Question about Burr Brown DACs

Generally speaking DAC chips have "evolved" over the last ten years but mostly to make them less expensive and easier to use. I sold both Crystal Semiconductor and Burr Brown for a lot of years. The Burr Brown PCM63K (military grade) was an awesome DAC but was expensive ($50) and required a POT and other circuitry to tweak it. Burr Brown did not have a volume consumer-market mentality, they were a bunch of very talented analog designers with a quality-first and cost-be-damned approach to things. That all changed when Texas Instruments purchased Burr Brown for $8 billion roughly five years ago. TI was a volume manufacturer with very good digital designers (their TMS32x DSP's are the most widely used in the world) but they lacked analog expertise, it made sense to acquire Burr Brown (although the price tag is still bewildering to me).

TI knew that in order to compete effectively with Crystal Semiconductor they would have to make devices that were less expensive and required no additional supporting circuits to allow for a much simpler end-customer design. The key motivator was getting design-ins on true consumer electronics to drive significant manufacturing volumes. Economies of scale allow for agressive average selling prices while maintaining a healthy margin. DVD players and the like have driven DAC design for the most part. Simpler/more elegant design techniques coupled with more advanced semiconductor processes, such as 130nm or the newer 90nm, and 12" wafers have driven costs down big time.

Has this made today's DAC's better than the more esoteric monolithic chips of years back? I'd have to say that it has and it's also made good digital gear very affordable. Need a CDP for a bedroom or office system? Pick up a decent DVD player for under $150 and you're off to the races. Jeff

Listen to the BB1751 dac in the little toshiba 3950 dvd player. This is their cheapest dac but the sound quality from cd on this player is amazing for the money. Will it beat your big rigs? No, but it has a relaxing, engaging manner of presentation that is simply fun and musical. Dacs are improving at a terrific rate. If you crank the toshiba the sound falls apart. What surprises me here is the toshiba sound is just on a smaller scale, less weight, less fullness but listenable....amazing. remember I'm not claiming this to be the holy grail, just a good little player...cheers

The awful truth is that pretty much everything here is a compromise, 'cause every company who's trying to make a living from what they do has to do it at an overall profit. For this question, that may not necessarily mean that each successively higher numbered chip is more sonically wonderful. Each is designed/manufactured to a price point - features for a price (that's also where the yield, etc. come into the equation). The '92 or '96 chip may or may not be improvements that you can hear at all - they may be manufacturing gains (more-chips-for-less) or packaging (chipset) gains - in profit.

Slappy may or may not be at risk for a big snack - we all probably won't know unless a B-B engineer spills the beans. And even if B-B did gain performance with the higher numbers, as Audioengr & Bombaywalla accurately point out there may not be an improvement to you if the overall system solution isn't better. You can easily bury the best of chips w/ lousy boards and/or support chipsets.

Bombaywalla:

I did notice that Denon's less expensive unit, the 2200, uses the 1791 while their 2900 and 5900 universal players use the 1790. I assumed that the 1791 was a less expensive and perhaps "less good" DAC. I don't know what the benefits of the 1792 is but Mr. Michaelson mentioned in Stereophile that it was implemented in their new DAC so it's obviously his preference.

Why I care about any of this I don't really know. I have limited technical knowledge so I'm just making myself more neurotic! The bottom line is how the unit sounds to me. The Anagram Technologies DAC that is in the Audiomeca (and Audio Aero + Orpheus) sounds remarkable. I guess I am afraid that I will make the wrong move and trade off to an inferior sounding player.

But then again if I do make the wrong move it's nice to know there is always Audiogon!

Thanks to all for the feedback.

Ken Golden

Agreed with Audioengr, Slappy & Sean.

In general, semiconductor manuf. add I.C.s to their portfolio all the time so that their product offerings always provide the best profit margin to them. I just looked @ Burr-Brown's website. I don't see the 1790 DAC there (in the Audio D to A section) but I see the 1791 DAC. Maybe the 1790 is too "old"?? However, if we compare the 1791, 1792 & 1796 then we see that the 1791 offers the lowest SNR(=113dB) at the lowest power consumption(=90mW). The 1796 seems to be placed right in the middle where it offers better SNR than the 1791 (but not as good as the 1796) at a modest power consumption increase over the 1791 (but not as much as the 1796). Looks like they offer a low power version (1791), a medium power version (1792) & a high power version (1796). I'm guessing from my experience as no further data is available that the 1796 is fabricated in some later technology than the 1792 & 1791 to realize the benefits of better SNR at a modest power increase. Semiconductor manuf. generally do this - migrate their portfolio gradually to the better technology 'cuz they can reduce die size hence realize a better profit margin.

All this does *not* necessarily mean better sound. The designer of the audio gear must know how to use the part expertly to squeeze every ounce of perf. from it. I know this from my experience wherein working w/ the customer on how to use our parts is a vital part of our relationship w/ them. 99% of the time, if we do not interfer, the customer totally botches the implementation of our part in their final product! Not surprising as our ICs are complicated but all too often we assume that if we put the part on the technology shelf someone will come & pick it up & use it correctly! Wrong!

Can't say whether or not there is an applications group in Burr-Brown that the audio manuf consults when they use their part. However, implementation on the audio manuf's part is key to making a good DAC design sound excellent.

Long answer/explanation to Audioengr's succinct & correct (IMHO) answer.

Audioengr: Thanks for posting a simple and to the point answer.

Slappy: You should only make such an offer if you have a really good appetite or to those that have severe hearing loss / don't know how to fully utilize the hearing that they have : )

Lasercd: Obviously, technology in this area is moving ahead at a great rate. Whether or not a designer / manufacturer knows how to fully impliment that technology is another matter.

Having said that, it is not uncommon for even the designer / manufacturer of an integrated circuit to be up in the air about what makes their design perform optimally and what hinders performance. This is what R&D ( research & development ) by the company using this product has to determine. There are many products using the same or quite similar parts that produce very different results, both sonically and measurably.

If you wait for the latest and greatest, you'll be waiting forever. Technology and use of that technology is ever expanding and ever changing. Sean
>

A newer model does not always mean better sound, it might produce identical sound but do it more effieciently, cost effectivly, and with less power. There is also a good thance that the improvement is so slight that nobody could tell.

If you had a system with 3 identical sources with identical setup, but one had the 1970 DAC, one with the 1972, and one with the 1976, and some Golden-Eared superguy sat there and took a listen, i doubt he would be able to tell the difference each time. If he was actually able to nail which DAC he was listening to with over 70% accuracy, i would eat all three sources.
:)

There is no guarantee that a better chip will sound better in a particular design. A poor board layout and parts choice can easily make a better DAC sound worse than an older generation DAC.