Life after Yggdrasil: Watering the Ash

@ultrabike, I used one of those several years ago (probably still have it in a box somewhere) and it was my first NOS DAC. Got it from eBay for about the same price it is now on Amazon. It drops off a cliff after 9-10k and has some odd distortion. However, it uses 4x 1543's so it must be 4x more better.

Edit: I was only referring to the Muse DAC. Never heard the other thing, but 8x > 4x.

 

My understanding is that TDA1543 comes with an on-board NE5532 opamp.

http://www.docethifi.com/TDA1543_.PDF
(go to page 3)

I've heard of designs like the SPS DAC3 which too use the TDA1543 in NOS mode and with no analogue section to speak of, taking the output directly from the D/A IC and to an external amp. I suspect this is possible thanks to the on-board opamps located right prior to D/A out on the schematics.

 

The opamps in that diagram for TDA1543 are out of the chip itself (they lie outside the 'square' TDA1543 demarcation) . It does output voltage directly instead of current, so you don't need an IV stage, and normally the reason why you see multiple chips is, besides the benefit to INL, to boost the output voltage so that you "don't need" output stage. remembered wrong chip, you can do resistor discrete IV stages like: http://myweb.tiscali.co.uk/g8hqp/audio/TDA1543IV.html or http://www.pavouk.org/hw/modulardac/en_10xtda1543.html so you can claim no op amp used.

 

Marvey

Can you give us an idea what you are running? What are your components? In terms of "aspects", can you further break these things down (other than dynamics and timbre which you already covered - for example what about resolution?) for Yggdrasil and the Lampi in relation to the BP DAC that you auditioned in your system? Also what model Lampi?

Might be worth loaner program?

--More below. I did ask Torq about including it in 'The List', and he kindly reached out to BP, but Gary offered him the 'purchase & try for 7 days', which he understandably declined.


cskippy,

I'm not trying to diminish your thoughts .....

--True. But, I don't think so at the moment. I will post updates in due course.

Winders,

There is nothing really comparative in your "review".

--More below.

Torq,

Number of observations.

--More below.

Introduction
I posted the detail of my system above, so I won't repeat it here.

I decided to start listening to some alternative DACs as since I moved to the Sonore microRendu (mR) I had been hearing sibilance and what I describe as edginess with a sub-set of my music. Having tried a number of tuning steps, which generally improved clarity, I was becoming frustrated.


I will compare and contrast the Bel Canto 3.5vb (BC), the Schiit Yggdrasil (Y), the Lampitizor Level 4 Gen 5 (L) and the Border Patrol (BP).

In fact the bar has moved in the last three weeks as Sonore distributed an OS update (2.5) and this substantially removed the issues I was hearing with the BC. However, having heard the Y I was happy to press on.

All four are excellent DACs, especially since the OS update! I could live with any of them, but I can see that system and personal taste will lead to different people choosing different DACs.

I borrowed the Yggdrasil and was suitably impressed; my review is posted above. However, since then I have made a further discovery. Both the Y and the BP had a tendency to sound bright with some music, whilst the L did not. I had deliberately left my system as stable as possible during this exercise, with the exception of the mR OS change, which i couldn't resist - and I am glad I did not. I had been using a Mark Grant Oyaide SPDIF from Singxer-F1 to DAC. When I replaced this with a good quality copper SPDIF the brightness simply vanished. I suspect this would address this issue with the Y as well as the BP.

What I am left to think about is a selection of attributes that are exemplified by each DAC within my system. I will discuss these below with reference to the listed music.

Music Used:

1. Torq's List
As far as I can replicate it using Quboz.

2. My List


Artist - Album - Quality - Track:
Eric Clapton - Unplugged - 4416 - Signe
Jeff Buckley - Grace - 4416 - Grace
Rolling Stones - Brussel Affair - 4416 - Brown Sugar
Mary Black - Live at Olympia - 4816 (DVD rip) - Columbus
Nils Lofgren - Live - 4416 (Quboz) - Keith Don't Go
Peter Gabriel - Compilation - 4416 - Sledgehammer
Peter Gabriel - So - 9624 - Sledgehammer
Rachmaninov - Symphony No1 - 2496 - Ormandy, Philadelphia Orchestra
Simon & Garfunkle - Old Friends - 4416 - The Dangling Conversation
Stevie Nicks - The Other Side of the Mirror - 9624(LP) - Long Way to Go
Phil Collins - Tarzan - 4416 - Son of Man
Pink Floyd - DSOTM - 4416 - Time
Elkie Brookes - Pearls - 4416 - Superstar
Carole King - Tapestry - 4416 - I Feel the Earth Move
LSO - Star Wars - 9624(LP) - Main Title
Dusty Springfield - Dusty in Memphis - 4416 - Son of a Preacher Man
Dusty Springfield - Dusty in Memphis Remastered - 4416 (Quboz) - Son of a Preacher Man
Roger Waters - The Wall - 4416 (Quboz) - In the Flesh?
Pat Metheny & Anna Maria Jopek - Upojenie - 4416 - Cichy zapada zmrok

