• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Sub 1V SINAD?

S/N ratio always decreases when the Signal is lower and the Noisefloor of the system remains the same.
When a different gain is selected the noise floor also changes.
Sure one can used very low noise pre-amps but would have a notch the measurement signal or make the signal small enough so it doesn't clip the pre-amp.

Amir doesn't have (or at least use) one of these so in the measurements shown here we are often looking at system noise floors.

There are tricks like averaging a signal so the noise lowers opposite the signal. Unfortunately this is just interesting for measurements our ears can only do that up to a point and use other techniques.
 
@DonH56 On the topic of noise, could I get your thoughts on the schiit yggdrasil? It uses instrumentation DACs, but it performed rather poorly in Amir's review of it. Was that due to implementation/poor design, or is there an inherent issue using R2R DACs for audio?
 
or is there an inherent issue using R2R DACs for audio?
This.

It is very difficult to make a low distortion R2R dac. Even if they don't make the stupid design choise to make them filterless.
 
It is very difficult to make a low distortion R2R dac. Even if they don't make the stupid design choise to make them filterless.
Interesting. So then why would they boast better ENOB even though SD has lower distortion figures?
 
Interesting. So then why would they boast better ENOB even though SD has lower distortion figures?
For the same reason all the other vendors of snake oil and audiophile jewellery lie about "creatively market" their products.

To make a sale.
 
For the same reason all the other vendors of snake oil and audiophile jewellery lie about "creatively market" their products.

To make a sale.
Wow. Thanks Texas Instruments :rolleyes:
 
@DonH56 On the topic of noise, could I get your thoughts on the schiit yggdrasil? It uses instrumentation DACs, but it performed rather poorly in Amir's review of it. Was that due to implementation/poor design, or is there an inherent issue using R2R DACs for audio?
I am not Don. Amir's Yggdrasil (Gen 1 & 2) review was from 2018. The current "More is Better" version that uses TI's DAC11001B R-2R DAC chips, per Schiit's AP measurement, isn't too bad.

https://www.schiit.com/public/upload/PDF/MIB more better.png
MIB%20more%20better.png
 
@DonH56 On the topic of noise, could I get your thoughts on the schiit yggdrasil? It uses instrumentation DACs, but it performed rather poorly in Amir's review of it. Was that due to implementation/poor design, or is there an inherent issue using R2R DACs for audio?
No idea, I do not own nor know anything about it. There could be any number of reasons, including whatever DACs, output buffers, and so forth. R2R DACs are generally harder to make and compensate than a delta-sigma design for various reasons but there is no theoretical reason they should perform more poorly. Most of the R2R DACs I have seen for audio top out around 16 bits max, however, compared to 24 bits for delta-sigma DACs, but those are very different architectures. That does mean in general R2R DACs will not match the performance of a delta-sigma due to the lower resolution (and thus higher quantization noise floor, though that is only one component of SINAD).

Note SINAD, and ENOB which is derived from SINAD (at least using the IEEE Standards I follow), do not distinguish noise from distortion or other artifacts like clock or power supply spurs. It is possible to have low noise and high distortion or vice-versa with equivalent SINAD (ENOB) but very different spectral signature (and thus potentially audible differences).
 
Last edited:
I am not Don. Amir's Yggdrasil (Gen 1 & 2) review was from 2018. The current "More is Better" version that uses TI's DAC11001B R-2R DAC chips, per Schiit's AP measurement, isn't too bad.

https://www.schiit.com/public/upload/PDF/MIB more better.png
MIB%20more%20better.png
GoldenSound measured the "More is less" version using TI DAC11001A chips. It does decently at 1 kHz but less so at higher frequencies.
1737079494522.png

1737079868783.png

1737079538539.png


Also distortion depends on sample rate. Here it is at 192 kHz instead of 44.1 kHz as above.
1737079565156.png

1737079587026.png
 
GoldenSound measured the "More is less" version using TI DAC11001A chips. It does decently at 1 kHz but less so at higher frequencies.
And that is with the R2R resistors implemented on chip where all the resistor matching and tolerancing is taken care of for the designer. Which is probably why it performs better than other R2R designs where the ladder is implanted in discrete resistors on the PCB.
 
Back
Top Bottom