Exasound E32 DAC Review and Measurements

[jtwrace]

I'd try to send it back or ask George for the PlayPoint to try and see if it really makes much difference as he says that's why it's all locked.

 

[tomelex]

its more proof that money does not equal quality in the audio world, and that message needs repeating. I really enjoyed your work Amir. As this site becomes more discovered over time these kinds of threads will be golden.

 

[watchnerd]

I also think for that money there are pro devices that would give good DAC and ADC results. I didn't think of it when you were asking for suggestions, but this RME ADI2 Pro is a good example.

https://www.sweetwater.com/store/detail/ADI2Pro

http://www.rme-audio.de/en/products/adi_2-pro.php

There are other choices too. Some of the Antelope gear, but special drivers and support are iffy.

I mentioned the RME ADI 2 in one of the threads on the topic.

I haven't bought one...yet. I don't really have a role for it in my living room, but in my being-revamped home office / mixing setup, it would serve a purpose.

On the other hand, the new iMac Pro is calling me as new DAW...

 

[watchnerd]

its more proof that money does not equal quality in the audio world, and that message needs repeating. I really enjoyed your work Amir. As this site becomes more discovered over time these kinds of threads will be golden.

Sadly, as good as the Behringer measures, it has -100 points on the audio bragging rights scale, so to most audiophiles it doesn't matter how good it is.

 

[Blumlein 88]

I don't know if you did anymore with your Exasound DAC. I did notice this interesting tidbit on their page with their test results.

• Unless the type of the power supply is specified, a 12V Lithium battery is used as a power source for the DAC

That seems disingenuous to me. They should test the unit as it is sold or sell it with a 12 v lithium battery.

 

[Kal Rubinson]

I don't know if you did anymore with your Exasound DAC. I did notice this interesting tidbit on their page with their test results.

• Unless the type of the power supply is specified, a 12V Lithium battery is used as a power source for the DAC

That seems disingenuous to me. They should test the unit as it is sold or sell it with a 12 v lithium battery.

I've been using a 12v sealed lead acid battery to run mine for quite a while.

 

[amirm]

I don't know if you did anymore with your Exasound DAC. I did notice this interesting tidbit on their page with their test results.

• Unless the type of the power supply is specified, a 12V Lithium battery is used as a power source for the DAC

That seems disingenuous to me. They should test the unit as it is sold or sell it with a 12 v lithium battery.

Oh, I had not seen that. That would certainly hide the mains related noise I measured with its in-built supply. Definitely not proper. They should show it both ways and as you say, offer the battery as an option.

 

[Kal Rubinson]

Oh, I had not seen that. That would certainly hide the mains related noise I measured with its in-built supply. Definitely not proper. They should show it both ways and as you say, offer the battery as an option.

FWIW, it is factory supplied not in-built.

 

[c1ferrari]

While some companies are willing to have their products tested and even offer them without me asking, those seem to be quite an exception. So you are right: expecting manufacturers to sending products for review is not a strategy that will work.

AP would sometimes have suites at audio shows and make their gear and expertise available. I'd visit and inquire about traffic...slow they would observe.
I was not at all surprised.

 

[amirm]

AP would sometimes have suites at audio shows and make their gear and expertise available. I'd visit and inquire about traffic...slow they would observe.
I was not at all surprised

I see and talk to them at every show. I have yet to see them measure anyone's gear despite the offer you mention.

 

[c1ferrari]

Went to a head-fi type show -- Bruel & Kjaer had one of their Head and Torso Simulators (HATS) available along with relevant measurement gear...there was a queue.

 

[oivavoi]

Two things were disappointing here:
1. It comes with a switchmode power supply as an external wall-wart. As with my other tests, this also leaks AC mains into the DAC and comes right out of its analog outputs. In my testing below, I opted to test with my lab supply to eliminate that.

Great review, Amir.
This got me curious: is there some validity to the audiophile belief that switching power supplies are bad for amps as well? Asking since I've been considering using a crown CTS amp with smps for my next setup.

 

[Jinjuku]

Thanks for doing all this and just confirming that $79 buys you a more than competent device.