Some specific comments below:

Eric Clapton - Unplugged - 4416 - Signe
This track includes some great percussion, which includes a triangle centre stage.
BC - Triangle present, but it can sound distorted, it doesn't sound as though it is ringing true at times.
Y - Music is well presented and I was aware of elements in the music that I hadn't heard clearly before. I felt that the BC was slightly opaque, this issue was addressed by the mR OS update. The triangle seemed to have decided to turn up occasionally. While the high frequencies were better presented they didn't extend as high and what I heard was subtley recessed in the soundstage. But, the detail and width of the soundsatge were excellent.
L - Triangle present and correct. The soundstage is almost super-sized. The bass via L reminded me of my LP12, a sloght upper bass lift. Enjoyable but perhaps not correct.
BP - Triangle present and correct, high frequencies beautifully rendered. Soundstage precise but smaller than Y and L, on a par with BC.

Jeff Buckley - Grace - 4416 - Grace
I picked this as in my system this could be bright, and there is a level of distortion towards the end of the track that allied to that reduced my enjoyment.
Solved with mR OS update.

Rolling Stones - Brussel Affair - 4416 - Brown Sugar
BC - The bass seemed to be present but more by feeling than clearly hearing the notes. Improved by mR update.
Y - I was immedialetly struck with the clarity of the bass and the ability to follow the bass line. The transparency of the Y is excellent.
L - Bass as BC post mR update.
BP - The transparency of the BP is very good. Not sure if it is quiet up to the Y. However, the richness of the bass adds to the feeling I get from listening to the music, although I think it is a tad less clear in following the bass line.

Rachmaninov - Symphony No1 - 2496 - Ormandy, Philadelphia Orchestra
I love some of the flourishes in this symphony. If Rachmaninov could have patented them he would have earned a stream of money from modern movie composers.
BP - All the DACs play this well. Y with its emphasis on detail and soundstage is very good. The BP adds to this with slightly more timbre.

Phil Collins - Tarzan - 4416 - Son of Man
In my system this has been a problem to all the DACs. Now much reduced, but still sibilance and spitting.

LSO - Star Wars - 9624(LP) - Main Title
BC - This recording has always sounded disjointed through the BC.
Y - First time I had heard it sounding cohesive. Still not the best recording.
BP - On a par with the Y.

Dusty Springfield - Dusty in Memphis - 4416 - Son of a Preacher Man
Dusty Springfield - Dusty in Memphis Remastered - 4416 (Quboz) - Son of a Preacher Man
Y - First time I heard this through the Y I was bowled over. I had never realised how brilliant the bass line was.
BP - Not quiet up to the Y .....but, that was with my CD rip. With the remastered version, which is what I fed to the Y, much clearer.

Roger Waters - The Wall - 4416 (Quboz) - In the Flesh?
Y - Detail and soundstage, but not the impact of the BC.
BP - Best of Y & BC.


Conclusion
I hope I have given a flavour of what I am hearing. Since the mR update I could live with any of these DACs. What I enjoy about the BP is that it gives me the detail of the Y with most of the richness of the L. Is it quiet as detailed as the Y, I suspect not. But, there were times when the Y could edge towards being a bit academic to me, it was forensically precise but perhaps missed a smidgen of heart.

With respect to Timbre, the L played almost all my 4416 files with aplomb, which I greatly appreciated. Now that I have swapped the spdif cable and heard a similar response from the BP I expect it would solve that issue for the Y as well. The L has a richness to it that makes classical music in particular very nice. When listening to Beethoven's 9th 9624 (Barenboim) I finally got what a great recording it is, it had always been a bit flat via the BC, and not really improved upon by the Y. The BP may actually be a bit more accurate, and is certainly a step up on the Y and BC, but I did not feel quiet the excitement I got from the L. However, with the Elgar Cello Concerto (Du Pre) I preferred the BP, which produced the power & rasp of the piece the best that I have heard through my system.

The two R2R dacs have an energy in the mid-range that the BC frequently doesn't approach. The L get very close much of the time, but misses out in the transparency.

Is the BP a technical marvel? Almost certainly not. Is it using all the data it is fed? I don't know. However, I do not associate high resolution with great music. I do have some hires files that are the best versions of a piece that I have, those are still the best version via the BP.

Four good DACs that may well appeal to different people / tastes. But, the BP DAC works well for me.

M

 

Random question concerning the zero-crossing glitch thingamabob from a guy who is admittedly an armchair wizard with no practical experience in this topic...

Some dac makers will stack chips, claiming it helps linearize any slight wobble from individual chips (something something -3dB per doubling?). Now doing this with the multibit chips would probably just exacerbate the glitch... but would it be possible to digitally create an offset in one chip then correct it later before stacking the results? So yeah the glitch will still happen, but at different points on each chip, kinda sorta swamping the effect out when you take the averages.

Maybe? I dunno, just armchair thinking. The article mentions using two chips to handle each half of the signal like class B; I was just thinking in a slightly different direction.

