New Aftermarket PSU For A Squeezebox Touch [Archive]

View Full Version : New Aftermarket PSU For A Squeezebox Touch

icehockeyboy

06-06-2011, 21:34

I noticed David from Mains Cables R Us mentioned his after market PSU made a significant difference to his Touch, and just wondered if anyone else is using a different one to the one supplied to good effect. :scratch:

Do I recall the experts on the Logitech forum saying don't waste your time as the original one is as good as they get? :eyebrows:

AlexM

06-06-2011, 22:08

Hi,

Apparently the existing PSU is a lot less noisy than that supplied with the SB2/early SB3. I think a lot of the improvement ascribed to linear PSUs is caused by the susceptibility of the rest of the system to high frequency noise that a poorly designed switch mode PSU will put back into the mains. If your equipment has good filtering in it's power supply then you will probably see less of an improvement.

I think also that it depends on whether you use the touch's analogue outputs or SPDIF via a DAC in that the analogue output will benefit more as the SPDIF interface already has good jitter performance.

It is just a shame that the Teddy Pardro PSUs are so expensive because it's noise levels are 45Db below the stock PSU, with no HF crap. There may be other decent alternatives for less than €270 though...

I did tried a maplin regulated linear PSU and noticed no improvement whatsoever although I can't say that this is representative.

Unfortunately you can't escape the effects of an SMPS altogether as the touch has one internally to power the display back-light Having said that there are hacks that disable the screen, disable unused output, WiFi etc that DO show an improvement in my system, and are reversible, free and well worth a try.

Regards,
Alex

icehockeyboy

06-06-2011, 22:19

Thanks Alex, I don't use the Touch's analogue output, I use a QED optical cable into my Theta anti jitter box, which in turn gets fed into the Theta dac, powered by it's own upgraded Goliath psu.

That all said, do you think I would benefit from a better Touch psu?

AlexM

06-06-2011, 22:46

IMHO I don't really think that you will see a big improvement. I use the stock PSU into my tweaked Cambridge 850C and the sound is indistinguishable from the original CD, or possibly better. The DAC section has good jitter rejection and I think your Theta interface will do a good job recovering the bitstream. If your system is susceptible to HF noise then you might see some improvement.

Do try the SB Touch software mods - they do help in my system. Have a look at soundcheck's blog Here (http://soundcheck-audio.blogspot.com/p/squeezebox-touch-great-base-for-network.html) for more info.

Regards,
Alex

Fi-Wi

07-06-2011, 06:34

Maybe you should try the rather inexpensive S-Booster (http://www.squeeze-upgrade.com/index.php?language=en) Klaus refers to in his Soundcheck's blog.

I have a Slimdevices Squeezebox (with the original psu) and am thinking of trying one.

WAD62

07-06-2011, 07:58

These are excellent value, I've been using them on my SB receivers and beresford DAC with great results, unfortunately there's no 5v setting to correspond with a Touch...maplins linear psu

http://www.maplin.co.uk/ac-dc-multi-voltage-20w-power-supply-32754

N.B. For a period of time the above PSU was advertised as providing 5V, but it appears that they've reverted back to the original spec...or more likely they'd got the wrong product details on their DB

AlexM

07-06-2011, 08:38

Hi Will,

That link is for a variable switched mode PSU - is that the one you use?.

Regards,
Alex

WAD62

07-06-2011, 09:34

Hi Will,

That link is for a variable switched mode PSU - is that the one you use?.

Regards,
Alex

It's a linear one, VN10L, which I use on both SB receivers and my beresford...It's not clearly marked as such.

The variable switched mode one is this one...

http://www.maplin.co.uk/ac-dc-multi-voltage-4a-switched-mode-power-supply-48517

Which I also have on my SB3, as I can't find a 5V linear at an affordable price, the SB3 is in the bedroom so it's a secondary system, it's an improvement on the stock one but not as good as the VN10L. I've compared them on the beresford (both do 12v) and the VN10L wins hands down.

Ali Tait

07-06-2011, 09:41

I use a better PS on my Touch, which feeds an AN dac, and the improvement with the better PS is obvious.

icehockeyboy

07-06-2011, 09:42

Thanks for the replies guys, I reckon the S Booster might be the way I will go, and as much as I'd like to try the mods suggested........I'm a bit apprehensive! :scratch:

PS Back when I had a Beresford I also used the Maplin PSU being recommended here, despite it having Stan's recommendations, I'm not 100% sure that it improved it......

WAD62

07-06-2011, 09:48

icehockeyboy

07-06-2011, 10:07

Stan's recommendation was the switched mode one, I prefer the VN10L over it...

Ah, ok, I do remember him saying to run it at 13.5v as opposed to 12v.

