The new generation UltraCap LPS-1.2: USER IMPRESSIONS and QUESTIONS thread

[JohnSwenson]

I can't prove it yet (I'm working on it) but there seems to be two primary sound impactors from the network side: high impedance leakage from SMPS attached to network equipment, AND the phase noise of the clock that clocks out the data bits on the line.

 

The optical interface completely blocks any leakage coming from upstream network devices, but it ADDS the phase noise of the clock in your FMC. It is certainly possible that this extra phase noise has a higher impact on sound than the decrease leakage the optical brings.

 

BTW the difference in "picking up" noise between a copper cable and the optical is quite small.

 

John S.

[Superdad]

5 hours ago, cat6man said:

For some reason, I felt it was time to order a second lps-1.2 to power the ultraRendu and wow, it only took one

track to convince me this was long overdue.  Listening to Joan Baez sing "It's All Over Now, Baby Blue" showed instantly the improvements, improving vocal depth (her voice moved forward out of the mix and had more micro-detail) with no trace of harshness (Baez' voice pierces my eardrums when not done right, so this is one of my definitive tests tracks for that reason) and separates instruments more clearly. 

 

NICE!!  

(And many thanks for your multiple orders.  Still I can never remember who's CA member name is which person in real life...)

[cat6man]

7 hours ago, JohnSwenson said:

I can't prove it yet (I'm working on it) but there seems to be two primary sound impactors from the network side: high impedance leakage from SMPS attached to network equipment, AND the phase noise of the clock that clocks out the data bits on the line.

 

The optical interface completely blocks any leakage coming from upstream network devices, but it ADDS the phase noise of the clock in your FMC. It is certainly possible that this extra phase noise has a higher impact on sound than the decrease leakage the optical brings.

 

BTW the difference in "picking up" noise between a copper cable and the optical is quite small.

 

John S.

 

I guess the thing that surprises me is that the phase noise of the clock in the FMC can make a difference when it is being reclocked not once but twice more!  (first in ultraRendu and then again in digital-digital converter to aes/ebu).

 

How can multiple re-clockings still be sensitive to the ethernet clock?

[agladstone]

John:

So, without the Optical FMC in the chain (as discussed above), would the “phase noise of the clock “ then be coming from the router (or switch) ? 

Would an in-line Ethernet isolator have any impact on improving the phase noise of the clock from a router or switch? 

[JohnSwenson]

43 minutes ago, agladstone said:

John:

So, without the Optical FMC in the chain (as discussed above), would the “phase noise of the clock “ then be coming from the router (or switch) ? 

Would an in-line Ethernet isolator have any impact on improving the phase noise of the clock from a router or switch? 

Note that I do NOT have detailed explanations of everything right now, this is conjecture based on preliminary findings, there is a LOT more research to do before I can really say what is happening.

 

What I am going to be mentioning is what seems to be happening, NOT detailed explanations of mechanisms at this point, I can't give those yet.

 

All oscillators, some much more than others have what is called phase noise, this is a different way of looking at jitter, which seems to have a higher correlation with sound than traditional jitter numbers. The oscillators in inexpensive equipment (such as inexpensive network gear normally used in homes) are usually on the very large phase noise side of things. The data coming over any digital stream (Ethernet, USB, whether copper or optical) is clocked out of the source component by the oscillator in that component, thus the phase noise of the oscillator gets "baked in" to the data stream, the arrival time of the data edges varies slightly, ie jitter which is determined by the output clock. When that data goes into another device, some how some of that noise winds up adding to the phase noise of the local oscillator in that component. As I mentioned the exact mechanism for this is not known, I'm working on it.

 

So even if you have a very low phase noise local oscillator it can get swamped by what is coming in from upstream sources. This adding of upstream noise to the local clock seems to vary from device to device. I have no idea what causes the difference, that is one of the things I'm trying to figure out.

 

The passive Ethernet isolators are just transformers and have no impact on any clocking issues. They have a slight impact on leakage currents.

 

So the phase noise that make up this additive upstream noise can come from a router, a switch, and other devices plugged into the switch. In a normal "home network" it can be coming from all kinds of things, adding together in various amounts.

