[mb3]

Info about the ess9038pro chip filter settings.

 

[b0bb]

I found this mod worked well for me - placing one heatsink across two transistors. The unit is now significantly cooler. see https://www.jaycar.com.au/to-220-clip-on-heatsink-6043-type/p/HH8503

That looks good a bit more metal than the one I used.

 

[mb3]

yes they were pretty good - and also insulated so no problem with shorting

 

[vick_v]

yes they were pretty good - and also insulated so no problem with shorting

What did you do with the snap-on ridges on the flat side? It looks like those would greatly reduce the contact surface.

 

[mb3]

What did you do with the snap-on ridges on the flat side? It looks like those would greatly reduce the contact surface.

I just used a flat head screw driver and fine pliers to flatten out the ridges - I then put a little kapton tape to cover that side of the heat sink.

 

[mb3]

I recently purchased the lks 004 amanero usb version. This is a very brief summary of my experience with setup, which I hope would be helpful to those interested.

It's pretty good to have the first dual ess9038pro chip dac. Sound quality is excellent, no doubt.

Setup took a little research and time.

For playback from PC by usb, driver install is needed - amanero 1059 for win10, then pc audio output selected as amanero usb. If this is not done correctly, then I get no sound at all.

For playback - foobar.

DAC filter setting, info shown above - usually set to minimum phase fast roll-off.

For DPLL settings - I've been unable to obtain specific info from either ESS or LKS about what the DPLL setting equals for the ess9038pro chip, and as it follows, for the LKS 004. From my limited experience, as expected, DPLL 1 provides the best sound quality by far.

For PCM and DXD file playback - I've had no problem using DPLL 1 for all file types and sample rates (LKS shows PCM). Playback of all PCM and DXD files and also DSD to PCM (set in foobar) was perfect, no matter what, with DPLL 1 setting.

Note the following discussion only relates to DSD file playback. For DSD files played back in native format (LKS shows DSD), DSD64 files with DPLL 2-5 setting is usually pretty good. For DSD128 or DSD256 results vary - some files play stably at DPLL 3. Others have 'static' even at DPLL 15.

Playback of DSD files seems extremely sensitive to a variety of poorly defined factors. From my experience this static issue for native DSD file playback is a real issue. Maybe this is why so few DACs offer this feature, or limit playback to DSD64. To clarify - by static I mean anything from, the snare drum in the left speaker sounds distorted (but sounded clear on the PCM file) to hearing white noise only.

Causes for dsd file playback issues can be summarized along these lines,
1. Power supply factors - to the lks itself. Due to the household wiring, eg. brought on opening the refrigerator door, using the washing machine, etc. Unfortunately, I found that using a line/surge protector to power the lks actually increased the chance of static noise during DSD playback.

2. Computer and usb related noise - there are many guides about optimizing a computer for audio playback, I found that making a few changes was worthwhile, see the yt video below as an example of how to do this. In general it seems that using the computer for anything other than music playback can be detrimental. There are various reports about disabling computer features, eg. i7 turbo mode, and generally disabling computer functions until windows barely works properly - I'm not sure if taking things to this level helps. Maybe using a laptop powered by battery, or using a sotm or antipodes device would give better results. I did find that using a usb 'decrapifier' probably reduces the occurrence of static.

3. Software factors - there are a variety of programs available for audio playback, so probably there is no 'best' option. I found foobar was best for me. In terms the setup that allowed dsd files to play - after trying several different configurations... I found best results using foobar set to ASIO playback, using ASIOProxyInstall-0.9.4, and foo_input_sacd-0.8.4 (links below). This allowed bit perfect DSD file playback for all file types and sample rates.

Then the following changes to foobar probably made a minor improvement in sound quality. I set ASIO output with around 10000ms buffer, then set to no replay gain, use 64 bit drivers, playback with priority, and use full file buffering from memory. Use a 40kHZ filter to limit artifact noise intrinsic to dsd files - which could in theory damage equipment.

4. Problems with the music introduced during the recording / processing +/- upsampling process seem amplified in DSD files.
Files recorded in PCM and up-sampled to DSD seem to have a higher rate of static, as expected.

5. Issues with the design of the device - eg. temperature, dual chip sync, dpll setting, power supply isolation, not allowing the usb input to touch the case. Ensuring optimal component temperature seems important. There are reports recommending both increasing and decreasing device temperature being helpful - allowing the device to warm up vs. keeping the device cool by using heat sinks on the transistors or keeping the lid off. Who would know? No way to be sure with the first device of its type.

Having tried to address several factors to improve dsd file playback, I'm not too sure it matters anyway. At the end of the day, I've found that native PCM and DXD file playback is outstanding, and setting DSD files to playback in PCM format will provide excellent results.

