SMSL M500 DAC Only “Thermal Issue” not “Golden Sample”

[WolfX-700]

When I started publishing measurement results in ASR, I said that part of my work would be devoted to "double checking" on Amir's measurement results. To avoid a situation where a vendor sends a "golden sample" to ASR

Of course, this part of my work depends on luck. The key is the equipment my friends can provide me, or whether the equipment I buy is measured by Amir.

Amir and I got results within a margin of error in a longer measurement. We followed basically similar temperature conditions and the end results were similar. But the trend is reversed. (The SINAD I have on hand goes up, while Amir's THD + N gets worse). You can see pages 4 and 5 of this post.


At present, the probability of "golden sample" issues is decreasing, while the probability of "thermal issue" is higher.

I have placed an order for a new unit. Please stay tuned.

2019/12/11 Update:
I received the unit I purchased.

Dashboard（Cold room temp=72℉ )：

Long-term testing:At the time I marked, I attached a warm pad to the M500 and covered it with clothes.

The final measurement results match Amir's earlier measurements! It can be considered that the M500 is only an issue in which the measurement result changes greatly with temperature changes.

There is no "Golden Sample"




Just today, I received a SMSL M500 from a friend.

After measurement, I think this is a big problem ... because the measurement results are very different from Amir.

Here are the measurements:

I think I used the same settings as Amir (ASIO4ALL USB 48K, Volume Set to 34 4V Out)

But the measurement results are very different from @amirm ：https://www.audiosciencereview.com/forum/index.php?threads/smsl-m500-dac-and-hp-amp-review.9606/

So if there is a chance, I think Amir can try to get another M500 from a third party channel.

I will also try to find another one.

If the "golden sample" is confirmed ...


Update :

Just now I read Amir's post carefully. It was found that his dashboard was using a sampling rate of 44.1KHZ.

To avoid controversy, I did another quick measurement with 44.1KHZ.

Here are the results:

 

[JohnYang1997]

Hmmm interesting. Did they really do this? Is it cherry picked or they changed design for the production model? I think the latter.
Can we get a teardown please?
@amirm

 

[Veri]

When I started publishing measurement results in ASR, I said that part of my work would be devoted to "double checking" on Amir's measurement results. To avoid a situation where a vendor sends a "golden sample" to ASR

Of course, this part of my work depends on luck. The key is the equipment my friends can provide me, or whether the equipment I buy is measured by Amir.

Just today, I received a SMSL M500 from a friend.

View attachment 41653


After measurement, I think this is a big problem ... because the measurement results are very different from Amir.

Here are the measurements:

View attachment 41656


View attachment 41657View attachment 41658View attachment 41659View attachment 41660
View attachment 41662


I think I used the same settings as Amir (ASIO4ALL USB 48K, Volume Set to 43 4V Out)

But the measurement results are very different from @amirm ：https://www.audiosciencereview.com/forum/index.php?threads/smsl-m500-dac-and-hp-amp-review.9606/

So if there is a chance, I think Amir can try to get another M500 from a third party channel.

I will also try to find another one.

If the "gold sample" is confirmed ...

SINAD, linearity, multitone and jitter are still decent. But how did they revert to IMD hum ????

Output stage must be different. @WolfX-700 can you test at maximum volume? Is performance same? Or worse still?

In any case this re-test is very valuable, looking at this: Sabaj D5 and Topping DX7Pro are better in objective performance. Thank you,

 

[JohnYang1997]

SINAD, linearity, multitone and jitter are still decent. But how did they revert to IMD hum ????

Distortion harmonics are also much worse. Perhaps they changed some critical component that the production model performs worse.

 

[WolfX-700]

SINAD, linearity, multitone and jitter are still decent. But how did they revert to IMD hum ????

Output stage must be different. @WolfX-700 can you test at maximum volume? Is performance same? Or worse still?

Take a quick measurement before going to supper:

MAX Volume SINAD = 109

 

[Veri]

Distortion harmonics are also much worse. Perhaps they changed some critical component that the production model performs worse.

@SMSL-Mandy any chance you could investigate?

 

[MusicNBeer]

Must be the silver paint. The black paint must be sucking up the residual distortion and noise. Mine's silver too. Bummer.

 

[Veri]

Must be the silver paint. The black paint must be sucking up the residual distortion and noise. Mine's silver too. Bummer.

Insert 'once you go black' joke here

 

[mi-fu]

Wow. I didn't expect that!!!

I think measuring a third (random) sample will be important.

 

[WolfX-700]

Just now I read Amir's post carefully. It was found that his dashboard was using a sampling rate of 44.1KHZ.

To avoid controversy, I did another quick measurement with 44.1KHZ.

Here are the results:

 

[BaaM]

I think measuring a third (random) sample will be important.

I think refusing a gift sample from the company should be important too.

 

[mi-fu]

I think refusing a product from a company should be important too.

i think we need a bit more evidence to draw that conclusion

 

[BaaM]

i think we need a bit more evidence to draw that conclusion

Regardless of the conclusion on this particular case, relying on the companies is already a significant risk that the tests will be biased.

 

[maxxevv]

Think there were some key component changes somewhere along the line as production went on over the months.

 

[WolfX-700]

Regardless of the conclusion on this particular case, relying on the companies is already a significant risk that the tests will be biased.

I don't think it's possible to completely avoid the situation where companies provide samples.

Many times, reader samples are not always readily available.

If possible, ASR can introduce a red and yellow card system for companies.

If you repeatedly find that the sample provided by the company is very different from the actual retail. Make certain marks or do something else ...

 

[MusicNBeer]

Regardless of the conclusion on this particular case, relying on the companies is already a significant risk that the tests will be incorrect.

Agreed. I'm no longer going to put any weight into the review unless it's a random production unit being tested.

 

[mi-fu]

Agreed. I'm no longer going to put any weight into the review unless it's a random production unit being tested.

Well.. we need enough members to send in samples to measure (it means they need to purchase it before seeing any measurement). And sometimes people want to have early (pre-production) measurements before making their purchase decision.

There is no perfect solution

 

[JohnYang1997]

Well.. we need enough members to send in samples to measure (it means they need to purchase it before seeing any measurement). And sometimes people want to have early (pre-production) measurements before making their purchase decision.

There is no perfect solution

Generally I don't think this would be an issue. If one manufacturer does this, the whole forum will be very cautious about this manufacturer. The reputation would suffer.

 

[WolfX-700]

Generally I don't think this would be an issue. If one manufacturer does this, the whole forum will be very cautious about this manufacturer. The reputation would suffer.

Yes, it may be a separate incident. I have recorded the serial number of this M500 and I will try to find another M500 for testing.

It would be a shame if SMSL did send a golden sample.

 

[Jinjuku]

I personally wouldn't accept a company provided unit. If a manufacturer wants a unit benched allow for an in the wild unit to be purchased and the company reimburses or they provide pre-payment for one to be obtained.

It's what Consumer Reports does.