Output Selector?

[Mike Hanson]

My friend has a PS Audio DirectStream DAC (not the latest Mk.2), with both XLR and RCA outputs. He's planning to use it as DAC and pre-amp, because all of his sources are digital.

I built his Avondale SE200 power amp, which has RCA inputs, so he's planning to connect it that way.

He also has a headphone amp, which also takes RCAs. I suggested that he just get an XLR to RCA adaptor cable (like this one from Benchmark), with pin-3 floating. However, the DirectStream docs suggest that it's not a good idea to use both XLR and RCA outputs simultaneously. Additionally, he would need to turn off the power amp when listening to headphones.

As an alternative, I was pondering an output selector (like an input selector on a pre-amp). Could I use something as simple as a DPDT switch, or is that too crude? I have a Neurochrome Input Selector on hand (for an eventual pre-amp build), but it seems like overkill.

One of his sources is an Eversolo DMP-A6, which has analog outputs that could drive the headphone amp. Its DAC is good, but inferior to the DirectStream. My friend is always trying to squeeze the last drop of juice out of his system. Given headphones are perfect for critical listening, it makes sense to try to use the DirectStream, if possible.

Your thoughts?

 

[martin clark]

DPDT switch should be just fine!

My headphone amp uses a DPDT signal relay from a small toggle on the front, switch between two inputs. Your case is no different.

( An aside- the note about 'not a good idea to use both XLR and RCA outputs simultaneously.' is because doing that, you run the risk of bonding one polarity of output on the XLR to mains earth via the output-connected other units internal arrangement - which may be harmless, but also - might make for expensive fizzy noises. DP/DT switching fixes that for single-ended input/outputs. Call back if you get gross clicks on trying this - a couple of extra resistors will mitigate that.)

 

[Mike Hanson]

Thanks, Martin. I'll rig something up, and take it with me when I visit him next month.

 

[martin clark]

actually - if you are doing that with time to spare - re clicks - just stick say, 47K across the input and output side signal-to-0v terminals. 6no resistors required.

It'll have no effect on signal output / 'drive from an active device; and will minimise clicks if for some reason there's a leaky input coupling capacitor in one supplied device, or the other (by hold all down to 0v) R value is really , really not critical, whatever you have, 22-100K.

 

[Mike Hanson]

Thanks for the additional info. I'm working my head around this.

If I have that on all the 6 input and output connectors, assuming 47K, that will put 23.5K ohms in parallel with the chosen output load. I believe the input impedance of a typical amplifier might be 47K, so this additional resistance in parallel drops that significantly. This won't impact the performance?

 

[martin clark]

Yes, the effective input impedance 'seen' by the source drops
- but if a modern solid-state piece of equipment cannot drive say, resulting as low as 10K output load without any detriment, it's a POS and should be discarded anyway

Even weedy antediluvian triode pre-amp 'designs' can go there ...

 

[Mike Hanson]

I built the box, but I'm not sure if those resistors will be able to stay. To clarify the current configuration:

• The signal wires are fed through the DPDT switch.
• The shield wires are all joined near the RCA connectors.
• Each RCA connector (input and output, so all six of them) have a 68K resistor between shield and signal.

Consequently, due to those resistors, all signal pins are connected to all other pins, with one or more resistor inline depending on the path. Therefore, things are never fully switched.

If I removed those resistors, then the problem would be resolved, but then I might suffer with clicks on switchover.

Alternatively, I could probably use a 4PDT switch, so that the shields could be switched too.