Review and Measurements of Soekris dac1421 Multibit DAC

[Blumlein 88]

I find it odd so many complaints about what the Soekris is being compared to in the review. The comparison with the Yggy makes sense as both are multi-bit. It could make sense to compare to the RME dac due to price being about the same.

Still all one needs to do is open two browser windows with the review of each DAC in it and compare any two you want to compare. We can do some thinking and comparing of test results for ourselves can't we?

 

[mindbomb]

Clearly, these numbers are an absolute fantasy aren't they?

Well, he said his sample didn't have the power supply hum and those harmonics, and using a -60dbfs 1khz signal to reduce the harmonic distortion, they might be legit?

 

[gvl]

Could this be another case of dynamic range vs s/n measurement confusion?

 

[mindbomb]

I find it odd so many complaints about what the Soekris is being compared to in the review. The comparison with the Yggy makes sense as both are multi-bit. It could make sense to compare to the RME dac due to price being about the same.

Still all one needs to do is open two browser windows with the review of each DAC in it and compare any two you want to compare. We can do some thinking and comparing of test results for ourselves can't we?

I don't want to make it seem like it's a huge deal, but I do agree with flipflop in that when the comparison is between 2 similarly priced things, I think they have more utility since I can imagine someone being torn between the 2 items. The multibit comparison was also interesting though.

 

[soekris]

Could this be another case of dynamic range vs s/n measurement confusion?

Apparently.... S/N ratio is the ratio of full signal to residential without signal, not to be confused with dynamic range....

And I said that the dac1421 do not have any power supply noise, and that goes for all of them, not just "his sample", they're all tested with FFTs at different levels.... It's pretty easy to pick up power noise at such low levels, just moving a unrelated power cable can change things.

Btw, while I do know the very old tda1541 chip, the naming is just a coincidence, I happens to use 4 digit product numbers and sooner or later they will hit something....

 

[Timbo2]

Don't you think it's a bit cheeky using the model number of arguably the most famous and revered audio D/A converter of yesteryear, the Philips TDA-1541...

Hey, for me I think of disk drive that liked to overheat and go out of alignment. I think it may come before the D/A!

 

[soekris]

A quick googling: Commodore 1541 disk drive from 1982, TDA1541 chip from 1984.

But I was never a Commodore guy, did my early computing on Z-80 running CP/M.

 

[mindbomb]

Apparently.... S/N ratio is the ratio of full signal to residential without signal, not to be confused with dynamic range....

When I thought of S/N, I was thinking of full scale thd+n. So that's why I was thinking it must have been a dynamic range measurement. I was wrong!

 

[soekris]

Actually, S/N Ratio is something everybody agree on, Dynamic Range there seems to be different definitions about how to measure it....

 

[restorer-john]

S/N ratio is the ratio of full signal to residential without signal, not to be confused with dynamic range

I get plenty of residential noise at my place

Signal to noise is full signal (your rated 2v) to the residual noise in V, expressed in dB. No-one is confusing anything here.

Unless you are deliberately muting the output when you get runs of zero data (which is essentially cheating), there is no way on earth you can get noise levels of less than <1.4uV in an active stage that can swing 2.0V RMS from 20-20KHz.

Amir's APX555 has a residual (20-22K bandwidth) of 2.8uV rated, and typically 2uV. Are you honestly telling us that a US$28,000 industry standard analyzer has a greater residual noise level in its D/A driven analogue generator than your entirely unshielded, single board discrete resistor D/A converter?

Sorry, I don't buy your numbers at all. It's a pretty looking board and congratulations on being brave enough to build a discrete D/A, but until proven otherwise, I'll take the (measured) mid 80s of S/N(+D) as being accurate and that isn't remotely SOTA is it?

 

[derp1n]

This is a friend's dac1541 measured in REW with a cheap Asus Xonar U7 by someone who doesn't know what they're doing (me), so take with a grain of salt. Power supply issues are present as Amir's graph, although a bit lower.

Topping DX7s measured at the same time:

So either I'm doing something wrong (likely!) or the DX7s is pretty disappointing.

