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

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
https://www.steampoweredradio.com/pdf/api/manuals/api 2520 op amp.pdf
This appears to have 2520 current draw specs. Surely you're driving something of 600 ohms or greater? Says max is 18mA driving 600ohms with +/-15V supply. So, if it's 72mA per stereo ACA, let's double that for some cushion. 144mA. Or a little less conservatively, 1.5X at 108mA. So, 432mA for 4 more ACA's. Hmmmm. Might want to take a close look at your power supply. Of course those 2520's in practice will not be drawing that much current.

Now how many IC's can you pull in the unused busses that will be replaced by ACA's? What is the max draw for those TLO72's (?) 5532's (?) or whatever is in there, added up?

This is how I'd proceed to think about it, but who knows what the current draw comparison between your IC chips and 2520's actually is in practice for their duties. I'm just going by max. Someone else knows way more.
 
Last edited:
Hello!

Just want to let you know, wow what a sound!
I'm currently modifying a small 8 channel Yamaha PM430 console with adding direct outputs. I'm using two Inv-ACA-revB boards with 8 CA0252 opamps and EA2623 output transformers.

The psu rails for the ACA boards are taken from the unregulated +-22V yamaha psu and regulated with a small homemade board with a few caps diodes and 78/7915 regulators. The PSUgnd for the ACA boards is taken from the yamaha psuGND.

The signal from all the channel strips is being picked up in between the EQ amplifier and the channel fader where i used a 47k bus resistor. I'm playing with connecting/disconnecting the shield of the signal cables at the channel strip sides, not sure yet because I'm still playing with the grounding scheme to get the console as quiet as possible
Essentially all is working and sounding great.

I have the problem of opamps are motorboating when not all of them are inserted, like a sawtooth-like oscillation. If all of them are inserted it appears to be quiet. Anyone have any idea about that problem?

Thanks!

Jeroen

IMG_20211026_143432.jpg
 
Last edited:
I just finished mine yesterday as well. Killer sound. I have a Yamaha PM2000 console. If you see my posts earlier in this thread, it was the first time I installed an ACA in it. I subsequently removed it after realizing the volume was too low compared to my other buses. I didn't know the circuit as well at the time and after one try at replacing the feedback resistors, I was still unsatisfied with the output volume and went back to my old buses. Silly me. Years passed and I eventually used Jeff's 2S-LA cards to build ACA busses directly into my bus channels. The setup has +/-24V at the moment, so I'm using Pierre's APP992 to sum and the Hairball 990 to boost. I have big Carnhill's at 1:2 for the first stage and then the Litz 2503's (I want more!) for the final outputs. Very much wow!
 
Are the regulators reducing the DC rail voltages, or just rectifying from the AC lines, and ignoring the dual 22VDC?
 
https://www.steampoweredradio.com/pdf/api/manuals/api 2520 op amp.pdf
This appears to have 2520 current draw specs. Surely you're driving something of 600 ohms or greater? Says max is 18mA driving 600ohms with +/-15V supply. So, if it's 72mA per stereo ACA, let's double that for some cushion. 144mA. Or a little less conservatively, 1.5X at 108mA. So, 432mA for 4 more ACA's. Hmmmm. Might want to take a close look at your power supply. Of course those 2520's in practice will not be drawing that much current.

Now how many IC's can you pull in the unused busses that will be replaced by ACA's? What is the max draw for those TLO72's (?) 5532's (?) or whatever is in there, added up?

This is how I'd proceed to think about it, but who knows what the current draw comparison between your IC chips and 2520's actually is in practice for their duties. I'm just going by max. Someone else knows way more.
Thanks JW. This is helpful, and I think operating from "max" seems like best bet in my head as well.

The console currently has 4 5532's and 6 TL072s per stereo group per group. Max draw at <10k is 13.5mA, and 10mA for 072s and 5532s respectively, if I'm making since of the Texas Instruments site. If I am thinking this through (and reading the schematic correctly) I should be able to get rid of all but 2 072s per module. So that should free up roughly 107mA per module?

I am also planning to loose 4 input modules to clear up space in the chassis for the ACAs and they each have 4 tl072s and 1 5532. Though I'm not sure that loosing power draw there directly tranlates over to the draw from the groups. I want to recognize my knowledge is thin here. I do have a tech helping me through, but I am trying to make sure I understand as much as possible, and get as much of this math out of the way ahead of time.

Thanks again for your response!
 
I'm considering using this with my soundcraft 600. Would anyone who has done something similar be able to recommend this? Also, how best to implement?
 