While the UMC204HD outputs are +3dBu the Behringer FCA 610 (24/96) interface is +8dBu out.

I wonder if it's kept together as well as the UMC204HD...

 

[Phison Audio - Sonny]

I can confirm the 3rd harmonic on the ES90x8pro dacs.

Behringer has both just the same and I showed those results:

I called it "RCA" because more people are familiar with that term than unbalanced.

 

[amirm]

I can confirm the 3rd harmonic on the ES90x8pro dacs.

That's good to know Sonny. How is the AKM DAC in this regard?

 

[amirm]

While the UMC204HD outputs are +3dBu the Behringer FCA 610 (24/96) interface is +8dBu out.

I wonder if it's kept together as well as the UMC204HD...

Given the limit of 96 Khz, I don't think it makes a good target for consumer DACs.

 

[amirm]

Great review, Amir.
This got me curious: is there some validity to the audiophile belief that switching power supplies are bad for amps as well? Asking since I've been considering using a crown CTS amp with smps for my next setup.

The issues I am finding with switching supplies in DAC are due to use of off-the-shelf power supplies. These have EMI mitigation that causes that issue. An internally designed unit can avoid this specific problem.

Outside of that I have nothing authoritative to offer there. I guess I need to get into testing these types of amps to gather more data.

 

[DonH56]

Great review, Amir.
This got me curious: is there some validity to the audiophile belief that switching power supplies are bad for amps as well? Asking since I've been considering using a crown CTS amp with smps for my next setup.

I agree on the great review. On SMPS, not Amir but my microcent... Also note I do not design these things for a living, but have done a few and been around quite a few.

There are some pros that SMPS (switch-mode power supplies) have including higher efficiency, smaller size and weight, less heat, etc. A high switching rate means fundamental noise is well outside the audio band, unlike conventional supplies that typically inject around 100/120 Hz and its harmonics. It also means much smaller capacitors can be used, as they are "topped off" at several hundred kHz instead of 120 Hz (here in the U.S.A.; 100 Hz in some parts of the world). You can design them to be as clean as linear supplies, but there are some drawbacks (natch).

That same high switching rate means fast edges (relative to audio) and harmonic energy extending well past the AM radio band (can hit tens to even hundreds of MHz). That means the SMPS, and potentially components using and/or around them must be well-shielded against RFI. The energy may be well above the audio band, but any sneak paths where it can be rectified or modulated down to audio frequencies will be Bad News. This is no different than for other digital circuits, like DSPs in the box (e.g. AVRs, digital processors, even just digital displays). Aside: I have seen marketing sheets for various AVRs and pre/pros touting their linear power supplies, ignoring the massive amount of internal noise from the control, display, and DSP circuits. Whatever sells...

Because a SMPS switches "rail-to-rail" there are large current spikes into the incoming power supply and into ground. Design and layout is critical to control these spikes and maintain the loop currents locally. Linear supplies also generate switching spikes as diodes turn on and off, but at a much lower rate and with much more limited frequency range (*). Get the layout wrong and you're going to have broadband noise injected "everywhere".

Isolation techniques can be similar for the two designs, but because of their HF content need to be more broadband for SMPS'. There are a variety of off-the-shelf input filters these days but again how they are used in the circuit and how ground and supply leakage paths are handled in the layout are critical. They need a lot more attention paid to HF paths as well as LF paths. And, since they are HF devices (again, relative to conventional power supplies or audio circuits), IME sometimes the LF paths get neglected, particularly for off-line supplies. Without good grounding and shielding you can end up with a very clean, low-ripple DC output riding on a little 60 Hz line leakage (leading to "It's a 500 kHz SMPS, where did that 60 Hz spur come from?")

Note some designs, though no power amps AFAIK, use an SMPS to get the voltage down before applying it to a linear regulator. That keeps loss down in the linear regulator and provides one more level of isolation from the SMPS HF energy. How well, or if, it helps depends on the design and layout, of course.

SMPS have built-in soft-start capability and regulation, which in addition to their small SWaP (size, weight, and power) and fundamental noise well out of the audio band, should make them very desirable in audio circuits. But, they are harder to design, especially for companies rolling their own, require additional and perhaps new skill sets for the design and layout teams, and have that "digital" stigma attached to them.

