Which DOA for active mixer design?

Hello out there , after spending some weeks on general active mixer design its now time to to think about an DOA which can be used for the busses as for the aux-sends also..(and eq amp etc) I have read the Buttachunk thread in the mixer-meta but all the schemos are removed..?
I was thinking of diy-ing somthing but
after reading (all?) the threads in the DOA-meta i ´ve gathered some questions:
Which DOA Design would be your choice considering price (there are going to be quite alot of them) , Diy-ability (there are going to be quite alot of them) and maybe Heat/Current Issues ( there are going to be...!!) ??
..also should be fast enough taking all those transients together..

what about CMRR + PSRR , i think this could also gets serious even in a 16-24 channel design..?

I know there is no general "one does it all the best" but please give me your thoughts as I´m still quit a noob and would like to have just one topology to learn / understand so that i will be able to configure it for all (?) my purposes

Thnx alot ,

Axel

big fetbloak. Current feedback. sounds outstanding.

ok .. kept an eye on this one already , just read that the transistors could get quite hot here..should do my box not too tight then :wink:
... will stumble over more questions as the implementation goes on , now
let´s go for some Pcb then..
greets
ax

If you are worried about heat and you're going to have plenty of channels it might be worth you potting the bloaks if you stick with the 990 layout.

OKW has enclosures bout £0.60p each in small quantities. The potting compound can be had from the majority of standard suppliers like Farnell (I think)...just make sure it is thermally conductive or as I understand it, you will be mixing with a recipe for disaster as the compound will retain all heat and fry your amps...

As far as I'm aware the best cooling systems include plenty of room for unrestricted air flow around said heatsources as well as increased surface area for good dissipation.

You could re-layout the bigbloak like this if thats what you wanted...


There may be errors in that as its been a while since I looked at it, discarded it for my apps, but hey its an idea. The output darlingtons are heatsinked for extra cooling - should be cool. Also the output diodes are located near the o/p Qs and heatsinks and could possibly be thermally coupled too with some silicone glue...

Cheers Tom

Oh also, are you on about using the bloak as a summing amp or a buffer? I dont think its suitable as an active virtual earth summing amp in its current form (maybe I'm mistaken but the addition of a differential input might make it more suitable?).

If you need a buffer a simple emitter follower might be more appropriate.

Tom

thanks Tom..
mmh that could get even easier then.., but i think i have read somewhere that its possible just to take the 2.(?) stage of the bloak to have an inverter? i will need somtime to get into that stuff..i have to reread the whole thread

the Layout looks good ,space will be no problem , sure i would like to keep the 990 pinout to check it out sometime...potting false could make it worser as you mentioned , so i will go the heatsink way.
I am going to breadboard one of the bloaks and play around alittle..
Having some volts and ohms for my next questions...!!

Hi Axel,

I sent you a pm,
but also wanted to let you know that I have little pcbs for 4 different DOAs. All from the forum.

2520bc
1731
M990
SGA-SOA

PM me if you are interested..
take care
Gil

The 1731 might be abit much in a mixer, it is a colorful little thing and great for a couple preamps, but I think I would get sick of it very fast if I had a mixer full of them.

adam

If you want a methodical approach for your choice you have to draw up a few key specifications:
* maximum offset (anything above 20 mV prohibits the (reasonable) use of DC servos and direct coupling and needs AC coupling)
* maximum input bias current (above a few hundred nA you may need to compensate the input bias current at certain locations, e.g. in the fader buffer to avoid scratchy faders)
* output stage bias and drive capability (affects distortion and the ability to design with low impedances)
* rough direction of sound: clean/coloured (influences the need for low distortion, high CMRR and perhaps even high PSRR)
* supply voltage and current draw (has an impact on headroom, heat and the size of your PSU)

One of the key decisions is really the offset spec--with a high offset opamp you simply need to design pretty different.

