As is a Noddy / Big Bird / Barney guide to R2R DACs.
What are they?
Advantages /disadvantages?
Are all R2R DACs the same?
and why might someone want one?
.sjb
1 Like
The name R2R, was originally written as R/2R, and refers to a resistor ladder network consisting of resistors of two values, R and 2R. Using only those values a ladder is constructed of resistors (either as discrete components or within a silicon chip). The ladder results in a voltage output of a fraction of a reference voltage depending on the bit value of a digital input, which is all you need for digital to analog conversion.
This paper isn’t a Noddy explanation, but it sets out things slightly more approachably than Wikipedia.
Most modern DACs use a delta sigma design rather than an R/2R implementation. This Wikipedia article describes the different DAC architectures under the Type heading.
An R/2R DAC can be implemented in silicon (TI make some) but there are a number of discrete component (usually surface mounted) designs. Some of them replicate the ladder in order to eliminate non-linear distortions. I think one output is subtracted from another and the difference inverted and applied to cancel such difference out of the final output. This may not be feasible in silicon, but I’m not sure.
As described the R/2R converts a multi-bit PCM input, but it is possible to do something similar for a DSD input. I’m listening to one now, the Holo Audio Spring. It has 8 resistor networks. 2 channels/PCMx2/DSDx2 where the second network is used to correct linearity as above.
As to why have they become popular, I think some people never bought into Sigma Delta. Certainly the Schitt guys converted a lot of folks back to R/2R with the Yggdrasil.
Edit: the R/2R architecture advantage over a switched resistor binary weighted DAC is that by using only two resistor values greater precision is possible than by attempting to match varying resistor values.
1 Like
R2R operates on a principle of switch resistors network and there’s no feedback loop to do the correction. Therefore R2R requires good matching to get good linearity. Because it operates in full set of bit range, e.g, 24-bit and does not require any form feedback loop to do the correction, the results is better time response and does not require noise shaping circuit. R2R DAC is inherently very quiet.
Sigma delta 1 bit converter operates in a feedback loop constantly correcting itself from errors. The delay of which feedback loop keeps correcting itself from errors has a great impact how fast these converter can response in a time domain. Because it operates in 1-bit fashion, it requires noise shaping to push the noise away from the audio band. Sigma Delta is quite noisy considered if one open up the whole spectrum of frequencies.
1 Like
Schitt Gungnir Multibit (Gumby) here. Little brother to the Yggy. Very nice and affordable DAC.
Some of us just used our ears and heard the much improved instrument realism from even the lower-end Schiit multibit DACs compared with even higher-end delta-sigma DACs. For the music I listen to, small ensemble modern jazz, solo/chamber classical, small ensemble world, some voices, even the relatively inexpensive Schiit Bifrost multibit is better than any delta-sigma DAC I’ve heard that is not more expensive than the Yggdrasil. The Yggdrasil, which I also own, is pretty much end-game for me. But I’ve just ordered a Holo Spring Level 3 as well, very curious about discrete ladder NOS vs the upsampling+exact filter solution + integrated ladder in the Schiit R2R DACs. It might replace my Bifrost on my headphone setup.
1 Like
I think there is difference between "regular " R2R DACs, such as Schiit Multibit DACs, and R2R Ladder DACs, like Holo Spring and Denafrips Ares. What the difference is, I am not sure.
List of known R2R Ladder DACs:
Holo Audio Spring, Total DAC, MSB Analo/DAC V, Aqua HiFi Formula, Metrum Menuet/Pavane, Denafrips Ares.
Learning my lesson, I suppose what I really wanted to know, what is the perceived sonic differences between R/2R DACs and delta sigma(?) DACs i.e. non R/2R DACs.
In what way do they sound different?
And then a related but separate questions is, from what I could understand the R/2R (and I may be wrong here) DACs use hardware to convert the signal and so there is no noise shaping etc. Why then use these DACs with HQP which add the noise shaping etc? Is it because the do no up sampling/noise shaping that can interfere with Miska’s magic?
.sjb
Yes, that’s it. A non-oversampling DAC that can accept a wide variety of digital inputs is ideal for HQP. It is like a blank canvas. There is also a cost saving in not implementing oversampling, but many DACs still have an option to switch oversampling on.
Is T+A DAC8 DSD (not R2R) sort of NOS DAC ?
