Topping E50 Review (Balanced DAC)

[Dac-man]

I don’t have any RCA cables. I never use them. The Meizu DAC was measured here. But the results were much worse because Amir was using his dirty PC USB port to power it. I measured 3 of the MiniDSP Flex units. And all 3 had far worse performance from 2 of the 4 channels. All 3 much worse than Amir’s review sample on 2 of the channels. Guess I’m just not getting golden samples.

My Merging Anubis measures exactly to spec on all 8 channels. To 3 decimal places. So my analyzer is perfectly calibrated.

 

[dsnyder0cnn]

Just got an E50 from Amazon. Hooked it up to my audio analyzer, and powered it from the USB port on my computer. With a 24/192 PCM 1khz sinewave at 0dBFS (4v same as this review) I was getting 110 SINAD from left channel and 111 from the right. So I figured it was due to the dirty power from my USB port. So I powered from my ultra low noise supply with 0.2uV ripple noise. This improved the result to 111 on the left channel and 113 on the right. So I figured maybe it’s conducting noise through the audio input USB port. So I connected it to the Toslink output of my analyzer and ran the test. it remained unchanged at 111 and 113.

Very disappointed with these results. I have a $20 USB dongle that measures better.

Bummer. I guess you could send your E50 in to Amir to have him re-test. Or, you could return the E50 and request a replacement.

 

[dominikz]

I don’t have any RCA cables. I never use them. The Meizu DAC was measured here. But the results were much worse because Amir was using his dirty PC USB port to power it. I measured 3 of the MiniDSP Flex units. And all 3 had far worse performance from 2 of the 4 channels. All 3 much worse than Amir’s review sample on 2 of the channels. Guess I’m just not getting golden samples.

My Merging Anubis measures exactly to spec on all 8 channels. To 3 decimal places. So my analyzer is perfectly calibrated.

Strange you are getting lower figures with E50, perhaps it is a faulty unit?

My E50 unit measures around 115 SINAD when measured by my E1DA Cosmos ADC in stereo mode (i.e. without ADC channel summing for lower noise). This is with a few dB lost due to slightly different sensitivity of the ADC and note also that SINAD of E1DA Cosmos ADC is in the first place similar to that of the E50, so around 3dB is lost on that too - so with that taken into consideration it seems my E50 unit performs relatively similar to the one Amir reviewed.

Of course if I sum the E50 outputs as well as the Cosmos inputs to get lower noise the setup can in that case measure down to 120dB SINAD.

These figures are with E50 powered from a generic USB port on a power strip and Cosmos ADC powered from a PC USB port - so nothing special there.

 

[Dac-man]

Bummer. I guess you could send your E50 in to Amir to have him re-test. Or, you could return the E50 and request a replacement.

No need to send to Amir when I have my own analyzer. Yes will return. But won’t be getting another one. 100% of them should measure on spec or they shouldn’t leave the factory.

 

[Gradius]

MEAN WELL GST18A05-P1J AC-DC High Reliability Industrial Table Top Power Supply, 2.1mm Plug Level VI, 5V 3A 15W 3-Wire https://www.amazon.com/dp/B01LZ0LJXQ/ref=cm_sw_r_cp_api_i_HPTKDP78PAW96KD330WV

This one?

Yep. That one is more than enough.

 

[Dac-man]

Subjectively, this DAC is in another universe when powered by my ultra low noise supply. Far more drastic of a difference than the measurements suggest. I think I’m going to keep it now, regardless if it doesn’t meet spec. But had I only heard it powered by my computers USB port, it would be getting returned.

 

[Dac-man]

Yep. That one is more than enough.

Certainly more than enough ripple noise!

 

[dsnyder0cnn]

Certainly more than enough ripple noise!

View attachment 196609

I'm using this power supply from Allo with my E50: https://www.pishop.us/product/audio-power-supply-5v-3a/

Great results so far, but I've only run it for an hour or so.

 

[Dac-man]

I'm using this power supply from Allo with my E50: https://www.pishop.us/product/audio-power-supply-5v-3a/

Great results so far, but I've only run it for an hour or so.

You don’t know what you’re missing until you hear it with an ultra low noise supply. Buy one of these regulator banks and it will drop the noise from that supply down to 0.8uV ripple

LT3045 Ultra Low Noise Power Supply

Ultra Low Noise Low Dropout Voltage Regulator, Best Power Supply for your HiFi and HiEnd equipment

www.ldovr.com

My supply has 4xLT3045 LDO’s in parallel. With an LT8640 pre-regulator. It has lower noise than any battery available. And extremely fast dynamic load response. It sounds like a completely different DAC than when powered by the USB port.

