+ Reply to Thread
Page 14 of 32 FirstFirst ... 4121314151624 ... LastLast
Results 131 to 140 of 320

Thread: Linear vs Switched mode PSU

  1. #131
    RothwellAudio Guest

    Default

    Quote Originally Posted by Starterman View Post
    I suggest getting yourself a Naim amp or an NCC200. Listen to it with the 22K TR2 collector resistor in place and shorted.
    https://get.google.com/albumarchive/...xpnj7TRdibQkKT

    This resistor subtly alters the distortions profile, albeit well below the supposed threshold of audibility (0.1% according to H Leak). Then you mght understand that amps can be voiced.
    Quote Originally Posted by Starterman View Post
    Putting the defeinition of "voicing" to one side, I think you missed the point about the extra resistor in the Naim/NCC circuit. It adds even order distortion (actually it's rather clever, it makes the LTP act more like a singleton input at ac whist retaining the excellent dc characteristics of an LTP) and alters the sound significantly. All this is at a level below the supposed audibility level for distortion. I cettainly the prefer the higher distortion version and I suspect >90% would if they did the experiment.

    How does that fit with your research?
    I believe an amp with a differential input (aka long tail pair) is entirely reliant on the performance of the LTP if it's going to sound any good. This concurs with Douglas Self's findings, so I'm not alone - though it should be obvious that the LTP is at the very heart of the amp. If I'm not mistaken those two 100R resistors coming off the emitters of the LTP are Les Wolstenholme's addition and are designed to balance up the two transistors through local feedback (emitter degeneration). In my opinion the 22k collector resistor is a sideshow (if not a plain mistake) and there are other parts of Naim's circuit that I would concentrate on.
    BTW, it's possible to do more than just have two transistors forming the differential input and effort spent there does pay dividends.

  2. #132
    Join Date: Oct 2008

    Location: Glasgowshire

    Posts: 9,656
    I'm Gary.

    Default

    In my experience,
    any switching supply I've swapped out for linear has been an improvement in sound quality.
    used to run a digital system with 5 of 'em in place!

    Its not about 'adding improvement' imo, but taking away a source of crap on the mains supply, which when you think about it is what becomes the music you hear.
    AC POWER
    Hardwired 10kVA balanced mains powering entire system
    AMPS
    Meridian 557 power Amp (Modded) / PS Audio BHK Preamp (Modded)
    SPEAKERS
    Wharfedale Evo 4.4
    DAC
    PS Audio Directstream (Modded)
    TURNTABLE
    Pro-Ject X8 balanced output via XLR / Ortofon Quintet Blue cartridge
    PHONOSTAGE
    Pro-Ject DS3 B balanced Input (TT and Phonostage powered by Pro-Ject Power box RS2 linear psu)
    DIGITAL
    OPPO 203 (Modded: Linear PSU, i2s output to Dac) - Roon Endpoint, HDMI input used for all things Streaming/ PS5 /AppleTV ... also good for movies apparently?
    MUSIC PLAYBACK
    Tweaked AP-Linux based Roon Server into Oppo 203 as Roon endpoint
    Ipad Roon Remote.
    Apple Music/ YouTube via AppleTV, fed to Dac via Oppo HDMI input/i2s output to Dac.
    SPEAKER CABLES
    Biwired: Duelund DCA10GA (Bass) Duelund DCA16GA (mid & treble) Duelund 12DCA used as jumpers (On
    "Blackcat Cable" Chris Sommivigo's advice - yup, even with biwire it sounds better - and it does)
    INTERCONNECTS
    All Balanced: Ghost+ recording studio XLR cables

  3. #133
    Join Date: Sep 2011

    Location: Kilmarnock, Ayrshire, Scotland

    Posts: 533
    I'm stuart.

    Default

    If everything was properly designed and then manufactured to that design we might not be having these debates but unfortunately not everything is properly designed, and even if it is, it's quite often not properly made.

    The vast majority of products get designed to a price - lower quality parts gets used than might be preferred and lower tolerance parts are used to get the right 'price'.

