New Project -- 8/8 BA mixer, help wanted w/buss design

Support for JLM Audio Kits

Moderator: Joe Malone

poserp
Posts: 40
Joined: Thu Mar 08, 2007 10:11 am
Location: Los Angeles, California

New Project -- 8/8 BA mixer, help wanted w/buss design

Post by poserp »

I've been thinking about this for awhile now and I'm ready to start buying some of the parts. However, i'm still working out some of the bussing topology. Here's the basics:

Configuration is like a Manley 8x8: 8 mic inputs w/pan, 8 line inputs w/volume and pan, into two selectable mix busses. I am going to do stereo pairs of different types of preamps for the mic channels for maximum flexibility, with direct inputs on each channel. I plan to use Baby Animals for all the inputs, with different combinations of input transformer and opamp; this will make powering much easier since I'll only need one supply (the 2.2Amp supply listed on the Baby Animal page). Depending on how I do the bussing, I will either use two BA neutrals or two DINGOs and two PEQ1s on one buss for "master" eq.

The bussing is where I'm still trying to figure out what to do. I'm going to order some mixing transformers and try that first -- no resistors, just six transformers for an 8-to-2 summing box. If that sounds o.k. then that'll be the first buss.

However, I was also thinking about a "hybrid" approach; using two mixing transformers to tie together four two-channel passive mixers made with resistors. How would the math for something like that work out? You'd have four resistors on the inputs of one transformer, two on a side. Looking back through the transformer would you "see" two sets of parallel resistors in parallel? If this seems feasable and sounds good to me, I'll set it up as the second buss.

If those ideas won't work out, I'll do traditional passive mixer networks. Finally, I want to put this together in such a way that all the parts can be detached and used seperately; the mic pres, the mix busses, and the master section w/eq. I'll post updates as I progress, the first thing to do is get the transformers and wire up the transformer-only mix buss for testing. Comments or suggestions appreciated!

poserp
Posts: 40
Joined: Thu Mar 08, 2007 10:11 am
Location: Los Angeles, California

Post by poserp »

...Ooops, realized that should be 16 l/r inputs total (8l, 8r) for 8 pannable channels. Not thinking straight there, or thinking more about an outboard summing buss for my DAW (which is how I'll test these topologies). Anyways, same principals with a few more transformers. Dunno if I like having signal through three, I'll see how it sounds with two in the signal path and go from there. Ordered my trannys the other day, awaiting arrival so I can set this up for testing.

User avatar
Joe Malone
Site Admin
Posts: 1644
Joined: Wed Jan 24, 2007 11:35 pm
Location: Brisbane, Australia
Contact:

Post by Joe Malone »

HI
We have another DIY guy building something very similar to this with some 14 or more BA's at the moment. So I have asked him to put his finding up here as his build progresses.

You can use normal BA's as the input mic/line preamps and go straight from the +out to a unbalanced volume and pan passive mixer into a couple of transformer or transformer less input BA's or more as output makeup amps with master volume control easy.
The bussing is where I'm still trying to figure out what to do. I'm going to order some mixing transformers and try that first -- no resistors, just six transformers for an 8-to-2 summing box. If that sounds o.k. then that'll be the first buss.
Even for multi winding transformer mixing to work you need resistors or the output impedance of the other transformer inputs will load/short the transformer. I would just go the resistors and no transformers except maybe at the summing point like 1272 amps were used in a Neve console.
Joe :-)
JLM Audio
Capturing Audio without Injury

poserp
Posts: 40
Joined: Thu Mar 08, 2007 10:11 am
Location: Los Angeles, California

Post by poserp »

[quote]Even for multi winding transformer mixing to work you need resistors or the output impedance of the other transformer inputs will load/short the transformer. I would just go the resistors and no transformers except maybe at the summing point like 1272 amps were used in a Neve console.[/quote]

Thanks for that info; I kinda figured that might be an issue. Hopefully the trannys I ordered come with some spec sheets or application notes or something, if not I'll keep searching. I know from a post on Prodigy Pro that mixers used to (way back when) use mixing transformers, but otherwise I can't find much info about the subject, so I figured I'd just breadboard and see if it worked. Glad to hear someone else is making a JLMed mixer, I'll be following that project with interest.

poserp
Posts: 40
Joined: Thu Mar 08, 2007 10:11 am
Location: Los Angeles, California

Post by poserp »

Received my transformers, four Edcor WSM6410 and two Edcor WSM6400. I emailed the manufacturer and asked them the same question about the necessity of resistors -- they said I wouldn't need any, but I don't know if they were considering the transformer-to-transformer junction.

