[BUILD]CAPI 2-ACA-Bo~Official Support Thread

[horvitz]

I myself see nothing wrong with this especially if you don't have really long cable runs. PRR has said many times around here that there is nothing wrong with unbalanced cable runs in a studio.

FWIW, the majority of the connections in my console are unbalanced. Maybe they did not have the same concerns back then with interference but my desk is pretty damned quiet, even now. Anyone who has built the 2-ACA-Bo knows that it is unbalanced busing. I know that balanced busing was one of the things that Paul Wolff introduced when he took over API.

Best, Jeff

Oh, I definitely agree.  If I have 50' of cable from a mic to a board it's a different matter but I certainly feel comfortable with the noise level in my contained little environment here which is FAR smaller than your desk.  I guess the only thing to do is to take some measurements when it's all together.  We'll see.

  Brian

 

[horvitz]

The inputs will get a +4 balanced sig from a pro t, Brian have you seen how dangerous 2 bus does the switch? It's all stereo pairs and when you hit the switch it's 2 monos(for kick sn bass center voc)
So one bank of 8 would be all you need for mono switch 4 swt = 8 monos..
I want 24 because the more chs you sum out the better the results..
I am still not clear on how many ins per doa , the diagram show 8 ch per ?

Yes, I had a look around at the different channel assigning mechanisms and I wanted to build something as flexible as possible, so I made every single channel assignable L-C-R.  The way Dangerous does it is probably fine for just about everything you might need to do.. look at your mixes and see what you might require.

And yes, I will have a maximum of 8 ch per DOA - 16 L and 16 R per 4-ACA board.  So if you don't want to exceed that number, you could have a total of 64 L and 64 R (4 boards of 4-ACA) bussed into the 2-ACA-Bo.  Thus, my question regarding sizing the PSU.  I'm sure these aren't hard limits, but it's the way I'm putting it together and will be FAR more than I'll ever really need (famous last words!).

Once I get the whole thing cobbled together in the next week or so, I'll be happy share what I did.

  Brian

 

[jsteiger]

That diagram was basically for my own mods and use. Sure anyone can use it though. The 2 mix on my console, which is essentially a single 2-ACA-Bo card, was summing 32 input channels plus 4 echos with no problems. I decided to add the sub-summing Inverting ACA boards since I am now also summing the 32 monitor inputs and those echos to one stereo mix bus. Now I am up to 72 channels at the 2-mix. I had read about the massive O'Henry API console designed and built by Harold Kilianski. Steve Firlotte also did some work on this huge board. In one of my few conversations with Steve, he told me they employed "bucket summing" much like Neve and SSL. They have some 88 inputs at mix. The Inv-ACA was my API like solution to "bucket summing".

Cool O'Henry Mix article

Dale Manquen article on O'Henry's API

Very interesting stuff, IMHO.

Cheers, Jeff

 

[Johndcx]

Yeah the cable runs it most cases be short, mine will be about two ft interface to mixer, but i am wondering if the transfo or line card input is where the fatness comes from,like you said in another post Jeff, running though the API sounds like a record..

 

[jsteiger]

Yes, this is a big part of the API all discrete signal path.

For a typical line input to 2-mix on my console, it goes something like this:

-2520 balanced receiver DOA on input channel (DOA count: 1)
-2623-1 x-fo balances input channel's 1st stage output or "send" (X-fo count: 1)
-2 2520's on typical 500 slot eq (DOA count: 3)
-2503 x-fo balances output of eq (X-fo count: 2)
-2520 post fader booster of input channel (DOA count: 4)
-2520 on my Inv-ACA (DOA count: 5)
-2520 on 2-Mix ACA (DOA count: 6)
-2623-4 x-fo balances output of ACA (X-fo count: 3)
-2520 on 325 post 2-Mix fader booster (DOA count: 7)
-2503 x-fo balances output of 325 booster (X-fo count: 4)

So, that's 7 discrete opamps and 4 transformers. Granted, it takes a lot more than this to "make a record" like a great song, great artist and a great performance but, it is definitely a mixing signal chain that is hard to beat.

Cheers, Jeff

 

[Johndcx]

Hey Jeff
could I use your 4ch aca cards for the line inputs, by changing a few caps & resrs
the 325 line card looks real close,that would make it complete..
Line doa to aca to booster= fatness??

 

[Tubemooley]

The answer to that would be yes. I am doing the same thing. A few trace cuts and careful mounting of components without a thru-hole. That should work fine. DW.

 

[SIXTYNINER]

HEY Jeff
do you mean the "****" console in the picture attached?

That diagram was basically for my own mods and use. Sure anyone can use it though. The 2 mix on my console, which is essentially a single 2-ACA-Bo card, was summing 32 input channels plus 4 echos with no problems. I decided to add the sub-summing Inverting ACA boards since I am now also summing the 32 monitor inputs and those echos to one stereo mix bus. Now I am up to 72 channels at the 2-mix. I had read about the massive O'Henry API console designed and built by Harold Kilianski. Steve Firlotte also did some work on this huge board. In one of my few conversations with Steve, he told me they employed "bucket summing" much like Neve and SSL. They have some 88 inputs at mix. The Inv-ACA was my API like solution to "bucket summing".