    I work in the electronic manufacturing industry (no hifi items but very high end computer equipment - multi $Million stuff) and have done for 30 odd years. Parts get substituted all the time - different vendor, different spec - and that's building stuff in house! Nobody calls the design engineer to get his approval or builds 20 units and sends them off for testing. The parts get used and if everything passes test that's it.
    Getting your product made by a 3rd party? - welcome to the world of counterfeit parts as they buy parts from whoever is selling them the cheapest. Ever tried to figure out who the manufacturer is of surface mount passives after they've been placed on the boards? Or the value or tolerance of the parts used?

    Yes there are some very good off the shelf SMPS's but they tend to be very expensive. There are lots of very bad SMPS's available, built to a price point, with poor quality parts and poor quality control. I see them every day returned to us and depending on the model you pretty much know what's going to be wrong - same thing failing again and again because the parts are being used outwith or right on the limit of their spec.

    Looking at DIY power supplies, SMPS's are not something your average DIY'er would attempt to build, too many components and difficult to fix if you make a mistake and get magic smoke. Linear supplies are far easier to build and far easier to fix. If you have the skills you can build a very good linear supply for not a lot of money. There are some very good linear designs out there for the DIY'er- SuperTeddyReg, AMB Labs Sigma 11 and Sigma 22 and Sjostrom to name a few where you can buy the bare circuit board and build it yourself.

    As they say, horses for courses. If you're going the DIY route build a linear. If you're buying off the shelf buy either type from a very reputable company. At least then you're more likely to get a properly designed, properly manufactured power supply built with the parts the designer intended- and if you buy a non hifi branded one it'll probably be at least half the price Will it make your hifi sound any better? It might but it might not. Like all things hifi, there's no definitive answer just a load of people with their tuppence worth of thoughts.

  4. #134
    Join Date: Jan 2015

    Location: Leeds

    Posts: 164
    I'm Dave.

    Default

    Quote Originally Posted by RothwellAudio View Post
    I believe an amp with a differential input (aka long tail pair) is entirely reliant on the performance of the LTP if it's going to sound any good. This concurs with Douglas Self's findings, so I'm not alone - though it should be obvious that the LTP is at the very heart of the amp. If I'm not mistaken those two 100R resistors coming off the emitters of the LTP are Les Wolstenholme's addition and are designed to balance up the two transistors through local feedback (emitter degeneration). In my opinion the 22k collector resistor is a sideshow (if not a plain mistake) and there are other parts of Naim's circuit that I would concentrate on.
    BTW, it's possible to do more than just have two transistors forming the differential input and effort spent there does pay dividends.
    Apologies to all for being OT.
    We should probably start a new thread.

    The 22k is anything but a mistake. It is is very enlightening to listen to the amp with it shorted and with it in place. Until you have tried this you are speculating.
    I thought the same as you for many years, but this opened my mind to some new avenues to explore on "why do amps sound different".

    I wouldn't get too hung up on Selfie, he has done some great research but he's obsessed only with minimising THD. He can't see the wood for the trees.

  5. #135
    montesquieu Guest

    Default

    Quote Originally Posted by Starterman View Post
    The resistor effects a more natural balance of harmonics in the distortion profile - more like valve amps tend to have. As far as I can tell Naim has always provided this for no extra charge.
    Also, FWIW the modest output impedance of Naim amps means they might have a lot more in common with your STA100 than you realise.
    Can you explain? Damping factor is fairly mainstream for a valve amp, though the distortion figures are rather special:







  6. #136
    Join Date: Jan 2015

    Location: Leeds

    Posts: 164
    I'm Dave.

    Default

    Quote Originally Posted by MrRadish View Post
    I feel like I'm drifting off topic answering this, given where the rest of the thread has gone :-)

    I've had or have some Maplin test bench LPSU, HDPlex LPSU, iFi iPower SMPS, Russ Andrews PowerPak 2 SMPS, Longdog Audio LPSU, cheap Maplin SMPS and SoTM mbps-d2s battery pack. I used them to power either a PPA v2 card, SoTM PCIe card or a Hi-Face EVO converter. I didn't have them all at the same time, but the Maplins and HDPlex provided a constant reference point. I had two of most of the units to handle 5v and 9v, but found the results (comparisons) were the same regardless of what I was powering. I mostly swapped them over in pairs, where feasible.

