ISO Regen performance Improvement Cheap!

[Monge]

I tried the kwmobile POE set and 1 m uGreen cat7 round on my Netgear modem/router with a linear Sbooster 12v. And to my big surprise, it just sound wonderful, when I stream Tidal hi-fi.

i also tried the Supra cat7+ and did’nt sound as Nice as the uGreen cat7 26awg round.

I have more POE adapters and uGreen cat 7 26awg flat on the way.

[BigGuy]

8 hours ago, Abtr said:

A screw is best tightened with a screw driver...

 

...and the driver of the screw can also be gotten tight with screw drivers!  ?

 

[Superdad]

17 hours ago, genjamon said:

My first thought when I first swapped to the Kemets was "where'd my dynamics go?!".  Now, I'll acknowledge a certain smoothening of any sibilance in my system, and I can see how those with tinnitus and/or bright systems could benefit from the haze the Kemets added.  In Charles's case, using a LPS-1 and MicroRendu, which I am very familiar with, I wonder if it's not a better synergy.  And maybe the Kemets are providing more benefit over the LPS-1, but the LPS 1.2 may not need the Kemets due to better output regulation.  So, I'm not trying to say the Kemets are trash or anything, and I'm still interested in trying them when I put my LT3045 board in between LPS 1.2 and the UltraRendu.  But at this point, the Kemets are not going back into the system in this configuration - no way, no how.  I'm getting magic with the straight 12awg romex and will be happy to come back to this configuration if results are mixed with other experiments.

 

You are 100% correct and here is why:

Linear Technology LT3042/3045 regulators are completely unlike any other on the planet.  Their very high PSRR and ultra-low impedance across an extremely wide bandwidth mean they are “faster” that any capacitor.  So putting big caps on their output (even low ESR 560uF ones like Cornan is using everywhere) is totally the wrong thing to do.  There is no cap that has a lower impedance across a wide band than what comes right off the 3042/3045.  For any other regulator sure.  But for the LT3042/45, anything more than the required 10uF “stability” cap is technically going to reduce their performance.

 

And the above jives with your report about better results adding a Kemets to an original LPS-1.  That model utilized the 1 amp Ti TPS7A4700 regulator (the LT3045 did not even exist when we were doing the original UltraCap design).
 

[Abtr]

26 minutes ago, Superdad said:

... for the LT3042/45, anything more than the required 10uF “stability” cap is technically going to reduce their performance. ...

Do you mean a 10uF cap is required for the Iso Regen, i.e., it is already build into it?

[mansr]

40 minutes ago, Superdad said:

But for the LT3042/45, anything more than the required 10uF “stability” cap is technically going to reduce their performance.

The data sheet tells a slightly different story:

"Given the high PSRR and low noise performance attained using a single 10µF ceramic output capacitor, larger values of output capacitor only marginally improves the performance because the regulator bandwidth decreases with increasing output capacitance — hence, there is little to be gained by using larger than the minimum 10µF output capacitor. Nonetheless, larger values of output capacitance do decrease peak output deviations during a load transient."

[Superdad]

51 minutes ago, Abtr said:

Do you mean a 10uF cap is required for the Iso Regen, i.e., it is already build into it?

 

Not speaking of the ISO REGEN at all.  Just discussing the LT3042/LT3045 regulators--which do require a 10uF cap on their output.  

But yes, we do have an 10uF X7R cap on each of the five LT3042s used in the ISO REGEN--and of course right on the pins of the LT3045s in the UltraCap LPS-1.2. 

 

[Superdad]

46 minutes ago, mansr said:

Nonetheless, larger values of output capacitance do decrease peak output deviations during a load transient.

 

To the detriment of other performance parameters.  Believe what you wish, but how much time have you actually spent designing with the LT3042/45?  

 

Frankly, given the fact that you always post retorts within minutes of each of my posts, I think you just like to follow me around and be contrary to anything I say. ?

