Thanks heaps.
Still top notch (as expected)
Makes for easier comparison to other DACs on SINAD chart
-
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!
Mola Mola Tambaqui DAC and Streamer Review
- Thread starter amirm
- Start date
Thanks heaps.
Still top notch (as expected)
Makes for easier comparison to other DACs on SINAD chart
OP
Mola Mola Tambaqui DAC being returned to the kind ASR member.
Hopefully that kind of insured value warrants a personal "Handle with care" note from a shipping company manager to the employee.
Feyire
Active Member
Veri
Master Contributor
- Joined
- Feb 6, 2018
- Messages
- 9,597
- Likes
- 12,039
Good to see the headphone output is actually good! I believe the XLR right next to the TRS is a balanced headphone output, too bad you didn't test it but it's probably samey good performance.Here are headphone measurements. It was a pain to do because by default headphone out is disabled and there is no way to enable it via the remote or the front panel! Had to download the app to do it. Get rid of the bloody fish logo and use the space to implement a proper UI to control the entire set for heaven's sake.
Power into 300 Ohm load:
View attachment 45096
Extremely low noise and distortion but limited by the maximum 2.7 volt output.
Similar story at 33 ohm:
View attachment 45097
So not powerful but very clean.
Signal to noise ratio at 50 millivolts is good but not top of the class:
View attachment 45098
View attachment 45099
Output impedance is comfortably low:
View attachment 45100
That's it. She is going back in the box so I can review the mountain of gear waiting!
- Joined
- Jul 4, 2019
- Messages
- 8
- Likes
- 361
Thanks Amir for doing this review!
It's kinda funny, it feels almost like a blast from the past as I did this DAC design in 2013... It wasn't a single handed job btw. I did the schematics and prepared all the algos in MATLAB and my mate Bart van der Laan then rolled the circuit boards and did some heroic assembly language coding.
In case anyone's wondering why I decided to go discrete, I actually started testing existing sigma-delta DAC chips first but could find none that didn't have idle tones. I suspect that is still the case. Chip manufacturers usually manage to move these out of the band at mid-scale (i.e. zero or small signal), but they show up in a THD vs level graph as a small increase in apparent noise typically starting at -20dBfs. Basically this "noise" are tones that are swept in and out of the audio band, frequency modulated by the signal. The simplest way of testing for this is to do a noise level vs DC input plot. The tones, when they appear, are well above the noise floor, even as integrated over the audio band. Using PWM as a conversion format solves this tone problem, but nobody is doing that on an IC. Hence the discrete design. I won't speculate on the audibility of this phenomenon but anything that is measurable is fair game for me. If people are going to shell out serious moolah for a DAC, least thing you can do is show an objectively provable benefit. Low jitter is also something I like to that's why we ended up coding our own ASRC algorithm.
Other than that I only wanted to comment on the FFT and notch filter. AP threw a bit of a curveball with the APX555 as compared to the previous SYS2 range. You see, they had been trying hard to get people to understand this idea of notching out the fundamental before digitizing the residual. And still it turned out that most folks (including chip manufacturers in their data sheets) preferred posting FFT's with the fundamental in place, presumably because their clientele couldn't read graphs with the fundamental notched out. So, for the APX555 they decided to do the following (if you check the "high performance sine analyzer" box on the control panel):
1) Notch out the fundamental
2) Convert the residual
3) In the digital domain, put the fundamental back in
4) Plot the FFT.
Presto, an FFT including the fundamental.
So now they have to explain to their more informed users that yes, don't worry, we do use a notch filter only you don't see it. It's a case of damned if you do and damned if you don't. Multitone tests do rely fully on the performance of the ADC in the analyzer but that's always been the case.
The upshot of all this is that Amir's HD FFT's do correctly use the notch trick as nature intended it. It's just cleverly hidden by the analyzer.