 

I can't give any definitive rules at this point, I don't know what the mechanisms are yet, so all anybody can do right now is try different things.

 

One thing to be careful about is coming up with your own theories and making large buying decisions based on those theories. There is a high probability any theories that are come up with right now will be wrong, that's just the way these things go. In a couple years things should be different. But for now it is the wild west.

 

John S.

[AlanTan]

hi,John: Does it need to turn off the power when it is not used ? Will it affect the life span if it stays open? tks!

[agladstone]

2 hours ago, JohnSwenson said:

Note that I do NOT have detailed explanations of everything right now, this is conjecture based on preliminary findings, there is a LOT more research to do before I can really say what is happening.

 

What I am going to be mentioning is what seems to be happening, NOT detailed explanations of mechanisms at this point, I can't give those yet.

 

All oscillators, some much more than others have what is called phase noise, this is a different way of looking at jitter, which seems to have a higher correlation with sound than traditional jitter numbers. The oscillators in inexpensive equipment (such as inexpensive network gear normally used in homes) are usually on the very large phase noise side of things. The data coming over any digital stream (Ethernet, USB, whether copper or optical) is clocked out of the source component by the oscillator in that component, thus the phase noise of the oscillator gets "baked in" to the data stream, the arrival time of the data edges varies slightly, ie jitter which is determined by the output clock. When that data goes into another device, some how some of that noise winds up adding to the phase noise of the local oscillator in that component. As I mentioned the exact mechanism for this is not known, I'm working on it.

 

So even if you have a very low phase noise local oscillator it can get swamped by what is coming in from upstream sources. This adding of upstream noise to the local clock seems to vary from device to device. I have no idea what causes the difference, that is one of the things I'm trying to figure out.

 

The passive Ethernet isolators are just transformers and have no impact on any clocking issues. They have a slight impact on leakage currents.

 

So the phase noise that make up this additive upstream noise can come from a router, a switch, and other devices plugged into the switch. In a normal "home network" it can be coming from all kinds of things, adding together in various amounts.

 

I can't give any definitive rules at this point, I don't know what the mechanisms are yet, so all anybody can do right now is try different things.

 

One thing to be careful about is coming up with your own theories and making large buying decisions based on those theories. There is a high probability any theories that are come up with right now will be wrong, that's just the way these things go. In a couple years things should be different. But for now it is the wild west.

 

John S.

Thanks John! 

This is very interesting information. 

So, it seems to me from what you’re saying, that it would be best (for sound quality), for me to unplug the Ethernet cable that goes into my Aurender server / streamer from my home router when I’m listening to music stored on its internal hardrive (the Aurender connects to my DAC via USB), and only plug it in when streaming from Tidal ? 

I know at this point you’re only making educated guesses on what causes the phase noise and how it effects downstream components, but if I’m not mistaken, it seems logical to at least infer from your findings, that my simply removing the Ethernet cable going into the Aurender while listening to locally stored music, I would be eliminating the potential for any negative effects from the network and router upstream? 

[R1200CL]

Hm....it seems the development of the EtherRegen will still take some time ?

 

 

[R1200CL]

14 minutes ago, agladstone said:

it seems logical to at least infer from your findings, that my simply removing the Ethernet cable going into the Aurender while listening to locally stored music, 

 

I can’t see the logic ?

 

Cause the source is your HD, not your Ethernet.

 

[agladstone]

35 minutes ago, R1200CL said:

 

I can’t see the logic ?

 

Cause the source is your HD, not your Ethernet.

 

John was saying (from my understanding) that anything attached downstream of a home network device (router, switch, etc), will most likely be negatively effected by the Phase noise from the oscillators in the home network device. 

So the logic I was using is that if I’m only listening to music stored on the Aurender’s internal hard drive, seems like if I disconnect the Ethernet cable that I currently have attached from my home router, that it would increase the sound quality. 

I always keep it attached (as of now), but I rarely even remember that I have Tidal!! 

[Superdad]

6 hours ago, R1200CL said:

Hm....it seems the development of the EtherRegen will still take some time.