Samuel

Ok. thanks everyone..
1. i´m not daggling with recording (almost not) this mixer should be alittle more then a summing box , but a lot less than a real console ..already build some nice gear from this forum and want to implement it(and the coming!!)


i would like to make something like 8 summing channels with L/C/R and routing ability to 4 busses/master and start with another 8 channels which should have aux sends also , for this all no pan , just a (stepped?)attentuator.inserts only for the busses/master.
master/busses/aux/returns should be just fast and uncoloured , and i was thinking of doing some different Input modules so that i can just swap around
tracks in the daw to colour to taste..
so mostly dealing with line levels from beat boxes and daw..but there is coming more pro level diy-gear..
output drive capability: do not really know how low it should go..all the sends should be shortening proof....yet i don´t have any gear that has to be driven at 24db/75ohms or lower .so what is practical here nowadays?
maybe all of my further DIY will have moderate impedance specifics
(i do not really understand yet how output impedance is affecting the input impedance of an OpAmp and so design of mixbuss resistance (over the feedback-path?-affecting avaliable gain etc?)
Samuel,how can i approximate the max.offset that will occure?
the Psu will be extern and can get as big as the projekt demands , i thought of separate regulators for each board or is this a bad idea?
voltage+-24V so supply in the range of+-28-30 v..

so lot of thx for your suggestions , i know all this has been talked about several times but getting things together can be really hard for the beginner.

p.s:Gil , i will contact you about some 2520 and maybe 1731 pcbs for the
inputs , but could take some time..
AX

Reading your above post it seems to me as if the 990 would be a good choice. It is easy to use and has very good overall specs which makes it an excellent all-round audio opamp.

You might consider bying them ready-to-use from www.johnhardyco.com, as you will realise that otherwise you quickly need half a dozen years to finish this project.

Otherwise there is the GainBloak which has very high offset, limited CMRR and some stability problems yet is very cheap and easy to build. Or you could go with my SGA-SOA-1 which has slightly worse overall specs compared to the 990 (yet has lower input bias current and current noise which comes in handy for many situations and is much cheaper to build). The 918 would fit as well, but there is no PCB design ready IIRC.

Local regulators are a fine thing, but you probably should get low drop-out versions to keep the dissipation within reasonable limits.

Samuel

Ok , after re-reading the the bloak threat i will defintifly breadboard some for testing.I think this one is quite easy to learn and cheap also , so i can make two or three of them and check in different conditions without taking care about smokin away some 50?or more..there are better ways doing this....
Have not looked at your SGA-SOA yet (sure i will right now), nor at the 918
Of course the 990 would be the best choice(just checkked out the listeningforums.com - and the 990apps were always on the winner side to my ears . As Samuel mentioned going prebuild there , but for all my amps this is much too expensive for me..maybe just as master summing amp/driver..
..when going the bloak way there will be that offset -issue , i read something like 1V at 22 V+/- , so there i have to use Caps for decoupling!.. a question:
how much offset would i expect with the other Opamps and (if fast to tell..)
if they have less offset under the same conditions how is this done...??)
So now let´s see what my lokal supplier has there for me...
I have time , but sure even if i would take readymade amps this will get not be done in weeks , but something under half a dozen years would be fine :grin:
Greets , Axel

How much offset would I expect with the other opamps?

Click to expand...

Below 20 mV with the SGA-SOA-1 or 918 and below 1 mV for the 990C.

If they have less offset under the same conditions how is this done..?

Click to expand...

They have different topologies and in the case of the 990 precision matched components.

Just to make sure: Offset is nothing inherently bad for audio, it just cuts some freedom in the design and may require more surrounding components.

Samuel

Samuel , just saw your excellent documentation for your SGA-SOA..just great reading!! no matching!
that should be worth trying also..In your documentation there is an interesting summing amp ..are there restrictions for using in inverting mode?


btw.the 990 kits ,do they solve most of the problems or is this one still too hard to setup..many people around ordered them so maybe someone can tell me here?
Ax.[/img]

I would recommend Samuel's opamp for this task also. The Bloaks are best as micpre gain units like the API312 or simple make up gain applications in a compressor, eq, etc. For the record, stability is only an issue with the darlington output versions that have way too much open loop gain.

Are there restrictions for using in inverting mode?

Click to expand...

Didn't understand your question? Are you asking why I used a noninverting make-up gain stage?

The 990 kits, do they solve most of the problems or is this one still too hard to setup?

Click to expand...

You can simply stuff all parts and be done with it, I think.

Samuel

Sorry Samuel , just forget the question (looked only at the "upper" half of your balanced schemo due to windowsize..)
I have to ask at the Blackmarket for some Kits to play.Maybe just a few left..
Up to now still big thanks for giving me all this additional thoughts..
For sure there will be more questions in the future...
Ax.