Overall Isolation - network, USB, and power

[fob69]

1 hour ago, fob69 said:

First, i took them away 6 monthes ago and replace them with an etalon isolator with a huge benefit

(-> I was talking about MFCs)

 

Here is the MC110Cs switching supply bypass.

 

depending on diods, voltage drop out can be 0,5V to 0,7V. Here with my 5V linear supply and 2 diods final voltage is 3,5V and the MFC works fine. He did not with 5V (no diods) and 2, 9 V (3 diods). 

 

[jelt2359]

Can't wait for your other experiments! As requested, here's what you PM-ed to me:

 

Hi, lps is very expensive. 

It is possible that the inner noise of the tp link (all component noise) would be much higher than the output noise of the lps1

I am going to go further in noise reduction tests with the tp link and it would be usefull for you before you spend a lot of money.

1) i am going to replace the 200uV noise existing LDO with a 4uV noise LDO to see if it has an impact.

2) i will do a supply test with a 3,3v lifepo4 battery that the lowest noisy supply that can be.

3) i will see if i can remove the green light diods

 

I could tell you if it worth or not to use a low noise expensive supply.

 

I will depend too on your player and usb card. If they are badly supplied, it will ruin the tp link upgrade.

[fob69]

I smoked my TP7A4700 regulator, so i did not test the MC110Cs LDO replacement (only switching regulator replacement). 

After several days of use without the switching regulator, i have to admit that there is still some high frequency noise with the MC110CS, and the sound change when i switch half / full duplex optical parameter. I think this HF noise is generated by the lasers and not the supply (and removing the green leds won't of course change anything).

 

I did open my D-link DGS105 switch and a dexlan hub : they both have 350khz switching regulators.

 

I found and test an other MFC that has no regulator at all (direct 3,3V input) :  http://www.delock.de/produkte/G_86443/merkmale.html?setLanguage=en  In half duplex mode, the SQ is much better than the tplink and all others network équipements i did test upstream my player.

It is more expensive than the TP link MC110CS but it it compliant with it so i did only buy one for the downstream side.

I did supply it for the test with a A123 lifepo4 cell that has a too short capacity for all day use. I will use it with a high capacity lifepo4 battery (+ regulator or drop out diods) of linear supply depending on the effect on sound. (The sound is horrible with the included wallwart switching 3,3V supply).
 

B.

[mourip]

On 1/31/2017 at 2:46 PM, Cornan said:

 

This is what I thought as well before trying out the FMCs.

I did'nt want to add them to my chain until I got some interesting reports that 100Base-TX reduced the power consumption of Auralic Aries. I decided to give the TP-Link MC100CMs a try with a couple of short DIY 100Base-TX Ethernet cables. I was'nt over the moon initially but I noticed that it sounded overall more "real" which gave me the will to carry on experimenting. I clearly noticed that powering the uphill FMC by battery unconnected to mains improved things further in my setup. The thing that actually blew my mind was when I installed a galvanicly isolated lab PSU to power my router. Wow, that thing really added all the pieces of the puzzle for me. EVERYTHING changed for the better! I have to pinch my arm everytime I listen to music now...and I still haven't "got use" to the very REAL feeling of the music which normally happen with updates. I am just keep on being amazed!

 

I have experienced the same thing as you with the USB GND lift. It actually sounds better with GND in my setup now and my Entreq Minimus grounding box have left my primary setup since it no longer makes any improvements.

 

My two best investments have been those that galvanicly isolate my AC mains connected devices. The Voltcraft FSP-1134 and the Peaktech2240.

 

Here is a spagetti image of my current setup to give you an idea :

 

 

Bottom line: It is the sum of all isolations that really matters...and AC mains isolation and limited ground paths is the primary target IMO.

 

 

 

Sent from my Nexus 7 using Computer Audiophile mobile app

 

Late to the table here...

 

Did you find that the DC Blocker improved sound quality or just minimized IT transformer physical humm?

[sgr]

20 minutes ago, mourip said:

 

Late to the table here...

 

Did you find that the DC Blocker improved sound quality or just minimized IT transformer physical humm?

I’ve had similar results with the fiber and FMCS. I use Uptone lps-1 bear DAC and battery at the other end. Adds more realism, depth, and liquidity for sure. 

[Cornan]

37 minutes ago, mourip said:

 

Late to the table here...

 

Did you find that the DC Blocker improved sound quality or just minimized IT transformer physical humm?

 

I still use the DC Blocker trap filter, but now in two spots. Pre my floating LPS powering my router and pre my balanced isolation transformer with floating center-tap. I did'nt have any hum problems on these spots. It increase SQ though.

 

I did try to cure my Ultra ITs hum with a DC Blocker trap filter. It did'nt reduce the hum though, but increase SQ. I heard that a DC Blocker trap filter is more efficient in 220/240v countries and less so in 120v countries.