U7 loopback:

 

[yummy]

TDA-1541 is not pure R2R, for christ!

nor 1547

which means, these ain't r2r

 

[DonH56]

Most if not all of the 16-bit (and higher) "R2R" types use a segmented architecture with unary MSBs and binary (R-2R) lsbs. The 1541 uses Rudy van der Plassche's dynamic element matching scheme, a pretty nifty idea for providing highly-accurate current sinks (or sources). A good description is in his book, Integrated Analog-to-Digital and Digital-to-Analog Converters (Kluiwer, eg. https://smile.amazon.com/Integrated...G-TO-DIGITAL+AND+DIGITAL-TO-ANALOG+CONVERTERS). Most folk these days lump data converters into delta-sigma and conventional (R-2R, segmented, slope, etc.) and somehow we manage to understand each other anyway.

 

[mindbomb]

Sorry, I don't buy your numbers at all. It's a pretty looking board and congratulations on being brave enough to build a discrete D/A, but until proven otherwise, I'll take the (measured) mid 80s of S/N(+D) as being accurate and that isn't remotely SOTA is it?

Tbh, I can buy that his measurements were in that range. The s/n of the rme adi2 is 117db unweighted, and comparing the j test here to that of the adi 2 in the dac 3 review, the noise floor on the soekris looks lower, except for the power supply stuff, which he says were not present in his measurements.

 

[gvl]

Yeah, noise seems to stay under -120dB for the most part and the main tone is at +5db. As the designer suggested lowering the volume should bring those harmonics down, and if the mains hum can be lowered somehow the numbers ought to go up.

 

[Blumlein 88]

Tbh, I can buy that his measurements were in that range. The s/n of the rme adi2 is 117db unweighted, and comparing the j test here to that of the adi 2 in the dac 3 review, the noise floor on the soekris looks lower, except for the power supply stuff, which he says were not present in his measurements.

Are you just quoting RME specs? Amir's results here for the Jtest don't look like 117 db s/n. The look more like 100db or a touch more.
https://www.audiosciencereview.com/forum/index.p

 

[mindbomb]

@Blumlein 88
yes, i'm going by the rme specs. s/n measurements don't take into account the noise floor modulation from the signal, so they can be lower... is like 12db lower out of the question?

 

[Grave]

Yay another DAC with 80 dB performance. Why would you recommend it?

 

[Blumlein 88]

@Blumlein 88
yes, i'm going by the rme specs. s/n measurements don't take into account the noise floor modulation from the signal, so they can be lower... is like 12db lower out of the question?

You probably know this, but just to be clear. SNR is usually a high level signal notched out to show the ratio of that signal to residual noise. If you have noise floor modulation usually SNR is lower than dynamic range. The standard for dynamic range is a 1 khz tone at -60 db notched out to see what noise level is left. It usually is a few db higher than SNR though sometimes they are the same. The -60 db signal is to prevent turning off the output when a silent signal is sent to the DAC chip and getting an artificially inflated number.

So in the example above the noise floor is bouncing along at -134 db which with FFT gain would be equivalent to -100 or 102 db. I didn't notice before that the signal is peaking at some plus 19 db or something. So that would actually be in the range of their claimed spec. I do think Amir should be more careful about such signals lining up with 0 dbFS as he displays them.

 

[mindbomb]

You probably know this, but just to be clear. SNR is usually a high level signal notched out to show the ratio of that signal to residual noise. If you have noise floor modulation usually SNR is lower than dynamic range. The standard for dynamic range is a 1 khz tone at -60 db notched out to see what noise level is left. It usually is a few db higher than SNR though sometimes they are the same. The -60 db signal is to prevent turning off the output when a silent signal is sent to the DAC chip and getting an artificially inflated number.

So in the example above the noise floor is bouncing along at -134 db which with FFT gain would be equivalent to -100 or 102 db. I didn't notice before that the signal is peaking at some plus 19 db or something. So that would actually be in the range of their claimed spec. I do think Amir should be more careful about such signals lining up with 0 dbFS as he displays them.

I thought this was true, but then soekris's description sounded like the high level signal wasn't notched. And that made me look up the official audio precision definition of an s/n test, and they also seem to indicate no notch, just residual noise with no signal. Am I wrong on this?

But then you're right about me overlooking the dbv scale. rme was definitely going with the tougher measurement that included noise floor modulation.