Cool O'Henry Mix article

Dale Manquen article on O'Henry's API

Very interesting stuff, IMHO.

Cheers, Jeff

 

[Johndcx]

I think I have to follow your lead on this Tubemooly, I am to much of a newb...

What's your plan on pwr supply?
The x51 may work with a bigger transfo,
with 24 ch I will have about 32 doa's in mine..
I had a vintage Amek that had a big supply so I am somewhat fimilar with console supplies.

 

[PATAPETER]

Jeff : I have 2 modules 528, 1 is a vintage ,circa 1972 , and the other is the datatronix era that you've saw in the pictures . The datatronix one has a different 2622 footprint, the room is for a jensen JE 6110K-A PC . And that footprint is the same as the ED's xfmr . Following the tracks it's seems that there's a little mod on them. the way that the pad switch is connected , and there's an R 5 that it isn't on the schematics that us have.( try to find R5 on yours).i will post you a pict tomorrow, now is too late.
regards.
pedro.

 

[PATAPETER]

here the picts: see the bottom switch tracks.they are different. and R5 is with R34 and on the other leg to gnd.

 

[Tubemooley]

I think I have to follow your lead on this Tubemooly, I am to much of a newb...

What's your plan on pwr supply?
The x51 may work with a bigger transfo,
with 24 ch I will have about 32 doa's in mine..
I had a vintage Amek that had a big supply so I am somewhat fimilar with console supplies.

Johndcx - Where are you located? If I can't just purchase an off-the-shelf linear power supply (such as Power-One), then I build one. I believe you are correct that the 51x power supply board will handle more current than it's designed for now using the standard build. But we should confirm that. I'm working on a discrete mixer of my own (slowly). It's going to be line input only. I don't need mic pre's for this mixer. It will have 16 input channels and then 8 FX returns mixing down to a stereo bus. I have most of the chassis hogged out and six or seven of Jeff's boards mounted in there and a bunch of transformers mounted too. I've been a little bogged down figuring out / getting / building all the pushbutton switches for the routing and such. I need 5 DPDT switches per channel. I can't remember what all 5 are needed for right now. I will dig out my notes. So getting these switches figured out and mounted in my chassis has bogged me down considerably for a few months. But I think I have that figured now. Send me an e-mail at tubemooley_at_yahoo.com and I can forward you a few photo's if you want. I'm more concerned with helping answer your questions on power requirements for your project. Do you plan to use DOA's which run off of +/-16vdc or DOA's which run off of +/-24vdc? Or both? Let's get a few parameters laid out here and then we can do some figuring. DANA.

 

[Johndcx]

Yeah Jeff said the 51 power can be modded up and just built with the + - rails,
at this point iam not sure what v I 'll run at..
I thinking of doa's on the lines (325  style) to aca to booster ,so with 24 ch I'll have about 36 doa's after control and headphone outs?
I may just keep this line only and do pres later,I am thinking of a descrete style Danger 2 bus,also with only 1 bank of 8 mono switches, all rest stereo pairs..
I d like to have some differant sounds on each bank of 8 chs,
by using differant Doas, maybe some trsfos,and maybe that neve line card that's around here...
My front panel will only be 8 mono switches and maybe signal LEDs and
master vols ..
I have to get a stereo pair going first, then go from there, I have a long way to go..and the woods are dark and deep...

 

[herrmann]

Hi
I'm testing my small mixer. Sound is clean, but I notice that the negative 2623 output is twice the positive. Is there any reason ?

Thanks

 

[jsteiger]

How did you orientate the 2623-4?

 

[herrmann]

The blue mark to the 1.

 

[jsteiger]

The blue mark to the 1.

Yes, this is correct.

Funny enough, I have never made the same measurement as this, from Hi to gnd and from Lo to gnd. I have only measured between Hi and Lo, the normal way.

I checked one of my PCB builds and found similar results to yours. Not only on the ACA side but the 325 Booster side as well.

After some investigating, I see Bill Whitlock has written some things on this subject. This is quite normal I guess. If you do a Google search, you will find many interesting things from Bill on this, like this reprint here. I guess there is a big to do regarding the myth of the balanced signal conductors having symmetrical voltages when having the same impedance is of much more importance. He has some nice info in chapter 32 of The Handbook For Sound Engineers too. To quote, "The idea that balanced interface is somehow defined by signal symmetry is simply wrong!".

It almost seems like this is a topic for a dedicated thread! A very interesting one indeed.

Cheers, Jeff

 

[herrmann]

Funny enough, I have never made the same measurement as this, from Hi to gnd and from Lo to gnd.

Cause I only have one probe on my scope ! I have to get another one someday.

So this is not something to worry about. But in this case, we are loosing signal (and increase the noise floor) in the master fader, no ?

 

[jsteiger]

Should be no increase in noise as the impedance (not signal level) between each conductor to ground should be equal.

 

[edanderson]

measuring on the secondary side of output transformers, you have to remember the that the hot and cold ends of the transformer are floating.  the signal is not relative to ground (as established inside your circuit or anywhere else) so if you measure with one meter lead to the hot end and one to ground, you could get just about any fraction of the nominal value.  this will change if you ground either end of the winding when you measure -- you could also ground the center tap, which would "force" the hot and cold ends to give you the results you expect.

so, this is not a problem, this is exactly what should happen...

ed