Siemens V275b Summing Mixer

Got a question regarding a little project of mine.
I have two Siemens Sitral V275b modules each being a 12:1 summing amp.
I would like to build a 24:2 summing box where I can sum all the 16 direct outputs of my Mackie Onyx 1640i mixer.
Now the question is, how do I have to wire it so that the panning/stereo information of my mixer channels doesn't get lost?  ???
If I have, for example, a stereo track on channel 1 and 2 on my mixer, one panned hard left and one hard right. Would I have to send one channel to the first V275b and the other channel to the second V275b?

On a lot of summing boxes I see mono and stereo inputs. What exactly is the difference between them? Are the stereo inputs just two mono inputs each panned hard right and left?

most summing boxes have  it set up so channel 1 is panned hard left and channel 2 is panned hard right. It continues from there for the full amount of channels. 

Some boxes do offer the option of taking any number of the stereo pairs and summing them to mono on an individual stereo pair bases.

So if you look at it like this all odd inputs go to the left amp all even inputs go to the right amp and all mono inputs go to both.
Plenty of information around here on summing, read the meta's.

Hey Pucho, thanks for your reply!
Yeah, that sounds logical.
I will search the meta.
One more question though.
I read somewhere, that you can change the unbalanced input to balanced by just cutting the connection of the input transformer primary to ground ( pin 2) and connecting the inputs cold legs through a 1k resistor to that transformer leg.
Would that work?
The V275b version I have is not identical to that one but the input arrangement is the same.

https://drive.google.com/file/d/15VjtgLxuiQusDIFWmgIrdjUGzIvRTh6c/view?usp=sharing

Murdock said:

I read somewhere, that you can change the unbalanced input to balanced by just cutting the connection of the input transformer primary to ground ( pin 2) and connection the inputs cold legs through a 1k resistor to that transformer leg.
Would that work?

Click to expand...

Indeed it would; taking advantage of the input transformer to operate balanced bus.
The link you added (schematic?) is a temporary one. No one can access it.

abbey road d enfer said:

Indeed it would; taking advantage of the input transformer to operate balanced bus.
The link you added (schematic?) is a temporary one. No one can access it.

Click to expand...

Sorry. I attached a google drive link to the schematic.
Thats good to know. So no drawbacks by disconnecting one of the primary legs from ground?
I'm wondering why they didn't do it like that in the first place then?

Perhaps it was designed to be driven from an unbalanced source in which case making it fully balanced would provide no benefit and probably result in some loss in level actually. But if you know the source is balanced then making it fully balanced would work and not only that but more than overcome said loss because debalancing gives you +6dB.

One thing I would check though is SNR. How much signal can you put it before distortion is too high? Then you will see how much noise there is. That's a really old transistor running on 0.1mA so I have to wonder what the noise floor is.

squarewave said:

Perhaps it was designed to be driven from an unbalanced source in which case making it fully balanced would provide no benefit and probably result in some loss in level actually.

Click to expand...

Mixers of the time used unbalanced bus, relying on very sturdy "ground" bus for signal integrity and low longitudinal noise. Balanced bus were not common because balanced pan-pots did not exist. German engineers thought that balanced bus should be driven by a balanced source.

But if you know the source is balanced then making it fully balanced would work and not only that but more than overcome said loss because debalancing gives you +6dB.

Click to expand...

I don't understand that.

One thing I would check though is SNR. How much signal can you put it before distortion is too high? Then you will see how much noise there is. That's a really old transistor running on 0.1mA so I have to wonder what the noise floor is.

Click to expand...

The input transistor runs at 50uA. The input transformer presents an impedance that is very clode to the OSI.

abbey road d enfer said:

I don't understand that.

Click to expand...

I know you understand that when you sum a balanced signal the result is +6dB so you must be wondering about the "loss" that would occur because the resistors and load make a voltage divider. If you have 3 elements in the divider instead of 2 you get more loss. How much exactly depends on the load represented by the transformer. If it's 2:1 step down into 33k then maybe it's nearly nothing. If it's 1:2 it could be 2 dB loss.

squarewave said:

I know you understand that when you sum a balanced signal the result is +6dB

Click to expand...

+6dB compared to what? Again I don't get it; I may be thick, though.

