[sdiy] OT: Audio line levels

[Byron G. Jacquot]

>Thank you Peter, and happy holidays to you!
>
>What I am have having to contend with is the choice between using the 
>floating transformer isolated inputs or the +4 unbalanced recieve inputs. 
>This is a real dinosaur:
>
>http://i108.photobucket.com/albums/n13/somnium7/gear/ToaBoard01.jpg
>http://www.toaelectronics.com/disc/manuals/RX-7-164_248_328_IM.pdf
>
>According to the specs I have a fairly large amount of channel headroom 
>(clipping at 12.3V-24dB) and there are 3 levels of padding for the tranny 
>inputs (-20, -40, -60).
>Do you think I will have decent performance if I use the transformer inputs 
>and ground the cold pin at the instrument output?
>Or will I absolutely need a DI box?

How much of a dinosaur is this?  Is it old enough that there's some room in the chassis to as some new circuitry?  And is technology under the hood something you could reverse engineer?  Those transformer-isolated inputs are designed for microphones, not line outputs, and have a rather low input impedance...the 1.2K and 1.7K settings are on the edge of hostility to some output stages.

If it's a common transformer into non-inverting opamp stage, one way to add a line input is to add a jack that bypasses the transformer, feeding the input opamp.  This also avoids the possibility of accidentally putting phantom power into a synth output.  The jack could be a switching one, so it bypasses the transformer input, or you could add a toggle switch.

Of course, the RCV jack is already built out to 20K input impedance, so it's maybe a better place to interface.

If you'd rather not mod the thing, hunt around (Ebay...) for a "bump box," -10 to +4 converter.  They're pretty common in broadcast installations, because most broadcast gear is strictly +4, but -10 front end gear is becoming more & more common.  ART and Henry Engineering are two vendors I can think of right now.

>In the past I've connected -10 outputs to +4 inputs on effects boxes and 
>sometimes it was very noisy and other times it didn't seem to make a 
>difference.

Do you know the input impedances?  Many solid-state outputs get noisy when asked to drive a 600-ohm input.

Also, correctly balancing & unbalancing can prevent some noise headaches...but needs to be handled on a case-by-case basis, depending on the type of circuit.

Byron Jacquot