cisco 2960 8tc/ vs paul pang tcxo

i had the opportunity to try the dlink tcxo switch ( paul pang) powered by hdplex linear ps.  The price is around 150GBP.  

I had now the cisco 2960 8tc. Alone, it is a bit edgy on my ears. I put an acoustic revive lan isolator between the router and the cisco and also a high quality power cord:  the edginess had gone away.

The tcxo powered by my high quality ps gave a little more details on some tracks but very minor ones. The cisco , as tweaked as described, gives more focus, body and bass, by a little margin.  So i decided finally to keep the cisco.

I also see that there is a more recent version on the 2960 8tc:  2960 8pc from 2015.  Anybody knows it?  is it better or the same?  i am sure Simon will tell us...

With these refurbished cisco, who knows how hard they run,  how long will they work correctly....Sometimes when i plug off then on the lan cables, my server is not found easily.  I have to power off my serve and nds and reinitialize all.   Perhaps it is because the switch is more complex....

FR

Original Post

The Cisco 2960 Catalyst switches are not made anymore, i use two 2960  port switches , one with inbuilt quality shielded PSU (SMPS) and the other externally 48 volt powered and a 3560 8 port Layer 3 Catalyst switch with inbuilt SMPS. My switches..

WS-C3560 8PC-S is revision  A0 

WS- C2960PD 8TT-L is revision C0 (external 48 v DC powered)

WS-C2960 8TC-L is revision C0 ( internal  SMPS)

I connect my 3560 to my NDX currently via a shielded low cost Cat 6A cable.. I find with these switches the sound is clearer and cleaner and less smudgey or blurred (compared to Netgear switches I own) ..  on my setup at least

You might find that very short Ethernet patch leads are not optimum for connection to audio equipment.. also by default the Cisco professional switches will delay enabling a port when a device is powered up or connected as they are doing loop detection analysis that could otherwise bring down your network which is why your NAS might take a little time before appearing... consumer switches don't normally have such safeguards . If you know how to program the Cisco switch you can configure the removal of loop detection on a particular port for almost instant activation of a port.... its very straightforward once you are in Configure Terminal mode.

Pierre, these switches are almost bomb proof, they may well outlive you....

Simon-in-Suffolk posted:

The Cisco 2960 Catalyst switches are not made anymore, i use two 2960  port switches , one with inbuilt quality shielded PSU (SMPS) and the other externally 48 volt powered and a 3560 8 port Layer 3 Catalyst switch with inbuilt SMPS. My switches..

WS-C3560 8PC-S is revision  A0 

WS- C2960PD 8TT-L is revision C0 (external 48 v DC powered)

WS-C2960 8TC-L is revision C0 ( internal  SMPS)

I connect my 3560 to my NDX currently via a shielded low cost Cat 6A cable.. I find with these switches the sound is clearer and cleaner and less smudgey or blurred (compared to Netgear switches I own) ..  on my setup at least

You might find that very short Ethernet patch leads are not optimum for connection to audio equipment.. also by default the Cisco professional switches will delay enabling a port when a device is powered up or connected as they are doing loop detection analysis that could otherwise bring down your network which is why your NAS might take a little time before appearing... consumer switches don't normally have such safeguards . If you know how to program the Cisco switch you can configure the removal of loop detection on a particular port for almost instant activation of a port.... its very straightforward once you are in Configure Terminal mode.

Pierre, these switches are almost bomb proof, they may well outlive you....

ok simon, thanks, if they may outlive me i am reassured. my is 2960-8tc, but i find 2960- 8Pc( so pc not tc) from 2015 year on amazon us. I was searching which models were the more recent.  What is the most recent 8 ports cisco in the league of the 2960 ?   it is difficult to find this information.  Thanks.    I am just curious about.

I am a big fan of the paul pang stuff but have to say his d-link switches are not anywhere near as good as the original zyxels. If you can get a comparison with one it would be well worth it. As I have said previously these produce the best streaming playback I have heard.

RICHYH posted:

I am a big fan of the paul pang stuff but have to say his d-link switches are not anywhere near as good as the original zyxels. If you can get a comparison with one it would be well worth it. As I have said previously these produce the best streaming playback I have heard.

have you compared to cisco 2960 catalyst?   i put a good quality power cord on it and found the sound with more bass and body vs the paul pang dlink with a big linear ps.

