I've been a long-time lurker but just recently made an account. I want to know why doesn't TI start making the PCM1704 chips again? Or Philips with the TDA1541.. Looking at the current demand, they can probably sell a lot of units..
P.S. I'm not an expert in how chip manufacturing works so sorry if my question was bad or I posted this on the wrong section..
P.S. I'm not an expert in how chip manufacturing works so sorry if my question was bad or I posted this on the wrong section..
A chip manufacturers definition of "a lot of units", is not the same as yours....
And anyway, those old chip processes are not used anymore.....
There is no demand for those outdated bulky chips. Just like vacuum tubes. They don't fit in laptops and smartphones.
They don't re-start producing them for the same reason they stopped in the first place. Simple economics. It's much less costly to mass product Sigma-Delta-Modulator (SDM) D/A chips than to mass produce multibit types. This is because SDM D/A chips are either completely or mostly, all digital. Meaning, without elements requiring precision trimming to achieve accuracy/linearity. So, the industry rapidly switched to SDM type chips once that technology was proven. By proven, I mean, on an purely objective specification basis.
Neither large semiconductor vendors or their large customers are concerned about anything besides objective performance and production economics. Subjective sound performance is a different matter, and mostly of concern only to small specialty D/A chip customers and the relatively small customer base they serve. So, the specialty audio industry has done what it always has done, adapt and create for itself what it can't easily obtain from industry suppliers. This is the reason for the rise of discrete D/A converter (non-single chip) based DACs, and the utilization of costly multibit D/A chips intended for industrial applications (not consumer audio). The main drawback for consumers, unfortunately, is the resulting high cost of such products.
Neither large semiconductor vendors or their large customers are concerned about anything besides objective performance and production economics. Subjective sound performance is a different matter, and mostly of concern only to small specialty D/A chip customers and the relatively small customer base they serve. So, the specialty audio industry has done what it always has done, adapt and create for itself what it can't easily obtain from industry suppliers. This is the reason for the rise of discrete D/A converter (non-single chip) based DACs, and the utilization of costly multibit D/A chips intended for industrial applications (not consumer audio). The main drawback for consumers, unfortunately, is the resulting high cost of such products.
Last edited:
I contacted TI about the same and they said that their DAC11001A might be a good starting point if I'm looking for a similar R2R design as that of the PCM1704. Has anyone used this chip on the forum for a diy design?
Edit: I found the thread and following it. Thank you all for the replies!
Edit: I found the thread and following it. Thank you all for the replies!
Last edited:
Unless you know how to program a microcontroller, the DAC11001 does not bear further investigation.
If you are set on going the route of non-oversampling multibit, there is one bright light of good news. There remains a multibit audio DAC chip from the golden age of multibit DACs still in current production. This DAC does NOT require programming. More good news, while 'only' 16-bit, this DAC exhibits THD lower than other top vintage multibit DACs, including the 24-bit PCM1704 and the 16-bit TDA1541. This is according to a contributor here by the name of 'Bernhard'. Bernhard has posted many spectrum analyzer photos here at DIYAudio which support his claim.
The DAC chip I'm refering to is the Burr-Brown PCM56. Which is still available from the big electronic component distributors in the U.S. (Mouser.com, for example) for $27 in single unit quantities. To read much good DIY information on the PCM56, just search here for posts on the term PCM56 by member, Bernhard.
If you are set on going the route of non-oversampling multibit, there is one bright light of good news. There remains a multibit audio DAC chip from the golden age of multibit DACs still in current production. This DAC does NOT require programming. More good news, while 'only' 16-bit, this DAC exhibits THD lower than other top vintage multibit DACs, including the 24-bit PCM1704 and the 16-bit TDA1541. This is according to a contributor here by the name of 'Bernhard'. Bernhard has posted many spectrum analyzer photos here at DIYAudio which support his claim.
The DAC chip I'm refering to is the Burr-Brown PCM56. Which is still available from the big electronic component distributors in the U.S. (Mouser.com, for example) for $27 in single unit quantities. To read much good DIY information on the PCM56, just search here for posts on the term PCM56 by member, Bernhard.
You lose about a bit of SINAD with the PCM56, but it's quite a venerable replacement for the TDA1541A.
The major thing are customers. Customers decide according to parameters vs price. It's impossible to explain, that a 20-bit R-2R can sound better than a 32-bit 768kHz delta-sigma with 10 digital filters. Chip producers and DAC manufacturers will adapt to customers. Producers are happy, because it is much more economical to produce a delta-sigma chip than a high quality R-2R.