-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
SMSL M500 DAC Only “Thermal Issue” not “Golden Sample”
- Thread starter WolfX-700
- Start date
I’m actually very strongly considering returning my unit, unfortunately I did not buy directly from SMSL so I anticipate some issues. Provided the Topping engineers resolve the issues @WolfX-700 found, I will buy the D90. SMSL has zero excuses for ignoring this issue, will definitely avoid them in the future if they don’t step up and shed some light.
OP
- Thread Starter
- #203
I have always believed that --- finding problems--researching problems--solving problems = better products and better manufacturers, and more consumers benefit.
And in fact I don't have any unfriendly behavior towards SMSL. I paid for the SP200 with an independent survey to verify whether the SP200 had a "golden sample" problem and (I think to a certain extent) further strengthened the SP200's credibility.
And this M500 is the first unit from SMSL that I received from a friend ...
And in fact I don't have any unfriendly behavior towards SMSL. I paid for the SP200 with an independent survey to verify whether the SP200 had a "golden sample" problem and (I think to a certain extent) further strengthened the SP200's credibility.
And this M500 is the first unit from SMSL that I received from a friend ...
Last edited:
pacman
Member
This year SMSL put at the market the Sabaj D5 with software glitches that requires a special tool to update the firmware.
Now there's this issues with SP200's pot and M500's heat problems (and other firmware glitches).
Kinda concerning.
Now there's this issues with SP200's pot and M500's heat problems (and other firmware glitches).
Kinda concerning.
Veri
Master Contributor
- Joined
- Feb 6, 2018
- Messages
- 9,598
- Likes
- 12,040
SMSL is a big company though. Sabaj probably has whole other team of people behind it.
Just wondering if you guys may be overreacting. We've seen from posts here that other manufacturer's products are having issues with heat from their ES9038PRO products, one even recalling their devices to fit a huge heatsink on top.
Until we get a look inside the device to better understand what's going on then I think we're a bit quick jumping to conclusions.
Shadrach
Addicted to Fun and Learning
- Joined
- Feb 24, 2019
- Messages
- 676
- Likes
- 980
Components should come with a tolerance and a temperature at which these tolerances are relevant. plus and minus 5% are reasonably common for electrical components, depending on type. Some capacitors for example also come with an estimated cycle life.
One can't count on the tolerances summing to a median value. It only takes a few components to be at the extreme of either the negative, or positive value to produce a quite noticeable difference from unit to unit.
It also needs to be born in mind that these tolerances are specified at a particular temperature. Change that temperature and the stated tolerances may no longer apply.
Components are sample tested in general. An example where sample testing may not be satisfactory in audio is driver impedance curves. Better loudspeakers should have matched drivers. I doubt many manufacturers test every component used in a product so it is quite possible to have an out of tolerance component in the chain which may produce a product measurement out of specification.
Doing a bit of science is great, but it needs to be born in mind that if the measuring techniques are highly sensitive, but the testing environment is not controlled differences between product measurements are inevitable. A combination of environmental differences and component tolerance differences are often what will account for differing same product measurements.
Opening a window, or putting a unit in a fridge for a bit until it feels cold isn't very scientific.
The site measurements are great, but they apply to that unit under those conditions.
One can't count on the tolerances summing to a median value. It only takes a few components to be at the extreme of either the negative, or positive value to produce a quite noticeable difference from unit to unit.
It also needs to be born in mind that these tolerances are specified at a particular temperature. Change that temperature and the stated tolerances may no longer apply.
Components are sample tested in general. An example where sample testing may not be satisfactory in audio is driver impedance curves. Better loudspeakers should have matched drivers. I doubt many manufacturers test every component used in a product so it is quite possible to have an out of tolerance component in the chain which may produce a product measurement out of specification.
Doing a bit of science is great, but it needs to be born in mind that if the measuring techniques are highly sensitive, but the testing environment is not controlled differences between product measurements are inevitable. A combination of environmental differences and component tolerance differences are often what will account for differing same product measurements.
Opening a window, or putting a unit in a fridge for a bit until it feels cold isn't very scientific.
The site measurements are great, but they apply to that unit under those conditions.
please document your teardown so that amir can duplicate. if there was a heatsink change or other seemingly minor change in later production models we may be able to isolate why the baseline measurements are lower on the units you've been testing.
amir should have his random production sample today as well.
amir should have his random production sample today as well.
As other folks have said, I do think we need to put things into perspective here. This is a low cost product, and for that cost, the real-world performance is truly amazing. It's useful to remember that, even in the worst case measurements presented here, all distortion products are below the resolution level of CD audio. Most loudspeakers & headphones are about 40dB worse than this.
I think what this thread shows more than anything, is one of the perils of attempting to categorize the performance of any engineered product, in an area way outside that of its normal usage requirements. Distortion products 100dB down from the signal are simply inaudible. Period.
However, what doesn't help in this particular instance is that SMSL themselves quote "SINAD up to 117dB" in their marketing materials. Though the term "up to" does beg the question... What experienced manufacturers learn is that if you quote a specification, you have to meet it. Every time. With "pro" equipment you often find that the manufacturer's spec is much more lenient than measurements on individual samples would suggest. On top of that, there may be several tiers of specification, based on operating temperature, power-up time and age of product.
