Jump to content
IGNORED

Multi-bit, but not R2R


Recommended Posts

Given the current interest in R2R dacs, audio nerds -- including professional reviewers (Stereophile, etc) -- can lump  classic Philips DAC chips (1541, etc)  into the R2R category.

The Philips chips were not R2R, however. The 154x family was DEM (dynamic element matching) or, for 1545 a relative of DEM known as CC (continuous calibration).

As this old diyaudio thread notes:

Quote

Yes, they are "multibit" Chips, however their operation internally drastically differs from R2R Architectures, by using a mixture of active switched current sources for the lower bits and multiemitter transistor current dividers for the upper bits. The operation methodes between R2R and the Philips devices could not be more different.
 

 

The reason for this Mr-know-it-all prattle is that designers who are chasing the multibit rabbit  -- HOLLO, Denafrips, TotalDAC, etc. -- might actually snare some meat f they try to emulate the orig. Philips design (maybe using discrete components, or FPGA).

I haven't heard any of the latest mega-$ discrete models. But, frankly, all else held equal, the R2R chip dacs of yesteryear  (PCM63, PCM1704, AD1862, etc)  usually lacked the punch and drive of the Philips 1541. With or w/o oversampling.

 

Refs:

Stereophile keeps fuckin' it up:

https://www.stereophile.com/content/borderpatrol-digital-analogue-converter-se

https://www.stereophile.com/content/holoaudio-spring-kitsuné-tuned-edition-level-3-da-processor

Link to post
Share on other sites

Its not correct to say that 'the 154x family was DEM' since really its not a family at all. TDA1543 is a bipolar IC technology DAC without any fancy DEM tricks. TDA1545 has, as you correctly point out, 'continuous calibration'. TDA1547 is an altogether different beast - a 'Bitstream' converter back-end. There is also TDA1549 which looks to be a cut-down variant of TDA1305, one of the first multibit Bitstream devices, also using CC.

Link to post
Share on other sites
16 minutes ago, opus101 said:

Its not correct to say that 'the 154x family was DEM' since really its not a family at all.

Yes, thx for that correction. I forgot about the 154x carryover to the Bitstream era. Somehow, I only had the tda1387 and 1305 in mind for the 1990s. Actually, the 1387 is the I2S version of the EIAJ-based 1545.

Link to post
Share on other sites

Without wanting to get excessively nerdy, there are more differences between 1387 and 1545 than just the digital interface. The output compliance range is a fairly significant one, the facility for increasing the Imax is another. I tried increasing Imax on 1387 and quickly ended up with a dead chip so that's not just a cosmetic difference.

 

Back to your original post, I agree some more attention in the mainstream for Philips' DAC designs would be welcome - why should R2R capture all the multibit limelight?

Link to post
Share on other sites
4 hours ago, opus101 said:

... - why should R2R capture all the multibit limelight?

Well, I think because: (a) the R2R circuit topology is "easier" to implement /tweak; (b)  more readily avail. schemas, design articles, etc. 

I have a few ancient Philips science articles -- pub. in various journals -- from the late 70s and 80s. But they are pretty scant on actual design detail. And I can find almost nothing from Sony. This is unlike the case for AD and TI/Burr-Brown -- both pub'd very detailed Application Notes, etc.

Link to post
Share on other sites
5 hours ago, analogue said:

Yes, thx for that correction. I forgot about the 154x carryover to the Bitstream era. Somehow, I only had the tda1387 and 1305 in mind for the 1990s. Actually, the 1387 is the I2S version of the EIAJ-based 1545.

 

Bitstream aka "DSD" DACs, like the Crystal (now called Cirrus Logic) CS4328 as well...

 

I really liked sound of the DF1700 + PCM63P, but that one is not functional right now, it is stripped and waiting for installation of JLSounds XMOS USB interface as front-end, so it has become NOS with front side including DF1700 removed. Plan is to run it in a modern way from computer based oversampling and see how it sounds then.

