A novel way to massively improve the SQ of computer audio streaming

[sandyk]

4 minutes ago, Em2016 said:

Not that it's a big deal but see the later discussion, including John's post.

 I was talking about John's original posts in this area.

 Some of you guys are fabricating cables that may be a whisker better, but at the expense of flexibility.

 You are also in many cases ignoring other well known technical methods that will pay even greater dividends, such as trying to eliminate as many cables as possible and avoid daisy chaining components, especially where the initial PSU has a VERY low Output impedance, and any additional lead lengths will substantially increase that output impedance. The lower the EARTH reference is for example, the better the results normally are.

[sandyk]

7 minutes ago, Em2016 said:

 

Yes and I quoted and linked to it..

 

Anyway it's there just a few posts above, for people to read and interpret their own way. No gains in arguing too much about it.

 

 

The fact remains that the shield is not normally earthed in these applications, it is normally left floating and the shield ends kept well clear of the outers of the plugs at both ends.

 

Quote

I guess I read "leaving the shield completely disconnected from the rest of the circuit is the best way to go" as the best way to go. - Lmitchie

 

[sandyk]

Has anybody measured the total resistance of their fabricated D.C. leads from the plug at one end, to the plug at the other end with both inner and outer properly short circuited in that plug?  To do this properly you would need a DMM that can accurately read very low resistance, or use a Milliohm adaptor with a normal DMM.

We need to take into account how the total resistance of the leads markedly increases the output impedance of the voltage regulator, which with a LM317 at a current of 100mA  for example, may be as low as 30 milliohms .

 

 This should show why D.C. connecting leads should be as short as possible, using the heaviest gauge wire that will fit into the plugs.

Using a JSSG can only assist with LF external rubbish getting into the leads.

 

For those into DIY, the attached link may be of interest.

http://www.tnt-audio.com/clinica/regulators2_impedance1_e.html

 

Alex

[sandyk]

9 hours ago, lmitche said:

I have successfully used the gx-16 aviation connectors on the sigma 11 power supply cases built here. They are a bit big, but easy to solder, and with embedded screw on cable locks are very reliable.

 

There are 12 mm versions as well, but I have never tried them.

 

Both also come in multiple pin versions from 2 and up.

 

They are ideal if the case has enough room for them. I am presently using 3 pin mini XLR connectors wherever possible, as they are also lockable. However ,it's still hard to terminate larger gauge wires., although you can bore them out a little if need be.

[sandyk]

8 hours ago, the_doc735 said:

Lightstream is ultrafast broadband. It connects your home to the internet using fibre optic technology, replacing the copper cables used by standard broadband. We’ve built a network designed for the internet and the way that people use it, so it’s more reliable, faster and future-proof.

 They are talking about the differences between FTN ( Fibre to the Node, and Fibre FTTP.)

In  Australia at least,  FTTP is connected by Fibre Optic cable all the way, and FTN is normally connected from a " mini  Telephone Exchange"  in the street  via a copper cable pair which may already be present in your house. FTN can be of a decent speed if you are reasonably close to the local Telephone Exchange, but FTTP has the potential for vastly improved speed.

 

[sandyk]

1 hour ago, rickca said:

Or why the 3A MPAudio (6xLT3045) regulator didn't sound good?

 Low noise isn't everything. You also need a very low output impedance that is virtually flat from D.C. to way past 100KHZ, preferably to 1MHZ.

Note also that the manufacturer makes no claims for it's suitability with HiFi Audio.

Applications
 RF Power Supplies: PLLs, VCOs, Mixers, LNAs, PAs
 Very Low Noise Instrumentation
 High Speed/High Precision Data Converters
 Medical Applications: Imaging, Diagnostics

 

In fact. it's need to use very low ESR capacitors virtually guarantees that HF detail will be accentuated UNLESS you also use some larger value normal (NOT Low ESR) parallel electrolytic capacitors. Using it straight after a typical PSU with a Bridge rectifier and large value conventional electrolytic capacitors will help to achieve this , as will using a parallel larger value electrolytic capacitor at it's output.

 However, this is at the expense of a reduction in speed which shouldn't normally be a problem with Analogue Audio .

 Several members have obtained a more balanced tonal presentation by doing this.

Micael, on the other hand, uses daisy chained LT3045s to obtain added HF detail.