All electronic equipment exhibits some degree of temperature related fluctuation in its performance, even if minuscule. But the change of operating parameters with time & temperature need not be a problem. I would venture to say that it is indeed not a problem here. The problem is in defining the limits of functionality, writing an appropriate specification, and interpreting that specification.
In a market place which is polarized by the extremes of magic phenomena and "anti-science" on the one hand, and the pursuance of "better measurements mean better sound, ad infinitum" on the other, objectivity is a difficult path to tread.
ASR does an excellent job of consumer advocacy in holding manufacturers to their claims, and in singling out the truly incompetently designed, and often overpriced products which are increasingly finding their way to market. The SINAD league table is quite interesting, but it's based on a single measured parameter &, in my view , should not become a goal in itself. There is a danger, as with all league tables that this will obfuscate some very real issues worthy of attention.
Beyond a point, better numbers don't necessarily mean better performance. In an audio product, SINAD of 110dB vs 120dB makes no practical difference, whereas factors such as build quality following good engineering practice, reliability (admittedly hard to measure with a sample of 1), and the availability of manufacturers' guarantees should realistically carry much more weight in any overall verdict on a product.
It's also well worth remembering, as the old saying goes, "the best is the enemy of the good".
I think what this thread shows more than anything, is one of the perils of attempting to categorize the performance of any engineered product, in an area way outside that of its normal usage requirements. Distortion products 100dB down from the signal are simply inaudible. Period.
However, what doesn't help in this particular instance is that SMSL themselves quote "SINAD up to 117dB" in their marketing materials. Though the term "up to" does beg the question... What experienced manufacturers learn is that if you quote a specification, you have to meet it. Every time. With "pro" equipment you often find that the manufacturer's spec is much more lenient than measurements on individual samples would suggest. On top of that, there may be several tiers of specification, based on operating temperature, power-up time and age of product.
All electronic equipment exhibits some degree of temperature related fluctuation in its performance, even if minuscule. But the change of operating parameters with time & temperature need not be a problem. I would venture to say that it is indeed not a problem here. The problem is in defining the limits of functionality, writing an appropriate specification, and interpreting that specification.
In a market place which is polarized by the extremes of magic phenomena and "anti-science" on the one hand, and the pursuance of "better measurements mean better sound, ad infinitum" on the other, objectivity is a difficult path to tread.
ASR does an excellent job of consumer advocacy in holding manufacturers to their claims, and in singling out the truly incompetently designed, and often overpriced products which are increasingly finding their way to market. The SINAD league table is quite interesting, but it's based on a single measured parameter &, in my view , should not become a goal in itself. There is a danger, as with all league tables that this will obfuscate some very real issues worthy of attention.
Beyond a point, better numbers don't necessarily mean better performance. In an audio product, SINAD of 110dB vs 120dB makes no practical difference, whereas factors such as build quality following good engineering practice, reliability (admittedly hard to measure with a sample of 1), and the availability of manufacturers' guarantees should realistically carry much more weight in any overall verdict on a product.
It's also well worth remembering, as the old saying goes, "the best is the enemy of the good".
worth saying this again, since as a scientist this is one of my mantras. this is a 360$ product after MQA royalties. that said, when you state up to 117dba sinad, you need to be that modulo a 2 to maybe 5% variance in production. if we see a mean 105dba sinad measurement as mean in our (albeit small N) with a 5% variance in production, they need to amend that advertising copy.
also note that the SP200 pot change is not on smsl. they have been transparent about the parts change, and that the unit they sent amir was a preproduction. they have also been completely open to replacing units with imbalances. i think they are currently making things right on that model.
OP
- Thread Starter
- #213
It looks like it's just a thermal issue ... Here's a long measurement
At the time I marked, I attached a warm pad to the M500 and covered it with clothes.
Please wait for follow up
At the time I marked, I attached a warm pad to the M500 and covered it with clothes.
Please wait for follow up
777
Senior Member
- Joined
- Jan 1, 2019
- Messages
- 486
- Likes
- 360
Like Sabaj D5. So, 9038Pro want to be hot.
OP
- Thread Starter
- #216
No. My first one is just like this one, and its performance improves as the temperature rises ...
maxxevv
Major Contributor
- Joined
- Apr 12, 2018
- Messages
- 1,872
- Likes
- 1,964
I have not dealt with the SMSL guys for RMA issues nor talked to their technical guys, but I certainly have chatted with the Topping and Matrix technical guys at CanJam. Those guys were in my opinion genuinely enthusiastic about their products and very interested in what the general user had to say about them. And as far as I can tell, they don't like the idea of their products being perceived as bad. The ultra-long delays of the D70 and DX7Pro were very good examples to trying to get everything right, as much/far as possible. Matrix don't churn out revisions in their products very often either. And as far as I can tell, both seem to have pretty lean setups from the frontline marketing to the backend technical development team.
SMSL seems a more complicated company due to the many sub-divisions and different branches of operations they have. As far as I know, they still have an OEM wing that does contract work for other brands. Sabaj and Loxjie have some kind of affiliation in their structure too. So, can't really say that they are worse but different company structures and different directives may drive them compared to Topping and Matrix. I usually reserve my opinions of the staff or company until I have clearer picture.
( I don't own an SMSL products for that matter, so real vested interest there ).
OP
- Thread Starter
- #220
Similar threads