 

I still have my early DAC designs functional. Like DF1700 + PCM1700 combo. And the CS4328 based one is not bad either. Not long ago I made also some measurements of these old things and they perform pretty decently!

 

I had Marantz CD60 with SAA7220 digital filter and TDA1541A DAC, but it always had too much digitus in the sound for my taste and has been recycled. IIRC, the analog reconstruction filter was quite fancy with LC/SK construction.

 

Signalyst - Developer of HQPlayer

Pulse & Fidelity - Software Defined Amplifiers

Link to post
Share on other sites
Quote

I had Marantz CD60 with SAA7220 digital filter and TDA1541A DAC,

I currently own a Philips CD60 (exactly the same as Marantz CD60) from 1989. Stock, it is okay. But i've modded it with a higher-grade 7220B and a double-crown TDA1541A. The output stage was replaced with a fully discrete "DIY" device. A better clock and better-branded passive parts (caps, etc) further improved sonics.

So it's pretty souped up .... and it remains one of my digital references to this day. I at least know it sounds better than my much-modded Theta Chroma 396; and modded Musical Fidelity A324 dac, that was Stereophile Class A back in the early 2000s.

Link to post
Share on other sites

I would love it if you two and/or others could elaborate further. Not being versed in electrical engineering. I've always felt my understanding of DACs and their number of bits was abstract at best.

 

I know that there are 1-bit/bitstream DACs that as Miska notes are basically DSD DACs. And I've read repeatedly that most current bitstream DACs are actually more PCM/DSD hybrids that are bitstream(?) but with 4 or 5 bits instead of 1 bit. Apologies if I've garbled that summary - as I say, I'm not sure I fully understand.

 

As for older multi-bit designs like the 1980s Philips 14-bit and 16-bit DACs, I'd always assumed all of those were R2R and had no idea that such a DAC could be anything else.

 

So... could anyone explain more about how a non-R2R but still multi-bit DAC is actually implemented?


Thanks!

Link to post
Share on other sites
3 minutes ago, tmtomh said:

I would love it if you two and/or others could elaborate further. Not being versed in electrical engineering. I've always felt my understanding of DACs and their number of bits was abstract at best.

 

I know that there are 1-bit/bitstream DACs that as Miska notes are basically DSD DACs. And I've read repeatedly that most current bitstream DACs are actually more PCM/DSD hybrids that are bitstream(?) but with 4 or 5 bits instead of 1 bit. Apologies if I've garbled that summary - as I say, I'm not sure I fully understand.

Most modern DAC designs convert normal 24-bit PCM data to a much higher sample rate with a relatively small number of levels, typically in the 5 to ~65 range. This can be converted to analogue using a resistor array with equal weighting, eliminating the need for extreme tolerances. A greater number of levels makes noise shaping easier while complicating the conversion stage. As with so many things in engineering, it's a trade-off. Some DACs (e.g. TI) use a hybrid approach with R2R (or equivalent) for the top few bits and sigma-delta for the remainder.

 

3 minutes ago, tmtomh said:

As for older multi-bit designs like the 1980s Philips 14-bit and 16-bit DACs, I'd always assumed all of those were R2R and had no idea that such a DAC could be anything else.

Many DACs from that era are switched capacitor rather than R2R designs.

Link to post
Share on other sites

There are just so many different variations in the ways a DAC can be implemented that it is not easy to draw a line or categorize.

 

Almost regardless of number of bits (either binary or unary coding) at the input, there are many ways you can design a converter to convert series of single bits or multiple bits into analog.

 

You can also drive existing designs in various different ways, especially if you have direct access to the actual converter.

 

Signalyst - Developer of HQPlayer

Pulse & Fidelity - Software Defined Amplifiers

Link to post
Share on other sites
5 hours ago, tmtomh said:

As for older multi-bit designs like the 1980s Philips 14-bit and 16-bit DACs, I'd always assumed all of those were R2R and had no idea that such a DAC could be anything else.