Perhaps the zyxel is better, i don't know, but the cisco is a 700GBP switch as new, with complex stuff inside.  But alone, this cisco is a little edgy. With lan isolator before and a good power cord, it has very good drive and dynamics.  The paul pang with big linear ps, is quite as good but a little lean.

Keler Pierre posted:
Simon-in-Suffolk posted:

Hi Pierre, the 2960-X switches are the latest available new

S

yes but what are the today cisco 8 port produced, their name?  perhaps their name are not 2960 or 3560?

As  a major vendor, Cico have several product lines. Look at the 296o-CX for their 8 port Catalyst switches, they also have a Meraki range where I think it's the MS220 that has an 8 port variant, be careful though, Meraki assumes some cloud management. There might be an 8 port Catalyst 2960-X variant, but I can't immediately spot it on CCO.

Eoink posted:
Keler Pierre posted:
Simon-in-Suffolk posted:

Hi Pierre, the 2960-X switches are the latest available new

S

yes but what are the today cisco 8 port produced, their name?  perhaps their name are not 2960 or 3560?

As  a major vendor, Cico have several product lines. Look at the 296o-CX for their 8 port Catalyst switches, they also have a Meraki range where I think it's the MS220 that has an 8 port variant, be careful though, Meraki assumes some cloud management. There might be an 8 port Catalyst 2960-X variant, but I can't immediately spot it on CCO.

i am curious of the 2017 cisco products, but it seems difficult to have these informations on the cisco website or another.

Hi the 2960X switches offer upgraded hardware over the original 2960 switches providing more switching capacity, energy efficient operation, net flow reporting capability and layer 3 capability subject to firmware as well as many other things.

The 3560 also offers energy efficient operation and supports layer 3 functionality..

A layer  3 switch is the term used to describe a router and switch combined (i.e. It can operate at layer 3 of the TCP/IP stack, as opposed to regular switches which operate mostly at layer 2)... though the routing functionality and routing performance on smaller layer 3 switches is usually less than full blown dedicated routers.. also these layer 3 switches aren't the type of devices that can really double up as your broadband ISP router.

Cisco 2960X FAQ.    http://www.cisco.com/c/en/us/p...s/qa_c67-728348.html

Simon-in-Suffolk posted:

Hi the 2960X switches offer upgraded hardware over the original 2960 switches providing more switching capacity, energy efficient operation, net flow reporting capability and layer 3 capability subject to firmware as well as many other things.

The 3560 also offers energy efficient operation and supports layer 3 functionality..

A layer  3 switch is the term used to describe a router and switch combined (i.e. It can operate at layer 3 of the TCP/IP stack, as opposed to regular switches which operate mostly at layer 2)... though the routing functionality and routing performance on smaller layer 3 switches is usually less than full blown dedicated routers.. also these layer 3 switches aren't the type of devices that can really double up as your broadband ISP router.

Cisco 2960X FAQ.    http://www.cisco.com/c/en/us/p...s/qa_c67-728348.html

ok thanks, all is clear now. thanks

Finkfan posted:

Has anyone compared the sound from a 2960 vs something like the little 110D with an ifi ipower? 

But surely these are quite different devices, what is the capability of the 110D? It's not the sound of the switch itself, it's the noise and clock accuracy of the switch itself feeding the network interface card of the streamer... also not quite sure what changing the power supplies hopes to achieve.... unless you are such a poor non compliant EMC device that is radiating from connected leads and polluting your mains...

Finkfan posted:

Has anyone compared the sound from a 2960 vs something like the little 110D with an ifi ipower? 

i had netgear gs105 with ifi power then with my uptone js2:  the cisco was more dynamic, more drive and bass. The netgear/ linear ps nicier, more natural.  After i put acoustic revive lan isolator before the cisco and add a good quality power cord to the cisco:   the edginess was gone.

For cisco alone vs your 110d with ifi power, cisco is better.

I wouldn't recommend the use of acoustic revive as my version appeared to push the Ethernet link to operate outside specification, and also interfere with some low energy line driver techniques on some switch devices which will work against reducing Ethernet line current and therefore noise.

if you want to use some sort of increased galvanic isolator for safety reasons etc ensure it's active.

i found on the Cisco 2960 devices no difference between 48 volt DC decoupled powersupply as used by PoE and integrated high quality (low emission) PSU [the 2960 EMC support appears quite extensive as I would expect for quality robust commercial equipment ]. I have not experimented with modular multiple redundant power supplies.

