5534 opamp based active summing with EV10862 and 10852 (added schems)

[mylesgm]

Does anybody have some info on how I might build an active summing mixer with 16x 5534 based line amps?

I have 4 neve 10862 cards each with belcere tf10016 tx preceeding a 5534 line amp/buffer.  Sadly I have no schematics but it looks like it is a simple circuit.  I also have 2 EV 10852 cards which are 2 channel discrete line amps driving willesden "pumpkin" 1:1 transformers.  I have schematics for these and it shows 5 non-inverting mix inputs per channel.  These come from a Neve V-series mixing desk.  I would like to build this into a 16 channel active summing amp using the 10862 cards as the front end, 1 10852 as main output and 1 as monitor out..  Any idea how I might do this or resources I might read up on to get an idea?

M

 

[Walrus]

Did you not receive my email with the 10862 schematics?

 

[mylesgm]

nope, I haven't received the schematics.  mylester  (@)  gmail.com  Would love some schematics...

 

[gemini86]

could you post these to the tech docs?

 

[Walrus]

OK, done. It's a big diagram so I've posted an overall picture and 3 detail pictures.

 

[mylesgm]

Thanks Walrus, I've just been reading them.  Any idea what the inductor is for?  Anybody have any thoughts for upgrading this circuit if necessary?

 

[mylesgm]

So of course the inductor limits high frequency response (not sure to what corner freq) but this needs to go so I'll remove it from the circuit.  Any suggestions on circuit for summing?

 

[mylesgm]

here's the schematic of the 10852 line amp.  There are 5 non-inverting mix inputs each with a 10k input resistor so this would allow me to mix 5 channels left and 5 channels right.  If I want to up that number to 8 channels left and 8 channels right should I simply add an additional 10k resistor for each channel or is it a little more complex?

I also want to add a master output gain control and a monitor output.  As I read the schematic I think I could use an off board pot either replacing RV1 or between the pins C31 and C29 and I assume it would be a 10k log pot here (I would use a stepped stereo).  For the monitor out I think I should use the unbalanced output and if the circuit is capable of driving a second EM20038 tx simply add another so that I have the exact signal that the main buss is outputting.  I have more EM20038 txs.  Or should I use the second 10852 card that I have as a buffer with the unbalanced output going straight into the non-inverting input C30?

Is the output return C16 simply ground referenced to the output and the input return ground referenced to the output with the 100r between them isolating the ground potential or what?  I've never seen that before.  Though as you can gather I'm not that flash with this stuff.

 

[mylesgm]

description 1 is here

http://www.groupdiy.com/index.php?action=dlattach;topic=44877.0;attach=9066

 

[mylesgm]

description 2 is here

http://www.groupdiy.com/index.php?action=dlattach;topic=44877.0;attach=9068

 

[PRR]

> the inductor limits high frequency response (not sure to what corner freq) but this needs to go so I'll remove it

NO!

5uH at 20KHz is 6.28*5u*20K= 0.6 ohms. "Nothing". If loaded in 600 ohms, it would not have large effect until 20 MHz. In 60 ohms, 2MHz.

This is NOT limiting audio response!

What it is, is a little complicated. Any good NFB "audio" amplifier has response far past the audio band. All loads (especially cables) have loading effect at high frequencies. The combination can sometimes get into trouble, oscillating hypersonically. The most common fix is a _small_ coil in series to the load. Look on many loudspeaker-amp plans. Find the docs on Jensen 990.

 

[mylesgm]

Thanks PRR, I was expecting it to be HF limiting factor like the neumann/lawo amps have to cut the response down to 18 or 20k.  Any thoughts on adding extra channels to the summing input?

 

[mylesgm]

I'm still working on this, I'm sure there'd be other interested parties out there with these cards so does anybody have any further advice on how to approach the summing on these cards?  Or even just doing summing through the 10852 card which already has 5 input (10k resistor) passive summing?  I'd like to do the summing balanced so perhaps a balanced resistor network (up to 16 channels) feeding a suitable transformer before hitting the makeup stage on the card.  Help or thoughts?

 

[mitsos]

Looks like an opamp? Take a look at the CAPI ACA-Bo card, you could probably do the same thing with those cards. Just take your channels through a resistor to the inverting input and use the trafo to correct polarity.

 

[quantyk]

Myles -

Working on a similar project based on this card, and have been playing around with for a while now. the mix inputs on the card work nicely for single-ended sources, but I'm also curious about a balanced configuration, as I plan to drive this with balanced sources.

in research on balanced summing circuits, I've seen it suggested to feed the summed + & - signal buses to the non-inverting and inverting inputs of a differential amplifier, which the 10852 seems to be.

i think this is the kind of arrangement David Self is referring to in the subchapter on summing systems, and perhaps closer to the intended use of this card.

I know this is an old thread, just getting back to grips with the 10852 project i've had on the shelf for a while now.

Would there be any problems with this approach:

Balanced summing bus, with separate resistor-fed buses for signal + & -, with the sums fed to the non-inverting and inverting inputs of this amp.

 

[quantyk]

HI - I started another thread just focusing on the 10852 card.

Neve 10852 Line Output Amp Card