Review and Measurements of Hifime UDA38Pro DAC

[restorer-john]

I paid with paypal. What is their policy on such things?

Product was misrepresented, no question. Not as advertised.

 

[xineis]

Let's hope this is a case of a DOA unit. These measurements remind me of the early engineering lab courses, where we would intentionally cause the sine waves to distort.

 

[andreasmaaan]

I have one of these - got it a few weeks ago - and don't think these measurements are representative of what I have - I think the test unit could be defective. The only issue I found is when it was powered only by USB, using line out, I was getting clipping on very loud parts of the music when at 100% (it is a variable line out). Once I added a 12 V power source, I have not heard anything negative.

I usually just listen with speakers. So, I just tried with usual test songs with my headphones (beyerdynamic DT 770 - 250 Ohm) and at 3 o'clock (12 v powered), they were quite loud (too loud) with no audible distortion. I turned up to full (i.e up to 11) for just a bit (way too loud), and did not hear any distortion. I played some songs with prominent bass and was not able to get any audible low end distortion.

In subjective listening tests, I can't hear any real difference between the UDA38Pro and the Topping NX4 DSD. Also, the UDA38Pro doesn't use many of the additional features (e.g. filters) of the ES9038pro and if DSD (or DSD upsampling) is important to you, it is limited by its USB interface to DSD 128 (which the Topping NX4 DSD and D50 USB interfaces are capable to 512).

I'm also surprised by the measurements TBH. I returned my UDA38Pro after a few days of listening because I didn't think it sounded like it was worth $250. But it didn't sound outrageously bad, just a bit hard and brittle to me, comparing slightly unfavourably to my partner's MacBook Pro that I tested it against (that was through a pair of Audio Technica ATH-M50x). Still, I didn't listen for long because I immediately didn't get a good feeling from what I was hearing.

I guess either @amirm got a defective unit, or I fortuitously kept my brief listening within that window in which THD+N stays below 0.1%.

 

[andreasmaaan]

PS lol it just occurred to me that there is so much distortion at 2V output that the intermodulation products of the fundamental and its harmonics are themselves high enough in level to outperform (and I obviously mean that tongue-in-cheek) just the harmonics of a typical DAC at full output

 

[CTRLM]

LOL - I have been intrigued by the wave of cheaper DACs and I have one of these and a Topping D50 on their way to me for a bit of a comparison in my setup with each other and my more expensive DACs. I guess these measurements don't change anything as far as expectation goes, as it was low to begin with and they are so bad that there could be a defect in the tested unit? We will see......

 

[maxxevv]

It seems like its a DOA item. Likely the power regulators or opamp chips in the circuit are compromised.

 

[Dialectic]

Purchased it beginning this year, been listening to it for about two weeks, wasn't totally overwhelmed with the sound and then when i got the Topping D50 never used it anymore.
These measurements clearly demonstrate why i preferred the Topping D50 over the UDA38pro. And i say preferred because the Topping died on me, halfway in a song it began to stutter and crackle and went dead. Currently using a Gustard X20pro (2 x ES9028pro) and this sounds 100 times better then the UDA9038pro.

As Amirm said, don't buy a product just because it has a top of the line dac chip, implemention is much more important.

This is the third dac Amirm tested with ES9038pro that goes south.
Is this particular ES chip so hard to implement properly?

I'd be most curious to see how that Gustard performs on the bench...

 

[Patrick1958]

I'd be most curious to see how that Gustard performs on the bench...

You and me both

 

[miero]

Maybe it is a defective unit, but it has passed their quality output "check", so I'd be very suspicious if they sent a "replacement" unit, which would be measuring better.

If this item passed, then other more worse items will pass too.

It would be interesting to find out, why it is such broken.

Amirm, can you show us photos of the PCB? I'm interesten which and how many voltage regulators did they use...

 

[oliveoilthansnakeoil]

WOW... Maybe this is what Schiit fanboys expect from chi-fi.
I almost bought this dac. But my DX7 is coming to me.

 

[amirm]

Amirm, can you show us photos of the PCB?

I am hesitant to open it until I hear from them. Hate for them to blame this on me.

I found a picture of it but it is low-res:

 

[rebbiputzmaker]

It seems like its a DOA item. Likely the power regulators or opamp chips in the circuit are compromised.

Sounds like it is defective/damaged in shipping.

 

[miero]

I found also another one, but not much better:

Source: https://hifimediy.com/news/New-Sabre-ES9038PRO-Chip

 

[amirm]

I found also another one, but not much better:

That one doesn't have RCA outs. I wonder if that addition was an afterthought and the cause for problems I am seeing.

 

[έχω δίκιο]

I paid with paypal. What is their policy on such things?

You have 180 days to contact the seller through PayPal's resolution center. The only disputes I know that they take are items not received or items "Significantly Not as Described."

One published spec on the UDA38 Pro is THD+N of -116dB @ 32ohm, which sure doesn't look like what you measured. Other than that, there's very little concrete with regards to performance. I'd claim that it's defective in design, construction, and/or components based on its performance, and that it specifically does not meet the stated THD plus noise spec they published. It doesn't do what it is advertised to do: Accurately convert digital audio to analog.

 

[Final]

Jesus

Definitely a record.... And at such high price....

Couldn't bring myself to press like on your post. I guess condolences are more appropriate

 

[έχω δίκιο]

Definitely a record.... And at such high price....

Someone has been advertising an Audio-GD NFB-2 on craigslist near me for $200. It's got all the signs that it could contend for that same record. From the Audio-GD web page:

• "Neutral but smoothness sounding..."
• "The ACSS is a non-feedback technology made with discrete amplifiers. Most people know the global feedback design can offer better specs in test measurements, and non-feedback can't do well in test measurements but can offer better sound for the human's ears."
• "That is why it is easy to find hi-end grade gears maybe without the best chips or amp stages, but with plenteous dedicated DC supply circuits."
• "...better human sound and neutral."
• "9 types digital filter users setting"

It's the ideal review product:

1. Audio-GD has an even more cult-like following than Schiit Audio.
2. If it performs poorly, Audio-GD fans can keep asking you to open it up and move jumpers to select one of the other nine digital filtering settings and rerun various tests, probably one at a time.

 

[yummy]

hey, I've said clearly before, don't buy these toys, now you (pural) wasted Amirm's money

sinners, apologize to Amirm, and do more homewroks

 

[έχω δίκιο]

hey, I've said clearly before, don't buy these toys, now you (pural) wasted Amirm's money

sinners, apologize to Amirm, and do more homewroks

I disagree. I think it's great when amirm's testing reveals overpriced, underperforming junk, as well as when it highlights budget equipment that performs to the highest standards.

 

[amirm]

Good news. The company responded to my inquiry and is asking their engineers for a response. They also said a refund would not be a problem at all.

FYI their response went into my junk folder. So perhaps the last inquiry went there too. If so, I apologize for leaving the impression that they may not be responsive to contacts.