So my router was the only remaining component—network or otherwise—with its OEM power adapter or supply in my audio system. I began reading about people reporting good results adding a linear or regulated power supply to their router, so figured I’d give it a try.
I got a mint second-hand Teddy Pardo (15vdc/2a) on the cheap, with its silver cable and a couple of MHz ferrites feeding my Google Nest router, and an AudioQuest NRG-Z3 for the PSU. The Google router is an odd one, rated at 14v/1.1a, but the difference in voltage seemed negligible enough and deemed safe enough to try.
My network resides in a utility room two floors below my audio room, and is now as follows:
Wall (rated to box @20a)—>Topaz Ultra-Isolation Transformer (rated @1kVA output, model 91091-31)—>Tripp-Lite Power Strip—>Paul Hynes SR4 feeding English Electric Switch in receptable #1 on strip; Teddy Pardo 15/2 PSU feeding Google Nest router in receptacle #2 on strip.
It has been installed for about a day now on the router, and after a listening session for a few hours late into the night, I have come to the conclusion that this is a worthwhile effort, especially as an additive means in what is already pretty-much overkill for my system. Though, like the EE8 was a nice addition that I continue to subtly appreciate, adding a well-designed supply to the router brought about what I feel is an equal subtlety upon listening. It is most notable in apparently removing a last bit of ‘glare’ or ‘glow’ or ‘sheen’—for lack of better descriptors—that sort of bled into some higher frequency program material, in effect masking some detail that came through quite nicely last night when auditioning familar material. This led to an underpinning of weight to tracks I historically felt were slightly unbalanced, tipped ever-so-slightly upward but which are now very smooth and detailed.
Now I’m very curious about others’ experiences. In the networking world of subtle improvements, this to me was well worth the $200 exploration, and is now the final piece put into place for me.
Good deal. I can appreciate not purchasing an expensive, over-engineered PSU for the router. It seems so long as one can get rid of the OEM adapters or SMPSs, the better the results. Thanks for your insight.
Having measured the output over time from a wallwart compared to a decent LPS and also EMF at 15cm I can say that the following generic observations are almost always true.
The voltage from a walwart is actually very stable. You are certainly unlikely to see any fluctuation of 0.01v over a 24hr period.
But wallwarts do tend to issue on the hot side of the stated voltage. You’ll probably find a 12v device spits out 12.2v. Wheras a LPS will be closer to 12v. Usually 12.08 is enough to cope with loss on the DC lead.
You invariably swap one type of EMF for another. A cheap wallart is generally in the region of 40-100 V/m and 0 uT @15cm. But that is tiny compared to the wifi antenna anyway. A good LPS will reverse that to <1 V/m but maybe <2 uT. Effectively, you swap electrical radiation for magnetic radiation.
So if the wallwart is in the vacinity of the hifi (within 1m) or on the same circuit, a LPS definitely could yield benefits.
In other words, it’s very little to do with quality of DC supply (I wouldn’t say nothing - I’d expect smoothing caps in a LPS to smooth out mains frequency spikes you won’t ever see with a multimeter). It’s a lot to do with the type of EMF. And common mode noise energy is greatly reduced by having a smaller V/m.
If the switch or router is on the other side of the room on a different circuit, I’d just use the wallwart and put ferrite chokes on both ends of the LAN cable.
All great points. I attempt to address common mode noise with the IsoTranny (~145dB reduction, with a reduction in traverse noise by ~70dB according to the Topaz specs).
I use an Sbooster to power my Vodafone router. To my ears it made a positive difference. I used an iFi iPower before, which I also found to be an improvement from the standard ps.