[tboooe]

On 11/17/2017 at 11:20 PM, fob69 said:

 

I found and test an other MFC that has no regulator at all (direct 3,3V input) :  http://www.delock.de/produkte/G_86443/merkmale.html?setLanguage=en  In half duplex mode, the SQ is much better than the tplink and all others network équipements i did test upstream my player.

It is more expensive than the TP link MC110CS but it it compliant with it so i did only buy one for the downstream side.

I did supply it for the test with a A123 lifepo4 cell that has a too short capacity for all day use. I will use it with a high capacity lifepo4 battery (+ regulator or drop out diods) of linear supply depending on the effect on sound. (The sound is horrible with the included wallwart switching 3,3V supply).
 

B.

Very interesting FMC.  Can you please provide your thoughts on the sound with the Delock?  

[fob69]

15 hours ago, tboooe said:

Very interesting FMC.  Can you please provide your thoughts on the sound with the Delock?  

 

Hi. I was first very happy with the MC100cs supply uprage and i confirm the benefit on image abd definition, but there is still a problem in high spectrum probably due to and hf noise source in the mc110cs. I thougth this would be resolved with hours of use but no.

Delock has a strong effect on image  (deft) and definition without the MC110cs high spectrum artefact. I definitely will keep it (in addition to my etalon isolator)

I quickly built a 3,3v supply with two lifepo4 and a loader to test it and the sound is so good i don't stop to load it ...

i am going to build a clean battery for 7 hours listening, replacing the 5v one i had built for MC110cs and that i don't use anymore (I don't use regulators).

 

i confirm that the delock has no inner regulator att all.

 

B.

[tboooe]

7 hours ago, fob69 said:

 

Hi. I was first very happy with the MC100cs supply uprage and i confirm the benefit on image abd definition, but there is still a problem in high spectrum probably due to and hf noise source in the mc110cs. I thougth this would be resolved with hours of use but no.

Delock has a strong effect on image  (deft) and definition without the MC110cs high spectrum artefact. I definitely will keep it (in addition to my etalon isolator)

I quickly built a 3,3v supply with two lifepo4 and a loader to test it and the sound is so good i don't stop to load it ...

i am going to build a clean battery for 7 hours listening, replacing the 5v one i had built for MC110cs and that i don't use anymore (I don't use regulators).

 

i confirm that the delock has no inner regulator att all.

 

B.

Thank you very much!  I will look to try one.  I noticed that there are no US dealers so hopefully I can buy one from a EU dealer and the shipping wont be too bad.

[tboooe]

1 hour ago, tboooe said:

Thank you very much!  I will look to try one.  I noticed that there are no US dealers so hopefully I can buy one from a EU dealer and the shipping wont be too bad.

After checking out the Delock website, this is the FMC I will try out (86440).  

http://www.delock.de/produkte/G_86440/merkmale.html

 

I assume these are compatible with Cisco SFP?  Or do I have to buy the Delock one?

[Tecnik1]

Grounding the router made a big difference for me in SQ, like the previous post have stated get rid of the noise further up the chain.

[Tecnik1]

Forgot to mention the router is also on a LPS.

[fob69]

18 hours ago, tboooe said:

After checking out the Delock website, this is the FMC I will try out (86440).  

http://www.delock.de/produkte/G_86440/merkmale.html

 

I assume these are compatible with Cisco SFP?  Or do I have to buy the Delock one?

 

They are several different ones and you have to check if it is compliant to your network (mono / multi mode and other options) but if caracteristics are ok it will be compliant. IME i use the MC110Cs on up stream side.

 

After new tests yesterday evening : i don't know why but the level of noise on this MFC change, IME, depending on the input voltage. It become magic with an input voltage < 3,2 V (and it cuts at 3V), and this is a brutal change.

Half / full duplex option on the upstream MFC change the level of noise too !!

 

B.

[sandyk]

On ‎13‎/‎11‎/‎2017 at 1:18 AM, jelt2359 said:

I am going to go further in noise reduction tests with the tp link and it would be usefull for you before you spend a lot of money.

1) i am going to replace the 200uV noise existing LDO with a 4uV noise LDO to see if it has an impact.

2) i will do a supply test with a 3,3v lifepo4 battery that the lowest noisy supply that can be.

3) i will see if i can remove the green light diods

 

 

1)  There is a very good chance that it will. See also, threads about improving low noise regulators such as the LT3045 with additional input and output filter capacitors.

2) There are voltage regulators available these days that are quieter than any type of battery, and with the very important fact that they maintain a constant voltage output, unlike the voltage of a battery that drops with use, and perhaps also sudden changes in load current.(internal resistance ,including connector resistances too)

3) Modern LEDs are not noisy, and are often used as low noise voltage references. Supplying them via a relatively noisy power supply may see RF/EMI radiation from their connecting leads though.

 

Alex

[fob69]

8 hours ago, sandyk said:

2) There are voltage regulators available these days that are quieter than any type of battery, and with the very important fact that they maintain a constant voltage output, unlike the voltage of a battery that drops with use, and perhaps also sudden changes in load current.(internal resistance ,including connector resistances too)

 

Hi Alex,

 

My suggest about removing the leds was stupid considering that lasers probably generates a huge noise compare to the Leds.

