SMSL M500 MKIII DAC & Amp Review

[777]

I listened to this dac. His sound is superb. SMSL made the best choice with the use of 8 opamps. The sound has strength, it is no longer soft.

 

[BR52]

Thanks Amir for the review great as always. Could you measure the DAC output impedance? Or do you have information about? Is essential to know. The max output voltage as well. Purifi direct mode you know…….

 

[VintageFlanker]

The max output voltage as well.

According to specs shared by @SMSL-Mandy, it is 4.1V BAL, so I guess SINAD was measured at max volume:

Line output amplitude
XLR 4.1Vrms
RCA 2Vrms

 

[JSmith]

Once this level of performance starts to seep into the $150,- territory, ASR can just become a blank page with a few affiliate links to those products and Amir can retire.

Nah, someone needs to keep an eye on the industry... one could even say the price of audio liberty is eternal vigilance.

The sound has strength, it is no longer soft.

Compared to which DAC's?


JSmith

 

[amirm]

Performance is out of this word. Thanks for the review, @amirm ! Is your dashboard at 4.18Vrms done at max volume? Previous version had higher output voltage (at least the mk1).

My pleasure. Yes, this was at full volume. Half a dB over nominal 4 volt but I let this bit of cheating go on as others are doing the same.

 

[VintageFlanker]

My pleasure. Yes, this was at full volume. Half a dB over nominal 4 volt but I let this bit of cheating go on as others are doing the same.

Got it. From my little experience measuring DACs, I found that ≈4.2Vrms is much more common than 4.0Vrms anyway.

 

[777]

Compared to which DAC's?


JSmith

Compared to expensive western DAC's. Chinese DAC's have always had undersized power and output stages. Without a high current supply a device does not sound good. The only criticism at SMSL M500 MK3 is the output stage opamps. OPA1612 is too weak for 10mA current. The resistance from the I/V converter is only 350 ohm. This opamp is not a perfect choice for that task. For the future M500, M500MK4, I would like to see OPA1656 instead of OPA1612. This opamp could drive 350 ohm better than OPA1612. Or, why not, a composite opamp, OPA1612+...., or OPA1612+power transistor. In this case the output stage it will be in Class A biasing.

 

[okok]

the clock is no good, multitone can be better

the 6120 hp amp is as no good as the mk2, can be better too

 

[VintageFlanker]

Compared to expensive western DAC's. Chinese DAC's have always had undersized power and output stages. Without a high current supply a device does not sound good.

What the heck are you talking about?

How do you explain that has not been seen in measurements so far?

 

[BR52]

According to specs shared by @SMSL-Mandy, it is 4.1V BAL, so I guess SINAD was measured at max volume:

The 4.1V would be good enough for me because of my high efficiency speakers. Output impedance is critical in this case.

 

[okok]

What the heck are you talking about?

How do you explain that has not been seen in measurements so far?

he's being very reasonable, means it can be better, but budget and bom

there's a su10 double the price got better output stage

 

[VintageFlanker]

there's a su10 double the price got better output stage

OK... What of it?

Will you guys finally understand that DAC measurements are made at the analog output, which means including the output stage? And that any influence of it (or the PSU or whatever) will be seen in the data?

 

[BR52]

Compared to expensive western DAC's. Chinese DAC's have always had undersized power and output stages. Without a high current supply a device does not sound good. The only criticism at SMSL M500 MK3 is the output stage opamps. OPA1612 is too weak for 10mA current. The resistance from the I/V converter is only 350 ohm. This opamp is not a perfect choice for that task. For the future M500, M500MK4, I would like to see OPA1656 instead of OPA1612. This opamp could drive 350 ohm better than OPA1612. Or, why not, a composite opamp, OPA1612+...., or OPA1612+power transistor. In this case the output stage it will be in Class A biasing.

As you can see from the measurements the 1612 with 350 Ohm is the perfect choice for the I/V stage. 1656 will double the noise

 

[AudioSceptic]

Amir,

A couple of minor complaints though: the USB connector is too close tot he upside down XLR jack which made it impossible to push its pin in with the USB cable attached to right channel.

How marginal is this, IOW would this apply to *any* such connectors, or just the ones you happened to use? Can you actually use XLR with USB but it's just that the XLR won't lock? In any case, it looks like careless design.

 

[777]

What the heck are you talking about?

How do you explain that has not been seen in measurements so far?

Because of that : 200k ohm's is the measurement load. Inside in the M500 Mk3, the load of the OPA1612 wich is I/V converter, is only 350 ohm's. 350 ohm on every OPA1612, each separately. The load is too little. In fact, the current, 10mA from the es9038Pro is too high for OPA1612. Es9038Pro has too big output current for common opamps. 10mA is too big for that opamp if you want the best quality of the sound. At such a current you approach to the overload protection circuit of the opamp. You need a stronger opamp or if you want, output stage.

 

[777]

As you can see from the measurements the 1612 with 350 Ohm is the perfect choice for the I/V stage. 1656 will double the noise

Who care about that extra-tinny noise ? We are interested in realistic sound perception, we are interested in transients that require higher and clean currents.

 

[BR52]

Who care about that extra-tinny noise ? We are interested in realistic sound perception, we are interested in transients that require higher and clean currents.

Can you explain what does mean in measurement parameters?

 

[777]

1656 will double the noise

Can you explain what does mean in sound perception ? Can you hear that noise ?

 

[VintageFlanker]

We are interested in realistic sound perception

We all are. I do not "listen to graphs" either.

Yet,

we are interested in transients that require higher and clean currents.

I would still ask for actual proof that any of your claims could be verified when related to audibility.

 

[BR52]

Can you explain what does mean in sound perception ? Can you hear that noise ? R

I am interested in technology not in ping pong discussions. So i stop here.