Paul Hynes

07-06-2011, 10:10

I sell a lot of SR3 power supplies for Squeezebox products, mainly overseas, the majority of which are currently for the SB Touch. All customers are very pleased with the upgrade and none have used my return if not satisfied guarantee. There is a lot of scope for power supply upgrade with the Touch, but don’t expect miracles with simple linear, low cost, off the shelf power supplies typically using very low cost three terminal regulators. There may be an improvement over the stock switcher but there is much further to go for the best performance. You will get what you pay for. It may seem strange to some to have to pay more for a power supply, than the equipment to be powered, but the gains in sound quality can far exceed the investment.

Regards
Paul

WAD62

07-06-2011, 10:28

Ah, ok, I do remember him saying to run it at 13.5v as opposed to 12v.

Yes, which confuses the hell out of me given the supplied one is 12v :scratch:

I run my TC-7510 @12v on the VN10L, and it knocks the hell out of the SMPS one, likewise the SB receiver (@9V)

icehockeyboy

07-06-2011, 13:55

Well I won't be doing anything with the Touch until further notice!

I loaded some albums onto the USB attached hard drive, re connected it, and no power on light from the HDD......attached a pen drive, still no power light.

Long story short, called Logitech and they diagnosed a hardware problem, Arse! :steam:

Ali Tait

07-06-2011, 13:58

Fi-Wi

07-06-2011, 17:21

Thanks for the replies guys, I reckon the S Booster might be the way I will go, and as much as I'd like to try the mods suggested........I'm a bit apprehensive! :scratch:

Great!

In fact, I just ordered one myself to give it a try. I am sure it will beat the psu that came standard with the Slimdevices Squeexebox Classic.

MCRU

07-06-2011, 20:06

Hi Guys,
Just to give you a bit more information on my SB PSU, I posted a picture on the forum and have a fully operational model ready now, my PSU weighs 2.6kg, it was designed by a specialist in this field and we worked together to come up with a design that is not just a piece of cheapo cable and £5 worth of electronics packed into an RS components aluminium off the shelf box, rather it's a seriously engineered component designed to do a particular job.

The reason I embarked on this route was because I tried a maplin heavy duty PSU with my dac magic years ago and it was rubbish, I cut the box open and there was virtually nothing in there, I have looked for a good length of time at building my own PSU and whilst I used other makes of PSU for my Technics SL1210 I felt even that could be bettered so having purchased a SB Touch a few months ago I decided to kill 2 birds with 1 stone so to speak and hey presto I ended up with 3 products, SB PSU, Technics PSU and a very special DC blocker mains lead that I have trialled against the Isotek Syncro myself (and other friends are currently trialling it).

I look at nearly all the PSU's for sale around the world and none of them seem to use audiophile components in the visible mains path, in particular they mostly use a relatively in-expensive IEC inlet and 2 core un-screened cable for the DC lead, I will be using kimber pbj cable externally and oyaide silver solder internally, a furutech gold plated iec inlet with audiophile fuse is also fitted as standard on the latest verison, there will be a version of both PSU's with pure silver internal wiring available to order.

Thanks for reading.

David

Paul Hynes

07-06-2011, 23:19

Ali,

I sell the SR3 power supplies for £270 and the matching DC3C high performance DC lead for £35. Insured carriage and packing in the UK is £10. If you don’t like what you get you have 56 days to return it for a full refund including carriage costs both ways. This guarantee applies worldwide, not just the UK.

The SR3 can be configured for a wide variety of voltage and current requirements and the SR3-05 is the one for the SB touch. It can provide 3 amps continuously and >20 amps transient. More importantly the SR3 is designed for very low noise operation as well as fast transient response and settling time to maintain regulation with rapid load current changes.

The SR5 can also be configured for 5 volt operation giving a further upgrade in sound quality over the SR3. The SR5 currently costs £525 and the matching DC lead costs £45.

Custom options include Vishay Bulk Foil resistors in the regulator circuitry, a choice of energy storage/filter capacitors and Fine Silver wiring including the DC lead. The prices of Bulk Foil resistors and fine silver are currently volatile so I would have to get quotes from my suppliers before offering prices for these options.

If I can be of further help let me know.

Regards
Paul

Ali Tait

08-06-2011, 05:56

Ok thanks Paul.

icehockeyboy

08-06-2011, 09:24

I think the hard part of going ahead with a PSU of that cost is most people will not get their head around paying more for it than the actual SBT, despite it boosting the SQ to a new level. :)

Oh, Paul, the S Booster is not a PSU, it is a device that attaches to your standard one.
Sorry if you already knew that!