Are the regulators reducing the DC rail voltages, or just rectifying from the AC lines, and ignoring the dual 22VDC?
I upped the filter capacitors just after the bridge rectifier. That's where I pick up the DC and regulate it down to +-15V for the DOAs.
 
I'm considering using this with my soundcraft 600. Would anyone who has done something similar be able to recommend this? Also, how best to implement?
I have the ACA in the master bus of my Soundcraft 500 - which is essentially the same console. It made. a huge difference and works excellently. It took a little bit of planning for the grounding - but I am now working on plans to get the input section of the console as "clean" and "straight wire" as possible so I can put 4 more of these cards in all of the "group" outputs and use them as my record outs. My console has the "medium console" power supply at the moment and I've decided to search for a new supply that's got a little more power than I need, but at least the master section seems to be happy working off this one for now.
 
I have built simple summing mixer with Capi -aca- Bo (without inv-aca pcb’s). 24chanel summed before ACA passively with a resistor network.

Because ACA accept not balanced connections after summed network cold leg is grounded. Not favorite solution but works.

I will ask or one better solution is to put input trafo between passively resistor network and input for ACA?

If so can you advise good (price is not issue) trafos to do balance to unbalance and keep 47k network?

I have put trafo in the pictures for what I mean to ask.



Thanks in advance

1648550443432.png
 
You can also use the CAPI 2-stage line amp pcb as a balanced receiver instead of a transformer. You can use just the first stage if you want. But, there is not necessarily a problem with keeping it unbalanced. If you don't have noise issues there's no problem. Mine runs very quietly. If you are getting some noise, moving it or the other gear around it may solve the problem.
 
Hy thank you for reaction. I have not noise issues but if don't have noting connected at one of 24 inpunts I get oscillation issue out. For that mixer is must always have some device connected at the input. Thats reason that I wil use transformers to make more robust and not to stressed my RME fireface's. Problem is that i don't know which one wil be good for this setup and also 47k network must be stay.

Thanks in advance
 
Last edited:
I see now that CAPI 2-stage line amp pcb accept not balanced signal. Anybody have another idee?
Thanks in advance
 
Hi all ! I am in the process of rebuilding my summing mixer, built around the ACA summing. Anybody know what's its power consumption needs ? I use to power it with a cheap PSU delivering 1A per power rail. I want to replace this with a meanwell PSU and I don't want it to be too big, yet I know an overdimensioned PSU is never a bad thing :)
 
Hi all ! I am in the process of rebuilding my summing mixer, built around the ACA summing. Anybody know what's its power consumption needs ? I use to power it with a cheap PSU delivering 1A per power rail. I want to replace this with a meanwell PSU and I don't want it to be too big, yet I know an overdimensioned PSU is never a bad thing :)
Around 25mA per opamp per rail is typical. Of course figure some headroom in on top of that.
 
I am replacing the stereo master buss in a Auditronics Son of 36 Grand console with CAPI 2 ACA BO.I’m replacing the entire QMM module with the CAPI and only utilizing the Front left and front right channels to creat the Left and Right Stereo master . In the original auditronics design the Left front and Right front signal arrived at QMM Modules as a “A” and “B” buss . How should I deal with this with the ACA BO ?
Attached picture is Auditronics QMM Buss A and B on schematic
 

Attachments

  • 92E4A477-B8C4-4707-883B-AB8A83CFF3B1.jpeg
    92E4A477-B8C4-4707-883B-AB8A83CFF3B1.jpeg
    1.5 MB
Last edited:
Ok I'm running into what I'm assuming is an impedance issue. I set up an ACA board in my MCI JH528, running at 32v w/ JLM 99v opamps. Ideally, I want to be able to run both stock and API ACA summing options. However, If I have both hooked up and powered, the API bus is thin and trebly, and the stock bus is the opposite (rolled off high end). If I remove the stock summing board or power to the API board both sound great. Pretty sure nothing is wired wrong...tapped directly to motherboards for the new summing.
EDIT: I guess I never understood virtual earth ACN summing (https://www.forsselltech.com/media/attachments/summing_buss.pdf helped a lot)


So given the way the ACN bus works, is there a way to accommodate both simultaneously? Or would it have to be switching (assuming so)? Actually I breadboarded an SSL 4k summing section (pre vca) also so might try three options...
 
Last edited:
Ok so yes, to have multiple summing options available with virtual ground ACN summing you need to do 1) switching of summing amp inputs or 2) multiple summing busses. I'm using the multiple summing busses by repurposing my unused rear quad busses as the ACA bus. Modified routing on channels to split output before the summing resistors and it works and sounds great.

Glad I did this and also that I kept stock as well. On quick test I might use this new card more often, but there's a nice aggressive touch to stock too.
mci api.jpg
 
Back
Top