SMSL SU-1 Stereo DAC Review

[bogi]

I get varying levels of noise, a lot of it obviously centred around 'mains' frequency (50Hz) but also containing other random fizzing and crackling from the GPU (a problem for games) and USB mouse/keyboard. It's been intractable and obviously relates to the ground in the USB connection modulating the analogue signal, hence the desire to try a USB isolator.

I have very good experience with Topping HS02 isolator.

and 2-into-1 cable may be needed

With HS02 you don't need any unusual cables since it has 2 input ports.

The differences between HS01 and HS02 are described on HS02 product page https://www.tpdz.net/productinfo/953487.html

 

[anthop]

I just got this dac and it works perfect. 1 question tho. when im not using it, can I just leave it plugged into my outlet? Or is it better to unplug the power?

 

[Mark S.]

No. If third party items aren't shipped by Amazon you have to deal with the vendor. I try to buy only items shipped by Amazon. Then returns within the window are simple.

Well, now I know. Thankfully the 3rd-party Amazon seller I ordered my SU-1 from a few days ago has a clear 30-day return policy in the listing so presumably they'd be obliged by Amazon to honour it if the need arose (they claim to be UK-based but the delivery time of about two weeks implies otherwise ...)

 

[Mark S.]

I have very good experience with Topping HS02 isolator.


With HS02 you don't need any unusual cables since it has 2 input ports.

The differences between HS01 and HS02 are described on HS02 product page https://www.tpdz.net/productinfo/953487.html

The HS-01 actually has a USB-C port for power as well. The simple USB-A in/out item that I've ordered from Aliexpress would need the 1-into-2 cable if it's 300mA out doesn't power the SU-1. I've found a few generic ones, probably get one pre-emptively.

 

[raif71]

I have very good experience with Topping HS02 isolator.


With HS02 you don't need any unusual cables since it has 2 input ports.

The differences between HS01 and HS02 are described on HS02 product page https://www.tpdz.net/productinfo/953487.html

Do you think this will work? My intent is to use Shanling m0 with e1da 9038d but as it is right now, it's a direct connection between the m0 and 9038d and it's draining the m0 battery considerably rendering it useless after few hours. Will the e1da in the configuration above with the HS02, use the dc 5v for power and will not use the m0 battery? ie just audio data from the m0 and powered by the DC 5V?

 

[Mark S.]

View attachment 290844

Do you think this will work? My intent is to use Shanling m0 with e1da 9038d but as it is right now, it's a direct connection between the m0 and 9038d and it's draining the m0 battery considerably rendering it useless after few hours. Will the e1da in the configuration above with the HS02, use the dc 5v for power and will not use the m0 battery? ie just audio data from the m0 and powered by the DC 5V?

I'd be interested in knowing that as well.

BTW, re. the HS02, I'd be a bit concerned about the mechanical switches for selecting the in and out sockets. The HS01 is simpler (and cheaper, and smaller) and I'd go for that one (although for some unfathomable reason the USB out is one of those ancient, over-size type B's), I imagine that otherwise it's functionally the same as the HS02.

 

[little-endian]

In my mind, DSP is somewhat out of place in a DAC (which BTW is one of the reasons I haven’t considered an RME ADI DAC, as well engineered as they may be).

Given your assumption of computer playback, an entirely comprehensible argument of yours. However, it should not be forgotten that playback from a non-PC-system also has its charme as stuff like a DAT-, CD-player or whatever stand-alone source via S/PDIF besides the haptics, has one very nice advantage: it just works.

No driver quirks, no crackling, no resampling stuff, no non-sense system sounds ringing your ears all of the sudden, no other potential ear shattering elements which might play along at higher levels, etc.

Hence, together with classical sources, it is nice to have all that tone control stuff in one well-designed package.

Required in conjunction with a PC or not - it is rather absurd that hardly any DAC in fact offers such possibilities since this would be the virtually only thing left to distinguish itself from others anyway, given entire audible transparency these days for a few bucks.

 

[bogi]

Do you think this will work?

When you connect external power supply with enough amperage to HS02 USB C input port, why should it not work? It is designed for that case.

I believe it will not use computer power at all if external power is proivided on USB C. For me it wouldn't make sense. But you can ask iFi directly.

An alternate way is to use something like this to provide +5V power on independent way: https://www.audiophonics.fr/en/dc-p...e-for-external-usb-a-power-supply-p-8389.html In such a case computer power is disconnected for sure (that's the purpose of the adapter cable).

 

[bogi]

The HS-01 actually has a USB-C port for power as well. The simple USB-A in/out item that I've ordered from Aliexpress would need the 1-into-2 cable if it's 300mA out doesn't power the SU-1. I've found a few generic ones, probably get one pre-emptively.

I still don't inderstand why do you need 1-into-2 cable or 2-into-1-cable with HS02. External power supply connected HS02 USB C input port can be for example 5V 1A or 5V 2A. HS02 manual says it can provide up to 800mA for USB powered device.