WAD62

08-06-2011, 10:02

Here's a thread from the SB site about the 's-booster'...

http://forums.slimdevices.com/showthread.php?t=86542

...it's the subject of some debate ;)

icehockeyboy

08-06-2011, 10:15

Here's a thread from the SB site about the 's-booster'...

http://forums.slimdevices.com/showthread.php?t=86542

...it's the subject of some debate ;)

Hmmmm....food for thought......:scratch:

Message for Stanley Beresford, Stan, if you get chance can you take a peep at this S Booster and evaluate it with your wonderful expertise? :)
And of course if any other "techies" can take a peep and tell us your thoughts, that would be great!

Paul Hynes

08-06-2011, 12:28

I appreciate concerns about power supplies costing more than the item to be powered. However, with PHD products, you are not buying a handful of parts that have been heavily marked up with no added value or a blatant copy of a chip manufacturers application note. You are buying a ground-up design, which has been specifically optimised for high quality audio reproduction. These designs take time and money to develop and produce. Performance is a key part of the value added. I do not mind if people can’t get their head around this, as there are many out there who can, and they are consequently enjoying their music much more than the ones who can’t. Preaching is not really my scene so I just usually leave it to personal recommendation from my customers to generate interest and sales. This works very well and is keeping me very busy.

I do try and help those on a limited budget by supplying built and tested voltage regulator modules, so people can build a first class power supply, by doing the hardware sourcing and assembly work for themselves. DIY enthusiasts can build a Touch power supply using the PR3ACHD-05 (AC input, heavy duty, 5vdc output) regulator module, which is the same circuit used in the SR3 and this module only costs £100 plus insured carriage. There are hardware costs to add to this, of course, but with a bit of ingenuity, and a root through the electronic junk box, these costs can often be minimised.

I have not tried the S Booster so I cannot comment on its viability. If anyone has tried it let us all know what you found.

I have tried the LM1086 adjustable regulator chip used in the associated “Best of Two Worlds” linear supply and National state it is basically an LM317 circuit topology with a lower guaranteed dropout due to a modified output stage. There are other differences in performance too so they somewhat changed the overall performance, not necessarily for the better. Have a look at the datasheets of these chips, which are easily accessible on the Internet. These chip regulators were designed for general industrial use and were not designed with precision in mind. There are several parameters that are important for audio power supplies that these regulator chips do not address.

In no particular order :-

1) Supply line rejection (ripple rejection). This should be as high as possible over as wide a bandwidth as possible to provide as much rejection of incoming RFI (radio frequency interference) as possible.

LM317 100Hz 80db, 1KHz 77db, 10KHz 70db, 100KHz, 40db, 1MHz 20db, above 1 MHz not specified

LM1086 100Hz 75db, 1KHz 60 db, 10KHz 55db, 100KHz 30db, above 100KHz not specified

PR3 reg 100Hz >80db, 1KHz >80 db, 10KHz >80db, 100KHz 74db, 1MHz 56db, 10MHz 38db, 50 MHz 24db

2) Output impedance. This is a measure of the likely deviation from nominal output voltage of the regulator when presented with a load. This often varies with frequency and the target is to make this as low as possible over as wide a bandwidth as possible thus minimising voltage fluctuations on the regulator output with varying load current.

LM317 500Hz 0.002 ohm, 1KHz 0.003 ohm, 10KHz 0.03 ohm, 100KHz 0.2 ohm, 1MHz 1 ohm.

LM1086 Not Specified but unlikely to be much different from LM317.

PR3 reg 500Hz 0.003 ohm, 1KHz 0.003 ohm, 10KHz 0.003 ohm 100KHz 0.006 ohm, 1 MHz 0.048

3) Transient response and settling time. Both parameters need to be as quick as possible with minimum overshoot. This is the killer for most of the three terminal regulators currently available. This is the measure of how quickly the regulator reacts to a sudden load current transient before the resulting output voltage deviation is corrected. This is usually specified as the time taken to move from 0 load current to full nominal load current. Both the LM317 and it’s HV version are measured from 0 to 1.5 amps.

LM317 5 microseconds

LM1086 10 microseconds

PR3 reg 0.1 microseconds

As an example of poor transient response and settling time I did a test for one of my customers back at Christmas to test the transient response of his old SL1210 power supply (I will refrain from mentioning which one but I can say it uses the LM317HV regulator in the standard National low noise application as this is factual and so are the measurements, which are repeatable), against the SR5-21 I sold to him. To keep the test reasonably representative I used pulsed loads of 100ma and 300ma. The output voltage deviation at the output terminals of the power supplies measured on a 150MHz scope was :

LM317HV 100ma 0.1v, 300ma 0.3v

SR5-21 (PR3EHD regulator) 100ma 0.0007v, 300ma 0.002v

The LM317 has poor transient response and suffers from serious load transient induced overshoot as evidenced by the large power supply output voltage deviation to relatively low current changes. This gets worse with higher current pulse loads.

