LKS Audio MH-DA003

[demos]

I take off the cchd575 on the 004, put in a 950x, there is significant improvement on the 950x, WOW

 

[Whitigir]

I take off the cchd575 on the 004, put in a 950x, there is significant improvement on the 950x, WOW

I was thinking of doing this. Tell me more about improvements. The 950x is twice as good in spec

 

[demos]

I was thinking of doing this. Tell me more about improvements. The 950x is twice as good in spec

The 950x bring improvement on 3D imaging, instrument separation, sound staging and presentation, sound more natural and accurate, Treble is more smoother.
Like @b0bb say "004 falls apart on complex highly compressed tracks", this is what it is on 575, the 950x has better layering, 3D imaging, instrument separation.
It sound Slightly warmer on the 950x, bass has better texture, more extended

Really wondering how it would bring on a Pulsar clock, if there is 100mhz available

look like the 4.8V is way too hight for clock, it works on the crystek clock, looking to lower the voltage.

 

[Whitigir]

Lks004 uses LT1763. Isn’t this capable of 500mA ? That means OCXO will work ? Actually, at turn on, it needs 3.6W
https://www.mouser.com/ProductDetai...JVRBSBQ7qx1B6RYgm%2bL8c/KIWJXzTGwc7Dapf%2bw==

I will just use 950X when the time come instead

 

[b0bb]

I am looking to change the clock on my DA004. I measure the voltage supply of the clock on the cchd575 on my DA004 today. Quite surprisingly found out it has 4.83V on it. I measure serval time on Vdd to GND and output pin of the clock LDO it's all 4.83V no kidding. Is this kind of voltage safe for the clock? For long term operation?

The DAC work perfectly all the time. So the 575 clock is working under this 4.83v voltage on my dac. My 004 is just brought recently, it use LT3042 for power supply for the clock. It also has a earlier version, which use TPS7A4700 to power the clock.

Really wondering if all the 004 all has this hight voltage on the clock. I check the data sheet on the cchd-950 it has a Absolute Maximum Input Supply Voltage of 6.0V, 575 did not mention Absolute Maximum Voltage. And the PULSAR clock has the Maximum Voltage of 4.0V

The supply voltage to the XO is high, on mine the supply voltage is 3.3V between pins 2 and 4 on the XO and the clock output (Pin 3 on XO ) DC level sits at about 1.5V.
Supply pins are the ones marked "O" in your picture.

My DAC uses the TPS7A47 regulator, if you are planning to use the Pulsar the voltage output of the regulator on the board has to be changed as it is a 3.3V part.

 

[demos]

The supply voltage to the XO is high, on mine the supply voltage is 3.3V between pins 2 and 4 on the XO and the clock output (Pin 3 on XO ) DC level sits at about 1.5V.
Supply pins are the ones marked "O" in your picture.

My DAC uses the TPS7A47 regulator, if you are planning to use the Pulsar the voltage output of the regulator on the board has to be changed as it is a 3.3V part.

Thanks for your info b0bb, looks like the voltage needs to be lower. Check the LT3042's datasheet and the 004's PCB, its voltage is set by the resistor near the regulator. Just change it to a 33k ohm it will output 3.3V

 

[b0bb]

Thanks for your info b0bb, looks like the voltage needs to be lower. Check the LT3042's datasheet and the 004's PCB, its voltage is set by the resistor near the regulator. Just change it to a 33k ohm it will output 3.3V

Your DAC appears to have the 33k resistor already, The color balance of the photo is off. Is the color band on the MELF resistor orange-orange-black-red-brown ?
If it is you may have a dry solder join, reflow the connection to fix

 

[demos]

Your DAC appears to have the 33k resistor already, The color balance of the photo is off. Is the color band on the MELF resistor orange-orange-black-red-brown ?
If it is you may have a dry solder join, reflow the connection to fix

Believe or not, I check the resistor's leg more carefully and found out one leg is sitting there left unsoldered It's on the other side I usually facing. It connects to the PCB just by surface contact.

