strange SMSL M500 high 3rd harmonic

[MusicNBeer]

Looking at right channel, here is [Left plot] 1KHz in both channels and [Right plot] 1KHz in right channel, 2KHz in left channel. Volume set to 40. It's indeed been a broken design from the start.

Ignore spurs more than -106 dBFs or so, those can be caused by my soundcard.

 

[MusicNBeer]

All is not lost, here is performance with volume set to 30. Same plots as previous post description.

At least this performance makes the issue of downgraded op-amps moot...

 

[liu]

LOL!

So this has been broken for years.
Now I guess various other SMSL products such as the SU-9 (or even M400) might be broken too.
It's super easy to find this bug out in a channel crosstalk test. but the test rubrics here does not include that for DACs.
So it's amazing to see that such huge flaw never got noticed by any of the "objective measurements" by @amirm (and later professor wolf)
which makes a few bad performing dacs easily sit on top of the SINAD chart.

 

[MusicNBeer]

If you listen exclusively to mono recordings, this could be the DAC for you. I'll even sell mine to you!.

Would this show up in crosstalk test? I don't see leakage into other channel, just distorted main channel.

 

[liu]

Distortion compensation/cancellation in the DAC may have been incorrectly implemented by SMSL? Perhaps they are using the D/A in a wierd differential/dual parallel situation where the harmonics reinforce one another.

"Residual distortion from suboptimal PCB components and layout can be minimized using ES9038PRO’s unique THD compensation circuit".

My bet is they are using the THD compensation co-efficients to get the stellar numbers, but there is some mutual coupling (sub-optimal layout) between channels that manifests in the absence of the right channel signal (ie it doesn't cancel, it reinforces) and the fact that the co-efficients for cancellation are not channel specific (they can't trim one differently to another)

So, do we regard these distortion compensation "features" as the equivalent of VW DieselGate? When the unit is playing a 'test signal"(identical in both channels) as opposed to real world music...

I can confidently tell you this is not the case.
There's an internal secret menu this DAC has to adjust thd compensation values and I have tried that but get exactly the same elevated 3rd harmonic.
The DAC is fucked up by its design, not because of this ESS feature at all.

 

[jruser]

Can anyone check this on other DACs using Sabre? Wondering if it's an SMSL issue or a Sabre issue.

 

[liu]

Would this show up in crosstalk test? I don't see leakage into other channel, just distorted main channel.

in AP crosstalk test, one would clearly see sinad and graphs for both channels.

Example: This is the test for Topping A90, which has elevated crosstalk in balanced.

had amir did this test, he could not miss the flaw at all.

 

[liu]

Can anyone check this on other DACs using Sabre? Wondering if it's an SMSL issue or a Sabre issue.

clearly a SMSL issue. I do have other ESS9038 dacs and no other product shows this issue.

 

[MusicNBeer]

I wonder if other SMSL products have this issue? This is a huge problem. SINAD of 75 dB isn't even $20 DAC spec.

 

[JSmith]

I bought a used M500 … it arrives in a few days.

Great, urgently waiting on your testing and results of this possible issue.



JSmith

 

[MusicNBeer]

Nothing possible about it. 100% there on my V1 unit and 100% there on latest V2 unit. I ran exact same test on my Benchmark DAC1 and all is clean. As I said, the good news is you can reduce the volume to 30 or so and get THD < 0.001

 

[abdo123]

You guys are seriously invested in a harmonic below -100 dB.

 

[Veri]

Nothing possible about it. 100% there on my V1 unit and 100% there on latest V2 unit. I ran exact same test on my Benchmark DAC1 and all is clean. As I said, the good news is you can reduce the volume to 30 or so and get THD < 0.001

Good detective work =/ thanks.

You guys are seriously invested in a harmonic below -100 dB.

Bit of a waste for a 'PRO' chip like the ESS9038Pro no? Might as well buy a 9018k2m/9038q2m if -100dB is what we're shooting for

 

[JSmith]

You guys are seriously invested in a harmonic below -100 dB.

Not specifically... however if same represents a design or product fault, then yes I think many would like to know including SMSL.



JSmith

 

[MusicNBeer]

You guys are seriously invested in a harmonic below -100 dB.

It's not 100 dB, it 75 dB below the fundamental with volume set to 40. That's atrocious.

 

[Jimbob54]

Could someone just confirm the conclusions from all this good work to a layman like me please.

If one drives only the left channel the THD is way above spec. But if use it in stereo playback, more like normal levels? If one drives only left but at volume 30/40 or so, likewise, back to normal levels?

So it's technically broken but in stereo use, not measurably regardless of output level?

Something more for designer to worry about than end user with a 12 month or more old v1?

 

[MusicNBeer]

When signals are different on left and right channels, as with stereo playback, THD is bad on right channel on my unit. Reducing volume reduces the problem.

 

[Jimbob54]

When signals are different on left and right channels, as with stereo playback, THD is bad on right channel on my unit. Reducing volume reduces the problem.

Ok, so to avoid either issue, dial back to 30 or so. Annoying, quite possibly not audible but definitely not what one expects. Thanks

 

[MusicNBeer]

Ok, so to avoid either issue, dial back to 30 or so. Annoying, quite possibly not audible but definitely not what one expects. Thanks

Yep, exactly. It'll be transparent set to 30, but annoying for sure. It's nowhere near meeting its spec.

 

[liu]

Bit of a waste for a 'PRO' chip like the ESS9038Pro no? Might as well buy a 9018k2m/9038q2m if -100dB is what we're shooting for

AKM ak4452 is just $1.5, and has 107db SINAD. Super transparent.
ES9038Q2M costs 10x more. 9038Pro costs 50x more.