• 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!

SMSL SU-8 Version 2 Balanced DAC

badyard

Member
Joined
Oct 18, 2018
Messages
84
Likes
75
What about the lack of 100Mhz clock oscillator in the v2, does the lack of it improve or affect some aspect of the DAC sound? if no then why was it implemented in their original design to start with?
Does it means V2 is synchronous while the old version had the choice to run asynchronous? (is the output selector in the asio drivers from thesycon under the "Format" tab, the option to switch synchronous/asynchronous ?
 
Last edited:

Hase

Member
Joined
Jan 5, 2019
Messages
9
Likes
11
The 100Mhz clock was attached to the MAX2 CPLD (kinda a FPGA in function). The MAX2 sits between the DAC chips and the USB interface (XMOS). I don't know the implementation details, but I think that the MAX2 CPLD takes the I2S data (stereo) from the XMOS and copies it at the same time to the DAC chips (dual configuration were each DAC chip is running in MONO mode). Most likely it will re-clock the I2S signal with either the 45 or 49 Mhz clock and will send it to the DAC chips. The data is probably passed with the help of a ASYNC FIFO between the XMOS clock domain to the DAC chip clock domain. The XMOS clock domain will contain the part, that recieves the binary I2S data from the XMOS chip over some kind of data bus (maybe a simple 8 bit data bus or just a 1bit I2C bus? Haven't analyzed the PCB throughly). This XMOS clock domain will need a clock, that is suitable for the XMOS to CPLD databus. They probably thought that they need a 100Mhz for this task (since those XMOS chips run most of the time at 500Mhz), but maybe they can run this parts with either the 45Mhz or the 49Mhz clock sufficiently, or maybe the XMOS chip can provide a suitable master clock signal for the MAX2.

Anyhow the 100Mhz clock wouldn't be needed in this case and can be optimized out without a negative side effect.
 

badyard

Member
Joined
Oct 18, 2018
Messages
84
Likes
75
The 100Mhz clock was attached to the MAX2 CPLD (kinda a FPGA in function). The MAX2 sits between the DAC chips and the USB interface (XMOS). I don't know the implementation details, but I think that the MAX2 CPLD takes the I2S data (stereo) from the XMOS and copies it at the same time to the DAC chips (dual configuration were each DAC chip is running in MONO mode). Most likely it will re-clock the I2S signal with either the 45 or 49 Mhz clock and will send it to the DAC chips. The data is probably passed with the help of a ASYNC FIFO between the XMOS clock domain to the DAC chip clock domain. The XMOS clock domain will contain the part, that recieves the binary I2S data from the XMOS chip over some kind of data bus (maybe a simple 8 bit data bus or just a 1bit I2C bus? Haven't analyzed the PCB throughly). This XMOS clock domain will need a clock, that is suitable for the XMOS to CPLD databus. They probably thought that they need a 100Mhz for this task (since those XMOS chips run most of the time at 500Mhz), but maybe they can run this parts with either the 45Mhz or the 49Mhz clock sufficiently, or maybe the XMOS chip can provide a suitable master clock signal for the MAX2.

Anyhow the 100Mhz clock wouldn't be needed in this case and can be optimized out without a negative side effect.
Wow, that's very interesting but also very well above my knowledge, make sense though. Basically your opinion is that it was probably overly engineered in the clock department, right?

I think it's at least a bit weird considering the cost of the unit that they took the luxury to slap a 100MHz clock oscillator on the production unit just for the sake of. However it is still very possible.
 

Hase

Member
Joined
Jan 5, 2019
Messages
9
Likes
11
Wow, that's very interesting but also very well above my knowledge, make sense though. Basically your opinion is that it was probably overly engineered in the clock department, right?

I think it's at least a bit weird considering the cost of the unit that they took the luxury to slap a 100MHz clock oscillator on the production unit just for the sake of. However it is still very possible.

I don't know the exact reasoning why they included it in the first iteration. But when you reiterate a product because you want to fix other problems, you can also optimize the BOM and remove unneeded parts if they are not really needed for the quality of your product (in this case sound quality) and you can also get functional parity. It's a process well stablished in the industrie. It's often a good idea to start "safe" in your product design and optimize later. Time to market and optimizing R&D costs always plays a big factor.

The cost of the 100Mhz oscillator is most likely around 50 cent.
 

Vexed hex

Member
Joined
Jan 25, 2019
Messages
20
Likes
17
Hi there. Welcome to the forum. Briefly, no, CD music is limited to 96 dB anyway so unless you are playing exceptionally well recorded music, and at low level with high amplification, it is not an audible concern.

Likewise channel linearity differential is not an issue.

Welcome aboard by the way.

Okay thank you I ordered one off massdrop as it seems to be the best dac one can get right now unless you spend like 1k

I plan on pairing it with a THX AAA 789 amp once they drop again on massdrop it seems to be all the rage in the amp world right now very clean amp. I got my 58x jubilee headphones today so once i get the amp ill be set for long time.
 
