The true experimental tweak thread

[sandyk]

33 minutes ago, TubeMan said:

Hello Sandyk,

What is your experience o opinion about these

 

http://www.audiowind.com/pdf/A-270V2.pdf

 

When fed with DC

 

They can make a worthwhile improvement when used to power an internal SSD from the +12V SMPS.

I have previously recommended these to other members of the forum, who also reported an improvement when using them.

 However, a further improvement can be obtained when they are followed by the JLH PSU Add-on. ( A Shunt Regulator)

[Solid-State]

ok

I have used them fore i time  ,

but i have changed the capacitor to Mundorf Mlytic AG in and ELAN Simlic III out

[Cornan]

8 hours ago, sandyk said:

I saw that I had been quoted by Tubelover 2 in the link that was posted by austinpop. (above)
He also said this in his reply :

 

 I have been doing that for quite some time now, and my preference is a 12V Li Ion battery into a 12V to +9V Voltage Regulator
 followed by a modified John Linsley Hood designed Power Supply Add-on ( A.K.A. "Ripple Eater") which is a form of Shunt Regulator, and in my version has a simulated capacitance of > 2 FARADs, powering an Uptone USB Regen.
 This means that there is no inadvertent coupling back to A.C. Mains Earth.
 Yes, a well designed Linear PSU instead of the Li Ion battery can come close in performance when using an R-Core transformer,  but do not connect it's screen to mains earth, as this causes a minor SQ degradation.

I have mounted a toggle switch on  the rear of my PSU which readily enables me to compare, with and without, the screen connected back to Mains Earth.

Alex

 

Would you say that a LT3045 followed by a shunt regulator would be a good thing to try using two LT3045´s in series when using a Gophert csp-3205II? Remember that the Gophert is floating (the ac mains safety ground is not passed through to the DC negative output, but keep in mind that my DC negative output is connected to a ground receptacle as well). 

The reason I am asking is that as I understand it using voltage regulators in series can have its draw-backs with sharp pulses while charging which have a wideband spectrum. I wonder if a LT3045 followed by a shunt regulator would improve things by reducing leakage back to the Gophert, ie. possibly better than using LT3045´s in series?

[sandyk]

13 minutes ago, Cornan said:

Would you say that a LT3045 followed by a shunt regulator would be a good thing to try using two LT3045´s in series when using a Gophert csp-3205II?

 

The problem with your typical Shunt Regulator, including the Salas etc. , is that they drop further voltage, and usually need to be specifically adjusted for their output voltage. The JLH PSU Add-On is different, in that if the Current Limiter section is bypassed, the only voltage drop is a mV or 2 due to copper track resistance .

[sandyk]

39 minutes ago, TubeMan said:

ok

I have used them fore i time  ,

but i have changed the capacitor to Mundorf Mlytic AG in and ELAN Simlic III out

 

 A big advantage with this PCB, is the larger than typical heat sink ,compared with most ebay offerings, and the ability to be used with either A.C. In, or D.C. In.

[Cornan]

12 minutes ago, sandyk said:

 

The problem with your typical Shunt Regulator, including the Salas etc. , is that they drop further voltage, and usually need to be specifically adjusted for their output voltage. The JLH PSU Add-On is different, in that if the Current Limiter section is bypassed, the only voltage drop is a mV or 2 due to copper track resistance .

 

That will not be a problem for me, since I will special order the shunt regulators with specific output voltages for my ISO Regen and Aqvox switch. Do you think it is a good idea to use the LT3045 in series with a shunt regulator though?

[sandyk]

Just now, Cornan said:

Do you think it is a good idea to use the LT3045 in series with a shunt regulator though?

 

 I have no experience with the LT3045 at this point in time, as the added JLH PSU add-on gives me <4uV noise and a very low output Impedance. Personally, I doubt that the noise difference between the LT3045 and the JLH is anything more than  academic at these very low noise levels.

I guess that there is only one way to find out ?

 

Have you had a chance to try 2 x 1A LT3045 in parallel , taking into account current sharing requirements ? 

 

[Cornan]

19 minutes ago, sandyk said:

 

 I have no experience with the LT3045 at this point in time, as the added JLH PSU add-on gives me <4uV noise and a very low output Impedance. Personally, I doubt that the noise difference between the LT3045 and the JLH is anything more than  academic at these very low noise levels.

I guess that there is only one way to find out ?

 

Have you had a chance to try 2 x 1A LT3045 in parallel , taking into account current sharing requirements ? 