Further info:
Latest amanero drivers; https://amanero.com/drivers/
ASIO proxy - I found this was needed; https://sourceforge.net/projects/sacddecoder/files/foo_dsd_asio/
Foobar asio output support; http://www.foobar2000.org/components/view/foo_out_asio
Foobar sacd component; https://sourceforge.net/projects/sacddecoder/files/foo_input_sacd/
Computer setup example; search 'how to optimize Windows 10 anniversary update'

 

[justbenice]

I have a oppo 205 and it ES9038PRO DAC chip is not too hot. So may be there are something wrong with the LKS 004 ?

 

[mb3]

Cheers, good to know - board design would be a factor. Transistor heat sinks solved this issue for me. Out of interest - how does the oppo handle setting dpll and also native dsd file playback back?

 

[abartels]

Hi b0bb,

Ever tried ultracaps? Maybe take a look at this thread, it's very interesting.
https://www.head-fi.org/f/threads/r2r-da-m1-ultracap-powered-build.839007/page-4#post-13466545

I have a Kali reclocker powered with ultracaps, and my AK4495 now also has all it's voltagerails (6x) powered by ultracaps.

Those psu's are LDO-free design, just needs careful calculation of runtime ultracaps to stay between voltage tresholds.
This gives us a pure DC supply which in my opinion outperforms EVERY conventional psu design.

These designs have lots of relay switching (for my dac 26 relays each ten minutes), but switching noise can be kept to a minimum within enclosure with little damping.

Since loading circuit is galvanic isolated from running circuit, there's no need to take care of clean power to feed and charge the modules, a laptop adapter is all you need.

Soundwise, how do those ultracaps perform? You have to listen to them to believe it. It just combines the best of the best of both worlds (analog & digital).

Since I first started to use them in psu for my Kali reclocker, I allready was convinced I wouldn't go back to LT3042 with R-Core and EMI/RFI filtering.
Since powering my whole dac, eh, networkstreamer it should be called, with ultracaps, I never COULD go back to conventional linear psu's, that's for sure.

Sad thing is, if I wanted to power my pre-amp with ultracaps in an audiophile manner, I would need at least 96 BIG ultracaps (350F Maxwell - D-Cell format), which would result
in an external enclosure of 430mmx120mmx410mm (lxhxd). But I am really thinking on building this in the future.


Just give it a try b0bb, you will be VERY pleased, promised

Cheers,
Alex

 

[penguin69]

Hi b0bb (and others)

Any further modding to the 004 to report?

I have been out of the loop for a good while due to bad health, but I hope to get a soldering iron back in my hand in the next couple of months.

pat

 

[fjc36]

Have anyone compare to or own a OPPO Sonic DAC?

 

[b0bb]

Hi b0bb (and others)

Any further modding to the 004 to report?

I have been out of the loop for a good while due to bad health, but I hope to get a soldering iron back in my hand in the next couple of months.

pat

The stock 004 is quite good, SQ wise several notches above the 003, even my modded one, it has caused a degree of complacency

I am almost done with house remodelling which has been taking almost all of my freetime. This is almost done so I will have time later in the year to add the OCXO

One of the things I added in the remodel are new dimmers for LED lighting which use reverse phase dimming aka IGBT dimmers.
Until LED lamps showed up, these were in the 4 figure price range used in stage and theater lighting.

These put out a lot less RFI compared to the older SCR or Triac dimmers,and therefore less RFI junk in the house wiring.

The 004 SQ went up by half a notch which was unexpected.

The 004 seems very sensitive to RFI, I already have 2 RFI filters in line going to the DAC in addition to the RFI filter on the old Triac dimmer.

 

[b0bb]

Hi b0bb,

Ever tried ultracaps? Maybe take a look at this thread, it's very interesting.
https://www.head-fi.org/f/threads/r2r-da-m1-ultracap-powered-build.839007/page-4#post-13466545

I have a Kali reclocker powered with ultracaps, and my AK4495 now also has all it's voltagerails (6x) powered by ultracaps.

Those psu's are LDO-free design, just needs careful calculation of runtime ultracaps to stay between voltage tresholds.
This gives us a pure DC supply which in my opinion outperforms EVERY conventional psu design.

These designs have lots of relay switching (for my dac 26 relays each ten minutes), but switching noise can be kept to a minimum within enclosure with little damping.

Since loading circuit is galvanic isolated from running circuit, there's no need to take care of clean power to feed and charge the modules, a laptop adapter is all you need.

Soundwise, how do those ultracaps perform? You have to listen to them to believe it. It just combines the best of the best of both worlds (analog & digital).

Since I first started to use them in psu for my Kali reclocker, I allready was convinced I wouldn't go back to LT3042 with R-Core and EMI/RFI filtering.
Since powering my whole dac, eh, networkstreamer it should be called, with ultracaps, I never COULD go back to conventional linear psu's, that's for sure.