 

This Dutch website is your best friend for Philips info:

https://www.dutchaudioclassics.nl/Philips-TDA1541_A-Monolithic-Dual-16-Bit-D-A-Converter/

Link to post
Share on other sites

What bothers me about brute-force conversion methods, like R2R, is Johnson noise (thermal noise). The more R's there are, the more this stuff adds up*. Not sure if Philips was thinkin' physics in their design decision (in addition to mass-scale economics). 

 

* It does matter in components, such as tradit. (analog) volume controls.

Link to post
Share on other sites

I can't recommend the following 50min vlog highly enough. 

 

Here, 

Quote

 Shahriar presents the theory, design and characterization of Nyquist Digital to Analog Converters (DACs). After a brief overview of DAC operation and theory, the schematic of an 8-Bit R-2R DAC is presented. The R-2R DAC, which is driven by a dsPIC30F6014A Microchip microcontroller  is capable of producing ramps and arbitrary waveforms uploaded through an RS232 interface. The static integral non-linearity (INL) and differential non-linearity (DNL) is measured by using a Rigol DM3068 Multimeter through a Matlab interface program. The dynamic performance of the DAC is characterized using an Analog Devices AD6645 105MSps 14-Bit ADC evaluation board coupled with a USB FIFO interface board. By using the 'Visual Analog' software, the spectrum of the DAC output as well as the signal to quantization and distortion ratio (SQNDR) is calculated. Finally, the impact of component mismatches, operational amplifier non-linearity and timing uncertainty on the INL/DNL and SQNDR (ENOB) of the DAC is examined.

 

Note how good the 'scope metrics look as  Shahriar experiments with damaging the design. Point being (for audio, anyway): measurements only tell you so much.

Link to post
Share on other sites
  • 1 month later...

I'm still trying to figure out what the market is for this chip - the fact that it has an SPI interface tells me its more industrial than audiophile. 'Lab and field instrumentation' it says but what about those applications needs R2R rather than (significantly cheaper) S-D ? Could it be performance inside a control loop, given there's an example shown on the first page? 'Professional audio amp (rack mount)' looks like an outlier.. The maximum supply and reference of 40V suggests some seriously impressive SNRs might be possible, the output noise figure of 7nV/rtHz  implies 1uV in the audio band, signal may go up to 14VRMS,

Link to post
Share on other sites
8 hours ago, opus101 said:

I'm still trying to figure out what the market is for this chip - the fact that it has an SPI interface tells me its more industrial than audiophile. 'Lab and field instrumentation' it says but what about those applications needs R2R rather than (significantly cheaper) S-D ? Could it be performance inside a control loop, given there's an example shown on the first page? 'Professional audio amp (rack mount)' looks like an outlier.. The maximum supply and reference of 40V suggests some seriously impressive SNRs might be possible, the output noise figure of 7nV/rtHz  implies 1uV in the audio band, signal may go up to 14VRMS,

It's obviously aimed at industrial and scientific applications. For audio applications, the specified THD of -105 dB of this chip is surpassed by many a sigma-delta design. Even if the analogue noise level is low, you're still limited to -115 dB or so by the 20-bit resolution. At low signal levels, glitching could also be a problem that needs to be dealt with. Regarding the "professional audio amplifier" application, I can't figure out how they're envisioning this device to be used there. The linked page mentions the somewhat similar DAC8812 in addition to a regular audio DAC, but the reference designs do not include it.

Link to post
Share on other sites

There's some technical background on the device from one of the guys at TI over here : https://www.diyaudio.com/forums/vendor-s-bazaar/349809-dac11001-20-bit-2r-precision-dac.html

 

Seeing as the maximum update rate is around 500kHz (faster if the deglitcher is bypassed but that comes with its own issues) there's room for some oversampling/noise shaping to improve on the quantization noise limit imposed by the part being only 20bits.

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now


×
×
  • Create New...