[sandyk]

38 minutes ago, Cornan said:

First of all I’m amazed that you speak freely about things that you actually have’nt tried yourself. Your favorite expression have always been HF detail. You keep on using those words again and again. What does that really tell anyone? Nothing at all! Keep it real and let us know what you experience about things that you have actially tried IRL instead of giving everyone your personal opinion about how you think it should be in theory. 

 

I note that several members (see above as well) have already removed some of the very low ESR capacitors that you love so much in favour of a more correct tonal presentation. Uptone has also previously distanced themselves from your advice in this area..   

All that you seem to be presenting is a way to further improve the HF area of a streaming service by artificially increasing HF detail. That doesn't say too much about the rest of your system either if you need to go to such lengths. 

 

Just because several ebay "Entrepreneurs"  have jumped on the multiple LT3045 area to make quick $$$ doesn't validate this as the best way to do things. Large numbers of parallel LT3045s and LARGE heatsinks that won't fit in a typical case and still have adequate ventilation is simply the flavour of the month and "Butt Ugly", although 2 parallel LT3045s used in order to increase it's capabilities to 1A is a good idea if correctly implemented.

The LT3045 does an excellent job in the areas it was originally designed to be used in though , such as a +3.3V low noise supply for the Input device of a DAC and the +3.3V digital supply for a DAC chip as I have done, or as a very good regulator as others have used it for in some of their products.  

 

[sandyk]

41 minutes ago, Cornan said:

Why is your opinion worth more than mine in this matter?

 Unlike you, my comments in this area not only come from >60 years of practical experience in the construction of many different types of electronics devices, and learning from the various Electronic Magazine articles, but as well as regularly updated Technical training from Australia's Telstra where I was a Principal Telecommunications Technical Officer before retiring.

 Way back in 1954 when I was a trainee, we even constructed a 6VGT valve amplifier as a project at a Post Master General's Department Training School.

 

 Neither is it a good idea to " Daisy Chain" LT 3045s as you do as it degrades the earth reference and runs the risk of picking up unwanted RF/EMI.

 The output impedance of a power supply, especially for Analogue, should be as low and as flat as possible from virtually D.C. to >100KHZ, with 1MHZ even better. The LT3045 without additional parallel higher value non Low ESR electrolytic capacitors has a lower Z Out at  > 100KHZ which increases apparent HF detail and results in a tonal imbalance.

However it will help with 2 in parallel when used as a +5V regulator for powering internal SSDs in a P.C. from  the internal +12V SMPS supply provided that a Pre -regulator is used to reduce the voltage differential so as to prevent overheating of the LT3045s.

[sandyk]

4 hours ago, lmitche said:

The reason for a fanless case is both about the acoustic noise of the fan, and about eliminating the electrical noise made by a sparky fan motor diminishing SQ.

 

To solve the acoustic noise problem get a larger fan and heatsink, run a water based system or remove the fan and use a passive heatsink.

 

The electrical noise problem can be solved with line filters or removal of the fan and installation of a passive heatsink. 

 

Obviously the Akasa case is a passive heatsink. It also seems to provide some vibration control as the 4 inch square board has five touchpoints with the very rigid case.

 

Electrical noise from a fan is mainly due to motherboard PWM control of the fan.(typically 25KHZ " square"  waves/pulses with a high harmonic content) It can be markedly reduced by using a fan filter, or running the fan from the 12V rail using manual speed control with a rheostat, or better still a thermistor attached to the CPU.

Some aftermarket fans come with both options.

[sandyk]

1 hour ago, austinpop said:

Believe it or not, this is what most of us want very much. A DAC that is 100% immune to the upstream chain would save people a lot of money and effort.

Rajiv

Please read right through before telling me I am delusional.

 

 That's not likely to happen when the PSU area of the PC mainly, and how electrically quiet it is, mainly  governs how a digital Audio or Video file will sound or  look. 

Both Digital A and V are subject to similar PSU area differences that also apply to Analogue Audio, both at the creation and playback stages.  

For example, if the PSU area has a lower Z at say 100KHZ than at mid range, the digital files will have increased HF detail, both with A and V, just as with Analogue . Yes, according to currently accepted theory this should NOT happen.