It's kinda funny, it feels almost like a blast from the past as I did this DAC design in 2013... It wasn't a single handed job btw. I did the schematics and prepared all the algos in MATLAB and my mate Bart van der Laan then rolled the circuit boards and did some heroic assembly language coding.
In case anyone's wondering why I decided to go discrete, I actually started testing existing sigma-delta DAC chips first but could find none that didn't have idle tones. I suspect that is still the case. Chip manufacturers usually manage to move these out of the band at mid-scale (i.e. zero or small signal), but they show up in a THD vs level graph as a small increase in apparent noise typically starting at -20dBfs. Basically this "noise" are tones that are swept in and out of the audio band, frequency modulated by the signal. The simplest way of testing for this is to do a noise level vs DC input plot. The tones, when they appear, are well above the noise floor, even as integrated over the audio band. Using PWM as a conversion format solves this tone problem, but nobody is doing that on an IC. Hence the discrete design. I won't speculate on the audibility of this phenomenon but anything that is measurable is fair game for me. If people are going to shell out serious moolah for a DAC, least thing you can do is show an objectively provable benefit. Low jitter is also something I like to that's why we ended up coding our own ASRC algorithm.
Other than that I only wanted to comment on the FFT and notch filter. AP threw a bit of a curveball with the APX555 as compared to the previous SYS2 range. You see, they had been trying hard to get people to understand this idea of notching out the fundamental before digitizing the residual. And still it turned out that most folks (including chip manufacturers in their data sheets) preferred posting FFT's with the fundamental in place, presumably because their clientele couldn't read graphs with the fundamental notched out. So, for the APX555 they decided to do the following (if you check the "high performance sine analyzer" box on the control panel):
1) Notch out the fundamental
2) Convert the residual
3) In the digital domain, put the fundamental back in
4) Plot the FFT.
Presto, an FFT including the fundamental.
So now they have to explain to their more informed users that yes, don't worry, we do use a notch filter only you don't see it. It's a case of damned if you do and damned if you don't. Multitone tests do rely fully on the performance of the ADC in the analyzer but that's always been the case.
The upshot of all this is that Amir's HD FFT's do correctly use the notch trick as nature intended it. It's just cleverly hidden by the analyzer.
Any chance of pictures inside?
- Joined
- Feb 23, 2016
- Messages
- 20,695
- Likes
- 37,432
When I purchased Quad ESL63 speakers in Rosewood, it was a COD delivery from UPS. I was at home waiting on them. Hear the UPS truck roll up. I go out the door and out to the truck. Hear the rear door roll up. As I am approaching the rear the first speaker box is flipped out vigorously, bounces on the corners end over end 3 times and then flops over on the side on pavement. I rush to the rear and catch the second one as it too is flipped vigorously out. It nearly knocked me down, but I save it.
Driver is grumbling about those big boxes being in his way all day and he is glad to get them out. I'm complaining about them being very expensive packages clearly marked fragile and this end up. And not in too kind a tone or language either. His reply is reassuring: "eh! UPS pays for stuff all the time, don't worry about it". Being COD I refused to pay unless he lets me open and inspect them first. Two corners on the first box are crushed. Driver says he isn't supposed to do it, and won't. I tell him fine, just carry them back, before you get there I will call and explain why I didn't accept the shipment. I took a picture (film camera in those days) of the crunched box. Eventually he lets me open them. The one with crushed corners missed damaging the Rosewood panels by a half inch, but they were extremely well packaged and no damage done.
Then there was the air shipped Sonographe turntable which had no visible damage to the box, but which was pretty much totally destroyed. The box Sonographe used on those was the size of a 15 inch subwoofer with plenty of protection. They must have dropped it 50 ft onto the pavement from the cargo hold of the plane and it landed dead flat with nary a mark on the box. The shipping company ate that one completely.
Oh, and the Soundlabs that Delta Air lost for nearly a week. They sat outside in a 3 day rainstorm during that time. When finally they turned up, and made it on to my local airport it was 7 degrees that morning. The shipping crate was ice. Thank god, the owner put two layers of plastic around them internally, plus original plastic bags and no water/ice got anywhere to do any damage.