It always look like a yellow-yellow color band on the resistor to me. Until I check out the resistor color code and found yellow=4 so I think the resistor value could be 44something ohm, about right on LT3042's output voltage. But if calculate precisely 44k ohm resistor should make LT3042 output 4.4V. So It was really weird. I am just going to unsolder that resistor to measure what exact value it would be, after I receive the 33.2k ohm metal film resistor I ordered. So now It solve all the question It's no needed. Thanks b0bb for this!!

 

[demos]

Lks004 uses LT1763. Isn’t this capable of 500mA ? That means OCXO will work ? Actually, at turn on, it needs 3.6W
https://www.mouser.com/ProductDetail/ECS/ECOC-2522-100000-3FC?qs=sGAEpiMZZMsC2cQJVRBSBQ7qx1B6RYgm%2bL8c/KIWJXzTGwc7Dapf%2bw==

I will just use 950X when the time come instead

The SC-cut OCXO can find on mouser or digikey mostly has a 5 or 6 pin layout, there is a pin Vref, not sure how to deal with that. Probably need extra circuits to provide that voltage. May be not a direct replacement on the DAC

 

[Whitigir]

I take off the cchd575 on the 004, put in a 950x, there is significant improvement on the 950x, WOW

Alright, I did just this, a little tricky as the 950x is much larger than the 575.

The improvements are ? Seriously ! woW!! Can’t believe how much it improved: soundstage, transparency, clarity, extensions, are all improved

I wonder why lks004 don’t put in the 950X stock, but 575 instead. The price differences are only $20 give or take ?

 

[demos]

The 950x I try on my 004 is CVHD-950x, I getting this because this is cheaper and easier to buy at digikey. Want to try it out only. But turn out I am quite happy for how it sound on the 004 too, has instant improvement on layering, instrument separation more musical.




As expected after reflowed the unsoldered resistor leg, the clock voltage go back to normal, 3.28V on my dac. I Guess I never getting this dac's original performance until the LT3042's Vset resistor solder join has been fixed. The solder to solder poor surface contact probably generate lots of noise which ruin the LT3042's very low noise capability and become a noise power source for the clock. After the fix, it even sound less digital on treble and sweeter. Background feel a bit cleaner and cleaner micro dynamic.

 

[Whitigir]

Nice, you got the problem solved as I never heard lks004 to be digital or glary at all.

Anyways, what is the differences between cchd and CVHd ?

Found it: CCHD has 25PPm -165 and -160 where as CVHD has 20PPm -166 and -162. Doesn’t explain why cchd is more expensive

 

[demos]

CVHD is a Voltage Controlled clock, CCHD is a fixed clock. It's for different application. The 20ppm on CVHD is for Absolute Pulling Range, which has a different meaning for the 25ppm on CCHD. CVHD Has a Voltage Control Pin which need 0~3.3V I just solder it to ground for 0V.

 

[Whitigir]

CVHD is a Voltage Controlled clock, CCHD is a fixed clock. It's for different application. The 20ppm on CVHD is for Absolute Pulling Range, which has a different meaning for the 25ppm on CCHD. CVHD Has a Voltage Control Pin which need 0~3.3V I just solder it to ground for 0V.

Thanks for the clarification, I can see that voltage control pin being wired to ground there.

 

[b0bb]

The 950x I try on my 004 is CVHD-950x, I getting this because this is cheaper and easier to buy at digikey. Want to try it out only. But turn out I am quite happy for how it sound on the 004 too, has instant improvement on layering, instrument separation more musical.




As expected after reflowed the unsoldered resistor leg, the clock voltage go back to normal, 3.28V on my dac. I Guess I never getting this dac's original performance until the LT3042's Vset resistor solder join has been fixed. The solder to solder poor surface contact probably generate lots of noise which ruin the LT3042's very low noise capability and become a noise power source for the clock. After the fix, it even sound less digital on treble and sweeter. Background feel a bit cleaner and cleaner micro dynamic.

Looks like it is working, I would not worry, the poor performance was due to poor conductivity of a badly heated solder joint. The reflow fixes it..