 

Actually, development is proceeding nicely. This next week we receive our third round of PCBs to conduct the next set of tests on. John did make a topology change a few weeks ago—for greater overall performance of all ports—and this necessitated addition of a second special switch chip (and it took me two weeks to get through the NDA process to obtain the full datasheet for it).  So yes, we are a some months behind schedule—I was originally expecting we would begin shipments this month. But given how every technical aspect of EtherREGEN will be so completely different from any other switch presently offered, we thing it will be worth the wait. We intend to present a block diagram and some explanations closer to the launch (which we are pretty confident will be February). 

[Bruce Orr]

Sorry, if I missed this, Alex. Will your EtherRegen have a Reference clock input port?

[octaviars]

Could you @Superdad not re-open the EtherREGEN section and post the answers you have so they are not scattered all over the forum and then close the thread again if you wish?

[Superdad]

5 minutes ago, octaviars said:

Could you @Superdad not re-open the EtherREGEN section and post the answers you have so they are not scattered all over the forum and then close the thread again if you wish?

 

Funny, while in the shower this weekend I was thinking (isn't that where we all do our best thinking?) of doing that--and then promptly forgot.  Reopened the thread just now, thanks! 

 

[Superdad]

1 hour ago, Bruce Orr said:

Sorry, if I missed this, Alex. Will your EtherRegen have a Reference clock input port?

 

Yes, it will, though it will already have built in what is about the lowest phase-noise production XO on the market, the Crystek CCHD-575.

As important is how clocking is handled.  Because EtherREGEN will have a true digital isolation "moat"--with expensive high-speed digital isolators and differential flip-flops (as opposed to just Ethernet transformers modules for isolation), there has to be separate clocking on each side.

There are 5 clocks total (at 3 different frequencies), so we use the excellent (and again expensive and hard to get) Silicon Labs clock synthesizer (actually 2 of them).  Since those each have 4 outputs and 2 inputs, we will use one input for the CCHD-575 and program the other input to accept an external reference 10MHz clock if people wish to get extreme. 

 

All that said, excellent clocking is but one aspect of what will make this switch special.  A whole lot else will contribute to the performance.  As we get closer, we will likely publish a block diagram to explain a bit more about the topology of this unique switch.  

 

 

[Bruce Orr]

Thanks, Alex. Look forward to Feb when you hit the market and we can hear the results of this labor of love.  Since I’ve already invested in the REF-10 it will be easy to A/B and pick the best path for my system and ears! 

[feelingears]

14 hours ago, Superdad said:

All that said, excellent clocking is but one aspect of what will make this switch special.  A whole lot else will contribute to the performance.  As we get closer, we will likely publish a block diagram to explain a bit more about the topology of this unique switch.  

 

@Superdad and @JohnSwenson, when you have a moment, I'm curious as to if the digital phase noise propagates over Wi-Fi as well?

 

Due to my home's floorplan and ancient construction, my chain is all consumer-grade Macbook server over Ethernet to Orbi mesh network to Airport Express. From AEx it's optical out but the last step before the DAC is a Wyred 4 Sound Remedy (signal re-clock and upsample to 96kHz) powered by an LPS-1. I've heard a few great systems lately and my setup (including power and interconnect cables) easily hangs with them. Sure I'd like "more there there," which is why I'm asking/wondering.

 

I realize I could reposition the Orbi receiver to remove the Airport Express from the chain and thus run Ethernet from Orbi directly to the upcoming Uptone Ether-thingy, but from there I'd need a new renderer (strongest candidate for me would be Auralic Aries for both h/w and s/w). But I just don't know if the SQ would step up enough.

 

Anyway, to stay on-topic, I plan to add an LPS-1.2 to power my W4S Remedy soon! Too bad the Orbi is 12v/3A otherwise I'd try the LPS-1.2 on my wi-fi.

 

[wanta911]

I'm really liking the pairing of the LPS 1.2 with my Chord Qutest DAC, it gets really hot but it sounds great and no issues until last night.

 

After about 5 and a half hours of music playback the LPS 1.2 would stop powering the DAC for approx 15-20 seconds then power back on for the same amount of time and it kept repeating this cycle. It almost seemed as if the LPS had exhausted it's power reserve? Anyway I switched the LPS off and switched it on again this morning, so I'll see how it goes tonight. I know Rob Watts said that 1.1 amps would be fine but perhaps there is a limit with this pairing?

