Parasound Halo P 5 Stereo Preamplifier Preview

[Irvrobinson]

Comparing two identical circuits, one being single ended, one being fully differential, it's very easy to measure a difference in noise and distortion. It's a very trivial matter actually. Audibility is another story however.

Perhaps with signaling at 500MHz and above that's true, but at audio frequencies it is difficult to see that. I've been reading test reports for years on audio equipment, and I've never noticed a correlation between lower distortion measurements and equipment that is fully differential. Have you, at audio frequencies?

 

[haraldo]

IRV, do you have any integrated Krell's around there, would be interesting to hear if you can spot any differences between singled ended and balanced connection from the DAC1
I'm not the only one around here suggesting this, would be good to know if I'm just imagining this or .... what the heck this is?

 

[Irvrobinson]

IRV, do you have any integrated Krell's around there, would be interesting to hear if you can spot any differences between singled ended and balanced connection from the DAC1
I'm not the only one around here suggesting this, would be good to know if I'm just imagining this or .... what the heck this is?

I use the DAC1-HDR as a pre-amp and DAC, which is what attracted me about the design, so that I wouldn't need a separate pre-amp. I figured saving a set of cables and an analog gain stage might be worthwhile. When I tested the DAC1-HDR with the single-ended amp, the ATI AT602, I did it both ways, once with the single-ended outputs from the DAC1, and once with the XLR outputs and adapters at the amp side, and I couldn't hear any difference at all. Of course, these were fully sighted tests, so make of them what you will, but no matter how I've tried to hear differences I have been unable to. During one session, just to satisfy my curiosity, I did a test at about 3:30am (I couldn't sleep) when the ambient noise in the room was below 30db, thinking that in my previous test perhaps the room wasn't quiet enough, and I still couldn't hear any differences, even on solo violin. Frankly, I'd have to say, haraldo, that you probably are imagining any impressions of "more grain".

 

[slipperybidness]

Usually they use a differential input op-amp and then convert to single ended. Depending on grounding scheme, its possible to get some noise immunity benefits but not lower distortion.

Ahhh. That makes sense. Thanks for a quick, well informed response.

 

[gene]

Perhaps with signaling at 500MHz and above that's true, but at audio frequencies it is difficult to see that. I've been reading test reports for years on audio equipment, and I've never noticed a correlation between lower distortion measurements and equipment that is fully differential. Have you, at audio frequencies?

Of course, its a mathematical certainty at all frequencies.

See figure 4 of my measurements of the XPA-2 amplifier comparing single ended vs bridged mode:
Emotiva XPA-2 Measurements and Analysis | Audioholics

Differential amplifiers are absolutely necessary for small signal amplification where weak signal are contaminated by noise pickup, ie. signals transferred over long distances on twister pair, DC amplification where no coupling capacitor is used b/c the symmetric design is inherently compensated against thermal drifts.

Here is a good read:
Fully differential amplifiers remove noise from common-mode signals | EDN

Really puzzling why someone would want to downplay the design superiority of a fully differential design topology. No engineer would EVER claim single ended is equivalent.

 

[Irvrobinson]

Of course, its a mathematical certainty at all frequencies.

See figure 4 of my measurements of the XPA-2 amplifier comparing single ended vs bridged mode:
Emotiva XPA-2 Measurements and Analysis | Audioholics

Differential amplifiers are absolutely necessary for small signal amplification where weak signal are contaminated by noise pickup, ie. signals transferred over long distances on twister pair, DC amplification where no coupling capacitor is used b/c the symmetric design is inherently compensated against thermal drifts.

Here is a good read:
Fully differential amplifiers remove noise from common-mode signals | EDN

Really puzzling why someone would want to downplay the design superiority of a fully differential design topology. No engineer would EVER claim single ended is equivalent.

All right, first of all, I want to give you special credit for coming up with such a great illustration as the XPA-2 bridged versus unbridged cases. Nicely done.

That said, my original point was regarding line-level components, not power amps, and you happened to choose a power amplifier for your example that had single-ended performance that wasn't up to snuff by your own measurements compared to another single-ended amplifier from the same manufacturer (the RPA-1). My point was, specifically, that it seems possible for amp designs using only differential output stages, and not being differential in all stages, to have distortion and noise measurements that aren't as a group distinguishable from fully balanced designs. I'm not sure that's true, but as a fan of fully balanced designs I've always tried to see if I could notice that correlation, and I can't.

