Douk Audio U2Pro Digital Interface Upgrade is Being Planned and Waiting for Your Advice!

[somebodyelse]

The DDC and oscillator clock upgrade topic is not in the same category as the op-amp rolling debate. With oscillators, the target variables are clearly defined -things like phase noise, jitter, ppm accuracy, and temperature stability- and the specifications and measured results are data that can be used directly for design validation and quality comparison.

By contrast, changes caused by swapping op-amps can also be measured, but they are difficult to summarize cleanly into a single metric like SINAD. On top of that, the interpretation becomes more complicated because results can vary depending on the measurement setup and conditions, so it’s not something that’s often handled well in the typical way communities discuss measurements.

You're making a couple of unverified assumptions here:

1. That improved performance of the oscillator will result in improved performance at the coax/toslink output.
2. That the assumed change in clocking performance at the coax/toslink will result in improved performance at the DAC output.

Point 1 depends on what happens between the oscillator and the output - how the output rate is derived from the oscillator. As you've previously noted in your blog, the 24.000MHz oscillator frequency is for the USB side, not for the audio clock side. So does Douk's method of generating the audio-side clocks actually benefit from the improved clock or not?

On point 2 we expect a good DAC to be more or less immune to timing errors like jitter on the digital inputs, so you'll only see a benefit on a poor DAC.

 

[Alan O]

if I connect an NVDIA TV Shield Pro by USB to the Douk Audio U2 Pro, should I use the UAC 1.0 or UAC 2.0 setting?

I don't know anything about the NVDIA TV Shield Pro, so I can't tell you for sure -- but I'd suggest you do what I did connecting my old Mac Mini.

As I understand it, UAC 2.0 can do high bit rates and is supported more by newer products, while UAC 1.0 is more compatible with older devices (and some that aren't specific audio devices) but is limited to 96kHz. If you try UAC 2.0 with a device that doesn't support it, you can get errors including drop-outs.

I started with UAC 2.0 but got occasional drop-outs, so I switched to UAC 1.0 and it's all working fine - and 96kHz is more than I use. So I suggest trying UAC 2.0 first, and switch to UAC 1.0 if you hear problems like drop-outs.

 

[pat355]

I don't know anything about the NVDIA TV Shield Pro, so I can't tell you for sure -- but I'd suggest you do what I did connecting my old Mac Mini.

As I understand it, UAC 2.0 can do high bit rates and is supported more by newer products, while UAC 1.0 is more compatible with older devices (and some that aren't specific audio devices) but is limited to 96kHz. If you try UAC 2.0 with a device that doesn't support it, you can get errors including drop-outs.

I started with UAC 2.0 but got occasional drop-outs, so I switched to UAC 1.0 and it's all working fine - and 96kHz is more than I use. So I suggest trying UAC 2.0 first, and switch to UAC 1.0 if you hear problems like drop-outs.

thanks

 

[Dmaumau]

Anybody upgraded the clock chip and can talk about the process and the results ?

@Douk Audio any useful info regarding the femto clock upgrade on the U2 ?

 

[A Very Private Zootopia]

Anybody upgraded the clock chip and can talk about the process and the results ?

Douk Audio U2 Pro의 오실레이터를 업그레이드 해보자!

0. Douk Audio U2 Pro, 오실레이터 꼭 업그레이드 해야 하나요? 필자가 보았을 때 Douk Audio U2 Pro는 기본 오실레이터를 업그레이드해야 그럭저럭 쓸만한 물건이 될거라고 생각한다. 그렇지 않구서야 필자가 쓰는 Topping D90III Discrete 정도의 USB 입력부가 U2 Pro 따위보다 훨 좋기 때문에 비교가 되질 않는다.그래서 필자는 기본 오실레이터를 제거하고 쓸만한 오실레이터로 교체 작업을 하고 있는데 이게 생각보다 만만치 않다. 1. 쓸만한 Dip-14 타입 오실레이터는 없나요? 일반...

grancartzoo.tistory.com

Here, I share my record of the struggles I faced while attempting to upgrade the oscillator of the Douk U2 Pro.

Unfortunately, I have not yet been able to personally hear the results of the oscillator upgrade.

Although I have not succeeded in completing the upgrade, my records may still help others understand the direction they should take.

However, I was able to realize that using the U2 Pro with DACs at the level of the Topping D90III Discrete and Gustard R26 may actually be a disadvantage in terms of sound quality.