Last edited:

obc

Member
Joined
Dec 22, 2018
Messages
14
Likes
5
can somebody experienced please give some advice/impressions on the pcm filters? like what filter to use for which musicgenre or when gaming fast shooters or when many different things happen at the same time (like in battlefield 5).
 

badyard

Member
Joined
Oct 18, 2018
Messages
84
Likes
75
can somebody experienced please give some advice/impressions on the pcm filters? like what filter to use for which musicgenre or when gaming fast shooters or when many different things happen at the same time (like in battlefield 5).
I'm having hard times trying to hear a difference with the filters, but not so much for the different sound profiles, those I think are awesome, they are essentially harmonic saturation "curves", very subtle but effective.
"Rich 1/2/3" gives a nice pop to the mids (perfect for movies and dialogs), "Tube 1/2/3" my favorite, gives some solidity to the low mids, deeper but slightly loosened sub bass and soften the highs a notch (particularly great for people like me who are very susceptible to sibilant sounds).
"Crystal 1/2/3" it's sort of the opposite of "Tube 1/2/3" tighten the bass a notch and opens the highs, (you can get nice sparkling transients with this).
These are my impressions so far..
 

Knowledgebass

New Member
Joined
Jan 26, 2019
Messages
2
Likes
0
I bought a Cambridge Audio 651w. This model is discontinued but there's a few demo/floor units around. Here's some measurements from a poster I trust on another forum (actually another member on this forum posted this link in the amps forum)
http://www.htguide.com/forum/showth...xpensive-Cambridge-Audio-651W-Power-Amplifier

I'm planning to use the SMSL SU-8 as a digital preamp using the balanced outs into the 651w. Speakers are a DIY design, designed to measure flat (Ansonica is the design name, not designed by me). Sources are an HTPC (Logitech Media Server > squeezelite for audio, mostly CD rips to FLAC and Spotify; Roku Ultra > toslink, Netflix, Google Movies, YouTube).

I've read this thread and sounds like this should work. I don't wholly understand the limitation of digital volume control, but what I've gleaned, the louder the SU-8 volume is set, the less impact the volume control will have on dynamics. And really, my reference listening, the rare time I'm home alone, is set pretty loud at 90dB+ (measured in room with a simple calibrated mic and an Android program called AudioTool). Anything I've overlooked? Feel like I more or less understand what I'm getting into. Just need to sort out a 12V trigger.

Currently I'm running an SMSL Q5 Pro which was meant to be a stop gap when my previous AVR died about a year ago (Cambridge Azur 540r). The Q5 is severely underpowered and drives the speakers to distortion. Distortion quantified as painful at in room SPLs that were fine previously. Good enough for TV and casual listening. But not for serious listening.
 

GGekko

Member
Joined
Dec 23, 2018
Messages
18
Likes
10
I am trying to use my SMSL SU-8 V2 DAC to connect my Samsung QE55Q6F 2017 TV to my JBL LSR 308 Active monitor speakers and I would need someone's help as the sound keeps cutting off every 3 to 4 seconds.

Here's my setup:
- I set my TV to "Audio Out Optical" (see attachment).
- I set all my sound settings to "PCM" (see attachment)
- I connected my TV to the SMSL SU-8 with an opticable cable (see attachment)
- I set my SMSL SU-8 to optical input (see attachment)
- I connected my SMSL SU-8 to my JBL speakers with XLR cables (see attachment)

Any help would be greatly appreciated!
 

Attachments

  • IMG_1452.jpeg
    IMG_1452.jpeg
    488.7 KB · Views: 258
  • IMG_1453.jpeg
    IMG_1453.jpeg
    524.6 KB · Views: 294
  • IMG_1448.jpeg
    IMG_1448.jpeg
    1 MB · Views: 340
  • IMG_1450.jpeg
    IMG_1450.jpeg
    739.1 KB · Views: 327
  • IMG_1449.jpeg
    IMG_1449.jpeg
    712.3 KB · Views: 304

badyard

Member
Joined
Oct 18, 2018
Messages
84
Likes
75
I am trying to use my SMSL SU-8 V2 DAC to connect my Samsung QE55Q6F 2017 TV to my JBL LSR 308 Active monitor speakers and I would need someone's help as the sound keeps cutting off every 3 to 4 seconds.

Here's my setup:
- I set my TV to "Audio Out Optical" (see attachment).
- I set all my sound settings to "PCM" (see attachment)
- I connected my TV to the SMSL SU-8 with an opticable cable (see attachment)
- I set my SMSL SU-8 to optical input (see attachment)
- I connected my SMSL SU-8 to my JBL speakers with XLR cables (see attachment)

Any help would be greatly appreciated!
Try this in order:
1) Check your optical cable, make sure it's in good working conditions.
2) Try connecting your SU-8 to another source with optical output (another tv or receiver), if it works proceed with the next solutions.
3) Check the tv optical output port with other gear to see if it's the source of the problem.
4) If your speakers have an "auto-off" feature, try disabling that.
5) Reset and update your TV firmware.
6) At this point I'll return the unit as defective and get a replacement SU-8 Dac.
 