 

[bogi]

I'd be interested in knowing that as well.

BTW, re. the HS02, I'd be a bit concerned about the mechanical switches for selecting the in and out sockets. The HS01 is simpler (and cheaper, and smaller) and I'd go for that one (although for some unfathomable reason the USB out is one of those ancient, over-size type B's), I imagine that otherwise it's functionally the same as the HS02.

You use the switches (if ever) for initial setup only. Then you don't need to touch them anymore.

The HS02 specs are higher than HS01. I already wrote "The differences between HS01 and HS02 are described on HS02 product page https://www.tpdz.net/productinfo/953487.html". There is a comparison table. Audio signal isolation voltage is saying about level of galvanic isolation. The table values are mistaken. They should he 1.5kV / 1.0 kV or 1500V / 1000V.

 

[Mark S.]

You use the switches (if ever) for initial setup only. Then you don't need to touch them anymore.

The HS02 specs are higher than HS01. I already wrote "The differences between HS01 and HS02 are described on HS02 product page https://www.tpdz.net/productinfo/953487.html". There is a comparison table. Audio signal isolation voltage is saying about level of galvanic isolation. The table values are mistaken. They should he 1.5kV / 1.0 kV or 1500V / 1000V.

Yeh, apologies, I didn't take time to check the page out.

However, the parameters it betters the HS01 seem to be completely academic for our purposes - high-voltage isolation (something that might be more relevant for data-critical applications?) and latency of 0.29μS (0.00029 secs) and 0.079μS (0.000079 secs) which, to put it in perspective, are about 1/32nd of a movie frame and 1/120th respectively (both imperceptible). ETA >> maybe in a recording/production/mixing application it would matter?

Then there's lack of isolation for the secondary power input on the HS01, but again academic if one is using a floating ground PSU or charger - I've never experienced any noise whatsoever with amps or DACs from such a supply.

Both are claimed to be 32/768 and DSD512 capable.

I suppose the specs might imply better quality componentry generally in the HS02, but who knows?

 

[Mark S.]

I still don't inderstand why do you need 1-into-2 cable or 2-into-1-cable with HS02. External power supply connected HS02 USB C input port can be for example 5V 1A or 5V 2A. HS02 manual says it can provide up to 800mA for USB powered device.

In fact you don't - that was me getting confused when I was comparing it and the other options I was looking at (such as the simple USB-A in/out box I ordered without being aware of it's current limitations).

Anyhow, your concern is whether the isolator will use only the secondary supply to power devices. If I had to guess I'd say probably it will, I would have thought they'd be designed to simply switch internally from the source (your phone) when 5v (non-data) is detected on the secondary input rather than allowing them to run in parallel (and presumably having to use regulating circuitry of come kind). At least, that's how I'd do it. ETA >> that said, the HS02 isolating both its inputs does imply that it might need to use both to provide the current needed for many devices it's intended for. Hopefully we'll know eventually, but this might make the HS01 a better bet because presumably you get the full USB 3.0 1A from the non-isolated USB C power input (again, using a typical floating ground supply noise shouldn't be a problem, but if you wanted to go really purist you could use a powerbank).

 

[Mark S.]

Given your assumption of computer playback, an entirely comprehensible argument of yours. However, it should not be forgotten that playback from a non-PC-system also has its charme as stuff like a DAT-, CD-player or whatever stand-alone source via S/PDIF besides the haptics, has one very nice advantage: it just works.

No driver quirks, no crackling, no resampling stuff, no non-sense system sounds ringing your ears all of the sudden, no other potential ear shattering elements which might play along at higher levels, etc.

Hence, together with classical sources, it is nice to have all that tone control stuff in one well-designed package.

Required in conjunction with a PC or not - it is rather absurd that hardly any DAC in fact offers such possibilities since this would be the virtually only thing left to distinguish itself from others anyway, given entire audible transparency these days for a few bucks.

Quite so - although I watch and listen almost entirely from PC I do occasionally have to use the optical-out from my TV/monitor (its only audio-out) and it would be nice to have EQ available (for moar bass and taming the mountainous treble peak in my Beyer DT770's).

 

[bogi]

there's lack of isolation for the secondary power input on the HS01

HS02 filters and isolates the external power source.

However, the parameters it betters the HS01 seem to be completely academic for our purposes

Ground loop and computer noise impacts on audio chain are always specific for concrete setup, therefore no generalization can be made. Effect of isolation may be audible in one setup and inaudible in other. And since nothing is black and white, I believe in some setups difference between HS01 and HS02 could be audible.

 

[Mark S.]

HS02 filters and isolates the external power source.

Ground loop and computer noise impacts on audio chain are always specific for concrete setup, therefore no generalization can be made. Effect of isolation may be audible in one setup and inaudible in other. And since nothing is black and white, I believe in some setups difference between HS01 and HS02 could be audible.