Back to line-level components, now I really can't see the correlation between better measurements and fully balanced designs. Not because I don't believe differential signaling is advantageous, but because the advantages are apparently swamped by other factors, the advantage has fallen below the threshold of measurability, whatever, or the difference would be easy to see.

Finally, I am not arguing that signal-ended topologies are equivalent to differential topologies in general. That would be ridiculous. I am arguing that for line-level audio components I have yet to see proof of a measurable advantage among real-world products. Don't interpret my position any other way, because it isn't intended that way.

 

[gene]

All right, first of all, I want to give you special credit for coming up with such a great illustration as the XPA-2 bridged versus unbridged cases. Nicely done.

That said, my original point was regarding line-level components, not power amps, and you happened to choose a power amplifier for your example that had single-ended performance that wasn't up to snuff by your own measurements compared to another single-ended amplifier from the same manufacturer (the RPA-1). My point was, specifically, that it seems possible for amp designs using only differential output stages, and not being differential in all stages, to have distortion and noise measurements that aren't as a group distinguishable from fully balanced designs. I'm not sure that's true, but as a fan of fully balanced designs I've always tried to see if I could notice that correlation, and I can't.

Back to line-level components, now I really can't see the correlation between better measurements and fully balanced designs. Not because I don't believe differential signaling is advantageous, but because the advantages are apparently swamped by other factors, the advantage has fallen below the threshold of measurability, whatever, or the difference would be easy to see.

Finally, I am not arguing that signal-ended topologies are equivalent to differential topologies in general. That would be ridiculous. I am arguing that for line-level audio components I have yet to see proof of a measurable advantage among real-world products. Don't interpret my position any other way, because it isn't intended that way.

There will ALWAYS be a measurable difference between single ended and differential designs, especially for low line level signaling. This assumes the test gear you measure with is good enough to measure the difference. BTW, the Audio Precision APX-585 and all of their test gear uses fully differential designs from input to output. No good test measurement equipment is designed single ended. Be leery of the websites that use PC software and a soundcard to measure amplifier performance as in many cases that type of test rig won't be nearly as accurate or revealing. There is a reason why we use Audio Precision test equipment. It's the best on the market and its....fully differential

 

[Irvrobinson]

There will ALWAYS be a measurable difference between single ended and differential designs, especially for low line level signaling. This assumes the test gear you measure with is good enough to measure the difference. BTW, the Audio Precision APX-585 and all of their test gear uses fully differential designs from input to output. No good test measurement equipment is designed single ended. Be leery of the websites that use PC software and a soundcard to measure amplifier performance as in many cases that type of test rig won't be nearly as accurate or revealing. There is a reason why we use Audio Precision test equipment. It's the best on the market and its....fully differential

Gene, you're stating the obvious and you are missing my point. Let me ask you a question to bring this discussion back around. If you measured the P5 and found that it measured similarly to the Benchmark DAC1, meaning that the noise and distortion were on the edge of measurability even with the best AP test equipment, would you reject owning it, simply because you would know that the noise and distortion could have been several decibels lower if Parasound had gone to the expense of making the P5 a fully differential design?

 

[gene]

Gene, you're stating the obvious and you are missing my point. Let me ask you a question to bring this discussion back around. If you measured the P5 and found that it measured similarly to the Benchmark DAC1, meaning that the noise and distortion were on the edge of measurability even with the best AP test equipment, would you reject owning it, simply because you would know that the noise and distortion could have been several decibels lower if Parasound had gone to the expense of making the P5 a fully differential design?

Probably not but the P 5 will not be below measurability of the AP test rig.

 

[haraldo]

Frankly, I'd have to say, haraldo, that you probably are imagining any impressions of "more grain".

I do have a great imagination

 

[PENG]

IRV, do you have any integrated Krell's around there, would be interesting to hear if you can spot any differences between singled ended and balanced connection from the DAC1
I'm not the only one around here suggesting this, would be good to know if I'm just imagining this or .... what the heck this is?

I think it would be easy to say you are imagining it if you cannot pass a blind listening test, and I bet you can't. To conclude that such a blind test is a good enough proof would be a different story.

I may not have read as many reviews (with measurements) as IRV have, but for all those I read I also do not recall one that shows the truly differential ones measured better in terms of noise. As to distortions, either way they are typically much too low to argue about, I mean like 0.00X% THD+N.

