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

[maxxevv]

The green lines are prior measurements without extra warm-up. The Magenta line is the one with warm-up.

I don't see the "hump" on the green lines.

 

[JohnYang1997]

The green lines are prior measurements without extra warm-up. The Magenta line is the one with warm-up.

I don't see the "hump" on the green lines.

View attachment 41711

Because noise decreased.... and there was hint of broad hump if you look at the picture without lines of warming up.

 

[JohnYang1997]

Take a look at dx7pro. Does it have more hump than d50s. Of course not. Only that the noise is lower, the hump is showing more clearly instead of being drown in the noise.

 

[maxxevv]

That's beside the point.

I'm comparing apples to apples here with the D50S , without warm-up and with warm-up. You are talking about the context of noise versus signal, I'm talking about measurements with and without temperature influences.

Please get that into context.

 

[Rja4000]

Amir, could you measure its linearity?

Which leads us to next question: how much warm up time does the APx555 require?
(I'm kidding)

 

[JohnYang1997]

That's beside the point.

I'm comparing apples to apples here with the D50S , without warm-up and with warm-up. You are talking about the context of noise versus signal, I'm talking about measurements with and without temperature influences.

Please get that into context.

It's the same. I only see reduced noise instead of increased hump. If the hump goes over the line without warm-up, then yes I would agree that it's increased. But no, it's barley touching at two points and the noise is lower. There was a hint of imd hump before warm-up and there is a hint of imd hump after warm-up but not more than before warm-up.

 

[JohnYang1997]

Now I have a guess. The performance of ess dac is highly depending on the oscillator. In this case is the issue the thermal stability of the oscillator? Can anyone confirm this?

 

[WolfX-700]

Now I have a guess. The performance of ess dac is highly depending on the oscillator. In this case is the issue the thermal stability of the oscillator? Can anyone confirm this?

I got permission from my friend to open the M500. I will do this later today.

But before that, I have to complete another new product measurement job that I just received today ...

 

[JohnYang1997]

I got permission from my friend to open the M500. I will do this later today.

But before that, I have to complete another new product measurement job that I just received today ...

Awesome!

 

[pcdinh]

I got permission from my friend to open the M500. I will do this later today.

But before that, I have to complete another new product measurement job that I just received today ...

Good jobs. Cross check by independent people is a must to ensure that measurements are done correctly and efficiently

 

[Rja4000]

Okay, I'm back ~

I turned off the air conditioner in the workplace and opened the windows before going to bed, so the temperature in this room dropped sharply, and now it is only about 50 ° F. I believe it would be as low as 40 ° F earlier.

So I got an interesting long-term test result.

Shows the performance of M500 at low temperatures.

I still decided to find another unit to test again.

View attachment 41708

Looks like you have 2 parameters changed here: devices is warming up with time, but room temperature was also varying...
Hard to sort that out.
No plot of room temperature vs same time, by any chance?
Or an IR measurement of device temperature on same time plot ?

 

[JohnYang1997]

Looks like you have 2 parameters changed here: devices is warming up with time, but room temperature was also varying...
Hard to sort that out.
No plot if room temperature vs same time, by any chance?
Or an IR measurement of device temperature on same time plot ?

It's thd rised over time while room temp being constant. Then reduced room temp leading to drop in thd.

 

[Rja4000]

It's thd rised over time while room temp being constant. Then reduced room temp leading to drop in thd.

Yes, I guess so.
Main conclusion is that there is a link between device temperature and performance.
Performance seems to increase with device temperature, and to stabilize after 1h.
This effect actually seems to increase the gap between Amir's quick measurement and Wolfx' slow one.

 

[ninetylol]

What I still dont get is why Amirs SINAD measurements decreases with warmup while WolfX measurements get a Higher SINAD with warmup. I mean its the same device.

Also a whopping difference of 15db on SINAD from best case to worst case is a bit too much imo.

 

[Rja4000]

Ok, after 2 hours of soak time, this is what I got:

View attachment 41689

We have a degradation of 5 to 6 dB depending on channel. Note that the above is highly zoomed in covering only 10 dB of range.

I then re-rand the FFT in the dashboard:

View attachment 41690

We have very good matching of the distortion order/level. Second and third harmonic are right on the money.

The noise level is lower in mine but that is affected by FFT parameters.

Looking at dashboard sinad, I see:
View attachment 41691

This is versus 110 in L7's tests. This is within what I consider acceptable given different analyzers, units, etc.

In summary, large amount of difference is due to thermal conditions, not any hand selection of a unit/components.

Am I understanding correctly that @amirm sees SINAD to decrease by 6dB while temperature raises while @WolfX-700 sees the SINAD to increase by a few DB for the same?

 

[Rja4000]

What I still dont get is why Amirs SINAD measurements lowers with warmup while WolfX measurements get a Higher SINAD with warmup.

I mean its the same device. Also a whopping difference of 15db on SINAD from best case to worst case is a bit too much imo.

Crossed posts

 

[JohnYang1997]

Their measurements showed higher temperature = higher distortion. I don't see what's the issue. If you are talking about the one wolfx's post after he went to bed, that one is measuring the performance vs ambient temperature. Lower temperature gives better distortion....

 

[ninetylol]

Their measurements showed higher temperature = higher distortion. I don't see what's the issue. If you are talking about the one wolfx's post after he went to bed, that one is measuring the performance vs ambient temperature. Lower temperature gives better distortion....

Im strictly talking about their SINAD measurements over time (i.e. warmup). WolfX gets a Higher SINAD when the M500 gets warmer while Amirs SINAD decreases with warmup.

 

[WolfX-700]

Their measurements showed higher temperature = higher distortion. I don't see what's the issue. If you are talking about the one wolfx's post after he went to bed, that one is measuring the performance vs ambient temperature. Lower temperature gives better distortion....

Although my and Amir's curves are similar, he uses THD + N and I use SINAD.

This is a very interesting thing ... As the temperature of my device rises, the performance starts to increase (SINAD increases), but as the temperature increases (THD + N value increases), Performance also began to degrade.

Although the end point is within the tolerance of the measurement error, the process is exactly the opposite ...

 

[Rja4000]

Their measurements showed higher temperature = higher distortion. I don't see what's the issue. If you are talking about the one wolfx's post after he went to bed, that one is measuring the performance vs ambient temperature. Lower temperature gives better distortion....

That's not what I understood.

In Wolfx' measurement over time, we see a first raise in SINAD due to the device warming, then he opened the window, temperature "dropped sharply", which SINAD follows, then slow increase.

That seems like opposite effect vs Amir.