L.K.S. Audio MH-DA004 Dual ES9038PRO DAC - Van Damme's double impact?
Jun 10, 2019 at 7:57 AM Post #2,131 of 4,419
Things to aid with harshness reduction while allowing hi-resolution playback with the LKS:

- Get rid of the Crystek 575 clock. Mine now has a TXC low phase 100MHz crystal and is a lot less bright with better fine detail
- Implement one of the internal RFI/EMI shielding ideas presented here, especially around the digital toroid
- Place the LKS on good isolation as it is very susceptible to vibrations (I use an Electric Beach shelf with Black Ravioli feet)
- Use a regenerated or very well filtered mains feed with a good power cable
- Select your interconnects carefully (balanced sound better with the LKS) and don't use silver cables
 
Jun 10, 2019 at 2:45 PM Post #2,132 of 4,419
In this forum many times I have read that the sound of LKS 004 can be aggressive, for the oscillator Crystek 575. I've never heard of it before. The Ethernet cable between the switch and SOtM SMS 200 Ultra was AQ Cinnamon 0.6m, up to the Sinxger SU-1 USB cable AQ Diamond and LKS 004 HDMI cable AQ Cinnamon.
I wanted to try the HDMI 0.3 m hoping for even smaller jitter on DSD material. I got Wireworld Ultraviolet 7 0.3 m. At the same time, I got the WW Starlight 8 Ethernet cable. Both cables changed the sound, pointing the middle area to the clay. Jitter did not even decrease or increase.
But the sound image moved forward. The sound in the Middle Ages was acutely aggressive.
WW cables are not inferior to AQ but point out the mid range and the sound on these frequencies gets aggressive.
On the other hand, AQ has more transparency and air, the sound is retracted. There is no aggressiveness, but it is noticeable that the sound in the middle range is slightly thinner. The violin is very resolute, but there are fewer differences when a musician tends to pull more gentle and harder.
Although I've already purchased the Crystek 950X, I have not installed it yet, because my engineer has no time. Now I've decided to hurry him a little.
 
Jun 11, 2019 at 12:33 AM Post #2,133 of 4,419
In this forum many times I have read that the sound of LKS 004 can be aggressive, for the oscillator Crystek 575. I've never heard of it before. The Ethernet cable between the switch and SOtM SMS 200 Ultra was AQ Cinnamon 0.6m, up to the Sinxger SU-1 USB cable AQ Diamond and LKS 004 HDMI cable AQ Cinnamon.
I wanted to try the HDMI 0.3 m hoping for even smaller jitter on DSD material. I got Wireworld Ultraviolet 7 0.3 m. At the same time, I got the WW Starlight 8 Ethernet cable. Both cables changed the sound, pointing the middle area to the clay. Jitter did not even decrease or increase.
But the sound image moved forward. The sound in the Middle Ages was acutely aggressive.
WW cables are not inferior to AQ but point out the mid range and the sound on these frequencies gets aggressive.
On the other hand, AQ has more transparency and air, the sound is retracted. There is no aggressiveness, but it is noticeable that the sound in the middle range is slightly thinner. The violin is very resolute, but there are fewer differences when a musician tends to pull more gentle and harder.
Although I've already purchased the Crystek 950X, I have not installed it yet, because my engineer has no time. Now I've decided to hurry him a little.

That should not be happening, the symptoms you describe is an indication RF noise from the Singxer is leaking into the LKS004.
I find that to be the major failing of the stock SU-1.

You are relying on the expensive HDMI cable to filter out some of this junk, neither options you tried seemed to do a complete job.
It was never designed to do this, rather the cable stops external noise from getting in not reduce noise that is already there

Add a few ferrite RF absorber sheets inside the Singxer in the area after the isolator.
I find that this cleans things up.

Fix the problem at the source.

This is how I can use low cost cable from Monoprice and it gets the job done, no difference with cables costing 10-20x more.

From my perspective the 2 exotic HDMI cables you mentioned is sufficient to pay for 1 HQPlayer license or 1 year of Roon.
 
Jun 11, 2019 at 12:54 AM Post #2,134 of 4,419
It was only a small improvement but I found the cheaper Blue Jeans Series-FE HDMI outperformed the more expensive Audioquest Cinnamon HDMI. The BJ is outstanding value for its low cost.
 