 

 

Yes, I will probably give shunt regulators a try. I have just sent an e-mail to ATL Hi-Fi to see if he can make these for me. It will be interesting to try.

 

My two 1A LT3045 boards (15v + 14,3v) was sent out from Alexey two days ago. He assured me that a 0,7v drop-down was sufficient for the LT3045´s. I will only use them in series. As soon as I get them plus the screw terminal DC2.5/5.5mm plugs and sockets I will start to compare different variants for my Aries Mini: LM317 with burnt-in ELNA silmic II 1000uF caps, LT3045´s in series, LT3045 plus ELNA silmic II 1000uF cap at the output and possibly LM317 with burnt-in ELNA silmic II 1000uF caps in series with LT3045.

I will post my impressions here while trying them out. I will be on holiday next week and will have some free time to experiment. Hopefully all the parts will arrive early next week. Touch wood!  

[Solid-State]

"Have you had a chance to try 2 x 1A LT3045 in parallel , taking into account current sharing requirements ? "

 

I have checked all my 6 LT3025 5v
and they are next precic 4.97 / 4.98 unloaded,
so I do not think there would be any problem connecting them in parallel,

 

and I've been testing a little with volt in versus volt out,
If you get less than 0.4v, they lose their dynamics and start to sound thin when the load is about 0.5Ah

so I would recommend to have at least 0,8 v differ

[Solid-State]

4 hours ago, sandyk said:

 

 A big advantage with this PCB, is the larger than typical heat sink ,compared with most ebay offerings, and the ability to be used with either A.C. In, or D.C. In.

 

Yes

Audiowind is making quite good things

[Cornan]

While scanning through some "similar artists" on Tidal I came across this quite unusual band that I coud´nt decide if it was a keeper or not, but the more I listened to it the more I liked it. Well, it was a keeper in the end. It grows on you!

 

Slow Steve - Adventures

https://tidal.com/album/57615451

 

 

 

 

[Cornan]

9 minutes ago, TubeMan said:

If you get less than 0.4v, they lose their dynamics and start to sound thin when the load is about 0.5Ah

so I would recommend to have at least 0,8 v differ

 

I´ll guess I´ll find out soon enough if 0.7v (with max 660mA load though) is sufficient as Alexey told me!

 

[Cornan]

Since I now have gone back to ChromeCast Audio in my B setup again (leaving my pc based setup behind) I was thinking how to improve it further. Using the BubbleUPnP server as audio transcoder instead of CCAs is a great improvement, so first started to look around for something to use as a BubbleDS Sever for the CCA. Right now the BubbleUPnP Server is running on my Android tablet together with BubbleDS Next. Suddenly, while looking around for Raspberry Pi3´s and Windows 10 PC sticks it strucked me that my son have a Android tablet collecting dust most of the time. Since he also got a Ipad Mini Retina that he uses, he only reach for the Android tablet when the iPad Mini have run out of battery juice. Now, I could´nt just take his Android tablet and make it mine, but I can secretly install BubbleUPnP server on it and have it plugged into constant charging. A cunning plan to get a simple fanless BubbleUPnP server in the house if I can say so myself!  

 

I will try to install it later today to see if it will improve SQ a bit furher when separating BubbleIUPnP Server on my sons Android tablet and BubbleDS Next o my Android tablet together with CCA.

 

 

 

 

 

[Cornan]

Success! ?

 

 

My son's Android tablet running the BubbleUPnP server is locked inside the cubborn. My own Android tablet with BubbkeDS Next

 

 

Works great and sounds better too! ?

 

I also got my gold plated forks today that I will use to get improved contact to the DC negative sockets on my Gopherts for the grounding wires, so I will have a little something to experiment with later tonight! ?

 

[Cornan]

Bummer! The gold plated forks did'nt work. A tad too big to get a good grip with the screw terminal banana sockets on the Gopherts. ?

[mozes]

3 hours ago, Cornan said:

Success! ?

 

 

My son's Android tablet running the BubbleUPnP server is locked inside the cubborn. My own Android tablet with BubbkeDS Next

 

 

Works great and sounds better too! ?

 

I also got my gold plated forks today that I will use to get improved contact to the DC negative sockets on my Gopherts for the grounding wires, so I will have a little something to experiment with later tonight! ?

 

Nice use of resources, great that it sounds better too

[Cornan]

7 minutes ago, mozes said:

Nice use of resources, great that it sounds better too

 

Thanks, my son have'nt even noticed that his Android tablet is "charging" somewhere else! ?