To reiterate; a floating ground USB PSU or charger will never create a ground loop (they have a virtual ground not an actual one) and rarely any other noise in my experience, so the unfiltered USB C power input on the HS01 can be used to advantage, assuming it does indeed make the full USB 3.0 spec 1A available to the powered device. If the HS02 is filtering/decoupling the power on both its USB inputs it might struggle to deliver even 1A (most of these decouplers seem to be limited to ~300mA per socket, so possibly 600mA - ETA >> and you might still have the problem of battery drain from the phone, I'm suggesting that the HS01 might well power your DAC/amp entirely from the USB C input).

 

[bogi]

To reiterate;

a) 800 mA as I stated in my above post; since USB 3.1 is 900 mA it seems it consumes 100mA. I assume it uses/filters power only from one its input at a time.
b) unrestricted with Y type of cable as I stated in my above post

 

[Shadow12347]

At work we use Motu M2's to drive one of our products, and I needed a cheaper device with good sound quality to give to my test VMs (via PCIe Passthrough) and feed each VM's output into a Motu 16A. We bought four of these, and I just set up the first one. It seems to sound/work quite well. My only gripe so far is that whenever the device has been idle for a few seconds, there's an audible pop before any new sound is heard, which slightly cuts off the beginning of the audio, and then another audible pop a few seconds after audio has stopped. I haven't had this issue with a spare Motu M2, nor with an assortment of USB to 3.5mm adapters on the same setup, so I think I can safely attribute this behavior to the device, rather than my VM setup. I don't have time to fully build this out and test each one yet, but I did try a few different sample rates and bit depths on this first device, and it happened in each combination so far. I also have USB selective suspend disabled in the OS, and I verified that all USB root ports are set the same way within the VM as well. I also found that this happens with automatically detected Windows drivers, as well as with the SMSL provided ones. I'm sure I've heard of this happening before on other devices, and I would imagine this is something that can be fixed with firmware updates? Once I have time to test the others and set them up in the 16A (which admittedly may not be for a week or so), I will report back, and hopefully with recordings of this behavior as well.

 

[Mark S.]

Anyhow, back OT;

having lived with a Sound Blaster PLAY!3 (!) for about a year and a half (which doubtlessly would measure "very good" to "excellent" were Amirm to test it and be theoretically "all the DAC you need" to some people) having flogged my Xonar Essence STX, I decided to get back to something 'respectable', hence the SU-1.

But this will tickle some of the forumites - I bought a Cayin RU6 R-2R a couple weeks back and that's been my audio source since then. The rationale for getting both is that for a lot of the time I want to have digital attenuation (i.e. volume control) and realistically, even thought the RU6 is a 24-bit DAC it just doesn't have the precision to slide the signal down into the 'redundant' bits below 16, hence a 'proper' 24-bit in the form of the SU-1 which should be able to provide 40dB or more of attenuation with practically no measurable or perceptible loss at all (30dB is actually more than enough).

How's the RU-6 you may ask? Actually very good indeed when fed by Foobar using WASAPI (which allows using some EQ as well), this in NOS mode. It varies with the music I'm playing, but some recordings (that I've known for decades) really pop and grab my attention - as I type, Jeff Beck's Wired (16/44.1) has just finished and it was hugely enjoyable.

Comparing the RU6 and SU-1 is going to be quite a bit of fun.

 

[little-endian]

The rationale is that for a lot of the time I want to have digital attenuation (i.e. volume control) and realistically, even thought the RU-6 is a 24-bit DAC, it just doesn't have the precision to slide the signal down into the 'redundant' bits below 16, hence a 'proper' 24-bit in the form of the SU-1 which should be able to provide 40dB or more of attenuation with practically no measurable or perceptible loss at all (30dB of attenuation is actually more than enough).

The question remains for me what perceptible loss is supposed to ever occur even without such kind of headrooms, considering that our ears' dynamic range goes down by definition as well, based on their steady noise- (or rather "tinnitus"-) floor anyway.

Also, even with 16 bit "only", the noise floor at about -93dBFS should be the only side-effect to be expected, given proper dithering and even 8 bit wouldn't be less "precise" in terms of analog waveform reconstruction, but only noisier.

 

[Mark S.]

The question remains for me what perceptible loss is supposed to ever occur even without such kind of headrooms, considering that our ears' dynamic range goes down by definition as well, based on their steady noise- (or rather "tinnitus"-) floor anyway.

Also, even with 16 bit "only", the noise floor at about -93dBFS should be the only side-effect to be expected, given proper dithering and even 8 bit wouldn't be less "precise" in terms of analog waveform reconstruction, but only noisier.

A lot of early CD releases are VERY 'quiet', 10dB or more (in terms of RMS SPL's) lower than typical, compressed 'modern' ones, even if they peak close to 0dB (many don't - go back far enough and you find discs that only hit -3dB or even lower), and you really wouldn't want to attenuate those in a 16-bit DSP, a lot of the music is already waaay down in the bottom 8 bits.