TuneInSoul

Member
Joined
Sep 20, 2018
Messages
76
Likes
5
I am trying to use my SMSL SU-8 V2 DAC to connect my Samsung QE55Q6F 2017 TV to my JBL LSR 308 Active monitor speakers and I would need someone's help as the sound keeps cutting off every 3 to 4 seconds.

Here's my setup:
- I set my TV to "Audio Out Optical" (see attachment).
- I set all my sound settings to "PCM" (see attachment)
- I connected my TV to the SMSL SU-8 with an opticable cable (see attachment)
- I set my SMSL SU-8 to optical input (see attachment)
- I connected my SMSL SU-8 to my JBL speakers with XLR cables (see attachment)

Any help would be greatly appreciated!
check the cables and the ports.
 

MusicNBeer

Senior Member
Forum Donor
Joined
Jan 20, 2019
Messages
333
Likes
478
Are the units being sold at Amazon and Ebay version 2? I see no indication of that in the descriptions. Anywhere else to buy from?
 

GGekko

Member
Joined
Dec 23, 2018
Messages
18
Likes
10
check the cables and the ports.
After testing different configurations, it turns out the sound is fine when playing movies through USB or when streaming through Wifi. It's only when I watch TV through satellite that the sound is buggy. Weird O_O
 

TuneInSoul

Member
Joined
Sep 20, 2018
Messages
76
Likes
5
After testing different configurations, it turns out the sound is fine when playing movies through USB or when streaming through Wifi. It's only when I watch TV through satellite that the sound is buggy. Weird O_O
bug of software may cause this problem I think
 

Pluto

Addicted to Fun and Learning
Forum Donor
Joined
Sep 2, 2018
Messages
990
Likes
1,631
Location
Harrow, UK
...the sound keeps cutting off every 3 to 4 seconds
If the sound is clean between the gaps, it is probable that either the sampling rate being delivered by the TV is not close enough to 48kHz for the DAC to achieve a good lock or the digital output of the TV is itself a bit twitchy. It is far more likely to be a TV issue than the DAC, assuming the DAC is working as it ought when fed from other sources. Does the DAC remain locked and stable during the gaps or does it go off-lock, recover and the sound returns?

I assume that the black Samsung unit in your picture is some sort of breakout box from the TV? Does the thick white wire connect this box to the TV? Does the TV not have an optical socket to which you can connect your DAC directly (without using the breakout box)? Check that the connectors to the breakout box are properly mated at both ends.

Many TVs only reproduce as good (or as bad) a digital audio signal as they are given – GIGO – Garbage In, Garbage Out. In other words, if the audio coming from your satellite RX is twitchy, the TV will not do anything to improve things. If the audio is stable when playing sources other than satellite, I think that tells you that the audio coming out of the satellite RX is poor and/or the TV is not re-processing digital audio effectively. This is a design issue with many TVs and may not be easily solved. In general, the best source to test the TV's audio stability is terrestrial – which is the TV all by itself with no reliance on any external equipment. If this audio is stable through the DAC, it tells you that you need to look at the stability of the source audio.


Additional edit

Can you connect the digital audio output of the satellite box directly into your DAC (bypassing the TV)? Such a test will demonstrate the reason for the problem.
 
Last edited:

GGekko

Member
Joined
Dec 23, 2018
Messages
18
Likes
10
If the sound is clean between the gaps, it is probable that either the sampling rate being delivered by the TV is not close enough to 48kHz for the DAC to achieve a good lock or the digital output of the TV is itself a bit twitchy. It is far more likely to be a TV issue than the DAC, assuming the DAC is working as it ought when fed from other sources. Does the DAC remain locked and stable during the gaps or does it go off-lock, recover and the sound returns?

I assume that the black Samsung unit in your picture is some sort of breakout box from the TV? Does the thick white wire connect this box to the TV? Does the TV not have an optical socket to which you can connect your DAC directly (without using the breakout box)? Check that the connectors to the breakout box are properly mated at both ends.

Many TVs only reproduce as good (or as bad) a digital audio signal as they are given – GIGO – Garbage In, Garbage Out. In other words, if the audio coming from your satellite RX is twitchy, the TV will not do anything to improve things. If the audio is stable when playing sources other than satellite, I think that tells you that the audio coming out of the satellite RX is poor and/or the TV is not re-processing digital audio effectively. This is a design issue with many TVs and may not be easily solved. In general, the best source to test the TV's audio stability is terrestrial – which is the TV all by itself with no reliance on any external equipment. If this audio is stable through the DAC, it tells you that you need to look at the stability of the source audio.


Additional edit

Can you connect the digital audio output of the satellite box directly into your DAC (bypassing the TV)? Such a test will demonstrate the reason for the problem.
Yes the sound is clean between breaks and yes the black Samsung unit is a breakout Box connected with that thick wire. The TV has no ports of its own besides ethernet. The only audio output is the optical out on that little box. My DAC works fine with everything else by the way. Thanks for your help.
 
Top Bottom