Joechris
Posts: 2
Joined: Thu Apr 12, 2007 3:03 am
Location: Norway

Post by Joechris »

Joe, do you plan to make a volume, pan, aux, mute and solo board in the future? It will make my day :D

heath
Posts: 7
Joined: Thu May 03, 2007 10:14 am

Post by heath »

Bumpage.

How's this coming along, Poserp?

I'm not much help on the technial knowledge, but I'll certainly cheer this DIY on from the sidelines!

heath.

poserp
Posts: 40
Joined: Thu Mar 08, 2007 10:11 am
Location: Los Angeles, California

Post by poserp »

I haven't had much time to do DIY in the last couple of weeks, unfortunately. I'm hoping that will change in the near future. I plan on taking apart a cheap patch bay I had and using the jacks from it for inputs and outputs, so that'll take a bit of time. Maybe I can even reuse the patch bay chassis. My current plan is to use mix resistors in pairs to double the channel count from 8 to 16 so I can do 8 stereo channels of summing. I'm waiting for someone to build a DAW interface with passive summing included; maybe even just an ADAT or AES/EBU D/A converter mated with a passive mix buss. I'm kind of surprised it hasn't happened yet. I'm not quite ready yet to butcher my HD896 to make it happen, though.

chrisp
Posts: 184
Joined: Sun Mar 04, 2007 11:46 am
Location: Sydney, Australia

Post by chrisp »

poserp wrote:I'm waiting for someone to build a DAW interface with passive summing included; maybe even just an ADAT or AES/EBU D/A converter mated with a passive mix buss. I'm kind of surprised it hasn't happened yet.
One problem is that mix busses are inherently susceptible to RFI unless a lot of effort is put into shielding them, and the clock source for digital conversion usually emits a fair thwack of RF - and its inside the box, exactly where you do NOT want it.

Building a unit with no internal clock (and so requiring external clocking) might lessen the damage, but that's not such an easy product to sell. Its easier to sell a fully functional DA and a separate mix box.
Chris P
---------------------------------------------------------
I do lots of things. I believe eclectic skills are best.

poserp
Posts: 40
Joined: Thu Mar 08, 2007 10:11 am
Location: Los Angeles, California

Post by poserp »

Of course, leave it to Dangerous Music to make the box I've always wanted:

http://www.mercenary.com/damud.html

That's basically what I want to build, it even has two channes of D/A and two sets of speaker outputs. I'll still build my mixer/summing box, but if I had $1400 bucks right now I'd buy this in a second. Hats off to them for making this.

poserp
Posts: 40
Joined: Thu Mar 08, 2007 10:11 am
Location: Los Angeles, California

Success!

Post by poserp »

Well, four mono channels into one anyway, but it proves the concept. Frankenstein Lives!:

Image

What you see there are two Edcor WSM-6410s into one Edcor WSM-6400. The last black cable on the right is the output from the "mixer". Plastic bags to keep everyone safe from random metal surfaces. Finally got up the activation energy to put everything together, at least for testing purposes (I'll put these in a box later, add some decent connectors, yada yada yada). I tested just after soldering some el-cheapo connectors (torn from the guts of a cheap neutrix patch bay) to wires and making use of the handy quick-disconnect tabs on the transformers. How does it sound? To my unschooled ears, beautiful. Increased volume into the "mixer" produced a different yet pleasant sound, I can't really describe it but the bass essentially got "fatter". I need to do some more controlled experiments to determine what sort of makeup gain it'll need and how it really sounds, but it was fine running into the line input on my behringer, not quite hitting 0db but hanging in the -5 to -10 range. My plan is to mate the final version with an RNC and, hopefully, a couple of Joe's PEQ kits. When are those coming out, btw? I am looking forward to getting all of that stuff into a couple of boxes and (after building or buying some sort of monitoring solution) ditching the behringer forever.

poserp
Posts: 40
Joined: Thu Mar 08, 2007 10:11 am
Location: Los Angeles, California

Post by poserp »

And here's a pic of the full version (sans permanent box). I've used the corpse of the old patch bay strictly for organizational purposes. This will ultimately get a proper box:

Image

After further testing there's very little signal loss, I have no problem driving the meters on the behringer into the orange and/or red (right around the same time I hit the red on the MOTU. Neither sounds very good operated that way). Sound-wise the stereo field is much wider than, say, that presented by using the headphone jack on the 896HD. This is true even for simple stereo monitoring. It's almost disconcerting to hear that much "space" around things. Also on further testing I've realized that I can't really be sure of what happens when you drive the transformers hard. The sound is indeed thicker through the trannys, but any attempt to drive them hard puts the MOTU into the red thus I can't be sure I'm not just hearing the D/A distort. Since this is ultimately intended to be the mix buss sitting behind eight mic preamps/line inputs, I should have opportunities in the future to crank it up and see what happens. So, with the mix buss idea hashed out I'm going to move on to coming up with a better monitoring solution, and then figure out what to do for a pan circuit. At that point I will wire up a 4-channel stereo mixer/summing buss that will switch between mic and line input so I can record, mix, and sum in the same box. The other four channels will be added as I build them.