The difference in SQ is not huge, but clearly noticable in clairity, when the transcoding is done on another unit than the control device. It sounds really nice now with Peaktech and Gopherts powering the whole setup and a lot quicker to get up and running compared to the PC approach.

 

The funny thing is that it will probably not last for long. I will move to a smaller appartment in January. My B setup will probably have to melt into my A setup in matter of months! ?

[mozes]

12 minutes ago, Cornan said:

 

Thanks, my son have'nt even noticed that his Android tablet is "charging" somewhere else! ?

The difference in SQ is not huge, but clearly noticable in clairity, when the transcoding is done on another unit than the control device. It sounds really nice now with Peaktech and Gopherts powering the whole setup and a lot quicker to get up and running compared to the PC approach.

 

The funny thing is that it will probably not last for long. I will move to a smaller appartment in January. My B setup will probably have to melt into my A setup in matter of months! ?

Even though I am living in a 2 storey house, I never felt the need for 2 systems. Better to focus all resources on one. 

[Cornan]

1 minute ago, mozes said:

Even though I am living in a 2 storey house, I never felt the need for 2 systems. Better to focus all resources on one. 

 

I actually found it very useful with two systems. Without my experimental B setup a lot of things in my A setup probably would'nt have been realized. It is however a simple proceedure to merge them into each other and I already got a clear idea how to. I might look into a speaker upgrade while I am at it though.

[sandyk]

Are Voltage Regulators , in general , degrading the full potential of Digital Audio ?

 

A little while ago I tried using a 15,000mAH Li Ion battery without using any voltage regulators, to directly power my USB Regen with +12V via a newly modified JLH PSU add-on. It has been modified to give approx.4 FARADs of simulated capacitance instead of around 0.7 FARADs of the original published design in Electronics Today International.

I am able to use 12V directly into the USB Regen as the Corsair Voyager's current consumption is relatively low.

 

I  just converted the .flac of “Julie London –You'd Be So Nice to Come Home To,"  to a .wav file saved on a Corsair Voyager.

The best way to describe it when it's decoded from .flac , saved to and played back via the same JLH (above) , is it's very ANALOGUE sounding, and it goes deep !

Her voice is far better sounding too!  I didn’t think such an old recording could sound so good!

The original .flac when played directly from my OS SSD as a .flac file, sounds quite pathetic in comparison.

 

Alex

[Cornan]

3 hours ago, sandyk said:

Are Voltage Regulators , in general , degrading the full potential of Digital Audio ?

 

A little while ago I tried using a 15,000mAH Li Ion battery without using any voltage regulators, to directly power my USB Regen with +12V via a newly modified JLH PSU add-on. It has been modified to give approx.4 FARADs of simulated capacitance instead of around 0.7 FARADs of the original published design in Electronics Today International.

I am able to use 12V directly into the USB Regen as the Corsair Voyager's current consumption is relatively low.

 

I  just converted the .flac of “Julie London –You'd Be So Nice to Come Home To,"  to a .wav file saved on a Corsair Voyager.

The best way to describe it when it's decoded from .flac , saved to and played back via the same JLH (above) , is it's very ANALOGUE sounding, and it goes deep !

Her voice is far better sounding too!  I didn’t think such an old recording could sound so good!

The original .flac when played directly from my OS SSD as a .flac file, sounds quite pathetic in comparison.

 

Alex

 

A good question that also might need to be refrased IMO.

Before I started with my floating SMPS (Gopherts) I powered most of my gears with unregulated battery supplies and strongly beleived that voltage regulators was one of the main obstacles against good sound. When ever I was forced to use voltage regulators, even linear ones, the overall SQ became worse. Not sounding as natural as unregulated battery supplies.

 

The true eye opener for me personally was when I first started with a floating LPS (Voltcraft FPS-1134) and then the Gophert CSP-3205II´s. I then realized that there is much more to it than voltage regulators. Noise is just something you always need to battle in various ways. I had never succeeded to get that kind of natural sound with lots of air pumped into the music. Battery supplies can sound very natural with great presence and bass performance, but never seems to get the ambience/air/3D part completely right.

 

Both battery supplies and SMPS struggles with problems of their own, that actually have some "similarities". With battery supplies it is called self-inflicted noise and with SMPS it is called leakage. Both are really self-inflicted noises. The problem is how to reduce these self-inflicted noises in the best possible way? With floating SMPS the voltage regulators seems to work really nice. They do still limit some of the ambience/air/3D compared to no voltage regulators IME, but seem to improve almost everything else. According to Aleksandar at ALT Hi-Fi using shunt regulators without voltage regulator could be very beneficial due to various reasons. It seems to me that fighting SMPS leakages and the self-inflicted noises of battery supplies can be cured in similar ways by using shunt regulators. I will most likely try it on my Gopherts and on a unregulated battery supply to see if that holds true.