 You need to use a power supply in the PC that has a very low and fairly constant output impedance from " DC to Daylight"

New recently posted image links in My Profile will help to illustrate this point. They were taken from paused TV screen images of pairs of videos with identical .md5 checksums using a cheap tripod. The HF detail can even be increased excessively. You would have to be " Blind Freddy" not to notice the very obvious differences.

 If you use for example , a 12V to 5V regulator to power an OS SSD or internal music storage SSD, and put a very low ESR capacitor at it's output, you will notice an increase in HF detail with saved digital audio or digital Video files. (Both at the same time !)

 Some of you DIY guys should try adding parallel very low ESR electros at the output of your linear PSUs for SSDs etc.and see if you end up with a tonally balanced audio file. It's easy enough to verify this if you are willing to try it.

 Don't just dismiss this based on purely theoretical grounds, as this thread is more of a hands on thread for new ideas and techniques.

This is exactly the type of thing Micael (Cornan) and others were doing with  daisy chained LT3045s and very low ESR electros at it's output to get increased HF detail from Tidal.

No 1s and 0s were harmed in the process either ! 

 

Alex

[sandyk]

20 minutes ago, austinpop said:

 

Alex,

 

Not sure why you're preaching to the choir?  I was making an aspirational statement.

 

How many vendors of digital gear have declared that their device is fully immune from upstream effects due to a combination of galvanic isolation or buffering, or whatever else? I've lost count. Yet, we have yet to encounter a single one that is fully immune to the upstream optimizations discussed here - in PSUs, in clocks, grounding, isolation, etc, etc.

 

Far be it from me to discount any experimental finding! Keep at it.

Hi Rajiv

 Please take the time to look at the latest images in My Profile,. The cheap tripod couldn't completely stop the camera moving when the button was pressed, but the differences should be very obvious to most.

Incidentally, you guys are right on the money with improved clocking too. I replaced the 24.576MHZ oscillator in my old X-DAC V3 with a .3PPM TCXO and although there wasn't much difference with 16/44.1 there was a marked jump in SQ at 24/192. I also used a very low noise  PSU to power it.

 

 An ex workmate from Telstra said this yesterday in an email  :

Quote

Not quite sure of the technical side of things without doing a bit of research, but the difference between your “png” image files is pretty obvious, even to a blind person

Kind Regards

Alex

[sandyk]

11 minutes ago, LTG2010 said:

I believe Innuos play back from RAM to cut out some of the SSD noise.

 You can also cut out most of the SSD noise using the +12V supply and regulating down  to +5V with a low noise voltage regulator.

 This prevents most of it getting back into other areas via the PSU.

 The attached JLH PSU add-on separately powers my OS SSD and Music SSD (12V IN, and 2 x separate +5V Out )

[sandyk]

11 minutes ago, LTG2010 said:

Just that the sound is so much clearer when the SSD has been removed with playback from RAM,

 

Using JRiver and playing from System RAM via an Asus Xonar D2X internal soundcard via Coax SPDIF I can hear clear differences with the same high quality audio file played from either HDD, SSD or USB memory.

 

[sandyk]

4 minutes ago, nbpf said:

It is both possible, easy and actually very cheap if a user accepts that the replay of a track or of an album only starts after that track or album has been completely transferred to the DAC's local memory. It is just a matter of waiting a little bit, like when we were

using a CD player and we had to wait for the tray to close for replay to start.

 

 Then try and explain why the same bit identical tracks saved in different storage media, can still sound different when played from System Memory !

[sandyk]

59 minutes ago, diecaster said:

 

This is just plain false. You keep trying to sell this as fact when it is only a few people's opinion. It's as bad as flat-earthers....

 Try telling this to many of the participants in this thread, including the OP ! They are implementing changes that according to people like yourself can't possibly result in any SQ improvements.

  Try looking at the images I have supplied in My Profile . The comparison videos that the images came from were even located in the same folder of the USB memory stick.

[sandyk]

59 minutes ago, greenleo said:

If what you quote is true, then CD transferred to a DAC (assume bit-perfect) and USB to a DAC (again assume bit perfect) and a "better" USB cable to a DAC (again and again assume bit perfect) should sound identical.  But this is not the experience for quite a few users that I know.

 

 It is for members like Diecaster, mansr, esldude,ralf11 etc.  Their Expectation Bias won't let them hear any differences !

