Monitor management project

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello everyone,

Sorry in advance for the long read - I'd like to put together a simple-ish, modular-ish monitor management solution for my home studio and perhaps even turn it into an open-source project for other hobbyists' benefit if it works well enough and I'd like to ask for your help with some issues with my design.

The general outline of the project is the usual: a signal source selector followed by an attenuator followed by a destination selector. I'd like to be able to hook the device up to both balanced and unbalanced sources as well as destinations.

The interesting bit (I hope) is that I'd like to have full digital control over the whole thing. I'm imagining it as a box that sits inside my rack under the table hooked up to my speakers and soundcard and I'd have either only a software control panel accessible from my laptop or a small box on my table with a few buttons and a large knob or both. This (mostly) still isn't the complicated part, at least for me - I intend to use a Raspberry Pi to toggle a couple of relays to facilitate source and destination selection along with a PGA2310 or similar digital pot to provide attenuation. As I'm a programmer IRL and have some experience with a couple of small Raspi-based projects the digital and software parts of this project should be a piece of cake for me.

The issues I'm struggling with are all concerning the analog side of things. The PGA2310 IC expects a low source impedance and moreover works on a single-ended (unbalanced) signal, so I need a balanced-to-unbalanced input buffer stage before the PGA2310 and an unbalanced-to-balanced output driver stage after it. I also need to decide on the order of the building blocks in the signal chain - one option would be to have a buffer / driver stage for each input / output and the source / destination switching done within the unbalanced part of the circuit, the other option is to have a single buffer / driver stage and do the source / destination switching on the balanced side. I understand there would probably be a cost difference, but I have no idea what possible consequences each of these options could have for the audio signal.

I'm currently looking into the balanced line receiver and transmitter projects from Elliot Sound Products (see here for more info), but I'm unsure about a couple of things interfacing them to the PGA2310:

  1. The published schematic for the receiver has no AC-coupling capacitor on the output. Reading about the PGA2310 I found people experiencing audible pops during changes in the attenuation setting which were found to be caused by the source having a non-zero DC offset (albeit very small). Therefore I'd like to ask: Would it make sense to add an AC-coupling capacitor to the circuit output? I'm thinking something like 10uF would be a reasonable choice given the 10k input impedance of the PGA2310.
  2. The gain of the receiver circuit can be configured to some extent. Leaving out R7 and thus getting an overall gain of 2 seems like a reasonable choice from the "unbalancing" point of view ("invert -In and add to +In"), but I'm not sure if the PGA2310 will be able to handle that and I'm not sure how to go about computing it (assuming the selected source is a balanced output from e.g. a soundcard). Of course I could always turn down the output at the source, but I'd rather avoid having to do that - that's what this whole exercise should be about, and moreover I think it would raise the noise floor as well. So the question is: Would a gain of 2 at the receiver provide a suitable input level for the PGA2310? If not, how would I go about attenuating the signal from the receiver? (A simple resistor divider would change the source impedance as seen by the PGA2310, or am I missing something?)
  3. Which of the two signal chain options makes more sense? Buffers - switching - attenuator - switching - drivers, or switching - buffer - attenuator - driver - switching? Disregarding cost for now.

If I ever manage to finish this project, I'll put together a nicely wrapped package probably on GitHub with all the documentation and source code so that others may just go directly to etching their PCBs :)
 
Normally you cannot use both unbalanced and balanced inputs at the same time.
You would need to have selection between them, switched or physically disconnect.
Similarly a single balanced output cannot supply both balanced and unbalanced at the same time. You would need to Buffer the outputs so that the balanced does not get corrupted by the unbalanced.
 
Balanced impedance attenuation is not easy to do well.
Have a read of B.Putzeys article where he inputs balanced - converts to unbalanced - attenuate/gain - then converts back to balanced.
He uses relays to switch between 2 balanced inputs and has one balanced output.
 
Thanks for the replies and sorry for not getting back sooner, for some reason I didn't receive a notification..

I think there's some confusion about what I want to do, so let's clear that up :)
  • There will only ever be one input and one output active at any given time.
  • I'm not intending to use the PGA2310 on balanced signals - all inputs will be converted to unbalanced first (either each input by its own buffer followed by a relay network to select the active input, or a relay network followed by a common buffer for the selected input) and the outputs will be then converted back to balanced (again either buffer -> relays or relays -> buffers).
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.