SMSL SU-8 Version 2 Balanced DAC

[Veri]

Let us know if there is goose pumps

Lol

 

[svenz]

I own a 789 and HE6se and I can say that even at maximum output on the 789 with my DAC at +19dBu (6.9v) there is no clipping. The 789 can deliver 3800mAh into 50ohm which is far less than the 6W max and just below the 4W recommended by HifiMan.

I'd personally leave the SU-8 at maximum output and and use gain 2 on the 789 (unity gain). It's more than you'll need and both using high gain and lowering the output of the SU-8 lower performance.

E: Fun fact, the 789 can make my HE6se jump on the table from the excessive vibrations it produces at high volume. Planar drivers can take silly amounts of power, I've always wanted to see like a LCD-MX4 at 15W (literally like 140dB theoretically).

Why would lowering the output of the SU-8 affect performance?

 

[headwhacker]

Why would lowering the output of the SU-8 affect performance?

I assume its because of digital volume control in su-8, lowering output means lowering effective bit depth.

However, ESS chips use the full 32-bit depth for volume control and should not matter much to affect actual resolution on the DACs output other the voltage level.

 

[badyard]

I assume its because of digital volume control in su-8, lowering output means lowering effective bit depth.

However, ESS chips use the full 32-bit depth for volume control and should not matter much to affect actual resolution on the DACs output other the voltage level.

Volume control in SU-8 is lossless.

 

[BurritoJustice]

I only meant in that lowering the output voltage usually results in lowered signal to noise, thought obviously not to any audible degree, and it also lets you avoid high gain on the 789 which also leads to minor degradation.

Splitting hairs, but I can't see why you'd go for less voltage and then high gain.

 

[pos]

ESS volume control is nothing magical:

Volume control in SU-8 is lossless.

There is no such thing as lossless digital volume control, and ESS' implementation is no exception.
That does not mean it is not perfectly adequate, of course.
The DNR is specified for full volume, and any dB of attenuation (be it done inside the DAC or at the source) reduces the DNR by that much as the noise floor remains the same.

 

[Krunok]

The DNR is specified for full volume, and any dB of attenuation (be it done inside the DAC or at the source) reduces the DNR by that much as the noise floor remains the same.

And that remains so no matter if volume control is done in analog or digital domain.

 

[WondrousHippo]

Does the USB input stay active even when selecting another input, or does it deactivate when you've selected another input? This might make a big difference in how I use the device, as I want to use coaxial from my TV, optical for a Chromecast Audio, and USB for my Raspberry Pi, but if USB deactivates when not selected (like my Sabaj D4) then that setup might not work for me!

 

[gtb75]

Just a heads up, I got my v2 from Massdrop last week and mine exhibits the same power supply audible noise mentioned in the original SU-8 thread here: https://www.audiosciencereview.com/...nts-of-smsl-su-8-dac.3778/page-10#post-116372

The noise also goes away on my v2 after a "warm reboot". I noticed it changed in pitch when the display was turned on/off, so I would definitely guess it's power supply related. It does not seem to have any impact on sound quality though.

 

[ReelMcCoy]

Does the USB input stay active even when selecting another input, or does it deactivate when you've selected another input?

I picked up a SU-8 V2 from Shenzhen Audio last week and from what I can tell, the USB input is only active if it's the selected source. Otherwise, with Windows 10 at least, the USB audio device (and another bit I don't have a driver for yet but I'm using optical anyway) disappear.

 

[WondrousHippo]

I picked up a SU-8 V2 from Shenzhen Audio last week and from what I can tell, the USB input is only active if it's the selected source. Otherwise, with Windows 10 at least, the USB audio device (and another bit I don't have a driver for yet but I'm using optical anyway) disappear.

Thanks a bunch, that's the same way the Sabaj D4 operates. Was hoping it'd stay active but oh well.

 

[Jimmy]

If you are able to find the part making the noise (probably an inductor), potting it using hot glue or epoxy (better) usually can fix the problem, although it's just an annoyance which is common with many switching power supplies.

Just a heads up, I got my v2 from Massdrop last week and mine exhibits the same power supply audible noise mentioned in the original SU-8 thread here: https://www.audiosciencereview.com/...nts-of-smsl-su-8-dac.3778/page-10#post-116372

The noise also goes away on my v2 after a "warm reboot". I noticed it changed in pitch when the display was turned on/off, so I would definitely guess it's power supply related. It does not seem to have any impact on sound quality though.

 

[gtb75]

If you are able to find the part making the noise (probably an inductor), potting it using hot glue or epoxy (better) usually can fix the problem, although it's just an annoyance which is common with many switching power supplies.

I haven't gone digging in the unit yet to try and isolate since it doesn't really bother me much. I know SMPS often experience this so I pretty much wrote it off, but I saw that someone else had the same behavior with the original so I figured I'd mention the issue still exists with the v2 model. It is consistent though - always there on initial power up after being unplugged and goes away after a power cycle. I don't think think it is a serious issue that warrants replacement, unless @SMSL-Mandy feels otherwise.

 

[svenz]

I have su-8 + balanced interconnect + thx aaa 789 + balanced hd800 cable - I'm finding the combination is too loud even on low gain. So I'm lowering digital volume on smsl su-8 by 6 to make it more manageable.

 

[Roen]

What about switching to SE HD800 cable to keep everything in the analog domain?

 

[svenz]

What about switching to SE HD800 cable to keep everything in the analog domain?

I could do that, but it's nice knowing crosstalk is kept to a minimum with the 4 pin connectors - thx aaa 789 has some issues with crosstalk on the SE output at low volumes, although I am not sure they are audible.

 

[rajapruk]

Crosstalk might actually be beneficial for headphone use...

 

[svenz]

What am I going to do with my balanced cable than? Seems like a waste. Maybe turn it into a Christmas tree decoration.

 

[Dialectic]

Crosstalk might actually be beneficial for headphone use...

This is a mistaken view that gained currency because of the old Headroom amps. If you hear the imaging through headphones of BACCH-HP processing, which does not add crosstalk, you'll realize that crosstalk is not good for headphone listening.

 

[Veri]

This is a mistaken view that gained currency because of the old Headroom amps. If you hear the imaging through headphones of BACCH-HP processing, which does not add crosstalk, you'll realize that crosstalk is not good for headphone listening.

Indeed. A carefully constructed crossfeed is not the same as unintended crosstalk seeping through the signal..