[look&listen]

26 minutes ago, Superdad said:

Frankly, given the fact that you always post retorts within minutes of each of my posts, I think you just like to follow me around and be contrary to anything I say.

Not sound like wise move, since he getting many corrections from you & others, including Grand CA Poobah ?

 

1 hour ago, Superdad said:

So putting big caps on their output (even low ESR 560uF ones like Cornan is using everywhere) is totally the wrong thing to do.  There is no cap that has a lower impedance across a wide band than what comes right off the 3042/3045.

Thank you (& genjamon) for report/explanation. I add Kemets to new USB LPSU build, based on use by Stammheim in 3A & 5A LT3045 modules (look like attach to output from place on PCB?), not just Cornan use, plus old practice of film 'bypass' caps.

 

Now new LPSU not sound so great after 1 week burn in (??) Will give more days too test (not easy in heatwave ? ). If no change will remove Kemets & retest & report.

 

[beerandmusic]

I am just curious if after 46 pages if there is a consensus on an "off the shelf" cable solution available yet?

[Ralf11]

Ghent ??

[mansr]

31 minutes ago, Superdad said:

To the detriment of other performance parameters.  Believe what you wish, but how much time have you actually spent designing with the LT3042/45?  

 

Frankly, given the fact that you always post retorts within minutes of each of my posts, I think you just like to follow me around and be contrary to anything I say. ?

I happened to have the data sheet open and thought people might like to know it's not quite as simple as you put it. Sometimes additional capacitance can be called for. It all depends on the load characteristics and what the requirements are. You are of course correct that slapping a huge electrolytic capacitor on the output isn't likely to do much good.

[[email protected]]

2 hours ago, Superdad said:

 

You are 100% correct and here is why:

Linear Technology LT3042/3045 regulators are completely unlike any other on the planet.  Their very high PSRR and ultra-low impedance across an extremely wide bandwidth mean they are “faster” that any capacitor.  So putting big caps on their output (even low ESR 560uF ones like Cornan is using everywhere) is totally the wrong thing to do.  There is no cap that has a lower impedance across a wide band than what comes right off the 3042/3045.  For any other regulator sure.  But for the LT3042/45, anything more than the required 10uF “stability” cap is technically going to reduce their performance.

 

And the above jives with your report about better results adding a Kemets to an original LPS-1.  That model utilized the 1 amp Ti TPS7A4700 regulator (the LT3045 did not even exist when we were doing the original UltraCap design).
 

I do agree with Alex,  large capacitors need to be used before the linear regulator, but not after, and its better to follow manufacturers demo boards design... but.. let people try, when you post photos on instagram, your are adding some filters, changing colors, contrast etc., so the final picture is far from the original but looks much better..

[sandyk]

1 hour ago, mansr said:

I happened to have the data sheet open and thought people might like to know it's not quite as simple as you put it. Sometimes additional capacitance can be called for. It all depends on the load characteristics and what the requirements are. You are of course correct that slapping a huge electrolytic capacitor on the output isn't likely to do much good.

Many of these members are using LT3045 pre regulators so the uber fast response isn't needed.They are also using them with other products, many of which are likely to already have capacitors considerably larger than 10uf across their DC input.

[mansr]

59 minutes ago, sandyk said:

Many of these members are using LT3045 pre regulators so the uber fast response isn't needed.They are also using them with other products, many of which are likely to already have capacitors considerably larger than 10uf across their DC input.

In those cases, what good does the added LT3045 do?

[beerandmusic]

3 hours ago, Superdad said:

 

To the detriment of other performance parameters.  Believe what you wish, but how much time have you actually spent designing with the LT3042/45?  

 

Frankly, given the fact that you always post retorts within minutes of each of my posts, I think you just like to follow me around and be contrary to anything I say. ?

 

Me thinks since he is working on his own solution, and wants to ensure he is on top of all of your knowledge plus (smile)....I don't blame either of you...must be tough being a mfr, trying to share, but not too much...lol.  I do believe he has a better respect for you and your products now that he is entering the mfr world...whether he admits it or not.