[/quote] so you must be wondering about the "loss" that would occur because the resistors and load make a voltage divider. If you have 3 elements in the divider instead of 2 you get more loss. How much exactly depends on the load represented by the transformer. If it's 2:1 step down into 33k then maybe it's nearly nothing. If it's 1:2 it could be 2 dB loss.
[/quote] I see that you don't realize that these summing amps have a rather low input impedance, in order to minimize cross-talk between channels. Typically, a period summing amp would use a 1:5-10 input xfmr, so its impedance would be about 100-300 ohms. For said summing amp, the bus injection resistors of 1k indicate a bus impedance of 125 ohms, which would suggest an actual input Z in the same ballpark. For a nominal operating level of 0dBu, the attenuation would be about 18-20dB and the signal at the transformer secondary about -6dBu.
Data of the BCY66 show it is capable of a very low NF (<1dB at Ic=50u) for a source Z of 10kohms). That would suggest a S/N ratio of about 100dB.

I don't know if that helps. But the spec sheet of the  V275b states an input z of >1K Ohm...
And a fixed amplification of 10dB.

I wish I could post all the documents I have but I promised the seller not to share anything... Even had to write a declaration. 

One thing I would check though is SNR. How much signal can you put it before distortion is too high? Then you will see how much noise there is. That's a really old transistor running on 0.1mA so I have to wonder what the noise floor is.

Click to expand...

Spec sheets states a THD of less than 0.7% at 40Hz with a +21dB output signal. And less than 0.3% from 60Hz to 15KHz.
Max output level is stated as +22dB.
Noise is stated as equal or less than -95dB.

Murdock said:

I don't know if that helps. But the spec sheet of the  V275b states an input z of >1K Ohm...
And a fixed amplification of 10dB.

Click to expand...

That's interesting. It shows the 1k resistor is dominant in the input impedance.

Spec sheets states a THD of less than 0.7% at 40Hz with a +21dB output signal. And less than 0.3% from 60Hz to 15KHz.

Click to expand...

That is likely to be mainly the output stage and OT's distortion.

Noise is stated as equal or less than -95dB.

Click to expand...

That is in line with my guesstimation.

So I guess this is not a 0 Ohm technology amp, right? Do I have to terminate the unused channels than?

Murdock said:

So I guess this is not a 0 Ohm technology amp, right?

Click to expand...

Correct.

Do I have to terminate the unused channels than?

Click to expand...

It depends. If you want to operate the unit as it was designed, for 10dB gain, yes. If you can live with some more gain and possible crosstalk, you don't need. Just try, you can't destroy anything.

Murdock said:

So I guess this is not a 0 Ohm technology amp, right? Do I have to terminate the unused channels than?

Click to expand...

It looks to me as 0 ohm "knotenpunktverstaerker" with already inserted input resistors (1k).

moamps said:

It looks to me as 0 ohm "knotenpunktverstaerker" with already inserted input resistors (1k).

Click to expand...

No, if you analyse the circuit, you can see that voltage feedback is used, that actually bootstraps the input impedance. My simulation shows that the input impedance is governed essentially by the transformer ratio and the secondary load.

I see. What's the transformer ratio?
It would be useful if OP can do a test with a signal and post here voltage readings on input, transformer's primary and secondary.

moamps said:

I see. What's the transformer ratio?

Click to expand...

That's the big unknown... Simulation shows that the ratio is probably 1:5-10, which would be adequate for matching the input transistor's OSI (about 10k) with the bus impedance (125 ohm).

It would be useful if OP can do a test with a signal and post here voltage readings on input, transformer's primary and secondary.

Click to expand...

Definitely.

moamps said:

I see. What's the transformer ratio?
It would be useful if OP can do a test with a signal and post here voltage readings on input, transformer's primary and secondary.

Click to expand...

Just injected a 1Khz sinewave 200 Ohm source impedance and 500mV into the input transformer primary.
Hooked up at the two pins the input voltage then fell down to about 438mV and on the secondar I read about 1.12V.
So about 1:2,5 voltage ratio?
I did not disconnect the transformer from the circuit... Hope that does not screw the readings.

I also just checked the "Prüfvorschrift" ( Test specification) and there it states that one has to terminate the unused inputs with 300 Ohm resistors. I don't know if that only counts for measuring and testing purposes though.

Murdock said:

Just injected a 1Khz sinewave 200 Ohm source impedance and 500mV into the input transformer primary.
Hooked up at the two pins the input voltage then fell down to about 438mV

Click to expand...

That computes to 1.4 kohm input impedance.

and on the secondar I read about 1.12V.
So about 1:2,5 voltage ratio?

Click to expand...

Yes. Unless...

I did not disconnect the transformer from the circuit... Hope that does not screw the readings.

Click to expand...

But was the module powered? That may change what the secondary sees.

I also just checked the "Prüfvorschrift" ( Test specification) and there it states that one has to terminate the unused inputs with 300 Ohm resistors. I don't know if that only counts for measuring and testing purposes though.

Click to expand...

That is necessary for ensuring nominal performance (mainly gain-related). In practical use, it's up to you; you may like the extra gain afforded by not loading the unused inputs.