Review and Measurements of PS Audio PerfectWave DirectStream DAC

[Sukie]

This is actually a case of a mechanical device needing "break in," not electronic. A relay is an electro-mechanical device.

Thank you for the correction. I should have read @jseaber's response more fully.

I knew that there was a break in needed, just got the wrong bit! I rather enjoyed having one of the few that needed the extra attention.

 

[BN1]

Wow ! Now I understand McGowan's somewhat negative comment about Amir.

 

[scott wurcer]

Note that I personally don't put a ton of value on such degrees these days.

Several well known top designers never got a degree from the normal process (a few went back after decades of industry work "just to do it"). IRCC Bob Dobkin left MIT without his degree.

 

[RichB]

the measurement I have is my ears... haha

The transformer / capacitor coupled passive output analogue (DSD) state is what I think benefits from the burn in, but no proof other than my ears... and it appears the problem the OP pointed to was related to the output performance.

This might be good to watch on how Ted designed this thing...

Only the sharp filter measured here support the design choices described in the video.
The SN and power supply noise should have had better measured results.

- Rich

 

[SIY]

Several well known top designers never got a degree from the normal process (a few went back after decades of industry work "just to do it"). IRCC Bob Dobkin left MIT without his degree.

Wasn’t Jim Williams self taught, no degree?

 

[MakeMineVinyl]

if Burnin wasn't critical, then most companies would just assemble high grade electric equipment and just send out out for use... in my world, aerospace... rack mount electronics that are integral to airplane systems must be burned in for several hundred hours to both insure reliability and benefit from the burn in benefits to redundancy. Equipment is very sensitive to on / off power cycling, but less so after a long burn in period.

To weed out infant mortality. If a component is going to fail, it will do so sooner than later. That's why there is burn in after assembly.

 

[MakeMineVinyl]

How is that good business practice? If burn-in was real, wouldn't good business practice be to do it at the factory before the product was shipped?

If I were making a product that I thought benefitted from an extended burn in, I'd be damn sure the was sufficiently "burned in" before shipping it.

 

[992SAM]

How is that good business practice? If burn-in was real, wouldn't good business practice be to do it at the factory before the product was shipped?

They have a video where they talk about how they do a 24-36 hour burn in before leaving the factory but that it's in many cases no where near enough, but that if they were to do multi-hundred hour burn ins, the cost would go up.. so they just tell their customers to do it. I mean it worked.. what can I say.. it was good out of the box, but after the first 200 Horus it's great... and when I get back and it's been up and running 200 more hours, I expect it might sound better... can I explain why? not really, I just know what I hear.

Is it the fact this this DAC uniquely converts everything to DSD and it's output stage is direct DSD...? Has that been addressed in how the testing might be affected when contrasting to a PCM output DAC that then must convert PCM to Analogue before outputting?

 

[scott wurcer]

Wasn’t Jim Williams self taught, no degree?

Right, they say his father cut him off but I never knew where the money came from to purchase all the antique instruments even circa 1970. He also had 2 or 3 antique cars but I will say his idea of fine dining was a frozen beef patty with plenty of Heinz and yellow mustard washed down with whatever was at hand. Pease used to bring the jugs of Carlo Rossi.

 

[amirm]

They have a video where they talk about how they do a 24-36 hour burn in before leaving the factory but that it's in many cases no where near enough, but that if they were to do multi-hundred hour burn ins, the cost would go up..

They cost is already up! This is no cheap DAC. A month of electricity use doesn't cost much for a DAC anyway. Heaven knows once they cut out their dealer network, they now have plenty of margin to play with.

 

[bunkbail]

They cost is already up! This is no cheap DAC. A month of electricity use doesn't cost much for a DAC anyway. Heaven knows once they cut out their dealer network, they now have plenty of margin to play with.

The upcoming PS Audio Obsidian TSS DAC has a tentative MSRP of $20-25k. Let that sink in haha.

