2520 LOSSLESS COMBINING AMPLIFIER

altitude said:

Thanks for your answer John

I wired up an NE5534 on some strip board and popped it into the sockets for the one of the 2520's

The circuit works perfectly.

There is a slight difference between the Ne5534  and the 2520, but as a working solution you would be hard pressed to tell.

Will be fine until I figure a way to repair the faulty DOA. Any suggestions?

My only real issue now is to find a way to reduce the 'crosstalk' bleed from left/right to the opposite channel.

Any ideas?

Thanks

Dave

BTW.... Here are a couple of pics....

Click to expand...

I can't tell much from the pix...

Crosstalk comes from multiple vectors.

1- electrostatic where it directly couples between paths due to intrinsic capacitance.
2- magnetic where the signal current creates a magnetic field, that gets detected by another path serving as a winding intercepting that magnetic field.
3- ground corruption where signals commingle in the grounds and show up in the audio.

#1 is mostly a HF phenomenon so look for this (things like wire dress, shielding etc) if crosstalk tinny sounding.
#2 is more of a hum source than audio corruption but could happen.
#3 if crosstalk is full range look at grounding.

Of course this is a gross oversimplification of console design, but will get you pointed in the right direction.

JR

analogical said:

Toobmooley-

My recording engineer and I have compared several of the popular DAW software programs. Nuendo, Samplitude and a few others. There are audible differences in the end result of summing the same data.

Click to expand...

My layman's hunch is the algorithms used in computing the pan laws would have the biggest effect on the summed signals.

Click to expand...

+1

Also if there is any EQ, we don't have a universal standard for equalization 'Q' so EQ settings may not track between platforms.

I remain interested in this. For now my best guess, if there really is some there somewhere, is sloppy floating point math, but even that should show up down in the noise floor and not as a major difference on large signals.

JR