 

Boom. We've just created a new dac topology. Give it a fancy new buzzword and we're off to the races!

 

Not so fast! Just wait until you hear the patina a few months down the line. When it gets to be just how you like it, seal it in with wax. Want to start over? Scotchguard pad.

Best materials for shielding? To be honest, I haven't a clue... but if this really is a thing, wouldn't the providers to audiophools have hit this before? Is it a thing?

Mass... without checking... more than Aluminium, less than steel. A lot less than titanium.

 

The KI series from Marantz uses coppershielding as an "improvement" over the standard versions. In that sense it has been around for quite a while.
That gave me an idea. Cut a copperplate for shielding a sound card from PC components...

 

It depends on the implementation and also the EMI that is needing to be blocked. Generic "EMI shielding" is way too vague and typical of vacuous audiophile marketing. EMI is bad, so blocking it must be good. Copper is visually attractive so it must sound like it looks. Mkay.

 

Whoops! Missed point error

My theory on this one is that the better ones are actually engineered for the environment in which they will be put, ie, the engineers already thought about it.

I like that. And I like that it is on the inside, not just decoration: maybe it really is for real.

By the way, I do not underestimate the importance of appearance in audio gear. I can dribble over milled-from-solid but it is the pro-utility studio look that tends to start me on the deep breathing.

But I also like the stately-home look. I could match copper-with-patina cases with a pair of Tannoy Turnberries. Mmmm... in the light of a real log fire!

(and I know from experience that the speakers, at least, would sound wonderful)

 

The right materials, and proper application/configuration, for EMI, RFI and even magnetic shielding depend greatly on what you're shielding, what you're shielding it from and the nature of that "from", the operational environment, and so on.

For example, the frequencies you care about shielding something from play into what materials you use. As does whether you can rely on reflection as your shielding mechanism, or if you need significant absorption also. And if you need absorption, can you dump to ground or must you convert to heat?

Take something like the Holo Audio Spring DAC. It has a big, beefy, aluminum* chassis, with nice copper cheeks. This will operate in a mostly reflective capacity (due to the relatively good conductivity of aluminum and excellent conductivity of copper), and will usefully deal with frequencies from the mid kHz up into the GHz ranges. Thus it's very well shielded from external EMI/RFI. And, at the same time, any such interference/noise generated internally is kept firmly within the DAC chassis.

Look a bit deeper. The Spring DAC has an open interior, which is common to many components. Within that there are bunch of devices that emit EMI/RFI and magnetic fields. Those emissions are kept INSIDE the DAC due to the efficient reflective-type shielding and will be reflected back throughout the chassis until they decay. It's very hard to say what effect they'll have, either measurably or audibly, in doing so ... and even if that sort of shielding is beneficial to the final result or not (i.e. is keeping the internally generated EMI/RFI inside the unit a better trade-off than allowing external EMI/RFI in?

That's a rhetorical question and the "right" answers depend on your design goals and operating environment. If I'm building a satellite for operation in deep-space I almost certainly care more about keeping external EMI/RFI etc. OUT than worrying about the internal reflections of comparatively low-level internal emissions.

But what if I care about both?

Well, I could individually shield internal assemblies/sub-systems/components. Lots of approaches do this. Look inside a Linn Klimax DS (or the previously linked Pioneer Elite transport) and you see separate compartments providing internal shielding for such sub-systems. Not only are we keeping external EMI/RFI out of the unit itself, we're minimizing how much of that can get from one set of components to another.

This is not a panacea though.

Let's go back the Holo Spring DAC ... which has a big ass, noisy, XMOS board sitting right in the open internally. We could put it in it's own enclosure and keep it's spurious switching emissions from infecting the rest of the DAC, or we could leave it open as it currently is. Which do we do? This would depend on whether the XMOS' generated EMI/RFI was more detrimental to the performance of the system as a whole or to the operation of the XMOS board. My personal experience would say that XMOS board should be in it's own enclosure as, being a digital device, it's much less prone to re-receiving it's own artifacts from spurious EMI/RFI than analog bits are, but it's possible that's not the case (implementation specific).

The only way to know for sure is to try both and measure/listen to the results.

...

TL;DR; materials matter when it comes to shielding ... but at least in the known audio-universe, tin-plated steel (that nasty shit cheap-feeling shit on PC cards) covers enough of the frequency band, well enough, reflectively at least, that you're deeply into diminishing returns before you start ... and you'll need to worry about a LOT more than the selected material to get an optimally shielded implementation.

In other words ... copper is good ... and theoretically better than tin-plated steel (depending on what you're trying to accomplish) but if that's the only "special" part of the shielding arrangement then it's an expensive way to solve a probably-non-applicable set of concerns.
--

*This is, of course, aluminium ... I'm must not going to fight the US-centric auto-correct on it every time I mention it.

 

What a superb post. This is the sort of thing our audiophile daddies should have sat us audiophiles-to-be on their laps, saying, "Son... There's something I need to have a serious talk with you about." But they didn't. They wrote audiophool articles in magazines instead.

Great post. Thank you.

lol