Last edited:
- Joined
- Feb 13, 2018
- Messages
- 574
- Likes
- 720
Wow that is very long, potentially disqualifying this DAC for video synchronization or digital instrument players for example.
That is a very interesting metric to have in all DAC tests IMHO.
Does this DAC upsample all inputs to 1bit high sample rate?
Do you have a test on another DAC that shows much lower delay? That seems very long for a DAC.
My M500 has a _very_ noticeable delay over USB compared to SPDIF on some cheap DAC.
I use the Tambaqui, through its optical input for films, lip sync is fine, I don’t know the latency figures but can ask.
Keith
Keith
Different TVs can route audio differently, such that it's possible a small delay in a DAC will improve lip sync over zero delay, especially if the TV is doing trick video processing.
Different people have different tolerances to this in my experience, I'd say 50ms lip-sync offset would not be noticed by many, when I've had complete control over the sync and tuned it in there is around +/- 50ms wiggle room where the error is not obvious.
Having the real numbers for the delay is what we need, and almost any delay won't matter for music playback, outside of a studio.
IVX
Major Contributor
100.00% because of PWM = 1bit, hence requires high sample rate. I remember an old japanese DAC IC used PWM as well, and it wasn't bad for 1995 or so.
If latency is important for you, buy a pro dac (rme or others) they do have low latency (because it is important live and for mixing).
HiFi News:
"Mola-Mola’s software upsamples all incoming data to 3.125MHz, truncating the wordlength to 5-bits while using a 7th-order noise-shaper to retain full dynamic range right up to 80kHz. Each 5-bit digital ‘word’ is sufficient to describe one of 32 possible pulse widths that, in turn, describe the amplitude of the audio signal on an (over)sample-by-sample basis. The pulses vary in steps of 10nsec (the system clock is 100MHz) right up to a full width of 320nsec (0.32µsec). This Pulse Width Modulated (PWM) signal is fed into a 32-stage shift register, clocked at 100MHz, so a composite of 32 pulses ends up reproducing the full PWM signal every 10nsec.
The 32 outputs of the register are summed together so that the final DAC output is the moving average of the PWM signal over consecutive blocks of
32 clock cycles (ie, one PWM cycle), updating every 10nsec. The PWM signal is ‘conditioned’ by a comb filter whose teeth coincide exactly with the 3.125MHz repetition rate. Mola-Mola could have used the signal from any of the 32 outputs alone and simply low-pass filtered it. Instead, the moving average technique not only overcomes any slight mis-match in the summing resistors but it also removes the PWM carrier that could potentially demodulate clock jitter down into the audio band."
"Mola-Mola’s software upsamples all incoming data to 3.125MHz, truncating the wordlength to 5-bits while using a 7th-order noise-shaper to retain full dynamic range right up to 80kHz. Each 5-bit digital ‘word’ is sufficient to describe one of 32 possible pulse widths that, in turn, describe the amplitude of the audio signal on an (over)sample-by-sample basis. The pulses vary in steps of 10nsec (the system clock is 100MHz) right up to a full width of 320nsec (0.32µsec). This Pulse Width Modulated (PWM) signal is fed into a 32-stage shift register, clocked at 100MHz, so a composite of 32 pulses ends up reproducing the full PWM signal every 10nsec.
The 32 outputs of the register are summed together so that the final DAC output is the moving average of the PWM signal over consecutive blocks of
32 clock cycles (ie, one PWM cycle), updating every 10nsec. The PWM signal is ‘conditioned’ by a comb filter whose teeth coincide exactly with the 3.125MHz repetition rate. Mola-Mola could have used the signal from any of the 32 outputs alone and simply low-pass filtered it. Instead, the moving average technique not only overcomes any slight mis-match in the summing resistors but it also removes the PWM carrier that could potentially demodulate clock jitter down into the audio band."