Further if using an external PSU for what ever reason for an Ethernet PD (powered device), I would ensure you use a PSU specifically designed for that PD i.e. Switch and designed to ensure the EMI compliance is met for that PD otherwise you may well be creating EMI issues elsewhere. For these devices it's not really about having a PSU with a linear transformer, that's almost irrelevant. You should ensure that the PSU you use when connected to your switch/router etc has been confirmed to meet Class B EMI spefications. Most quality switch vendors will specify compliant PSUs that have been tested to meet this if decoupled from the device itself. Also on 48 volt PoE devices you would expect an incorporated quality filtered DC to DC buck converter to be used that meets the EMI specifications.

Simon,

I view this whole area with some suspicion.I have been trying to address edge in my system for the best part of a year. I have more or less solved this by addressing my earthing. I believe this was an issue introduced via floating earth in DC powered devices. As a part of testing the 2960 I am going to review all the bits and pieces I have introduced in the last seven months. But, I will start by simply slotting in the new (to me) switch.

M

Simon-in-Suffolk posted:

I wouldn't recommend the use of acoustic revive as my version appeared to push the Ethernet link to operate outside specification, and also interfere with some low energy line driver techniques on some switch devices which will work against reducing Ethernet line current and therefore noise.

if you want to use some sort of increased galvanic isolator for safety reasons etc ensure it's active.

i found on the Cisco 2960 devices no difference between 48 volt DC decoupled powersupply as used by PoE and integrated high quality (low emission) PSU [the 2960 EMC support appears quite extensive as I would expect for quality robust commercial equipment ]. I have not experimented with modular multiple redundant power supplies.

for me it works very good, between router and switch.  More fluency, softness of tones, and same dynamics.   But when i put the acoustic revive before the serve, the sound quality decreased...

Mr Underhill posted:

Simon,

I view this whole area with some suspicion.I have been trying to address edge in my system for the best part of a year. I have more or less solved this by addressing my earthing. I believe this was an issue introduced via floating earth in DC powered devices. As a part of testing the 2960 I am going to review all the bits and pieces I have introduced in the last seven months. But, I will start by simply slotting in the new (to me) switch.

M

i bought for around 50GBP good used power cord on bay for the cisco. It put off some edginess from the cisco.   But perhaps you will not hear this edginess....

Mr Underhill posted:

Simon,

I view this whole area with some suspicion.I have been trying to address edge in my system for the best part of a year. I have more or less solved this by addressing my earthing. I believe this was an issue introduced via floating earth in DC powered devices. As a part of testing the 2960 I am going to review all the bits and pieces I have introduced in the last seven months. But, I will start by simply slotting in the new (to me) switch.

M

Mr U if you are happy with your sound and performance I certainly wouldn't dabble here. However some cheaper consumer network equipment are not that great for radiated emissions... they are legally compliant (I'm sure), but EN55022 Class B allows quite a tolerance. Certainly if you read some of the TI engineering papers on reducing Ethernet (and USB) radiated emmisions you can see why there is a huge variance and reduced performance on cheaper devices... and why having a trailing DC lead from a powersupply to a consumer switch is inviting challenges.

But good luck with your trials, and yes most of the commercial mains powered Cisco devices that I am aware of are safety earth grounded... not floating... which would not be desirable at all.

2960 vs Netgear + LPSU

Well, this is all quiet subtle.

The 2960 is a nicely built little switch. In terms of what I am hearing:

2960
Music feel a bit more relaxed
Some detail comes through more easily
Some high frequencies a tad more strident

Netgear+LPSU
Slightly more dynamic
Slightly more transparent

Examples:
I have a number of 'difficult/interesting files' that I use for demoing. I will mention two here.

Grace - Jeff Buckley 16/44.1
From about 4.03 Buckley builds, almost screaming into the mic by the end.
He backs himself multiple times as the crescendos are reached.
2960 - I felt that the layers as Buckley backed himself were revealed slightly more easily
NG - as Buckley began screaming into the mic the distortion was slightly better handled here

Rachmaninov Symphony No 1, 1st Movement, HDTT, Ormandy 9624
Love this Symphony. This rendition can sound a bit lively.
About 6 mins in a xylophone joins the percussion top left.
2960 - Some notes are strident, zinging.
NG - Doesn't happen, notes remain controlled.

