The power supply I built is linear, using a a LM7815/LM7915, with six bypass bipolar power transformers EACH. That gave me a max of 18A for the +/-15v rails for a semiconductor standpoint, and the transformer I used was something massive I pulled out of the junkbox. It's a 40V center tapped transformer with 120V and 240V primaries, and can provide 30A. It's a huge, massive transformer. I designed it with overkill, with the intention of explaining much later how to calculate what power is actually necessary, then shaving off bypass transistors that aren't necessary. Or, one could leave them all and use smaller heatsinks. I'm not covering the power supply until the very end for this reason. Also, the power supply is designed to be seperately housed. You don't want a 35lb transformer near audio electronics. Same reason why many console manufactures make them rack mountable with long cables. We're going to do the same thing here.
When I built each of these modules, I tested them using an old industrial AT power supply, which is a switching power supply. The industrial switcher I had provided +12V at 25A, and unlike most computer power supplies it provided more than 500ma on the -12V feed - up to 5A actually - so for testing five years ago on the workbench, I found this acceptable. The 15V linear power supply coming later is a better choice. I'm not an engineer so making multiple power supplies ganged together is not within my skillset. I am good at making massive linear powersupplies with huge heat sinks, so those of you who have clues in this area are absolutely more than welcome to say "that sucks, here is a better design" and I certainly would not be offended in any way. As I attempted to say in my first post on this thread, I've built each of these modules, and tested them apart and together, but I've not built a massive 56 channel, 16 buss console. I'm sharing, because I thought this would be fun, a good learning experience for you guys
and for me, as has happened already. You guys (and gals) are quick in finding my mistakes, which is wonderful.
Speaking of mistakes...
Ms. Hilarious is correct, the schematic software picked voltages on the caps for me, which I don't like either. I'll manually edit the pages and put the appropropriate values, after I finish the switch module. It will work "as-is" as that's how I've built the modules, but I would
highly recommend no one buys anything for the first few modules until we get a little further down the "building block" line, so to speak. Should I say again this is for informational/learning only? And I offer no warranty? heh-heh.
L1 and L2 don't have part numbers. They are two donut-shaped ferrule beads with three turns of 16 gauge insulated, stranded wire spaced about 1/16" or so apart. They help with hum rejection on long XLR mic cables. They aren't critical if they aren't there. Just don't expect a 200' mic cable not to have a minimal amount of hum.
To invert the XLR connector you can use a DPDT switch. Slide, toggle, pushbutton, doesn't matter. The next module is the electronic, bouncless switching module, which you can also make a simplified version to do the XLR inversion, and have it digitally controlled. I'm presenting both schematics - mechanical switching and digital switching, in the next module.
As far as cost, about 5 years ago when I ran the bill of materials (BOM) on the "master schematic" I'm pulling and editing these building blocks out of, I hit about $8500 US in parts. This approximate price did not include things like knobs for the pots and faders or an enclosure... but it did include digitally controlled everything, as well as motorized faders for al of the input and buss channels (which are $35 each right there for something nice like Alps). Also, bulk quantities weren't taken into consideration either. Each resistor was priced out as if it was the only one, for example. That's when I dropped this project and started getting into other things.
Compressors - my inline design is
not an SSL equivilent. Not even close. It's a simple, basic, inexpensive compessor which "does the job" well enough for most things, and fairly noise free. If you want to build the amazing SSL, here is a clone project that works extremely well:
http://www.gyraf.dk/gy_pd/ssl/ssl.htm
The SSL compressor design also is a balanced design - most mixers (including my design) accept balanced inputs and outputs, but through the "middle" of the console things aren't balanced. Analog SSL's absolutely are - so keep that in mind. My compressor is very basic, and intended to help out with the understanding that I'm balancing affordability with having one for each input channel. There is a reason why analog SSL's were expensive.
And there are going to be mistakes in some of the diagrams, and I'll apologize up front. I'm manually re-drawing these from my master, large format hybrid console design (analog audio path, digitally controlled), so it's a bit tedius, time consuming, and I have a lot of other stuff drawing my attention away from this. So before you buy anything, critique what I post and hollar if you're unsure. Better to be safe than sorry.
I really thought cutting and pasting from the master design to building blocks would be significantly easier... but alas it's not. The hardest part is not missing anything when removing all the digital control stuff, as well as making sure I don't miss anything when making blocks. It's very easy to say "I'll put that component on the compressor buliding block" then when it comes time to draw that in-line compressor, to forget that particular part. I think we'll have to go over all of them side by side when completed, and see what's missing, and what's duplicated.
Also keep in mind these are building blocks - like legos. If you think my comrpesser design sucks, by all means, surf the web and find another design, rework what I've done, or do something of your own design. It's AOK, and why building blocks are fun. You can re-order things, change one or two blocks around, add features and subtract features. If I posted the master schematic of my design, most people would just say "oh fook" and not solder a thing... it's big, unweildy, and very cramped and difficult to read without constantly zooming around.
Again, no warranty expressed or implied