Analogue outputs on Directstream

Hi, I’m new to the forum, so go easy with me. I’ve just bought a Directstream and am extremely happy so far. I’m running it balanced straight into my poweramp but also want to be able to listen through headphones. To do that I’m using single ended outputs into my headphone amp. I’ve read that you shouldn’t use both sets of outputs at the same time but am I right in saying that as long as only one of the amps is powered up, that there shouldn’t be any impact to the sound quality? Thanks!

Welcome, Mike!

Unfortunately it does not matter if an amp is powered up or not; it presents the same load to the DS powered or not.

As a practical matter you will not hurt anything so try it and see.

Elk, thanks for the quick reply - I’ll do some experimenting.

Please report back!

Elk is right.

I think you should try it since it does no harm. Maybe you do not experience the same sound degradation as I do.

Mike Marks said Hi, I'm new to the forum, so go easy with me. I've just bought a Directstream and am extremely happy so far. I'm running it balanced straight into my poweramp but also want to be able to listen through headphones. To do that I'm using single ended outputs into my headphone amp. I've read that you shouldn't use both sets of outputs at the same time but am I right in saying that as long as only one of the amps is powered up, that there shouldn't be any impact to the sound quality? Thanks!
I have my DS connected to a Levinson 502 for my main listening/theater system and then I use the single-ended outputs to a 3-way switch that I use to control inputs to the multi-room amp that drives my whole house audio. I've listened closely to the DS through the Levinson with the single-ended connectors connected and disconnected and I can't hear any difference. Not even loudness is effected. Maybe someone from PSA can weigh in but I suspect these two circuits are isolated.

Karl

ksalno said ...I can't hear any difference. Not even loudness is effected. Maybe someone from PSA can weigh in but I suspect these two circuits are isolated.
Nope, the unbalanced is paralleled with the "top" of the balanced outputs. But with any reasonable loads the loudness won't change. Tho it would be convenient it would add quite a bit to the price for there to be two independent outputs. As others have said, try it, if it works well for you it's fine.

So you can’t run this audio line driver in single ended output mode:

http://www.ti.com/lit/ds/symlink/drv134.pdf

i guess it would be challenging to feed it from the output trafo, so I guess you need to use the primary feed.

I think it’s not the problem of the DS but of your main and headphone amp. In my system I use an ASR Emitter I battery powered main amp for my speakers and an amazing tube headphone amp Little Dot MKVIII se for my Sennheiser HD800. Both work oustandingly with the Directstream. However they won’t work in a parallel manner, because the ASR Emitter is DC coupled and the Little Dot doesn’t like DC coupled main amps.
In my case I was lucky because the Emitter has a very sensitive protection to avoid any damage of the amp and of the speakers. For playing vinyl I tried to loop the line through the headphone amp but it didn’t work. So I have to plug in the main/headphone amp into the Phono preamp line. It’s not comfortable but an compromise.
The NPC would be a way out of this dilemma and I’m considering to buy one.

The best would be to show a schematic of the problem, but in esssense the SE output (RCA) is connected to one side of the balanced output (XLR) transformer. If the XLR output is not used, the load seen on the SE output normally constitutes about 47-200kOhm which is quite high considering the output impedance of the DS is about 600Ohm if I am not much mistaken. The trouble begin when you connect a load to both outputs enabling current flowing through both outputs. There exist an iteration influence between the two as long as the DS does not see an infinite resistance (like air) on the SE output.

Now, if the load is high this should not impact the +6dB ‘volume’ boost for the XLR to a significant degree before the load would drop to maybe as low as 1kOhm. Also (in theory) the sound quality should not be seriously affected, however my impression is that it suffers ‘big time’ during serious listening. Also note that the load changes with frequency.

“I’ve listened closely to the DS through the Levinson with the single-ended connectors connected and disconnected and I can’t hear any difference.” - ksalno

I had exactly the same experience in my rig.

@Frode: I remember that you had pretty significant negative effects on the bass. Was there anything else? My system is multi-amped and I have extensive control over the bass so I am able to negate these effects with a digital crossover which is not in the signal path to the higher stuff (line level RC). Funny, though, I only listened to the mid/treble portion of the system while alternatively connecting and disconnecting the unbalanced load (with the volume turned down on the bass amps). I made my adjustments to the bass with the whole thing up and running and never thought twice about trying to test this in the opposite way.

btw, I’ve settled on some new baffles for my speakers, inspired by yours. They are just prototypes now but I’ll post photos when I make the final product. I can’t reproduce the complex contours that yours have, but experimenting resulted in the outer “wings” dropping back at a modest angle and the inside ones in a steeper forward angle as wave guides. Took what, a year or so to arrive at this?laugh

The point regarding the bass is that my active stereo subs are connected to the DS single ended outputs and without my subs the lower end is obviously affected.

Where I have a SQ problem is that I perceive the mids to be somewhat smeared/muffled and recessed opposed to the SE output being disconnected. The difference is not huge, but still noticeable. One theory could be that the deflection of the woofer instigate amp noise that propagates back to the DS (superimposed on the IC signal) and is then forwarded over the XLR output to my main speaker amps. In this case, the added (skew) load is maybe not to blame, as such.

I’ll check again!

Great Northern Sound ( while they were in business ) sold a device marketed as the “PASI” or Passive Audio Single Isolator".

It was the stand alone version of the output isolation transformers that they used to modify the outputs on Wadia players and DACs.

It had a pair of single ended and balanced inputs and outputs.

It was basically several 1:1 isolation transformers for each channel and output type.

I wonder if connecting a PASI to the balanced output would allow you to have a balanced and single ended output that are isolated from each other?

I have a PASI, I’ll have to check with a meter to see how well isolated everything is.

I also believe that you could use both the single ended and balanced outputs at the same time.

I was curious about using high quality transformers as well, using them in the process of dividing the balanced signals to feed both “arms” of my setup. Sounds like a great question for Ted.

A transformer can isolate the grounds of the inputs and the outputs which can get rid of some groundloops which is good.

It can also help let the balanced inputs and/or outputs do their jobs of getting rid of common mode noise which is good.

But there’s really no difference between inputs and outputs on a transformer, i.e. any noise on any winding affects all other windings, e.g. it doesn’t isolate the noise from one output interfering with another output. With two balanced outputs this isn’t as much of a problem, but it can be a bit of a problem with an single ended outputs (or not depending on your system.)

So, in my case, if I used txformers I would still have common mode problems when the single ended equipment/cabling got involved in the circuit; “back talking” across the trans former. This would then be only one leg of the balanced circuits.

Whether that’s a real problem or not depends on your system. Many systems don’t have enough noise pickup in the unbalanced interconnects to be a problem, especially considering that the transformer will be getting rid of a groundloop on that unbalanced connection. All things considered I’d rather have a balanced connection, but some designers prefer the opposite…

“Where I have a SQ problem is that I perceive the mids to be somewhat smeared/muffled and recessed opposed to the SE output being disconnected. The difference is not huge, but still noticeable.” - Frode

Gave this a re-try for an extended period and I can say that at my house the sound is too close to call. I can’t tell a difference.