Even when taking into account the price of the U2 Pro and the labor involved, purchasing a higher-quality DDC may ultimately be the better choice.

Since my blog is written in Korean, please use an automatic translation feature if needed.

 

[Schlippwhip68]

I have an issue with which I have reached out Douk Audio with. When using the Coaxial output and shutting down the PC which is connected via the usb, I get a huge amount of static crackle and popping. It’s silent over optical when shutting down.

Even after completely disconnecting the PC the crackle remains audible as the U2 trains connected to my dac via true Coaxial cable. I have tried different cables which had varied results but still the crackle remains, it’s not quiet either. It eventually dies down but sporadically will make a noise.

Do I have a faulty unit, it seems so. I have used it lightly and it is less than 6 months old. The optical is fine, no issues but the coaxial began to play up a month or two ago losing connection and now the ‘fire crackle’ on shut down.

It’s a shame as I prefer the coaxial using an Neotech 075ohm cable as opposed to the slightly less weighty sound of the optical. I bought it on Ali Express and the 90 day return is up so it looks like I might be stuck with it unless I can fix it. Not sure if Douk Audio offer a warranty on these but at least 1 year would have been fair.

Any ideas?

 

[somebodyelse]

I'd be suspicious of the DAC coax input. I don't see how the U2 could be responsible if the problem continues sporadically even when the U2 has been unpowered for some time. It looks like the coax and toslink outputs are driven from the same signal, so any fault would be in the final few components or tracks - connector, pulse transformer, U10 and the passives around it.

 

[Schlippwhip68]

I'd be suspicious of the DAC coax input. I don't see how the U2 could be responsible if the problem continues sporadically even when the U2 has been unpowered for some time. It looks like the coax and toslink outputs are driven from the same signal, so any fault would be in the final few components or tracks - connector, pulse transformer, U10 and the passives around it.

I think you may be right. It’s the iFi Zen One Signature that I am using. It’s fine when connecting the Fiio SR11 via coaxial, no issues at all. Connected to the U2 then the issues are as mentioned. Something between the two that is not obviously right.

 

[Douk Audio]

Anybody upgraded the clock chip and can talk about the process and the results ?

@Douk Audio any useful info regarding the femto clock upgrade on the U2 ?

Hello, regarding the U2pro crystal oscillator replacement issue, we have verified it, and the following is the updated information:
To replace the U2 Pro's oscillator, it needs to meet these specs:
24.000MHz/1.8-3.3V (active oscillator)
Can be SMD or through-hole (DIP)
SMD size: 5032
DIP spec: DIP14
Other than meeting those specs and size,the smaller the frequency tolerance (+ppm), the better. Choosing a high-precision TCXO would give a more noticeable improvement.
*We also tested fixed 3.3V,it can work normally as well. But we cannot guarantee that all 3.3-5V will work. But no worries, unmatched oscillator won't cause damage to the U2PRO, but only not work.

Our team has also been working on upgraded versions of the U2PRO temperature-compensated crystal oscillator and femtosecond crystal oscillator, and we expect to release these two new versions around April for all users to choose from. Please stay tuned.

 

[Douk Audio]

Would this apply to the Femto Oscillator as well? Many thanks!

Hi, sorry for my late reply. Yeah, it also applies to the Femto Oscillators.

 

[Schlippwhip68]

Hello, regarding the U2pro crystal oscillator replacement issue, we have verified it, and the following is the updated information:
To replace the U2 Pro's oscillator, it needs to meet these specs:
24.000MHz/1.8-3.3V (active oscillator)
Can be SMD or through-hole (DIP)
SMD size: 5032
DIP spec: DIP14
Other than meeting those specs and size,the smaller the frequency tolerance (+ppm), the better. Choosing a high-precision TCXO would give a more noticeable improvement.
*We also tested fixed 3.3V,it can work normally as well. But we cannot guarantee that all 3.3-5V will work. But no worries, unmatched oscillator won't cause damage to the U2PRO, but only not work.

Our team has also been working on upgraded versions of the U2PRO temperature-compensated crystal oscillator and femtosecond crystal oscillator, and we expect to release these two new versions around April for all users to choose from. Please stay tuned.

Thankyou.

 

[Schlippwhip68]

Hi, sorry for my late reply. Yeah, it also applies to the Femto Oscillators.

