Columns Retired Columns & Blogs |
Odd that none of the Digital Link products show up on the PS Audio website.
Is it vaporware? Doesn't sound like how I've thought of PS Audio in the past.
The rear panel features a pair of high-quality, gold-plated RCA output jacks, RCA digital input jack, and Toslink optical input jack. A four-pin jack accepts DC from the outboard power supply, a 4.5" by 2.75" by 2.5" black box.
Technology
Although I couldn't get inside the power supply, it reportedly contains a single large transformer. Once inside the Digital Link, the stepped-down AC is rectified, filtered, then regulated by four three-pin regulators. Like other PS Audio converters, the Digital Link has lots of filter capacitors (a total of 18,000µF) distributed around the board.
The circuitry is contained on two pcbs: one handles the digital inputs and input switching control, the other is the main board containing the digital and analog circuitry.
The input receiver is the 16-bit Yamaha YM3623B, but implemented with the jitter reduction circuit developed for the UltraLink. Digital filtering is provided by the 8x-oversampling Yamaha 3434, a small and inexpensive filter chip. A pair of Analog Devices AD1860 18-bit DACs perform D/A conversion. The AD1860s are the premium "K" version, selected for best low-level linearity and lowest THD. The DACs are matched between left and right channels, but no MSB (Most Significant Bit) trimmer is included.
Like the SuperLink, the Digital Link II uses passive current-to-voltage (I/V) conversion. PS Audio is the only company I know of that doesn't use an active I/V converter (usually an op-amp). The current-to-voltage stage takes the DAC's staircase current output and converts it to a voltage. In the Digital Link II, this stage is nothing more than a 150 ohm resistor connected to ground. The advantage of this technique is one fewer active device in the signal path, but at the expense of higher noise; by Ohm's Law, using a low-value resistor means a lower voltage, therefore requiring extra gain in the line stage.
The direct-coupled output stage is based on a Precision Monolithics OP37 op-amp (one per channel). A FET current source biases the OP37 into class-A operation. Because of the low voltage from the I/V converter, the op-amp is run at a very high gain (30dB). This op-amp is the only active device between the DAC and analog output; only one gain stage is used, with passive de-emphasis and a passive first-order low-pass filter. The amount of circuitry after the DAC is minimal, owing both to the passive I/V converter, low-pass filter, de-emphasis, and single gain stage/output driver.
A dual-channel relay shunts the audio output to ground when the unit isn't locked to an incoming digital signal or when the unit is turned off. Note that the front-panel on/off switch merely activates the shunting relays and turns off the front-panel LEDs; all other circuitry remains fully powered. This ensures that the unit will sound its best without its owner having to wait for warm-up.
Build quality is far better than what one would expect from a $499 digital processor. The chassis is nicely made, with countersunk screws, thick metalwork, and a ¼"-thick front panel.
Listening
The Digital Link II provided excellent sound (footnote 1). Although I didn't feel it equaled the California Audio Labs Sigma or Sumo Theorem, it was nevertheless impressive for its price.
In perspective, the Digital Link II's sound was very similar to that of other PS Audio processors: slightly forward in the mids, less than impressive soundstage depth, and a full-bodied low-frequency presentation. Starting with the bass, the Digital Link fell between the Sigma's fatness and the Theorem's lean, tight rendering. There was a nice sense of weight to the bass, with fairly good pitch definition. Although not as fast and articulate as the Theorem's, the Digital Link's bass was rounder and fuller, making the entire presentation warmer.
The Digital Link II's soundstaging was better than the FortéDAC 50's, but not up to the level heard from the Sigma and (especially) the Theorem. There was a slight opacity and congestion to the presentation compared with the Theorem. The impression of individual instruments hanging in space, a sense of air and bloom, and clarity were all good considering the $500 price, but in absolute terms were only fair. The Digital Link II put a thin veil between me and the music, slightly reducing the level of musical involvement. On Robert Lucas's Usin' Man Blues (AudioQuest AQ-CD1001), the Digital Link II lacked some degree of the transparency, resolution, and air heard from the Theorem, reducing the palpability of the images.
One area in which the Digital Link II excelled was its treble presentation, where it was smoother than the Theorem, and even the Sigma. This treble smoothness had its price, however: a slight obscuring of detail. The upper octaves lacked a sense of hearing everything that was going on. Going back to the Theorem revealed that the Digital Link presented less information to the listener than the Theorem. There was, however, a purity and lack of grain in the treble, though not to the same extent heard from the Theorem. To its credit, the Digital Link didn't affront the ears with a hashy, metallic treble. Long listening sessions without fatigue were possible with the Digital Link II. Moreover, the Link had a greater warmth than the Theorem, particularly in the mids.
