Switches and Switch PSUs with 1st gen streamer (a 272)

Power supply tests: (Keeping the switch constant and changing the PSU)

.

Netgear GS105 – standard unmanaged gigabit switch

Standard PSU > IFi iPower X
HF
Marginal improvement – less harsh, but also slightly “recessed” no improvement to detail
MF
no significant change
Disappointing, only marginal improvement.

.

Netgear GS105E: ‘Simple’ L2 “Smart Managed” gigabit switch,
Power saving mode enabled (Reduces voltage on the Ethernet cable)

Standard PSU > IFi iPower 2
HF:
Less HF distortion – noticeably reduced ‘splashiness’
MF
Slight improvement in MF – just a little bit smoother without loss of resolution, standard PSU sounds slightly harsher
.

IFi iPower 2 > IFi iPower X
HF
Further reduction in HF distortion, further noticeable reduction in splashiness, better HF ‘texture’
MF
Significant improvement in MF distortion & smoothness with improved resolution and much better separation of tonality. Vocals are more understandable. Easier to listen at higher volume without increased fatigue.
Greater difference than Standard PSU > IFi iPower 2, this is a very useful upgrade with the GS105E version

1 Like

Switch tests, Different switches with the same PSU

.

Netgear GS105 – standard unmanaged gigabit switch
Vs.
Netgear GS105E: ‘Simple’ L2 “Smart Managed” gigabit switch,
Power saving mode enabled (Reduces voltage on the Ethernet cable)
.

Netgear GS105 + Standard PSU > Netgear GS105E + Standard PSU
HF
Less HF distortion and better resolution – noticeably reduced ‘splashiness’, better texture & detail apparent
MF
Slight improvement in MF – just a little bit smoother without loss of resolution, standard switch sounds slightly harsher
.

Netgear GS105 + IFi iPower X > Netgear GS105E + IFi iPower X
HF
Smoother, less distortion and better resolution, better texture & detail apparent
MF
Smoother less distortion and better resolution, much better separation of tonality. Vocals are very much easier to hear. Standard switch sounds notably harsher and much more confused
A big improvement: this is a very worthwhile upgrade. This is a much bigger improvement than I have seen with any change of Ethernet cables.

1 Like

Switch tests, vs Cisco 2960

.

Netgear GS105 – standard unmanaged gigabit switch
Netgear GS105E: ‘Simple’ l2 managed gigabit switch, Power saving mode enabled
vs
Cisco 2960 (white) (default configuration).
.

Netgear GS105 + Standard PSU > Cisco 2960 (white)
HF
More recessed but less distortion, slightly less ‘splashiness”, but this is marginal
MF
Slightly less distortion & more resolution, better ‘organised’, less muddled. Vocals are easier to understand.
Overall GS105 sounds a bit harsher and slightly more muddled, this is a useful upgrade
.

Cisco 2960 (white) > Netgear GS105E + IFi iPower X
HF
Less distortion & ‘splashiness’, smoother without loss of detail
MF
Better resolution, significantly better separation of tonality and ‘organisation’. Vocals easier to understand.
Overall Cisco switch sounds significantly more muddled and also a bit harsher, this is a useful further upgrade

1 Like

Order of preference:

GS105 (any PSU)
GS105E (standard PSU)
Cisco 2960
GS105E iFi iPower 2
GS105E iFi iPower X

.

It’s unlikely to be confirmation bias as it’s not what we expected!

Particularly:

  1. GS105E + iFi iPower 2 slightly beating the Cisco
  2. How much better the iFi iPower X was with the GS105E
4 Likes

A further observation on the GS105E + iFi iPower X combination is that not only does it improve the apparent resolution in the amplitude domain but also a big improvement in the apparent time domain.

For the amplitude domain, I often use Case Lang Veirs: Georgia Stars as a critical test, as there are many layered vocal parts but also even more layered instrumental parts, it a very complex piece; it’s easy for parts to get ‘lost’.

For the time domain, I’ve found that Gabriel’s Red Rain is an exceptional test track as it needs good timing otherwise it sends to sound messy; with a system working well in the time domain it sounds together and precise instead.

.

It should be noted that jitter is a signal correlated noise in the time domain, and this interacts with signal correlated noise in the amplitude domain; so it’s not that surprising that improving one aspect will tend to give some degree of improvement in the other aspect.
Also note that these aren’t just ‘noise floor’, but rather noise that is correlated to the signal - it’s much harder for the brain to sort out correlated noise than uncorrelated noise.

2 Likes