24.000MHZ Gold Plated Rectangular Temperature Compensated Crystal Oscillator Golden 3.3V-5V 0.1PPM

10KHZ-200MHZ. We Can Customize Any Frequency.

This Model Is An Upgraded Version Of The Rectangular Crystal Full-Size Temperature Compensated Oscillator "TCXO-DIP-14"

3.3V-5V Voltage Can Be Supported, High Accuracy ±0.1PPM.

This Crystal Oscillator Is Compared To The Ordinary Silver TCXO-DIP-14.

The Shell Uses Gold Plating Technology. Made Of Real Gold. Improved Frequency Temperature Stability. And Temperature Index.

Frequency And Temperature Stability Increased From Ordinary ±1PPM To ±0.5PPM

The Temperature Resistance Coefficient Is Increased From -20 ~ +70℃ to -40 ~ +85℃.

Frequency Range : 10KHZ-200MHZ

Output Mode : Square Wave

Frequency Accuracy : ±0.1PPM ( Max )

Frequency Temperature Stability : ±0.5PPM

Operating Temperature : -40 ~ +85℃.

Aging rate ：±1.0PPM/YEAR (Maximum)

Frequency Load Stability : Maximum ±0.1PPM When Load Fluctuation Is ±10%

Frequency Load Stability : Maximum ±0.1PPM When Load Fluctuation Is ±10%



These are the specs of the one I ordered as finding an TXCO clock that supports 1.8v is near impossible, its an incredibly outdated voltage requirement. 3.3v to 5v which are common readily available voltage rates, 1.8v to 3.3 or 5v is not!

Can Douk Audio provide an link to an retailer whom sells 1.8v-3.3v TXCO clocks?

 

[Alan O]

The Shell Uses Gold Plating Technology. Made Of Real Gold. Improved Frequency Temperature Stability. And Temperature Index.

I'm curious how gold plating it (with real gold, no less) would do either of those two things.

 

[Schlippwhip68]

I'm curious how gold plating it (with real gold, no less) would do either of those two things.

It certainly wasn’t a deciding factor for me either.

 

[Schlippwhip68]

A couple of what looks like TXCO SMD options here. The importance of PPM measurements is arguably not as important as handling jitter and the measurements thereof for audio apparently.

https://www.mouser.co.uk/ProductDetail/ABRACON/ATX-H13-F-24.000MHz-G50-T?qs=HoCaDK9Nz5dJC4Pnpdgzqg%3D%3D

 

[MCH]

Regarding smd specially, I wonder if it even makes sense to chase the finest performance and then hand solder the part....

 

[Schlippwhip68]

Regarding smd specially, I wonder if it even makes sense to chase the finest performance and then hand solder the part....

Good point.

 

[Alan O]

It certainly wasn’t a deciding factor for me either.

I'm sure it wasn't. But the issue for me is that if they'll make up BS about gold plating to sell the thing, how far can we trust them on the claimed spec?

 

[Roland68]

I'm sure it wasn't. But the issue for me is that if they'll make up BS about gold plating to sell the thing, how far can we trust them on the claimed spec?

Several DIY enthusiasts, including some from China, have already tested these gold-plated quartz crystals, and the frequency stability was around ±25 ppm, far from the advertised ±5 ppm or ±1 ppm. The 0.1 ppm claim is simply laughable.
These are likely just cheap knock-offs of Rakon crystals, or discarded ones that didn't meet the specifications.

Compared to the original crystal, the performance is likely worse.

Since apparently too few people are buying this junk, these crystals are now appearing at increasingly lower prices, sometimes for €10 or €15 with a populated circuit board.
It's a shame for any user who falls for this and loses their money.

 

[Leporello]

Several DIY enthusiasts, including some from China, have already tested these gold-plated quartz crystals, and the frequency stability was around ±25 ppm, far from the advertised ±5 ppm or ±1 ppm. The 0.1 ppm claim is simply laughable.
These are likely just cheap knock-offs of Rakon crystals, or discarded ones that didn't meet the specifications.

Compared to the original crystal, the performance is likely worse.

Since apparently too few people are buying this junk, these crystals are now appearing at increasingly lower prices, sometimes for €10 or €15 with a populated circuit board.
It's a shame for any user who falls for this and loses their money.

View attachment 509276View attachment 509277

Maybe these nominally low grade oscillators actually make the sound more analog? Nature is imperfect, therefore imperfect components usually sound more natural.