UnitiServe: should it stay or should it go?

Nigel I'll be joining you at least in spirit.  My uServe's HDD has gone bad it seems (clicking and noises and frequent crashes can only mean one thing to me), so I'll go with the Synology as a server whilst the uServe is in hospital.

Hi Bart,

I had the clicking thing and it turned out to be the power supply. All is not necessarily lost.

Keith

I've now hooked up the Synology to the TP power supply, which has elevated the sound significantly, and it's now clearly better than the UnitiServe. The Synology Media Server is not quite as zippy as the Naim offering, and I need to do a bit of tidying of some metadata and file structures, but other than that I'm really pleased that I've ended up with better sound and a slab of cash in the bank. 

Does it supply the synology with enough current? I thought the voltage/current values were off..

What model synology do you have?

Thanks -- I'll check.  I'm using the Israeli ps, but it's certainly worth checking!

The Synology supply delivers 5.4amps, and the TP delivers 4amps. The Synology certainly works, but does anyone know whether the lower power delivery could cause any damage? If so, I will revert to the original supply, which would also mean that the TP could go and I'd have even fewer boxes.

The power supply will be rated for more current than the NAS will draw so need to check the data sheet for the power consumption for your NAS model - my DS213 is rated at 18.48 W which at 12V equates to current draw of 1.6 A.

Thanks - my NAS is a Synology 213j, so hopefully I'm ok. It certainly sounds better with the TP in place, but I wouldn't want to risk damaging anything. The NAS has started going to sleep when not in active use, which is something it didn't do before I loaded DSM, which is a bit weird. When it's sleeping the TP goes virtually cold. This is very different to the behaviour when powering the UnitiServe, when the TP was very very warm all the time.

Yes you will be fine with the 213. Most linear power supplies have some sort of thermal overload current limiting protection so will feel very hot when overloaded.

In sleep mode the NAS draws little power so the TP will cool down.You can set the hibernation behaviour of the HDDs in DSM under control panel / system / hardware & power. I run mine with no hibernation but HDDs are now rated for ample power down / up cycles so it comes down to personal preference. The only certainty is that a HDD will fail at some point so a backup scheme (not RAID) is essential.

This does not seem like a proper strategy to me -- relying on "some sort of thermal overload current limiting protection" to keep my house from burning down.

They should not overheat to that extent before the current-limiting kicks in but certainly house fires have been caused by cheaper 'wall-wart' type supplies failing, and also overloading of cheap powerstrips.

I would instead recommend a strategy of matching the power supply to the device, whereupon the power supply won't overheat.  Then, "some sort of thermal overload current limiting protection" can fall back to a just-in-case emergency role.

Absolutely agree which is what has been done above - a Synology 213 NAS with a current draw of 1.6A matched to a psu with current capability of 4A. Just to clarify I was not suggesting that one should rely on the power supply going into thermal overload to decide if it is a suitable match or not for the load, just pointing out that if the power supply is running very hot then it is overloaded, and at that point most well-engineered linear supplies should have thermal current limiting. With a new install you should always check the temperature of the units, and check regularly on an ongoing basis. Adequate ventilation is also important.

I didn't think TP used Switching supplies ?

I know TP are not SMPS

The early versions of the TP supply used a linear supply, but apparently it got very hot. The later versions use a high quality SMPS, with regulation. This later version is supposed to work better, and is the version I have, or rather had, as I've now sold it and am using the Synology with its original power supply. I think the US supply is the only one to use SMPS; all the others are linear to the best of my knowledge.

Interesting .........  I will take a closer look tomorrow

My previous inquiries to supplier confirmed the US TPSU utilises a 'medical-grade SMPS' to support the current draw. The TPSU still generates noise (with a different 'grain') on the mains and improves significantly with US filtered cable plus ferrites. The lower-current supplies utilise conventional transformers.

With the Synology drawing 1.6A the 12V/2A fully linear supply may be a viable option? I have seen some other stats that suggested the current draw is quite modest. However I vaguely recall other discussions in a Computer Audio forum re Synology experiencing a power- fail symptom during shut-down on a TPSU due to supply limitations during this peak activity.

I suspect the effects of 'other spur' electrical noise will vary significantly across home environments.

Interesting.Thanks for the info.

James

Yes interesting -40 ('F or 'C).   A friend of mine has a TP 12v 4.5A PSU & it for sure has a transformer & the TP reg,  

Whatever, TP might be advised to get his www info is updated .......... it says .......... The design is based on the experience gained with the TC family of power supplies for Naim Audio equipment, and uses the same technology.

Then that page contains a link to all the blurb on his STR & goes on to explain how his design is better that other linear & SMPS's ................  misleading methinks

Hence my question Mike. I'd never seen any mention regarding the use of Switching supplies in any of the TP products. It makes a lot of sense in this application,  but i'd agree it's not made clear on the website. 

As background, I spent some time investigating the in-market solutions. This supplier's answers to my questions were:

Q: Your power supplies claim very low noise on the DC side. What are comparative measurements re noise fed beck into AC mains side, and how is this kept to a minimum?

A: We didn't measure that.

Q: I saw a comment in jrivers/jplay forum that the higher- current 12/4 and Unitiserve PSU both use an SMPS module internally rather than conventional 50hz transformer/rectification. Could you please advise?

A: for power suppliers up to 2A we use transformers, for higher currents we replace the transformer by a selected high quality medical grade power supply. This power supply costs like 4 transformers, no intention to save here.....

Our tests and measurements show that for higher currents this supply followed by our high-current version of the SuperTeddyReg generates less noise than a transformer- based solution, and sounds better. The reason is that the ripple of the transformer-bridge-capacitor solution increases with current while the SMPS ripple is constant. At higher currents the transformers ripple becomes higher than the SMPS.

This is supplier has an aversion to large capacitors, and a 'stiff' SMPS reduces regulator drop/ dissipation, so the SMPS is quite logical choice when only focusing on the DC output quality.....

I have a Qnap HS-251 and I had a linear PSU from HDPlex that I didn't use any more (gone from a DIY PC streamer on base of C.A.P.S 3 project to an NDX). So I plugged the Qnap in the linear supply: no major SQ improvement but at least it looks awesome and very high end