FWIWFM - Don

(*) Rate and frequency spectrum in this context are not the same though can be coupled. A very narrow spike that occurs rarely will have very broadband spectral components but relatively low energy overall. A big spike that occurs more often but is broader may have more energy but frequency components that are not nearly as high. SMPS generate narrow spikes at a very high rate so have a fair amount of HF energy, typically way above the audio band. Conventional supplies have bigger, broader spikes that happen much less often (much lower rate) and extend much lower in frequency before they die away.

 

[Phison Audio - Sonny]

I have a few measurement that is not made by me but I think they are okay and i have not manipulated them. First is non isolated
Second is isolated USB

 

[oivavoi]

I agree on the great review. On SMPS, not Amir but my microcent... Also note I do not deign these things for a living, but have done a few and been around quite a few.

There are some pros that SMPS (switch-mode power supplies) have including higher efficiency, smaller size and weight, less heat, etc. A high switching rate means fundamental noise is well outside the audio band, unlike conventional supplies that typically inject around 100/120 Hz and its harmonics. It also means much smaller capacitors can be used, as they are "topped off" at several hundred kHz instead of 120 Hz (here in the U.S.A.; 100 Hz in some parts of the world). You can design them to be as clean as linear supplies, but there are some drawbacks (natch).

That same high switching rate means fast edges (relative to audio) and harmonic energy extending well past the AM radio band (can hit tens to even hundreds of MHz). That means the SMPS, and potentially components using and/or around them must be well-shielded against RFI. The energy may be well above the audio band, but any sneak paths where it can be rectified or modulated down to audio frequencies will be Bad News. This is no different than for other digital circuits, like DSPs in the box (e.g. AVRs, digital processors, even just digital displays). Aside: I have seen marketing sheets for various AVRs and pre/pros touting their linear power supplies, ignoring the massive amount of internal noise from the control, display, and DSP circuits. Whatever sells...

Because a SMPS switches "rail-to-rail" there are large current spikes into the incoming power supply and into ground. Design and layout is critical to control these spikes and maintain the loop currents locally. Linear supplies also generate switching spikes as diodes turn on and off, but at a much lower rate and with much more limited frequency range (*). Get the layout wrong and you're going to have broadband noise injected "everywhere".

Isolation techniques can be similar for the two designs, but because of their HF content need to be more broadband for SMPS'. There are a variety of off-the-shelf input filters these days but again how they are used in the circuit and how ground and supply leakage paths are handled in the layout are critical. They need a lot more attention paid to HF paths as well as LF paths. And, since they are HF devices (again, relative to conventional power supplies or audio circuits), IME sometimes the LF paths get neglected, particularly for off-line supplies. Without good grounding and shielding you can end up with a very clean, low-ripple DC output riding on a little 60 Hz line leakage (leading to "It's a 500 kHz SMPS, where did that 60 Hz spur come from?")

Note some designs, though no power amps AFAIK, use an SMPS to get the voltage down before applying it to a linear regulator. That keeps loss down in the linear regulator and provides one more level of isolation from the SMPS HF energy. How well, or if, it helps depends on the design and layout, of course.

SMPS have built-in soft-start capability and regulation, which in addition to their small SWaP (size, weight, and power) and fundamental noise well out of the audio band, should make them very desirable in audio circuits. But, they are harder to design, especially for companies rolling their own, require additional and perhaps new skill sets for the design and layout teams, and have that "digital" stigma attached to them.

FWIWFM - Don

(*) Rate and frequency spectrum in this context are not the same though can be coupled. A very narrow spike that occurs rarely will have very broadband spectral components but relatively low energy overall. A big spike that occurs more often but is broader may have more energy but frequency components that are not nearly as high. SMPS generate narrow spikes at a very high rate so have a fair amount of HF energy, typically way above the audio band. Conventional supplies have bigger, broader spikes that happen much less often (much lower rate) and extend much lower in frequency before they die away.

Thanks so much, Don! I stand very enlightened!