 They are so sure they are correct that they refuse to even try many of the things suggested in threads such as this one.

Quote

I do not want to argue but to learn.  Have you ever tried and compared?  - greenleo

 A very good question !

[sandyk]

10 minutes ago, nbpf said:

I have looked at the pictures but, to be honest, I do not understand which consequences we can draw from them. I am not criticizing you experiments, just trying to understand what the results actually are and what they could possibly imply.

 

Thus,  just to mention one example, it looks as if the first picture was spoiled by camera shake while the second one looks much better. Without further information about the f-stops, the shutter times and the stabilization mechanisms used, it is very difficult to draw any reliable conclusion, at least for me.  Perhaps I am missing something obvious.

I will take this to a PM to avoid derailing the thread.

[sandyk]

8 hours ago, LTG2010 said:

Jriver also implement playback from RAM as a selectable option although, as I understand this is due to user demand and the developers don't actually beleive their is any sound quality benefit in its implementation.

 

 There must have been an awful lot of customer demand for them to finally agree to provide this facility.

YES, it does result in a worthwhile improvement in SQ with JRiver for me.

[sandyk]

26 minutes ago, The Computer Audiophile said:

Then a guy sat down next to them and started smoking, claiming he should be able to do this despite the wishes of the cancer survivors that he please smoke in the designated smoking section. Does anyone thing the smoker's behavior is OK? I doubt it.

 

 This also applies to commuters. Even having a smoker sitting down next to you may cause an Asthmatic (I am) to start wheezing.

 We shouldn't need to get up and walk away to avoid a nasty confrontation.

[sandyk]

21 minutes ago, mormogi said:

studer warmer , little colored ,syrupy,more micro details.

 

Those clones on ebay appear to use quite old technology, such as the type of power transistor used and dated opamps.

The 25VA transformer in them would also markedly reduce the available current when you take into account that the available current from a bridge rectifier with capacitor input is only around 62% of the rated current from the transformer.

In other words, with a 12V output, the maximum available constant current would only be around 1.29Amps.

[sandyk]

3 minutes ago, greenleo said:

Did he really say this (I'm not questioning you)?

This seems quite an over statement.

 

We have been down this road before, and his post has been referred to Admin.

[sandyk]

1 hour ago, mourip said:

Soon I will move the music back to an internal SSD.

 

mourip

If you intend powering an internal SSD from an earlier version (?) HDPlex 200W PSU , do NOT adjust the adjustable 19V rail down to +5V to power it. ( . 9V (user adjustable between 5V & 19.5V) via XLR)

 There may be too much heat dissipation and the supply may fail and give the  FULL unregulated supply voltage (i.e. > 19V )  to the SSD unless it has an auto shut down protection feature as in later model PCBs .

I doubt that  HDPlex envisaged users wanting to power devices drawing up to 1A at +5V from the 19V rail, as it is never good a good idea  to have such a large voltage drop at a relatively high output current.

 

Alex

[sandyk]

33 minutes ago, mourip said:

Personally I would not use the 19 to 5v output for lower than 12v.

 A good idea.

 BTW, this happened to a friend of mine.

 However, HDPlex did the right thing by him, also providing a later model  replacement PSU PCB with protection.

[sandyk]

25 minutes ago, Chopin75 said:

ok, I am planning to get the latest Hdplex LPS as sold on web now. Would there be issue with this later model? I am planing to use internal SSD 2.5 Sata as boot drive (euphony OS).

 

 I would expect that current models all have overload protection incorporated.

 Normally, the manufacturer will prevent this happening in the design of the PCB by restricting the amount of adjustment possible, or at least provide warnings in the User Manual.

 

Irrespective, it is never a good idea to adjust the output of a voltage regulator PCB way below the maximum output voltage.

[sandyk]

6 minutes ago, Chopin75 said:

Strange the manufacturer would allow such adjustment if it would actually cause overheat issue. I would check with Hdplex later when I get the LPS.

 

  Use the 5V output and you will have no problems !

In many cases the adjustable 19V rail was left unused, so some who needed extra outputs tried to use it adjusted down to +5V,  which is fine if only using a low current device such as a Regen etc.

 HDPlex will have made 100% sure that this is no longer a problem.