Currently I am sticking with the NG, but, I am VERY tempted to strip out the SMPS from the cisco and run it using an LPSU - just doing some research.

M

 

P.S. Keler - I have one fancy main cable, never heard it make a difference to anything. I did try it with the 2960 and thought it may have made a smidgen of +ve change .....all VERY subtle, and would have hated to try it double blind.

Mr U - I have probably missed it - what are you streaming into? - and which 2960 version are you using .. on the 2960 one variant can run with a 48 volt DC power feed and uses a highly filtered internal buck converter to allow it to be powered by PoE or standalone power supply. I use one of these amongst others. I really would be cautious however in using a  regular linear power supply that has not been designed to work with that specific switch  with out the correct HF load filtering as it might well be  pushing the powesupply/switch setup out side the EMC/RFI specs. With out the right equipment you might not be able to determine this. A Buck converter, in case you don't know, is an internal DC to DC converter switch mode type power supply for transforming voltage to current and vice versa.

A slightly dirtier more RF loaded environment can remove detail and remove edge definition - some might and do actually prefer this - and way back in the early days of the awareness of RFI in hifi this was quite a common talking point..

Mr Underhill posted:

2960 vs Netgear + LPSU

Well, this is all quiet subtle.

The 2960 is a nicely built little switch. In terms of what I am hearing:

2960
Music feel a bit more relaxed
Some detail comes through more easily
Some high frequencies a tad more strident

Netgear+LPSU
Slightly more dynamic
Slightly more transparent

Examples:
I have a number of 'difficult/interesting files' that I use for demoing. I will mention two here.

Grace - Jeff Buckley 16/44.1
From about 4.03 Buckley builds, almost screaming into the mic by the end.
He backs himself multiple times as the crescendos are reached.
2960 - I felt that the layers as Buckley backed himself were revealed slightly more easily
NG - as Buckley began screaming into the mic the distortion was slightly better handled here

Rachmaninov Symphony No 1, 1st Movement, HDTT, Ormandy 9624
Love this Symphony. This rendition can sound a bit lively.
About 6 mins in a xylophone joins the percussion top left.
2960 - Some notes are strident, zinging.
NG - Doesn't happen, notes remain controlled.

Currently I am sticking with the NG, but, I am VERY tempted to strip out the SMPS from the cisco and run it using an LPSU - just doing some research.

M

 

P.S. Keler - I have one fancy main cable, never heard it make a difference to anything. I did try it with the 2960 and thought it may have made a smidgen of +ve change .....all VERY subtle, and would have hated to try it double blind.

if you find a way to put a linear ps on the cisco, i would be very interested. I made researchs but not found: it is 48v....

Remember professional switches like the 2960 are designed somewhat differently - although almost certainly they are utilising internal  buck converters (DC SMPS)  to create internally +12 volts and possibly -52 volts depending on the 48 Volts PoE /DC feed. So if you really want to try and remove any sort switching power supply I would look at adapting the Redundant Power Supply input and ensure you have a model that can take a modular  RPS. I note most of the RPS inputs are 12 Volts at 4 Amps with a few also specifying -52 volts at 8 amps. However this would be very much be in the space of experimentation.  I have not dabbled here......

Mr Underhill posted:

2960 vs Netgear + LPSU

Well, this is all quiet subtle.

The 2960 is a nicely built little switch. In terms of what I am hearing:

2960
Music feel a bit more relaxed
Some detail comes through more easily
Some high frequencies a tad more strident

Netgear+LPSU
Slightly more dynamic
Slightly more transparent

Examples:
I have a number of 'difficult/interesting files' that I use for demoing. I will mention two here.

Grace - Jeff Buckley 16/44.1
From about 4.03 Buckley builds, almost screaming into the mic by the end.
He backs himself multiple times as the crescendos are reached.
2960 - I felt that the layers as Buckley backed himself were revealed slightly more easily
NG - as Buckley began screaming into the mic the distortion was slightly better handled here

Rachmaninov Symphony No 1, 1st Movement, HDTT, Ormandy 9624
Love this Symphony. This rendition can sound a bit lively.
About 6 mins in a xylophone joins the percussion top left.
2960 - Some notes are strident, zinging.
NG - Doesn't happen, notes remain controlled.

Currently I am sticking with the NG, but, I am VERY tempted to strip out the SMPS from the cisco and run it using an LPSU - just doing some research.