Sad thing is, if I wanted to power my pre-amp with ultracaps in an audiophile manner, I would need at least 96 BIG ultracaps (350F Maxwell - D-Cell format), which would result
in an external enclosure of 430mmx120mmx410mm (lxhxd). But I am really thinking on building this in the future.


Just give it a try b0bb, you will be VERY pleased, promised

Cheers,
Alex

Been looking off and on at the thread mentioned, the LKS 004 consumes far too much power to make an ultracap PSU a practical reality.
I have another DAC (JOB/Goldmund DA96) which uses about 5W that is a candidate for the ultracap PSU.

However I find the dependence on relays clunky and affects overall reliability.
When DC is switched, the arc causes oxide to build up on the switch contacts.
On the other thread, you reported the contact resistance building up on some of the relays you tried.

Waiting to see if someone puts out a decent charge controller and voltage regulator for these caps.
A LDO could be modified to work for this purpose if it supports constant current output.

I think in the next 12-24 months we will be seeing much more refined ultracap PSUs hitting the market, hopefully without the relays.

 

[mb3]

The stock 004 is quite good, SQ wise several notches above the 003, even my modded one, it has caused a degree of complacency

I am almost done with house remodelling which has been taking almost all of my freetime. This is almost done so I will have time later in the year to add the OCXO

One of the things I added in the remodel are new dimmers for LED lighting which use reverse phase dimming aka IGBT dimmers.
Until LED lamps showed up, these were in the 4 figure price range used in stage and theater lighting.

These put out a lot less RFI compared to the older SCR or Triac dimmers,and therefore less RFI junk in the house wiring.

The 004 SQ went up by half a notch which was unexpected.

The 004 seems very sensitive to RFI, I already have 2 RFI filters in line going to the DAC in addition to the RFI filter on the old Triac dimmer.

Just checking if you could be so kind as to let us know which RFI filter you are using? This one looked ok to me. http://www.poweronaustralia.com.au/product/eaton-psfi/

Thank you

 

[abartels]

@b0bb,

I can understand your concerns about relay reliability, but you know as I know that there are very reliable relays which can do the job perfectly, like the Takamisawa NA12W-K or others.
The gold contacts really do not oxidate that much as you mentioned, and, they are bifurcated, and, those are DPDT, so you can parallel those and reduce resistance to 25 mOhms.

Those Chinese relays, used at the boards I bought, are far from reliable, and this isn't related to relay reliability in common, they just are very cheapo and crappy.....

To prevent oxidation within relays, they often use a so called "WET circuit", which simply is a bypass film cap across switching contacts, in this case 100nF to 1uF should do the trick.

Quote from http://control.com/thread/963420679 :

"> What is Oxidation of relay contacts and what can be done to prevent it ??

For both relay and metallic switch contacts for low current level (e.g. logic) You have two choices.

1) use Au plated contacts

2) Use a film capacitor which has low ESR to "wet" the contacts. Unless the contact is specifically designed for dry contact switching (eg telephony sealed switches) you MUST use a cap to make it a "WET" circuit. This means the current during switching is high enough to burn through a partial oxidation layer to lower the contact resistance ESR using the low ESR and stored charge of a capacitor to dump the current across the contacts. The capacitor must be close to the switch. Suitable values of Film caps start at 10nF and for say 30Amp relays use 10uF tantalum or low ESR alum. This will keep the contacts clean when used enough. Normally current must be 10% of rated current to prevent oxidation. Although the transient from a low ESR cap will be much higher than this, the duration will be so short that there is not thermal rise except at the surface oxidation layer enough to burn off the ofxide.

Tony Stewart EE '75 (retired PEng)"



When looking at the Takamisawa specs, Electrical lifespan is about 500.000 operations at 1A-30VDC
It's mechanical lifespan is about 100.000.000 operations.

In my case, switching every 10 minutes, relays should exactly last as much as long as the 100F ultracaps which, if used 24/7, would last for 18 years of continuous playing.........

For my new design I will use those Takamisawa's, and put them in simple DIP-14 IC-sockets, which is very handy to replace a relay in case of troubles.

There are some people who are designing charge controllers, which just makes it easier to use (no runtime calcultaion because of voltage sensing), but still would need relays, mechanical or SolidState, or mosfets to switch.
You know as much as I know, voltage drop of SSR's or mosfets ruins ESR of ultracaps, which makes all those upcoming "better" designs worse sounding.
And, using voltage regulators also ruins the perfect audio characteristics of those ultracaps.

I can understand if you wanted to use utracap psu's for your LKS, it wouldn't be a practical design, but, since we are no commercial guys, and want the best of the best available SQ,
this should be of some interest to you.

Just try to power a simple USB-I2S interface with ultracaps, I am very sure you'll get interested................


Cheers,
Alex