 

I would be interested in any other users impressions that are using the LPS 1.2/Qutest combo.

[pl_svn]

5 hours ago, wanta911 said:

I would be interested in any other users impressions that are using the LPS 1.2/Qutest combo.

 

been using it since late February without the faintest glitch ever, sorry

worried, at first, about the LPS getting really hot but as Alex explained many times it is nothing to worry about

[wanta911]

OK so there is a problem. It is not losing power for 15-20 seconds....that is how long the DAC takes to power up so the intermittent loss of power is a second or two. I have also now tried it on my MiniDsp unit that takes 5V and only draws 0.5A and the same problem is occurring. Now when I try to power up the Qutest, the startup light sequence is flickering and it won't actually get to run mode, the sequence keeps starting over. The Qutest is fine with 2 other LPSs I've just tested on it. It's strange, it behaves like when you have a faulty wire or a loose connection?

 

No flashing lights on the LPS-1.2 when all of this is happening - solid green light all the way.

 

It'll be interesting to see what Alex thinks. It's a good thing I have a plethora of power supplies on hand to carry on ?

[Superdad]

14 hours ago, wanta911 said:

OK so there is a problem. It is not losing power for 15-20 seconds.

 

Hi:

Sorry you are having trouble.

I have family visiting from out-of-town and also a bit of a cold.  So got behind on e-mail of the past two days.  Just spent the whole afternoon catching up, so if you already had sent a message then I am sure you got a reply about quickly replacing your unit. (I can't connect CA screen names to real names in e-mail.)

If you have not already sent a message to me, then please do so.  

Thanks and have a great weekend,

--Alex C.

[wanta911]

46 minutes ago, Superdad said:

 

Hi:

Sorry you are having trouble.

I have family visiting from out-of-town and also a bit of a cold.  So got behind on e-mail of the past two days.  Just spent the whole afternoon catching up, so if you already had sent a message then I am sure you got a reply about quickly replacing your unit. (I can't connect CA screen names to real names in e-mail.)

If you have not already sent a message to me, then please do so.  

Thanks and have a great weekend,

--Alex C.

 

Hey no sweat Alex, I haven't sent a message because I haven't given up yet and intend to give it another go later this weekend before hitting the panic button. As I said, there are no flashing lights so it might be something simple.

 

Enjoy your weekend....the JS-2 is doing a fine job in it's place ?

[littlecx]

regarding the switch, crystek has a new 575X, why still use 575?

how much it cost more if use pulsar clock?

[Superdad]

9 hours ago, littlecx said:

regarding the switch, crystek has a new 575X, why still use 575?

 

Crystek CCHD-575X is not new.  It is the same part at CCHD-575 without suffice, just rated for more extreme temperature range.

Regular 575 is rated 0-70 degrees C;

575M is -20 to 70C

575X is -40 to 85C

Phase noise performance an frequency stability within operating range is exactly the same.  Crystek confirmed this to me personally. 

 

9 hours ago, littlecx said:

how much it cost more if use pulsar clock?

 

Well the Pulsar clock--if they were even willing to make the custom frequency we would need for the EtherREGEN--retails for $332.  Even if they could make them in volume (at our frequency) and gave us 20% off, that is still $266.  So we would have to charge about $450 more for an EtherREGEN will Pulsar clock.  Considering we are trying to keep the price of the EtherREGEN down to less than $600, there is very low likelyhood of our ever offering such a version.  Besides, it is not necessary.  EtherREGEN is about far more than just great clocking, and if someone really wants to go overboard they can run an external 10MHz superclock into it. 

[Franatic]

Superdad, sorry if similar question has already been asked .........

I am interested in Singxer's new SU-6. It's normal operation current draw is well below the 1.2A threshold. However, at turn on it charges a 7.5f supercap which draws from 1.2A to 2A per the manufacturer. If the LPS-1.2 could continue working and limit current draw to its max level while the supercap is charging it could work well with this DDC. Does the LPS-1.2 have this continue working/current limiting capability, or does it just shut down?