[b0bb]

Further analysis of the reverse leakage current behavior Ir of the STPSC8065 SiC Schottky

This device is designed for high voltage applications and as such the specification data is done at the voltage.
The LKS004 operated at 30V, the Vishay 10H100 is a device specified to work at 100V.

ST suggests in their datasheet that the leakage current drops if the voltage across the device is reduced.
As this info is presented as "typical" behavior, this means the official maximum still applies as far as acceptable performance.

Ir for the device at 25C and 100V could be as low 0.8uA

However this is just an educated guess, this presents an alternative interpretation to those dead keen on adding SiC rectifiers to their LKS004.

However based on the official numbers, one could turn the LKS004 into a 1980's style DAC.

 

[tdrood]

Thanks b0bb - sounds like bottom line is use the mfg max values (or min, whichever critical to your parameter/application) unless you want to do a lot of testing...

 

[b0bb]

Thanks b0bb - sounds like bottom line is use the mfg max values (or min, whichever critical to your parameter/application) unless you want to do a lot of testing...

Yes.

Given the 50:1 spread of the Ir parameter, it could mean one may have to sort from a pool of 1000pcs instead of 100pcs to get the 8 lowest noise devices if one goes down the DIY route.
That assumes the person has constructed a measurement setup that can apply a 8-10ms reverse voltage pulse and accurately measure the leakage current in the nano ampere range.

 

[Gilberto 62]

Good morning guys. If the usb cable is connected and the lks is playing but this appears on the screen (... see image ...), does it mean that the Amanero board is broken ...?

 

[hpsxrb]

Good morning guys. If the usb cable is connected and the lks is playing but this appears on the screen (... see image ...), does it mean that the Amanero board is broken ...?

The DPLL you can use is dependent on the quality of the source (PC set, USB cable, etc.).
A smaller DPLL implies a better quality source, so if you reduce the DPLL too much, the LKS is unlocked because the source does not have the quality required for that DPLL.
A higher DPLL (so that the LKS is locked) is a lower quality source.
If you can adjust the DPLL so that the LKS is locked then this means (in principle) that the problem is not with the LKS (nor with the Amanero board).
If anyone has a better explanation, help is welcome.

 

[Gilberto 62]

It is true...! Now, I have changed the font (... before it was a very normal flac ...) and now I have put a dsd and yes, everything is fine ... Thanks "hpsxrb" for your valuable help ...

 

[VerloK]

From what I've read, USB cable is a transmission line: the impedance (sqrt(L/C)) has to be 90 Ohm
for the differential data to be captured correctly.
I can't find the L and C spec on the soldicoreaudio webpage.

I built my own USB cable like Kimber Cable does.
I dont think it has 90ohm but it sounds phenomenal.
How can u measure that?


Regards

 

[Gilberto 62]

Well speaking of cables, I have problems with the Audiquest Diamond 5 meters again. Yesterday, after a few days of peace and quiet, he suddenly cut the playback again, and from there it was no longer normal. Today, I have been testing again for six hours and after removing and replacing the Diamond and the stock cable a million times, there is something about the Diamond that creates problems. I think that, as you have told me, those 5 meters are problematic and I will have to adopt another solution since I need, more or less, that distance because between my pc and the audio equipment there are about four meters. Today, to do the tests with the stock cable, I had to carry the LKS next to the PC and use a deck as a preamplifier.
And I thought that my problems were already over ....!

 

[oldearwax]

I built my own USB cable like Kimber Cable does.
I dont think it has 90ohm but it sounds phenomenal.
How can u measure that?

Send a pulse down the cable and measure the degree of reflection .. ...from what I've read from
https://en.wikipedia.org/wiki/Time-domain_reflectometer

"High-speed USB" cable single-ended impedance is specified to be 45R.
I don't think you can measure characteristic impedance of cables using a ohmmeter.

 

[b0bb]

Send a pulse down the cable and measure the degree of reflection .. ...from what I've read from
https://en.wikipedia.org/wiki/Time-domain_reflectometer

"High-speed USB" cable single-ended impedance is specified to be 45R.
I don't think you can measure characteristic impedance of cables using a ohmmeter.

That is correct.
The device analyzes the forward pulse and reflected pulse and superimposes the 2 traces to form an eye diagram.

The intersection of the eye and its size and shape indicates the quality of the signal, bigger is better.

Notice the eye area of the 5m cable is much smaller.
In practice this translates to a less reliable connection.

The differential characteristic impedance is 90ohm.
If the trace touches the red area the compliance test fails.

Look what happens when the cable is not 90 ohms, this is a non compliant USB cable.

 

[AuxInput]

Well, after 3 weeks of burn in, my LKS Dac just never got there. The biggest problem was the upper mids had a glare.

I had a suspicion of the problem – KZ caps in the post-regulator section for digital power supplies. So I pulled them out and put in some low impedance 1000uf 25V Panasonic FR caps I had bought for a Krell project (ultra low impedance at 18 mOhms - Digikey part # P14425-ND). I put in a Nichicon KL cap for the electroyltic 100uf/25V next to one of the bigger caps, but I'm not really concerned about this because the 100uf is just there to stabalize the ground pin of the regulator.

Well, immediately I could tell a huge difference, after only 20 minutes of warmup and playing time. The upper midrange glare is gone and the bass actually hits harder and is stronger. The high frequencies have more air and resolution. The overall depth of the sound is significantly better. Pretty much solved my problems. It's been 7 days burning in now and it sounds excellent. The resolution of the tones is much higher and much more realistic. No extreme brightness or harsh click type sounds like with the KZ.

It could be that the higher impedance of the KZ audio caps does not work in this position or it could be that I had 25V KZ caps (25V version sounds bad). In either case, I think that these KZ caps have no business in a digital power supply anywhere. That being said, I think the KZ caps sound excellent when in an analog circuit.

 

[hpsxrb]

Can someone help me with HQPlayer settings to be able to upsampling to DSD128 / 256, I can only get PCM and only once I got DSD but I lost the setup (I looked on the internet but there are so many more recommended settings, the most being the most referring to an older version of HQPlayer).
These are the setups that I have at the moment while I am not a rocket scientist. Thank you in advance.

It was only today that I realized that after changing the settings in HQPlayer, it is necessary to leave the program for them to take effect.

 

[b0bb]

Can someone help me with HQPlayer settings to be able to upsampling to DSD128 / 256, I can only get PCM and only once I got DSD but I lost the setup (I looked on the internet but there are so many more recommended settings, the most being the most referring to an older version of HQPlayer).
These are the setups that I have at the moment while I am not a rocket scientist. Thank you in advance.

It was only today that I realized that after changing the settings in HQPlayer, it is necessary to leave the program for them to take effect.

Yes, exit and restart

Change the following
SDM Defaults box: Set Bit rate (/ limit) to 44k x256.

 

[Gilberto 62]

Good afternoon sirs. Sorry for my intrusion but I would like to ask a somewhat unusual question: would it be possible to cut a five-meter usb cable and leave it one meter long ...?

 

[b0bb]

Good afternoon sirs. Sorry for my intrusion but I would like to ask a somewhat unusual question: would it be possible to cut a five-meter usb cable and leave it one meter long ...?

Possible but not recommended.

Cutting the cable and adding a new connector requires careful preparation of the cable to avoid the impedance mismatch problems described a few posts above.
The new connection should be tested for proper operation using an instrument like a TDR mentioned above.

TDRs are not common in ordinary electronics repair shops.

You will be better served getting a shorter cable instead.