Jun 11, 2019 at 10:19 AM Post #2,135 of 4,419
That should not be happening, the symptoms you describe is an indication RF noise from the Singxer is leaking into the LKS004.
I find that to be the major failing of the stock SU-1.

You are relying on the expensive HDMI cable to filter out some of this junk, neither options you tried seemed to do a complete job.
It was never designed to do this, rather the cable stops external noise from getting in not reduce noise that is already there

Add a few ferrite RF absorber sheets inside the Singxer in the area after the isolator.
I find that this cleans things up.

Fix the problem at the source.

This is how I can use low cost cable from Monoprice and it gets the job done, no difference with cables costing 10-20x more.

From my perspective the 2 exotic HDMI cables you mentioned is sufficient to pay for 1 HQPlayer license or 1 year of Roon.
Maybe you're right about expensive cables.
I managed to conclude that with WW cables, AQ Diamond USB is too much silver. Now Belden is better, less agressiveness of midrange tones.
I tried to exclude the Singxer from the combination and realized that it stopped working the USB port on the DAC. Other inputs are working. The device has not been modified so far.
The device is under warranty but is purchased from another EU country via eBay. I'm not sure I will make a guarantee.
Which USB input failure is most common?
Thanx.
 
Jun 12, 2019 at 1:40 AM Post #2,136 of 4,419
Maybe you're right about expensive cables.
I managed to conclude that with WW cables, AQ Diamond USB is too much silver. Now Belden is better, less agressiveness of midrange tones.
I tried to exclude the Singxer from the combination and realized that it stopped working the USB port on the DAC. Other inputs are working. The device has not been modified so far.
The device is under warranty but is purchased from another EU country via eBay. I'm not sure I will make a guarantee.
Which USB input failure is most common?
Thanx.

Double check to see if the Amanero shows up when you plug it in on a Windows PC/laptop.
SMS-200 also reports the Amanero under the DAC Info if it is working.

If none of the above works, see if warranty replacement is possible.

Sometimes the flash image goes bad, re-flashing using the Amanero OEM117 tool might fix it.
 
Jun 12, 2019 at 1:50 AM Post #2,137 of 4,419
Seller promised to send a new USB. I hope that USB will have the latest firmware.
 
Jun 12, 2019 at 5:37 AM Post #2,138 of 4,419
That should not be happening, the symptoms you describe is an indication RF noise from the Singxer is leaking into the LKS004.
I find that to be the major failing of the stock SU-1.

You are relying on the expensive HDMI cable to filter out some of this junk, neither options you tried seemed to do a complete job.
It was never designed to do this, rather the cable stops external noise from getting in not reduce noise that is already there

Honestly SU-1 is overrated and I own one myself. It's much cheaper to get a MATRIX X-SPDIF 2 USB, and just use it without external power but only USB for both data and power, using a laptop on battery. Super low noise, no mods needed and it already has Assusilicon clocks. Incredibly hassle-free. The Singxer SU-6 is another contender but it is much more expensive for essentially same performance (SU-6 uses super capacitors for power which seems unnecessary/gimmicky, but cool).
 
Jun 12, 2019 at 10:52 PM Post #2,139 of 4,419
Honestly SU-1 is overrated and I own one myself. It's much cheaper to get a MATRIX X-SPDIF 2 USB, and just use it without external power but only USB for both data and power, using a laptop on battery. Super low noise, no mods needed and it already has Assusilicon clocks. Incredibly hassle-free. The Singxer SU-6 is another contender but it is much more expensive for essentially same performance (SU-6 uses super capacitors for power which seems unnecessary/gimmicky, but cool).

Below is the LKS004 I2S input pinout.
Pin14 in the diagram lets the DAC know that a DSD stream is present, it is non-standard and specific to LKS.
lks004_i2s.png

Next is the Matrix X-SPIDIF I2S output pinout from the manual downloaded from their website.
Note the DSDOE pin missing, instead we have an unconnected pin, left side 3rd from the bottom

matrix2_i2s.png

This is not fully compatible with the LKS004, as Native DSD mode can never be engaged.
With the Matrix, the LKS004 cannot be used to its full potential.

This is might be fixable with a new firmware flash if they ran a copper trace from that pin back to the FPGA
I would not use the Matrix on the LKS until there is specific clarification from the manufacturer on compatability.

PS
Matrix raised prices, it is $10 cheaper than the SU-1 on Amazon, I would stick with the SU-1, it is the known quantity here.
 