 

[Dac-man]

I suppose I should mention, you need to use a 6v+ supply to feed that regulator box in my above link. Because there’s probably a 1V voltage drop. So just use any cheap 1a+ 9V SMPS to feed it. Any higher voltage and the regs will run hot, as it uses a linear pre-reg stage.

 

[dsnyder0cnn]

It sounds like a completely different DAC than when powered by the USB port.

Hmmm, I was not able to get the E50 to power on with when connected only by the USB port. Perhaps my powered USB hub does not have enough juice to bus-power the E50? I just figured that the E50 requires external 5V power to operate at all. Did Amir mention how he powered it during his testing?

 

[Dac-man]

Hmmm, I was not able to get the E50 to power on with when connected only by the USB port. Perhaps my powered USB hub does not have enough juice to bus-power the E50? I just figured that the E50 requires external 5V power to operate at all. Did Amir mention how he powered it during his testing?

My ports are 2.1A capable. A standard USB 2.0 port maxes out at 500mA. Which is what this DAC draws. So likely too much. A 3.0 port is good for 900mA. So it will work. I also tried the 2.1a ports on my power bar, and same sub-par sound.

But even my $2500 Merging Anubis sounds drastically better with this supply, rather than their included one. So it’s not much of a surprise. The reason they’re able to sell this DAC for so cheap is because it uses mediocre power regulation. But the beauty of the ES9068AS is it has integrated LDO’s inside of it. So it can be fed with dirty power and still get decent results. However the same can’t be said for the clocks and opamps.

There’s loads of switching regulators inside. Although most switching regulators switch up in the ultrasonic range, they have very poor PSRR. So noise from a switching supply passes right through them. For some reason the SINAD measurement doesn’t pickup on everything that’s audible.

 

[Veri]

It sounds like a completely different DAC than when powered by the USB port.

A "night and day" difference ?

 

[dominikz]

For some reason the SINAD measurement doesn’t pickup on everything that’s audible

SINAD definitely doesn't show all that is audible - e.g. it doesn't contain frequency response, which is something we're very sensitive to (unlike SINAD).

Could you perhaps show some alternative measurements that show the effects you are describing?

Also, have you used controls when listening (level-matched, double-blind)?
Thanks!

 

[Dac-man]

The difference is so drastic that you’d need to be deaf not to hear it. Do you need special controls to tell the difference between a glass of clean, fresh spring water, and water from a mud puddle?

 

[dominikz]

The difference is so drastic that you’d need to be deaf not to hear it. Do you need special controls to tell the difference between a glass of clean, fresh spring water, and water from a mud puddle?

Then it should be easy to measure and demonstrate

 

[Dac-man]

Then it should be easy to measure and demonstrate

Tell me the measurement to use then. Not any that are used on this forum will cut it.

 

[777]

The PCB traces from power supply to op-amps has any 1.2 ohm !!! So, you can double that taces with very good sonic results. About the power supply, be aware to the power supply. E50 want 5v, it has not accept 4,9 or 5,1 volts. For undervoltage will display "IN ERR", for overvoltage will display "d3". Another power supply bottleneck come from current. The power supply must deliver 5v very quickly otherwise will see an erros warning. If you put a bigger capacitors, the power supply must has a bigger current. bigger power to loading the capacitors very quickly.

 

[dominikz]

Tell me the measurement to use then. Not any that are used on this forum will cut it.

If none of the (pretty extensive) measurements from this forum show any significant difference then I would again suggest placing standard controls on the listening test and see if these perceived differences in reproduction quality persist.

Human perception is easily tricked, and it is not something we can influence even if we're aware of it, here's one example I love:

If you trust your visual perception the circles move and change size. If you measure their location and size you will see they stay the same - but knowing this doesn't change the fact that you will still perceive them as changing. It is a completely automatic and involuntary perceptual mechanism.
Auditory perception is of course not exempt from such cognitive effects.

 

[Dac-man]

I have conducted 100’s of blind tests that have proven to many people that all of the measurements used here are only just a good starting point to achieving great sound. And it’s even possible to have horrible sound from a DAC with 120dB SINAD.

For me SINAD is only a QC test. I only use it to verify if a finished design meets spec. Because usually if the SINAD spec is met, the gear is performing as intended by the designer. But it’s not a guarantee that the sound will be good.

The only reason I bought this DAC is because the DAC chip is so integrated with built in LDO’s and voltage output. So it’s a hard one to mess up. And you can use an external power supply. So you’re not stuck with a garbage one that’s built in.