 

[haraldo]

Music and audio is not about what you can hear or not hear.....

Music is about pleasure.... It's how it makes you feel, how do you measure this

 

[Irvrobinson]

Music and audio is not about what you can hear or not hear.....

Music is about pleasure.... It's how it makes you feel, how do you measure this

I'm a measurements sort of guy, so I would say with an Audio Precision APx585.

If it sounds good but measures bad, it is bad.

 

[RichB]

I read somewhere that Amps that are fully balanced differential designs increase efficiency and can lower the noise floor.
The reduced distortion tends to be the even harmonics not the more objectionable odd harmonics.
Is there any truth to that?

Does full balanced circuitry reduce IM?


- Rich

 

[sharkman]

Wow, I'm puzzled at the prolonged discussion on balanced vs single ended design. The P5 is one step up from Parasound's entry level 2100. It has a lot of value at it's price point with bass management, a DAC and theater pass through, but balanced designs are not cheap! Having a DAC with digital input options should really be popular with the digital music and ipod crowd. It's intriguing that they've souped up the performance of this feature vs their Zdac model which is well regarded in reviews thus far.
This site supports the all-amps-sound-the-same school of thought, at least it's nice to see differences in sonic signature discussed. I've got the Halo A21 and have been greatly impressed with its reproduction of vocals and highs and I'm looking forward to trying the P5 to see what it adds to the A21.

 

[gene]

I read somewhere that Amps that are fully balanced differential designs increase efficiency and can lower the noise floor.
The reduced distortion tends to be the even harmonics not the more objectionable odd harmonics.
Is there any truth to that?

Does full balanced circuitry reduce IM?


- Rich

I agree with most of what you've read. The cool thing about doing a balanced power amplifier stage is that you essentially double your rail voltage as you can read in my XPR-1 review:

Emotiva XPR-1 Mono Amplifier Review | Audioholics

If I were building my own high performance audio amplifier, you bet for sure it would be a fully differential design. Not only is it measurably superior, its just damn cool from an engineering standpoint to do

Back in the day I designed an Integrated Red/Black Speaker for NAOC and guess what, the power amp was fully differential. I even used a high end Vifa 5" driver for the job instead of the cheesy paper Bose one they wanted to initially use. I recall the General's at the meeting room telling me the Speaker sounded better than their surround systems at home which incidentally were Acoustimass 5's

 

[gene]

This site supports the all-amps-sound-the-same school of thought, at least it's nice to see differences in sonic signature discussed. I've got the Halo A21 and have been greatly impressed with it's reproduction of vocals and highs and I'm looking forward to trying the P5 to see what it adds to the A21.

A few vocal so called "objectivists" forum regulars on this site do NOT reflect the opinion of the masses and certainly NOT the opinions of the site operatives, especially myself. You will NEVER hear state "all amps sound the same if not driven into distortion" crap. I've listened to too many products over the years in amazement that have removed such prejudices.

BTW, I really liked the A21 amplifier back when we reviewed it. Nice product.

 

[sharkman]

Thanks for that, perhaps I should watch my generalizing! It's good to hear that there is room for differing opinions here.

 

[Irvrobinson]

A few vocal so called "objectivists" forum regulars on this site do NOT reflect the opinion of the masses and certainly NOT the opinions of the site operatives, especially myself. You will NEVER hear state "all amps sound the same if not driven into distortion" crap. I've listened to too many products over the years in amazement that have removed such prejudices.

BTW, I really liked the A21 amplifier back when we reviewed it. Nice product.

Being in the agnostic-but-tending-towards-there-are-differences camp myself, at least when I'm making buying decisions, I'm curious, did you ever put some of these products that amazed you on the test bench to try and figure out what amazed you objectively? If so, did you get any interesting results?

 

[gene]

Being in the agnostic-but-tending-towards-there-are-differences camp myself, at least when I'm making buying decisions, I'm curious, did you ever put some of these products that amazed you on the test bench to try and figure out what amazed you objectively? If so, did you get any interesting results?

Yes the Pass Labs amp IMO didn't measure to stellar but it sounds awesome. One day I need to sit down and do more multi-tone testing to see if the lower feedback approach really does have merit to the types of distortion we prefer.

My brother bought over a few guitar amps that all measured like crap but the distortion profile some of them produced was very pleasing using his guitar of choice. I realize this is a different animal since he was using the amp as a sound producer not a reproducer like we often do for playback of prerecorded music.