For a voltage reference (dac chip) i agree that battery is not the solution, but with LDO you loose the very low output impedance and i think it has a big impact on the noise reduction. I did test different players (squeezebox, raspberry, Olimex) with lifepo4 batteries and with a LPS with a TPS7A4700 regulator (10uV), the battery did win each time and i don't think it was because of the 10uV or more LDO noise. A player or a MFC will not have big current load changes (as there are full digital there is not impulse response to manage) and if there are small current change they won't be significant due to the very low output impedance.

But yes, batteries are a bit complex to manage for an all day use and there are security concerns too.

 

What depress me at the end is that data transfert is noise by itself whatever will be the supply or upgrades ...  The name of this forum could be : "Computer audiophile, manage the noise of digital audio"

[sandyk]

6 minutes ago, fob69 said:

"Computer audiophile, manage the noise of digital audio"

 

 LOL !

 

Re the output impedance of typical Voltage Regulators. That is why I use the JLH PSU Add-on after voltage regulators.

 

[fob69]

There is really something to build for those noise concerns on data input level : a player with low noise optical TCP/IP input, or a minimalist player with a dual memory load, mecanicaly discoupling the one that is loading once the other is playing on the same way uptone did it with their supercap supply.

Manage the noise and isolation lower in the chain (usb and spdif) seams to me more hazardous and difficult because of real time / jitter concerns even if it gives the benefit on the paper not to take care to everything that is upstream.

 

[fob69]

I rebuild a battery supply proto for the delock.

It has very low noise when supplied at 3v <> 3,18 V.

My new battery is : Lifepo4 6,4V >> TPS7A4700 ouptut 3,3V >> two germanium low drop out diods in parrallel to get a 3V output.

With a 3,1V out TPS7A, i can heat it.

With a 3,1V out TPS7A + lifepo4 in low charge (..3,1V), i can heat the TPS7A.

With the 3,3V TPS7A + lifepo4 in high charge (3,3V) + germanium diods out 3,1V i don't hear the TPS7A anymore and the delock isolation is great.

 

I am gonna package it now.

 

[hurka]

My spaghetti done!!!

[charlesphoto]

Fob69,

 

SWitched my TP Link FMC’s to half duplex mode and that’s made a big difference - fuller, less hash up top, doesn’t sound as hollowed out. Will be interesting to try a delock on the recovering side. Will have to wait for our next family trip to Germany or my wife’s sister visiting, so at least six months out. But that way I can avoid shipping charges to US. 

 

Thanks for the tips!

[sgr]

2 minutes ago, charlesphoto said:

Fob69,

 

SWitched my TP Link FMC’s to half duplex mode and that’s made a big difference - fuller, less hash up top, doesn’t sound as hollowed out. Will be interesting to try a delock on the recovering side. Will have to wait for our next family trip to Germany or my wife’s sister visiting, so at least six months out. But that way I can avoid shipping charges to US. 

 

Thanks for the tips!

How do you switch FMC to half duplex load? Thanks,

and Merry Holidays!

[tboooe]

2 minutes ago, charlesphoto said:

Fob69,

 

SWitched my TP Link FMC’s to half duplex mode and that’s made a big difference - fuller, less hash up top, doesn’t sound as hollowed out. Will be interesting to try a delock on the recovering side. Will have to wait for our next family trip to Germany or my wife’s sister visiting, so at least six months out. But that way I can avoid shipping charges to US. 

 

Thanks for the tips!

Here is an interesting tid bit....recently I had a little electrical mishap that burned out my downstream FMC connected to my NAA.  This forced me to switch from the dual port fiber nic in my HQP server PC to the standard copper mobo ethernet port to connect my NAA.  To my great surprise, I greatly preferred the sound in this configuration. The music was more immediate, visceral, and realistic. I am not sure how much impact removing the FMC or changing from the fiber port to the copper port had to do with this.  I will be getting the Delock after the Xmas to see how this compares.

[charlesphoto]

7 minutes ago, sgr said:

How do you switch FMC to half duplex load? Thanks,

and Merry Holidays!

 

Switch #3. 

[fob69]

50 minutes ago, charlesphoto said:

 

Switch #3. 

Sorry i did not see the question.

 

Did someone test this delock ? I would be happy to have feedback from other listener.

 

I did a lots of tests since one month on the supply with mine. I am building another  lifepo4 supply for it. The result is magic but it is not trivial because the great noiseless sound is between 3v and 3,18v and all the linear regulators i tried (tps7a4700, lm317, lt1086) in floating charge gave at the end a not so good result than the lifepo4 alone.

i use parrallel lowdropout diods after mylifepo4 to decrease the voltage and with 4 lifepo4 3,2v in parrallel i have hours of very stable voltage.

 

Bernard

[fob69]