Seems to output good SQ when using with HQP, why ?
Non over sampling DAC tends to perform better in time domain response at the expense of signal to noise ratio.
All of those DACs use resistor ladders for the main D2A step. They differ in two things:
- ladder’s design and implementation
- whether they use oversampling (if not, they are called NOS for “no oversampling”)
For 1), Schiit multibit DACs use instrumentation DAC chips from Analog Devices that include the resistor ladder. Some of the others, for example the Holo Spring, use discrete resistor ladders.
For 2), the Schiit multibit DACs use a proprietary closed-form digital filter implemented on a SHARC DSP processor to upsample to a target high sampling rate, and them they send that to the chips containing the resistor ladders for A2D conversion. Some other DACs are strictly NOS, that is, they do the A2D conversion via ladder at the source’s sampling rate, or have a NOS mode (like the Holo Spring). In general, the point about NOS is that it avoids upsampling artifacts. Schiit claims that their approach with a closed-form filter has the best of both worlds: like NOS, it preserves the original samples, but by oversampling it relies less on downstream analog processes to filter out the high-frequency energy in the (approximate) analog step function created by the resistor ladder. Metrum has a good discussion of NOS.
I have two Schiit multibit DACs (Yggdrasil and Bifrost) and I have a Holo Spring on order, so I’ll get to make up my mind about NOS vs not soon enough. I ordered the Holo Spring because of a couple of very interesting reviews I read that put it on a par (but sounding subtly different) with the Yggdrasil, which I love. I may end up using it to replace the Bifrost (which I really like, although the Yggy is at a different level).
1 Like
The T+A DAC8 DSD does have oversampling capability as per these specifications. It can also be set to simply output the rate that is input, meaning that it would then operate in NOS mode.
The favourable reviews of the DAC8 DSD when using DSD 512 input from HQ Player would have been done using the DAC in NOS mode without the internal oversampling filters being applied.
Edit: I haven’t been able to tell from the specs or reviews whether the T+A uses an R/2R design. It seems to implement resistor ladders within silicon for PCM and refers to a one bit converter for DSD, which may or may not use an R/2R design. The main thing is that it does not appear to use a Sigma Delta conversion chip.
Check this out:
It’s not the most affordable R2R in the market, the Schiit Modi Multibit is cheaper. People get confused that “ladder” means “discrete ladder” but the chips in the Schiit multibit DACs contain resistor ladders as much as the DACs with discrete resistor ladders.
Fernando:
Did you get the Holo Spring yet? I am interested in your experiences in comparing the Spring with Yggdrasil. Thanks!
Still waiting, they have a bit of a queue for the custom Level 3 units which is what I’ve ordered.
Wow! It’s been exactly a month from your post / order. They must be swamped with orders. I guess I was lucky I found one used.
Fernando,
Did you see the write up on the Yggy from Robert Harley’s The Absolute Sound article?
He said the Yggy is only one of the Three best DACs he has heard ever! The Includes the dCS Vivaldi $36K and the $20K Berkley Reference !
His conclusion says it all though…
I don’t know how Schiit Audio has done it, but the $2300 Yggy
is in many ways competitive with any DAC I’ve heard regardless
of price. In some criteria—transient speed without etch, clarity
of musical line, whole-body involvement—the Yggy is as good
as digital gets. Yet the Yggy’s bold incisiveness may not resonate
with listeners who prefer a more relaxed and easygoing sound.
I, however, have no such reservation; this is a DAC I could
listen to and enjoy for a long time. In fact, there was something
different about the Yggy that pushed my buttons—I felt a musical
exhilaration that was experienced not as some intellectual
abstraction, but at a more fundamentally visceral level.
If you’re looking for a DAC that does quad-rate DSD, decodes
MQA, offers a volume control, and includes a headphone
amp, look elsewhere. But if the very best reproduction of PCM
sources is your goal, the Yggdrasil is the ticket. It’s a spectacular
performer on an absolute level, and an out-of-this world bargain.
The Yggy is not just a tremendous value in today’s DACs,
it’s one of the greatest bargains in the history of high-end audio.
1 Like
but does he like it?
.sjb
1 Like
Rhetorical ??? Lemme see…a $2300 DAC that keeps up/exceeds a $36K and $20K DACs???