[mansr]

7 minutes ago, beerandmusic said:

Me thinks since he is working on his own solution, and wants to ensure he is on top of all of your knowledge plus (smile)....I don't blame either of you...must be tough being a mfr, trying to share, but not too much...lol.  I do believe he has a better respect for you and your products now that he is entering the mfr world...whether he admits it or not.

I can assure you I am not trying to steal anything from anyone. If I were, I wouldn't be targeting Uptone.

[beerandmusic]

3 minutes ago, mansr said:

I can assure you I am not trying to steal anything from anyone. If I were, I wouldn't be targeting Uptone.

 

just toying with you, trying to get more info...lol....all the same, i am sure you are learning some of the pitfalls of mfg already (wink)...many different than software development.

[sandyk]

44 minutes ago, mansr said:

In those cases, what good does the added LT3045 do?

Let's just say that it's not something that I would do.

[mansr]

29 minutes ago, sandyk said:

Let's just say that it's not something that I would do.

For once, we are in agreement.

[Abtr]

On ‎7‎/‎5‎/‎2018 at 11:28 PM, [email protected] said:

... large capacitors need to be used before the linear regulator, but not after, and its better to follow manufacturers demo boards design ...

Hi Alexey,
I'm planning to use a single 15V LT3045-A board to power a pair of active crossovers that drive my subwoofer(s) and stereo power amp to main speakers (https://www.xkitz.com/collections/active-crossovers-and-bi-amplifiers-1/products/linkwitz-riley-2-way-active-crossover-fully-assembled-xover-2?variant=5286065569825). Each (100mA) crossover has an on-board LM317DCYR voltage regulator which is implemented as shown below:
 
The 15V output of the LT3045-A will be connected to VIN and GND. To bypass the LM317DCYR I was advised to short pin 3 and pin 2. Should I also remove resistors R63 and R64, caps C15 and C65, and any other capacitors across VIN and GND? Thanks. 

[Monge]

Hi,

I have now tried a kwmobile  POE set and 1 m uGreen cat 7 flat 26 awg on a TP-Link TL SG108E switch with SBooster (BOTW P&P ECO) 9V. It gave me a another lift In SQ.

So that made me wonder why it works so Well with the SBooster unit.

 

So I tried the POE tweak In another setup after an IFI iPower and no magic happened atall.

Hmm...

So I tried a SBooster single unit after the iPower and then the POE set + 1 m ugreen cat 7 flat 26 awg. Then the magic start happening.

My humble observation.

Regards Monge

[BigGuy]

Enthusiastic review of Paul Hynes SR4 LPS on Audio Bacon...

 

https://audiobacon.net/2018/07/11/paul-hynes-sr4-the-worlds-best-audiophile-power-supply/

[beerandmusic]

 

I "think" i believe that these "DIY" cables do improve SQ, but I have no desire to build my own cable...unless it is substantially cheaper, and "very easy" (less than 30minutes) to build?

 

Superdad said in another thread that he does offer a "quad" cable of some sort that is similar to these DIY cables, and someone else mentioned Ghent.

 

Can anyone confirm that these "off the shelf" cables are the same concept that many here are experiencing "notable" improvements with? 

 

Can someone please share actual links or specific cables that do the same?

That or if someone can summarize a 30minute step-by-step substantial savings DIY cable with parts breakdown?

 

 

 

[ducatirider]

I was curious if anyone knows of a source for silver stranded star quad cable that might be suitable to a DC cable.  Or do you think it makes any difference.  I ask because reviews of the SoTM SPS-500 shows that the silver cable dc cable is an improvement over the copper.

[Cornan]

57 minutes ago, ducatirider said:

I was curious if anyone knows of a source for silver stranded star quad cable that might be suitable to a DC cable.  Or do you think it makes any difference.  I ask because reviews of the SoTM SPS-500 shows that the silver cable dc cable is an improvement over the copper.

 

Ghent already do a silver plated starquad cable.

https://www.ghentaudio.com/part/dc-j01.html