-detented 20K dual linear pot, but I'm finding it impossible to source these. Mouser lists the part from Bourns, but doesn't stock it (and doesn't even provide a factory lead time for it (which would suck anyway-- I have clients awaiting mixes)). Forgoing a center detent, I'd settle for stepped panning (and appreciate the recall-ability of such), but I'm having trouble sourcing the right switch configuration in a shorting (MBB) variety.

How bad would it be if I used a non-shorting (BBM) switch for a pan circuit?

 

[ed rees]

Rs exports sell rotaries a-la carte, by the wafer.

http://export.rsdelivers.com/product/lucas-stability/457846ma/1p-11-way-mbb-wafer-rotary-switch/0352244.aspx

http://export.rsdelivers.com/product/lucas-stability/454486mk/12-clips-dummy-for-wafer-rotary-switch/0327844.aspx

 

[Ego Tripper]

Interesting links-- I was wondering if something like that might be possible.

Unfortunately, there doesn't seem to be much information on these; even after reading the PDF datasheet, I'm still confused about a few things, particularly where the additional rotors would come from for the extra poles (this application calls for a 4 pole switch). Also, the assembly on the datasheet seems to call for spacers and I don't see anything of the sort on RS's site.

And while the "dummy" wafer to which you linked would work, they only have 16 in stock, which is 56 fewer wafers than I'd need to make a 24 of these 4 pole switches for a 24 channel summing box. This is another problem I've been encountering on the few occasions I've run across a part that could conceivably work.

Does anyone know of any other sources for DIY switch parts? Or, even better, center-detented dual 20K linear pots??

 

[ruffrecords]

How bad would it be if I used a nfon-shorting (BBM) switch for a pan circuit?

Just add 4.7Meg resistors between adjacent contacts to eliminate clicks. Neve did this in all their classic EQs.

Cheers

 

[Ego Tripper]

What a clever dog that Rupert Neve was. Thanks for the tip, that really opens up the options for me!

 

[ed rees]

This wont affect the curves or the Q at all? Great to know!

 

[ruffrecords]

This wont affect the curves or the Q at all? Great to know!

No. The 4.7 Meg is very large compared to the impedances within the EQ and so its effect is miniscule.

Cheers

Ian

 

[Ego Tripper]

I was thinking about this, and this switch will already have 2k resistors across all the contacts, so I wonder if the 4.7M resistors would even be needed...

 

[keefaz]

The ideal part would be a center-detented 20K dual linear pot, but I'm finding it impossible to source these. Mouser lists the part from Bourns, but doesn't stock it (and doesn't even provide a factory lead time for it (which would suck anyway-- I have clients awaiting mixes)). Forgoing a center detent, I'd settle for stepped panning (and appreciate the recall-ability of such), but I'm having trouble sourcing the right switch configuration in a shorting (MBB) variety.

How bad would it be if I used a non-shorting (BBM) switch for a pan circuit?

Mouser has 20k lin dual potentiometer with center detent, conductive plastic from BI Technologies
(it's cheap, don't know about the quality though)

Part number : 858-P858-P160KN-1EC20B20K

Center detent is not specified in product description but the manufacturer part number says it has it, if you follow part number description in datasheet

(edit)
Sorry it's not a dual potientiometer  :

 

[ruffrecords]

I was thinking about this, and this switch will already have 2k resistors across all the contacts, so I wonder if the 4.7M resistors would even be needed...

OK, as this is effectively a fader it means you wan't get clicks due to caps charging but you may get momentary signal loss between steps. This is where shorting contacts really come into their own. If you do the maths you will find that as the switch shorts adjacent contacts, the attenuation produced is mid way between the two positions you are switching between so instead of a click you get a half step. I believe some of the very old stepped attenuators took advantage of this and put a detent at the shorted position effectively doubling the number of steps for the same number of resistors.

I am not sure is there is a fix for this that will allow non shorting contacts.

Cheers

Ian

 

[tonycamp]

I'd be very interested in seeing a schematic on this? I love passive summing, i hate not having pans and attenuators. How do you plan on handling the levels? In the box level automation will reduce bit depth and resolution at output, not to mention the response of your outboard gear....right, and if you use your make up gain properly out of a comp or eq, you'll end up with +4, u wont be able to control the level of each channel, with out compromising S/N ..right? Do you plan on using level controls? The only commercial unit i'm aware of that controls pan/level in passive mode is the SM pro PM8. I would love to build a passive device that answers these Q's, i've only been able to find LRC or LR passive builds w/no controls, it defeats the purpose of coming out da box if ya gotta go back in and come back out again...right ;D

 

[Ego Tripper]

Eh... I might just have to live with the clicks since going with non-shorting switches is going to save me hundreds of dollars. On the rare occasion that I need an active pan during a mix I guess I'll use a stereo tremolo or just record the part to the computer, automate the pan in the DAW, and bus it out to the summing box along a stereo pair. I don't mind doing that on those rare instances... I would mind having to do that every time I wanted a pan though, so I suppose this is an acceptable compromise.

If it bugs the other engineers that much they can plunk down the money for the shorting switches.

@ tonycamp: I do plan on including level control for the very reasons you listed, mostly the outboard reason. I'm basically building it according to NewYorkDave's plans in his balancedmixnetwork.pdf, available in the GroupDIY photo album (it's not titled, so search by his user name and page through, hovering yr cursor over the PDFs 'til you find it). The method for adding-panning/changing-the-mix-resistors is discussed in this thread:
http://www.groupdiy.com/index.php?topic=11077.20

I won't be adding the mute because that can be accomplished by unplugging a patch point that's normalled to itself. Unlike the panning (and to a lesser extent, level control), it can be accomplished from outside the unit.

 

[Ego Tripper]

I made this schematic at tonycamp's request, but I figured others be able to benefit as well since the originally discussed schematic with pan pots seems to be unavailable.

I should also clarify that the mute is not identically accomplished by unplugging a patch point that's normalled to itself when there is a level pot in the circuit, but when the patch point is unplugged and the level pot is all the way counter-clockwise then it will still function the same and you don't need to use a normalled jack.

Also, I found some shorting switches I could afford, so that's an added bonus.

*EDIT*
OK, the attached schematic is now correct.

 

[PRR]

> schematic is now correct.

I don't get it. Looks like it is a dead-short at either switch stop.

 

[Ego Tripper]

Perhaps I should have noted that in that position the switch is panned hard right. If I understood the discussion from the earlier thread (again, no schematic for the pan pot version is available), then to pan a channel hard to one side the other side is getting shorted to ground. In the center position there is 10k to ground for each side and the opposite extreme is the opposite input getting shorted to ground.

Is this not the way it works?

 

[SkunkFunk]

Dual 22k lin centertapped are available at audiomaintenance.com, both as carbon and conductive plastic.

 

[Ego Tripper]

Dual 22k lin centertapped are available at audiomaintenance.com, both as carbon and conductive plastic.

That's good to know. I've already purchased my parts for a switched-pan version, but I'm sure there are some other people out there looking to build a passive mixer for themselves.

Can anyone who has the original schematic for NewYorkDave's summing mixer with pan controls confirm that the wiper of the pot goes to ground like in the schematic I've posted above? And that it connects to the left & right buses like I have in the schematic?