 

[scott wurcer]

The upcoming PS Audio Obsidian TSS DAC has a tentative MSRP of $20-25k. Let that sink in haha.

We should probably just leave this alone at some point the negativity only drives up the appeal with the worshipers.

 

[cjm2077]

it's this and it's also the observation that burn in (with no on/of cycles) leads to a longer lasting, better working unit... especially as it relates to capacitors.

Explain to me how that works with with capacitors. I have more than one EE degree, including actually finishing undergrad at MIT, and the only changes I have heard involving caps are when electrolytics are "formed", but that happens pretty much instantaneously if you turn on the unit and supply full rated input voltage to it. It doesn't take 100 hours, and you don't do it by just turning on 120VAC (Or 240VAC). I guess this also applies to Tantalum caps, but nobody should ever use those since they provide little benefit compared to other types and tend to catch fire when they fail. Electrolytics also have incredibly poor tolerances and shot not be used in any location where a precision value matters. If your design's performance changes greatly because of the exact value of an electrolytic, you done screwed up.

Component part and product level failure rate is described as having a bathtub shape. You have a high rate initially from parts that may have been improperly made or handled, or products that were assembled incorrectly, and after a long time you have cumulative stress related failures at an increasing rate. Maybe somebody forgot to add the thermal compound to a part before installing a heatsink. Maybe they dropped a part on the floor or handled the board without proper static protection. This is why products are burned in. For amps you cycle power output levels from low to high to heat and cool the parts and bring out thermal cycling faults. Some products I've designed (for semiconductor processing) were placed in chambers while running at full power and heated and cooled while vibrations were applied. This would bring out part failures, but also assembly mistakes like forgetting a lock washer or not siliconing the electrolytics in the supply correctly. We did this to catch early life failures and improve the reliability of the shipped products. But nobody from any level of the company (marketing, engineering, test, reliability, manufacturing, executive) ever once said "this is definitely going to make this work better." There is a reason that all tests confirming performance is within spec are done before the burn-in tests.

My guess is that people saw that burn-in testing was done at some companies, asked about the practice and completely misunderstood what was explained to them. And then the lore was passed around.

 

[992SAM]

The upcoming PS Audio Obsidian TSS DAC has a tentative MSRP of $20-25k. Let that sink in haha.

We should probably just leave this alone at some point the negativity only drives up the appeal with the worshipers.

you guys haven't even heard it yet and you're already dooming it... try to be a bit more objective... And to start with, at least hear it's inventor explain what sets it apart.

 

[JohnYang1997]

you guys haven't even heard it yet and you're already dooming it... try to be a bit more objective... And to start with, at least hear it's inventor explain what sets it apart.

I bet you it won't measure nearly as good as 99 dollar d10s.

 

[scott wurcer]

you guys haven't even heard it yet and you're already dooming it... try to be a bit more objective... And to start with, at least hear it's inventor explain what sets it apart.

Taking bets? Objective and inventor explanations don't mix very well. PS Audio has a history of explaining away mediocre measured performance.

 

[JohnYang1997]

well, once you get your hands on one, copy it and build a clone in one of your slave shop factories .. maybe you can improve on it?

Ok. This is too far. @amirm

 

[cjm2077]

I actually kind of want to see where he goes from here. We debunked most of what he said, then he went to an insult. What stage is next?

 

[MediumRare]

you guys haven't even heard it yet and you're already dooming it... try to be a bit more objective... And to start with, at least hear it's inventor explain what sets it apart.

Let's do a head-to-head analog output comparison and see what the audible difference is: https://www.audiosciencereview.com/...o-measurement-of-lg-g7-thinq-smartphone.4468/

 

[amirm]

well, once you get your hands on one, copy it and build a clone in one of your slave shop factories .. maybe you can improve on it?

Watch your language! We don't tolerate this type of attitude whatsoever. I will leave it to Thomas to decide if formal sanction is necessary.