In summary, I would rank the Digital Link II behind the Theorem and Sigma, but ahead of the DAC 50. At $799, a recommendation would be iffy. Considering its $499 price and good build quality, however, the Digital Link II is a bargain. Recommended.
Putting it all in perspective
After getting to know each of the processors, I auditioned them in relation to the $1495 Bitwise Musik System Zero (favorably reviewed last month) and $399 Audio Alchemy Digital Decoding Engine. Where do the four processors fall compared with a $1500 processor and the inexpensive DDE?
Starting with the DDE, it was no contest; all four processors were superior to the DDE, particularly in the treble. Although the DDE had good clarity and resolutiongreater resolution than the Digital Link IIits hashy treble was a significant liability.
The most logical comparison is between the DDE and the $100-more-expensive Digital Link II. First, the Digital Link's build quality is far superior. Comparing the power supplies, for example, the Digital Link's is much larger and heavier, and has a high-quality locking connector. The DDE's supply has a mini-jack that momentarily shorts the supply rails to ground if the plug is inserted with the power supply plugged in. The Digital Link's full-sized chassis and high-quality RCA jacks add to its value. The DDE, however, has an absolute-polarity inversion switch and the ability to be upgraded through the I2S bus port.
In terms of sound quality, the Digital Link was clearly more musical. In virtually all areastreble smoothness, soundstaging, bass, lack of grain, and involvement in the musicthe Digital Link II was convincingly better. Again, the DDE had superior clarity and resolution, but its aggressive treble was its Achilles' heel. The Digital Link had a warmth and smoothness that made it more enjoyable. Further, the Digital Link II's presentation of detail was more refined, with finer gradations between the salient and the subtle.
Conclusion
I was generally impressed by the price/performance ratio of the three processors earning a recommendation this month. This was particularly true of the PS Audio Digital Link II, which, at $499, is nothing short of a steal. It is clearly better than the Audio Alchemy DDE, and in some ways competitive with more expensive converters. I felt, however, that the Digital Link II was bettered in absolute terms by the CAL Sigma and Sumo Theorem. Because the Digital Link II costs $300 less than the Theorem and $200 less than the Sigma, it is probably the best value of the group. If you want the best sound for the least money, the Digital Link II is the processor to buy.
Odd that none of the Digital Link products show up on the PS Audio website.
Is it vaporware? Doesn't sound like how I've thought of PS Audio in the past.
Odd that none of the Digital Link products show up on the PS Audio website.
It isn't odd, as not all companies list discontinued products on their websites.
John Atkinson
Technical Editor, Stereophile
May be JA1 could review the new Bryston BDA-3.14 multi-function DAC, streamer, pre-amp ($4,200) .... (may be a follow-up review)? :-) .........
The product that is the subject of this article wears the PS Audio brandname, but that product was not built and sold by the same business that is today's PS Audio. It is not a legacy product of the current PS Audio business.
Paul McGowan and Stan Warren co-founded PS Audio. Warren later left PS Audio. Later still, McGowan sold the business, and the product that is the subject of this review was something designed, built and sold by PS Audio under that other ownership.
That business later failed and liquidated, and McGowen bought back the PS Audio brandname, but not the legacy product line, and then built a brand new business under that name, the current PS Audio. The earlier products were of an older defunct business. I expect that Paul McGowan and PS Audio have been careful in not claiming any false relation to older products that were not theirs.
The old PS Audio was 'Mufasa', the old Lion King ......... The new PS audio is 'Simba', the new Lion King ....... 'Simba' is roaring :-) .........
John...
I'll confess to being confused when I saw Harley's name, but even with that clue, it took me a few tries to realize that this review is from 27 years ago...
It's a neat artifact, but particularly for a digital processor, I'm not sure it belongs in the same stream as the current reviews. It wouldn't be hard to mark it more clearly, even adding a sentence at the beginning of each of these that says something like, "This is a classic review from our wayback machine. It was first published in 1992".
Without the context, it's not only not worth reading, it's affirmatively confusing, no?
...but the year the article was first published is included in the byline:
Robert Harley | Nov 8, 2019 | First Published: Oct 1, 1992
I guess my point is that if you were reading a stream of articles from a newspaper, a sports site, the weather channel--you probably wouldn't expect to find an archival story mixed in. If the Times put "War in Germany!" in their feed, most people wouldn't stop to read the date next.
I'm certainly not complaining about all the goodness that Stereophile delivers, it's a fabulous resource. Simply pointing out that if the labeling isn't working 100% of the time (and the comments on this one post indicate that at least twice it didn't work) there might be a better way...
And I appreciate that these are added along with the new content.
That there were aliens sighted!
Some guy named Orson Welles reported on it..??