 

So, what about the refrase?

 

I don´t think that voltage regulators alone is degrading the full potential of digital audio. It is just means to reduce noises caused by power supplies. It is not a double-edged sword, but still a sword so to speak. I do not know yet, but shunt regulators might just be a better choise for that perticular battle. A double-edged sword? Probably not, but maybe a bit sharper?  

 

 

[sandyk]

1 hour ago, Cornan said:

I don´t think that voltage regulators alone is degrading the full potential of digital audio

 

 No, not by themselves, but it's the way they are implemented. The E.E. designers quote better, and better measurements, but what relevance do they have to what we actually hear ? The comments made in this thread about the increased air and separation when using the LT3045 modules suggest that they are biased towards a decreasing impedance at frequencies well above the audible range. With the types of capacitors used, that doesn't surprise me. Historically, the use of Ceramic capacitors with Audio has always been a No-No ! Try Googling "Capacitors for Audio".

 

In fact, using the JLH PCBs, I can make HF detail improved, even to the point of excessive HF detail, and  too big a soundstage just by using 2 low ESR capacitors in parallel in the Capacitance Multiplier section. Conversely, I can also  make the sound a little too dull sounding, by just by using 2 normal electros in parallel in the Capacitance Multiplier section. I use a combination of selected normal type and Low ESR types to get a more natural and balanced sound.

This also works for DIGITAL, but according to the E.E.s , this is not possible, is it ?

According to currently accepted theory it shouldn't happen.

I am even able to compensate for mpeg 4 bitrate that is a little below optimum for recorded HD TV, resulting with playback from a USB memory stick that looks more like  clearer mpeg 2 HD TV than the original direct .mpeg 4 HD TV transmission, using an optimised JLH for powering the USB memory stick.

Your typical E.E.s will refuse to accept this though, based on purely theoretical grounds.

 

Most DIY people do a lot more than just use basic modules such as the LT3045 without also using Audiophile grade electrolytic capacitors at their output, that have won a hard earned reputation with Audiophiles over a period of many years.

The problem with that, according to the Engineers who design the new type regulators,  is that it slows down the response of the regulators !

[marce]

Bloody engineers, don't know nowt...

[Cornan]

Reading this...

 

Quote

What we know as electric current is rate of flow of charge per unit time what we express as dq/dt. What happens when we apply a voltage across the electrodes of a capacitor is, the dipoles present in the dielectic media is get polarised and this in the form of displacement current is established in the circuit. Even when a DC voltage is applied to a capacitor which is not charged a current will flow till the capacitor is fully charged as in the process of charging there exist dq/dt. once it is fully charged no additional charge is pumped in or out of the capacitor and it blocks current. In case of AC because of change in polarity of the applied voltage there will be continuous polarisation and depolarisation in each cycle causing rate of change of charge stored in the capacitor and hence flow of current with a 90 degree phase shift with the applied voltage. Please feel free if you need further clarifications. Thank you

 

...makes me wonder why I should´nt try a DC2.1/5.5mm female socket>short wires between positive and negative terminals>ELNA silmic II 1000uF>DC2.1/5.5 male plug pre my Aqvox switch, ISO Regen and Aries Mini? In other words just letting the Gophert supply the regulated voltage and the capacitors to take care of the leakage. Would that actually work and block DC leakage into the device?

 

I hope you all realize that I am thinking out loud here, since my knowledge of capacitors is very limited!

 

[sandyk]

1 hour ago, marce said:

Bloody engineers, don't know nowt...

 

Most are too fixated by measurements that don't give too much of a clue how something will sound to the end consumer.

How easily (conveniently) you guys forget things like the Sony/ Philips claim of perfect sound forever from the CD, yet we need now to use a computer to save the CD contents to System Memory to appreciate the quality of the contents on a CD, assuming that over zealous Mastering Engineers haven't compressed the Bejesus out of it !

Now the Industry is trying to foist MQA on a gullible public, just like the original claims that .mp3 encoding wasn't audible as it was masked by loud passages, and also that most people wouldn't notice the difference with lower than optimum bit rate .mp4 HD TV.

 I suppose it still looks better than what Foxtel think is HD though ?

 

 

P.S.

Are you able to play both .flac and .wav files through decent gear using a good software player ?

 

G'night from Sydney Au.