From the specs of the LM regs I would expect a small but noticeable improvement in performance over the typical switcher supplied with a lot of audio gear. As far as I am concerned, a small improvement is not good enough and with a low resolution system it may not even be audible.

I have to get some regulators built now so will sign off for the moment.

Regards
Paul

AlexM

08-06-2011, 13:39

Paul,

Thanks for your very informative post. Your credentials and the quality of products goes before you and is in no doubt!. Your data suggests that any old linear power supply isn't necessarily going to be an improvement over the stock SMPS. Excluding regulation issues, can a SMPS output be put through a low-pass filter to get rid of the RF noise?. I presume this is what the 'S-Booster' is supposed to do, but it doesn't seem that there is any explanation (probably just a ferrite in a box :))

I am interested in whether the PSU will improve the performance of the SB Touch when used via the SPDIF interface, and am interested in what drives this improvement if the obvious influences of the d/a conversion and analogue output stages are eliminated?. How does the internal two-stage regulation of the Touch interact with a low-noise PSU?. Is it possible that these degrade the very low noise from a PSU such as yours and might it be beneficial to bypass?

Regards,
Alex

Paul Hynes

08-06-2011, 15:37

Alex,

There are other things to look for that I have not discussed yet but that will have to wait until I have some more spare time.

I have several switchers in my lab and I have briefly looked at removing the RFI noise they generate. RFI is quite difficult to deal with as it propagates through the air and gets everywhere. Because of this ability it can often present a common mode signal on ground and the supply line so normal regulation techniques do not prevent it passing through the power lines.

Any solutions must look at absorbing the RFI and dealing with common mode signals on the supply and ground. Vertex AQ have some good solutions for these problems.

There are a number of reasons why a better external power supply will improve the performance of the Touch regardless of which output is chosen. Mains driven switching supplies actually have two major noise sources, the first is the interference caused by the output stage switching frequency and all the non-linear harmonics this can generate and the second is the mains RFI that passes through the switcher. All the noise sources are cumulative so removing the external switcher from the equation removes one noise source reducing interaction with signals being processed in the equipment, whether digital or analogue. Not all linear power supplies are born with the same attributes so any deviation from the ideal power supply will have some effect on the equipment performance and will manifest itself audibly especially with high resolution systems. Transient response and noise are two key areas that affect sound quality and a better supply will be audibly obvious in these areas too.

A combined solution is to have a high quality external supply and upgrade all the internal regulation with high performance local regulators, specifically designed to match their load requirements, where space permits. This really does make big changes to the sound quality on the earlier SB products as it addresses all the regulation issues and I see no reason why this should not happen with the Touch. I am planning on taking my SB Touch apart in the near future for some major power supply surgery.

Regards
Paul

AlexM

08-06-2011, 15:57

Paul,

Thanks for that - do keep us up to date if and when you pull your touch to pieces!. I think that some of the internals will be a limiting factor, and it would be interesting to see what can be improved here.

I'm interested also in getting an impedance matching pulse transformer and BNC socket for the SPDIF output in there.. must get it opened when I get time..

Regards,
Alex

Fi-Wi

10-06-2011, 17:30

After having received the "Best of Both Worlds" psu with S-Booster a couple of days ago, I just replaced the original SB3 psu with it.

My very first impressions are absolutely positive! Better soundstage is my first thought. I experience the same (with the same smile) as when I first put a Caiman/Gator combo in my system. My opinion is that these are 135 euros well spent.

Especially because in my case it removes the distortion I had through my Caiman headhone output when playing lower (than 256k) bit mp3/streams. After experiencing that distortion, Stan replaced my Gator but with no positive effect. It actually stopped me from using my Caiman as a headphone amp. Now the distortion completely gone!

I am a convert. :eyebrows:

Paul Hynes

10-06-2011, 17:51

Fi-Wi

10-06-2011, 20:33

Steven,

I would expect some improvement over the stock switcher with the” Best of Both Worlds” and this sounds like a good result for the money.

Did you try the S-Booster with the switcher? The results of this would interest many on the forum.

Regards
Paul
Paul, l'll try do some testing this weekend but due to the holiday weekend I can't promise a report.
If I had a high(er)-end setup, I without doubt would go for a high(er)-end psu.

Paul Hynes

10-06-2011, 21:50

Steven,

Don’t worry about where your system performance stands because there will be others at the same level who are also looking for improvements. They will be able to relate to the equipment you have and what improvements you have made. Just report what you find and this will help others. Any improvement has value. It's just a question of how much this value is and is it affordable.

Regards
Paul

icehockeyboy

11-06-2011, 01:05

I have just received my replacement Touch, and apart from heaping praise upon Logitech on the rapid turnaround on the Touch, 3 days it total! It may be my imagination, but it sounds better than the old one, I did notice a slightly different set up, so maybe some internal bits are different/improved?

Anyway, one happy bunny here!