M

 

P.S. Keler - I have one fancy main cable, never heard it make a difference to anything. I did try it with the 2960 and thought it may have made a smidgen of +ve change .....all VERY subtle, and would have hated to try it double blind.

i want to say that i had similar impressions for sound quality. For the little edginess of the cisco on some tracks, i put an acoustic revive lan isolator between my router and the cisco. The edginess gone away for me.  I know simon doesn't share this , but we are a lot on this forum and other forums to experience very positive upgrades on sound quality with it. 

I strongly recommend it, it costs around 200GBP and you can return it easily on the bay...

Mr Underhill posted:

 

Most of this built up as I was trying to solve an issue which I resolved by getting my earthing right. I need to start stripping some of this out and seeing if it makes a difference.

 

 I'm sure simplification in the replay chains would give you a good starting point to reassess your setup. I can understand how you've got here given the problems you were trying to solve, but with this sort of complexity you may be creating more issues as you've introduce more components into the chain.

James

james n posted:
Mr Underhill posted:

 

Most of this built up as I was trying to solve an issue which I resolved by getting my earthing right. I need to start stripping some of this out and seeing if it makes a difference.

 

 I'm sure simplification in the replay chains would give you a good starting point to reassess your setup. I can understand how you've got here given the problems you were trying to solve, but with this sort of complexity you may be creating more issues as you've introduce more components into the chain.

James

i thought my network was complex, i finally have something more simple vs this ....

router>lan isolator>meicord lan>cisco 2960>audioquest vodka>fmc swich( powered by hdplex 5v) > 2 optical cables>fmc switch( powered by uptone audio js2)> audioquest vodka > nds/555dr

cisco> meicord > unitserve ( powered by uptone audio js2)

Hi Mr U - thanks for sharing -  as James has suggested I suspect useful gains can be made through simplification ... ideally you want to cut down on the number of discrete components and powersupplies in your replay chain at least - as each one will be introducing its own noise and possible inconsistencies - good old closed system theory. A good old systems design adage is if you are having to unduly add complexity to a solution for using a given set of components then  you are probably using the wrong components for that solution.

I would definitely drop those media converters with linear power supplies between the switches - to me that is just asking for trouble ... if you want to use fibre media use devices that support SFP transceivers that are powered by the host power supplies and interface to the device back planes

 

Simon-in-Suffolk posted:

Hi Mr U - thanks for sharing -  as James has suggested I suspect useful gains can be made through simplification ... ideally you want to cut down on the number of discrete components and powersupplies in your replay chain at least - as each one will be introducing its own noise and possible inconsistencies - good old closed system theory. A good old systems design adage is if you are having to unduly add complexity to a solution for using a given set of components then  you are probably using the wrong components for that solution.

I would definitely drop those media converters with linear power supplies between the switches - to me that is just asking for trouble ... if you want to use fibre media use devices that support SFP transceivers that are powered by the host power supplies and interface to the device back planes

 

this media converters with linear ps are the most important upgrade for sound quality i had in my system( for streaming with my nds), we are a lot to experiment that.  Theory can't explain all, sorry simon......but there are and were topics for that...so i will not continue 

Hi, A question for Simon in Suffolk again, please. Do you think the Cisco 2960PD-8TT-L with external power supply would benefit with a specialist PS or not and do you think it sounds any different  to the version with an internal ps.

I am asking as I want to give one a go and I am experiencing a problem with the Paul Pang Zyxel oxco with interference on my phono stage (I believe because the earth/ground are not connected between the switch and clock, if I connect them it stops working, any ideas there?)

Many thanks yet again Simon, in advance.

Rich 

Hi Rich, I suspect it won't make any difference at all. Most of the clever stuff with power regulation and noise decoupling will almost certainly be taking place close to the componentry in the switch itself..which I suspect would be a pre requisite for it to meet its EMC certification.

perhaps you can describe the noise interference you are getting with your current setup... it could be from bad Ethernet cabling etc. I had a similar issue once with poor cabling and a Netgear switch which was introducing 'birdies' on my Naim FM tuner. (NAT05). Birdies on FM from discrimination and stereo decoding are a symptom of RFI, and I fairly sure what I heard in the warble was low frequency jitter from the Netgear switch clock... this is not good in audio in terms of noise coupling and we don't always have a FM discriminator to readily hear it....... most low demanding home data networks however couldn't care less about this'll it makes no difference to standard data transfer.

Simon

Likes (0)
×
×
×
×