Review and Measurements of SMSL Sanskrit 10th DAC

[Krunok]

Hopefully this review insires SMSL to make a revision, especially since they've proven they can make better. I've been looking for a cheap well performing DAC that I can use with my video game consoles and this was almost it

Let's not jump to conclusion as this measurement surely looks atypical.

 

[SMSL-Mandy]

Dear amir,our engineer think maybe that is the power is not engough to the usb,the power consumption of USB input is biger than the S/PDIF. Can you change a PC power supply for testing?Thank you very much.If have any question can contact me.I'll do my best to help to find out the problem.

 

[Maki]

I think a multitone SINAD rating and ranked list would be useful, would it feasible?

 

[sssn]

SMSL is the erratic little brother of Topping.

 

[GGroch]

Dear amir,our engineer think maybe that is the power is not enough to the USB.......

If this solves it that would be great.
This is the 1st desktop SMSL DAC I have seen that uses USB power instead of a separate wall wart.

Why a separate USB power input? There is confusion on this point:

The user guide indicates the auxiliary USB power input is for when using with phones/low powered USB devices so as not to overdrive the phone's USB circuits. AND that "you may not use a Quick-Charge power supply because it may output over 5V, that is dangerous for this unit."

However, the specifications section of the manual indicate 1.5W power consumption which is a good bit above USB 2/3 spec.

A review of the Sanskrit 10 at audiophinics.fr describes the use of the USB power port differently. It states: "To isolate the audio signal from the noise coming along the power supply, two USB entries are present: one dedicated to data only and the other dedicated exclusively to the power supply. Thanks to this configuration, the audio source is not in charge of powering the DAC - which is generally poor in this function - and it is quite possible to connect this USB input to a real regulated linear power supply that meets the audiophile standards to maximize your listening experience."

I do not believe a USB power supply ships with the Sanskrit 10th. But if this indeed solves the issue that would be good to know.

 

[Krunok]

This is the 1st desktop SMSL DAC I have seen that uses USB power instead of a separate wall wart.

IIRC SMSL SU-8 also uses USB power.

 

[GGroch]

IIRC SMSL SU-8 also uses USB power.

IDTYRC I see an AC socket.

 

[Krunok]

IDTYRC I see an AC socket.

I don't.

 

[amirm]

Dear amir,our engineer think maybe that is the power is not engough to the usb,the power consumption of USB input is biger than the S/PDIF. Can you change a PC power supply for testing?Thank you very much.If have any question can contact me.I'll do my best to help to find out the problem.

Ah, that would explain why the USB output is lower level.

I hooked up the unit to my USB 3.0 port which has more power and the low frequency noise went away:

The low frequency noise is gone but we still have the 500 Hz spikes keeping SINAD at the same value.

Next I plugged in the switching supply for Schiit Modi 3 into the second USB port and got this:

Ah, problem solved! Note how the output now matches S/PDIF in level.

So it seems this should ship with an external USB adapter.

 

[Krunok]

Ah, that would explain why the USB output is lower level.

I hooked up the unit to my USB 3.0 port which has more power and the low frequency noise went away:
View attachment 18901

The low frequency noise is gone but we still have the 500 Hz spikes keeping SINAD at the same value.

Next I plugged in the switching supply for Schiit Modi 3 into the second USB port and got this:

View attachment 18902

Ah, problem solved! Note how the output now matches S/PDIF in level.

So it seems this should ship with an external USB adapter.

So, this littlle gismo at the end turned out to perform quite well.

 

[amirm]

I switched my USB cable with the one provided by SMSL into my USB 3.0 port and got this:

So lots of variability without external power supply.

 

[amirm]

So, this littlle gismo at the end turned out to perform quite well.

Only if you are the reader of ASR Forum and know to add the external power supply.

 

[Krunok]

Only if you are the reader of ASR Forum and know to add the external power supply.

Allow me to draw this conclusion from your statement: readers who haven't learned that much from you so far should reconsider visiting this forum.

 

[andreasmaaan]

Is this kind of variability typical for devices with USB power, or is something unusual going on inside this DAC?

 

[bennetng]

Ah, that would explain why the USB output is lower level.

I hooked up the unit to my USB 3.0 port which has more power and the low frequency noise went away:
View attachment 18901

The low frequency noise is gone but we still have the 500 Hz spikes keeping SINAD at the same value.

Next I plugged in the switching supply for Schiit Modi 3 into the second USB port and got this:

View attachment 18902

Ah, problem solved! Note how the output now matches S/PDIF in level.

So it seems this should ship with an external USB adapter.

Nice. Does it offer filter selection? Could you compare it with Modi3 in IMD vs level, jitter and linearity? Thanks!
[edit]
and dynamic range!

 

[amirm]

Is this kind of variability typical for devices with USB power, or is something unusual going on inside this DAC?

I have seen some DACs not power with my normal USB 2.0. I think the Chord Mojo was one. I have not seen one that runs and doesn't output right. And in all cases, using my 3.0 USB hub makes them all work. So this is a first.

 

[Krunok]

I switched my USB cable with the one provided by SMSL into my USB 3.0 port and got this:

View attachment 18903

So lots of variability without external power supply.

Huh.. This is again "atypical". It's hard to think of any reason why only 1 channel would be affected. Still, it's probably the best to have external 3.0 USB power supply for this unit. IMO it still presents great value for the money!

 

[amirm]

Nice. Does it offer filter selection? Could you compare it with Modi3 in IMD vs level, jitter and linearity? Thanks!
[edit]
and dynamic range!

At first I was going to say No due to incompatibility of Schiit Modi 3 USB implementation and Windows/ASIO4ALL. To my pleasant surprise, it seems that Microsoft has fixed this issue now and the Modi 3 runs fine with ASIO4ALL! So re-ran the tests for Modi which improved its IMD results substantially.

Here is how the two compare now, both using the Schiit Modi 3 external power supply (although Modi 3 doesn't seem to need it with USB):

They are neck and neck until the levels get high enough for the output stage of the Modi 3 saturates and distortion starts to rise. One channel in SMSL Sanskrit 10th has a bit more noise but both outperform Schiit Modi 3 still.

The tables turn when we look at linearity:

So Schiit Modi 3 is better at low level detail (objectively) but worse at the peaks.

Oh, you sneaked in the dynamic range. Will run that next.....

 

[amirm]

It's hard to think of any reason why only 1 channel would be affected.

One channel may be closer to the power supply and hence steals more of the current.

 

[Krunok]

One channel may be closer to the power supply and hence steals more of the current.

Yes, may be. But still, power supply sees electronics of both channels as impedance in parallel and I can hardly imagine that 1 inch (or so) of printed board lead could have such impact on it.