Are there numbers for rme? I expect most normal DACs are low latency, it's probably the conversion to PWM that introduces the delay.
- Joined
- Jul 4, 2019
- Messages
- 8
- Likes
- 361
All ADCs/DACs have some sort internal data format that is used to get from digital to analogue. The converter circuit and the internal format are chosen together, depending on the designer's judgment or expectation of which combo is optimal.
There are only two cases where the internal format coincides with an existing audio format. On one extreme are R2R DACs that directly convert the PCM data you feed them by controlling 24 switchable current sources, each of which has half the current of the previous one. On the other extreme sit 1-bit converters that switch a single current or voltage source but at a high sampling rate. Both these extremes have limitations and the most common choice these days is a kind of compromise. They use a small number of bits (typically 5) that are used to control 32 (2^5) current sources with nominally equal currents, plus some trick to make the conversion minimally sensitive to any imbalances in those currents. The Tambaqui sits more or less in this camp: the PWM signal has a switching frequency of 3.125MHz, and can take 33 discrete lengths from 0*10ns to 32*10ns. So it's basically a 5-bit, 3.125MHz converter. The choice for PWM was given by the need to get rid of those idle tones I mentioned.
In any DAC design the internal format is essentially fixed, as it's a design choice. So any type of digital signal first gets converted to that format. That conversion consists of two stages. The first is upsampling and that takes the sample rate up to the sample rate of the internal format. Then the second is sigma-delta modulation (or noise shaping, however you want to call it) which takes the word length down to the word length of the internal format while preserving the SNR inside the audio band. Unfortunately, a succession of audio product and sales managers have conspired to conflate the second stage with the first, thereby confusing just about anyone with a passing interest in the technical background.
---
The latency is indeed around 100ms, for all inputs. There's no real reason for it other than at some point we finally got the ASRC algorithm working glitch-free by making the input data buffer rather long. Much later we figured out how to get that down to less than a millisecond but never got down to applying that insight to the Tambaqui. Oh well.
---
I hope I'm not offending anyone if I don't take part in the discussion further, but I have amplifier and speaker work on my plate. As I said the DAC project was done in 2013... Before Purifi, and before Kii even...
There are only two cases where the internal format coincides with an existing audio format. On one extreme are R2R DACs that directly convert the PCM data you feed them by controlling 24 switchable current sources, each of which has half the current of the previous one. On the other extreme sit 1-bit converters that switch a single current or voltage source but at a high sampling rate. Both these extremes have limitations and the most common choice these days is a kind of compromise. They use a small number of bits (typically 5) that are used to control 32 (2^5) current sources with nominally equal currents, plus some trick to make the conversion minimally sensitive to any imbalances in those currents. The Tambaqui sits more or less in this camp: the PWM signal has a switching frequency of 3.125MHz, and can take 33 discrete lengths from 0*10ns to 32*10ns. So it's basically a 5-bit, 3.125MHz converter. The choice for PWM was given by the need to get rid of those idle tones I mentioned.
In any DAC design the internal format is essentially fixed, as it's a design choice. So any type of digital signal first gets converted to that format. That conversion consists of two stages. The first is upsampling and that takes the sample rate up to the sample rate of the internal format. Then the second is sigma-delta modulation (or noise shaping, however you want to call it) which takes the word length down to the word length of the internal format while preserving the SNR inside the audio band. Unfortunately, a succession of audio product and sales managers have conspired to conflate the second stage with the first, thereby confusing just about anyone with a passing interest in the technical background.
---
The latency is indeed around 100ms, for all inputs. There's no real reason for it other than at some point we finally got the ASRC algorithm working glitch-free by making the input data buffer rather long. Much later we figured out how to get that down to less than a millisecond but never got down to applying that insight to the Tambaqui. Oh well.
---
I hope I'm not offending anyone if I don't take part in the discussion further, but I have amplifier and speaker work on my plate. As I said the DAC project was done in 2013... Before Purifi, and before Kii even...
Similar threads
- Poll