poserp
Posts: 40
Joined: Thu Mar 08, 2007 10:11 am
Location: Los Angeles, California

Post by poserp »

Things are happening a bit faster on this project now. Next I'm building 4 channels of panning logic. All inputs will be switchable -- two (A and B) will be linked to the panning circuit, and two (C and D) will connect straight to the outputs and bypass the pan pot circuit. So a single module (containing one channel) will have four inputs, a L and R output, a pan pot, and a 4 pole 3 post rotary switch. The 4th position on the rotary switch will be "mute", or disconect the channel from the circuit.

For the panning circuit I'm using a design by NYDave over at the Prodigy Pro forums based on the Orban AES paper. The circuit is passive, as shown in his schematic it has about 6 db of insertion loss. One issue that I'm still working out is buss impedance. In order for the panning law to work correctly, the dual linear pot impedance should match the buss impedance. I haven't measured the impedance of my circuit, but I don't expect it to be much over 1k at most (I doubt the transformers present too much impedance). So my first shot will be to build the circuit using 1k as my estimated buss impedance, but that's just a guess. His schematic can be found in the third post in this thread:
http://www.prodigy-pro.com/forum/viewto ... torder=asc

I'll measure the impedance and construct accordingly when the parts arrive, it's entirely possible that I'll need a 1k resistor on each output (from the pan circuit) to correct for the low impedance of the buss in keeping the panning law correct. I ordered the parts for the panning box from futurlec today, so now I wait for them to arrive. For those wondering how much something like this costs, here's some approximate numbers (in USD):

mix buss:
4 Edcor WSM-6410 transformers: $40.00
2 Edcor WSM-6400 transformers: $20.00
100 cnt. .187" female quick disconnects: $7.00 (at Home Depot)

pan circuit:
4 3-pole, 4 position rotary switches: $4.00
4 1k dual linear pots: $4.00
8 knobs: $3.50
asst. resistors: $1.00

common components:
34 1/4" female jacks: $ 12.00

project total (without box or boxes): approx. $90.50

In my view that ain't bad, since you could cut out the panning circuit parts and save about $20 if you just want a two-buss summing circuit that doesn't need makeup gain.

Since the pan circuit will be driven by the outputs from mic preamps or line-level equipment (my plan is mic pres on the "A" inputs, keyboards and etc. on the "B" inputs) I think I can get away with not using makeup gain even with the panning circuit. The insertion loss of the mixing transformers is essentially negligible (.25db as rated by Edcor, my tests seem to support this). As designed the panning circuit would (just a rough guess) provide about 2 to 3 db of insertion loss. I wanted the keep the circuit passive so I could drive it fairly hard and not worry about clipping; I'll need to test to see exactly what the insertion loss will be overall, but I suspect I may want a trim knob on the main L/R outs since the panning circuit and bus impedance are so low. Besides, I'll be driving this with Baby Animals (and hopefully a couple of 1290s in the not-too-distant future) and an RNP; those should provide plenty of oomph. I've already tested the summing buss with my MOTU 896HD and it didn't need any makeup gain, so I think I'll be good to go. I'll post more once I have the parts.

poserp
Posts: 40
Joined: Thu Mar 08, 2007 10:11 am
Location: Los Angeles, California

Post by poserp »

Another update: after experimenting with the panpot design I'd originally picked, I figured out it won't work. There were various issues with the design, but the biggest is that the mixing transformer needs to see a constant impedance at input and output. So my next task is to test using two constant-impedance attenuators per transformer. This means that for a mixer with four stereo channels, I'll have 8 attenuators (2 per channel) to control the "balance" between the left and right sides of the channel. Ultimately I think it'd be cool to shoehorn three transformers and four attenuators into an API 500 series single space rack module for a simple four-channel mono mixer. At this point I doubt it'd fit into a single-space panel, but it might fit into a double. Thus with four single-space preamps and one mixer module you could turn a six-space lunchbox into a self-contained four-into-one mixer. There are easier (and cheaper) ways to do that using active circuitry, but what's the fun in that?

User avatar
Joe Malone
Site Admin
Posts: 1644
Joined: Wed Jan 24, 2007 11:35 pm
Location: Brisbane, Australia
Contact:

Post by Joe Malone »

Do you have a schematic ?
Joe :-)
JLM Audio
Capturing Audio without Injury

Post Reply