Last edited:
Jun 13, 2019 at 1:36 PM Post #2,140 of 4,419
USB was never designed for such hi-res streaming, no wonder it's fraught with issues.

I'm using the I2S input with streaming from Qobuz Studio and the sound quality is superb. I don't care about DSD and I know of no streaming service that provides it.
 
Jun 13, 2019 at 2:02 PM Post #2,141 of 4,419
USB was never designed for such hi-res streaming, no wonder it's fraught with issues.

I'm using the I2S input with streaming from Qobuz Studio and the sound quality is superb. I don't care about DSD and I know of no streaming service that provides it.
Following that thinking I2S was never designed for travelling more than a couple cm between chips
 
Jun 13, 2019 at 6:47 PM Post #2,142 of 4,419
Following that thinking I2S was never designed for travelling more than a couple cm between chips
USB requires both interfaces at both end that has to be able to compress and send the packages and then uncompressed and process then convert the packages into I2S.

After the USB host to USB slaves interfaces .....is I2S.

Therefore, going to I2S, you cut down many different extra processing. Not to mention that the USB interfaces in LKS004 suck...Amanero suck, period. You don’t have to take my word for it, it is better to find out for yourself. You are also correct regarding I2S that isn’t made to travel further than a few cm, and that is why many people opting for HDMI as it was designed and engineered the best for I2S to travel at length. However, the shorter, the better. No one is stopping you to put one device next to LKS-004 I2S ports and having a 4-5cm HDMI cables or i2s cables

The final conclusion is that going to I2S is the best if you don’t need DSD. The LKS-004 can take in DSD1024 over I2S btw....and there is no USB DDC that is capable of such yet....talk about I2S and it great performances eh ? The only limitation is that I2S chip as the interface has never been exposed to such demands, and hence they are stuck at 192Khz. Until the demands are enough, you will see a lot more I2S that is able to do ways more than USB can handle
 
Last edited:
Jun 13, 2019 at 8:53 PM Post #2,143 of 4,419
USB was never designed for such hi-res streaming, no wonder it's fraught with issues.

I'm using the I2S input with streaming from Qobuz Studio and the sound quality is superb. I don't care about DSD and I know of no streaming service that provides it.

I think you got it the other way around.

Your I2S streamer is based on the Raspberry Pi from your sig

Below is the system block diagram of the RPI3.
https://www.element14.com/community.../blog/2017/01/16/raspberry-pi-3-block-diagram
pi3-block-diagram-rev4.png

The ethernet interface is a USB device plugged into the USB port on the BCM2837 SoC.

Taking a closer look at the LAN9514, we see this ethernet device is a dongle plugged into an internal USB hub inside the LAN9514
http://ww1.microchip.com/downloads/en/DeviceDoc/00002306A.pdf

lan9514.png

Your streamer box has not one but two USB links before it reaches the SoC
The presence of USB in your streaming path does not seem to affect its performance as you seem quite happy with the end result.
 
Jun 13, 2019 at 9:18 PM Post #2,144 of 4,419
The final conclusion is that going to I2S is the best if you don’t need DSD. The LKS-004 can take in DSD1024 over I2S btw....and there is no USB DDC that is capable of such yet....talk about I2S and it great performances eh ? The only limitation is that I2S chip as the interface has never been exposed to such demands, and hence they are stuck at 192Khz. Until the demands are enough, you will see a lot more I2S that is able to do ways more than USB can handle

There is no 192kHz restriction on the LKS I2S input. Both the LKS004 and LKS003 with handle DSD512x48, this equates to 768kHz. This is with the Singxer SU-1
This is only a consideration with devices like the Pink Faun I2S card which cannot go higher.

There is no need to give up on DSD if you want to use I2S, SU-1 is an example.

This whole issue of I2S vs USB is quite overblown @MartinWT quite happily using this I2S network streamer and that thing used a USB ethernet dongle connected via a USB hub.
 
Last edited:
Jun 14, 2019 at 2:05 AM Post #2,145 of 4,419
Can the latest Amanero Combo work with the DSD 512? I'll get a new one, because old Amanero does not work. Someone mentioned the new firmware with which DSD 512 works. The LKS Web site in description 004 DAC continues to support DSD 512 support only through Windows.
If that worked well, then I might need to buy the LT3042 and connect the SMS 200 Ultra directly to Amanero?
 
Last edited:

Users who are viewing this thread

Back
Top