fully balanced signal path?

0 Members and 1 Guest are viewing this topic. Read 2617 times.

radarnyc

fully balanced signal path?
« on: 5 Mar 2015, 01:51 am »
Hi Vinnie
I see there is an option for a balanced headphone output. Is the entire signal path on Lio's backplane balanced? Is the answer dependent on the options (DAC, tube buffer, phono)? Is this a dumb question  :D?
Thanks

Vinnie R.

Re: fully balanced signal path?
« Reply #1 on: 5 Mar 2015, 04:04 am »
Hi Vinnie
I see there is an option for a balanced headphone output. Is the entire signal path on Lio's backplane balanced? Is the answer dependent on the options (DAC, tube buffer, phono)? Is this a dumb question  :D?
Thanks

Hi radarnyc,

Not a dumb question at all.  LIO's "backplane" (aka, LIO motherboard, or LIO MB) is all single-ended.  The balanced output of of LIO's HPA is accomplished via a SE to BAL conversion at the output stage, giving 2x the signal swing (better for harder to drive headphones when you want to play at louder listening levels).

On the same topic, the upcoming LIO BAL OUTPUTS module will do a similar conversion.  And the upcoming LIO BAL INPUTS module will
do a inverse conversion from BAL to SE.  Once the signal makes it to LIO's motherboard, it is all single-ended and referenced to LIO's Ultracap bank negative (-) as the GND. 

- Vinnie

TomS

Re: fully balanced signal path?
« Reply #2 on: 2 Dec 2015, 02:10 am »
Hi radarnyc,

Not a dumb question at all.  LIO's "backplane" (aka, LIO motherboard, or LIO MB) is all single-ended.  The balanced output of of LIO's HPA is accomplished via a SE to BAL conversion at the output stage, giving 2x the signal swing (better for harder to drive headphones when you want to play at louder listening levels).

On the same topic, the upcoming LIO BAL OUTPUTS module will do a similar conversion.  And the upcoming LIO BAL INPUTS module will
do a inverse conversion from BAL to SE.  Once the signal makes it to LIO's motherboard, it is all single-ended and referenced to LIO's Ultracap bank negative (-) as the GND. 

- Vinnie
Vinnie,

I run 25' balanced DH Labs cables with XLR/RCA adapters at the LIO end to Acoustic Imagery Atsah mono amps XLR inputs. Being Hypex NCore based, it has a differential input and they do recommend balanced connections (pins 2+3) from the source, even if it's a single-ended signal on RCA's.

Can you elaborate a bit on topology of the balanced line I/O you're using? Is it passive such as the Jensen example below, which provides balanced impedance, but doesn't drive differential signals (or add gain)? Or if it's active/differential, is it a discrete stage?




Vinnie R.

Re: fully balanced signal path?
« Reply #3 on: 2 Dec 2015, 04:40 am »
Hi TomS,

Good question!

After experimenting with a few different configurations (passive transformers, active SE to differential conversion, and passive balanced impedance), I decided on going with the passive balanced impedance method w/o the transformers (similar to your Jensen example)

It simply sounded better to me (like nothing was being added to or taken away from the sound) compared to the others.

So no added gain on the LIO BAL OUTPUT.

BTW - the rear panel plates for the LIO BAL INPUTS and OUTPUTS modules should *finally* be shipping tomorrow.  Hopefully I'll have
them in a few days and then will get building them!

Vinnie

TomS

Re: fully balanced signal path?
« Reply #4 on: 2 Dec 2015, 01:33 pm »
Hi TomS,

Good question!

After experimenting with a few different configurations (passive transformers, active SE to differential conversion, and passive balanced impedance), I decided on going with the passive balanced impedance method w/o the transformers (similar to your Jensen example)

It simply sounded better to me (like nothing was being added to or taken away from the sound) compared to the others.

So no added gain on the LIO BAL OUTPUT.

BTW - the rear panel plates for the LIO BAL INPUTS and OUTPUTS modules should *finally* be shipping tomorrow.  Hopefully I'll have
them in a few days and then will get building them!

Vinnie
Great! That seems like the best way to preserve the pristine transparency of the existing SE outputs.