    The cheap Maplin SMPS was awful, noisy and just yeeuch...
    The Maplin test bench were mediocre with poor transparency, poor liquidity and a slight treble roll off, though they were reasonably musical and had decent bass weight.
    The HDPlex was almost identical to the Maplin LPSU but a little more rolled off in the treble and slightly more musical. Pleasant but low-fi.
    The iFi were technically fine (no roll-off), reasonably transparent and didn't draw attention to themselves, but they weren't particularly musical. I just got bored listening to them.
    The Russ Andrews SMPS was beautifully sweet, open and transparent with excellent treble purity and very musical, but slightly lean in the bass. I only had a 9v one of those, so that was testing with the EVO.
    The SoTM battery pack was a good all-rounder with no deficiencies and reasonably musical, but not as transparent or enjoyable as the RA (testing with the EVO).
    The Longdog supplies are the best of the bunch, with a musical, liquid, well-balanced sound that doesn't draw attention to itself. I don't know if they're as open as the RA, but it's not something I can compare any more.

    I should also add a couple of riders:
    1. All my testing used substantial mains filtering or regeneration. The SoTM battery supply's main strength is that it doesn't need any of that (for itself), so the other supplies had a helping hand.
    2. I've used both SMPS and LPSU ATX power at various points and on one occasion devoting an AG1500 just to the SMPS ATX unit and running the USB card and everything after that from a different regenerator produced excellent results. That was with the Maplin LPSU in place, so the quality and cleanliness of the AC feeding these supplies seems to matter a lot.
    Good write-up. Thanks!

    I have no technical bias against SMPS but all the 'standard' ones like an iPad charger or SB wallwarts are easily beaten by a decent DIY linear.
    I would never pay more than the value of the parts for somebody else's PS because I would rather do it myslef and learn what affects the sound and how. Therefore anything above £50 is OoB for me.
    After quite a lot of experimenting I have found quite a good sounding solution for 5V digital supplies. I was sort of tempted by an iFi given all the positive reviews but your comments have removed my temptation. For that alone I thank you;-). I have enough gubbins in the junk cupboard!!

  7. #137
    Join Date: Apr 2015

    Location: Oxford

    Posts: 195
    I'm Andy.

    Default

    Quote Originally Posted by Starterman View Post
    I would never pay more than the value of the parts for somebody else's PS because I would rather do it myslef and learn what affects the sound and how.
    I wish I had the skills to do that, but I'd just electrocute myself!

  8. #138
    Join Date: Jan 2015

    Location: Leeds

    Posts: 164
    I'm Dave.

    Default

    Quote Originally Posted by montesquieu View Post
    Can you explain? Damping factor is fairly mainstream for a valve amp, though the distortion figures are rather special:
    The 0.22R resistor Naim uses means the DF is similar to that of the ST100 (and similar PP valve amps).
    Most SS amps have much higher DF.

  9. #139
    Join Date: Jan 2015

    Location: Leeds

    Posts: 164
    I'm Dave.

    Default

    Quote Originally Posted by MrRadish View Post
    I wish I had the skills to do that, but I'd just electrocute myself!
    Don't worry, you would only do that once! :-)

  10. #140
    Join Date: Mar 2017

    Location: West Sūžsēaxe

    Posts: 2,015
    I'm Edward.

    Default

    Quote Originally Posted by MrRadish View Post
    I've had or have some Maplin test bench LPSU, HDPlex LPSU, iFi iPower SMPS, Russ Andrews PowerPak 2 SMPS, Longdog Audio LPSU, cheap Maplin SMPS and SoTM mbps-d2s battery pack.

    <snip>
    Thanks very much Andy. Lots to chew on there. Overall a mixed picture I'm guessing. I've heard Longdog at some shows and they always impress. But perhaps it is 'cos they partner with excellent other stuff.

    I'm using an IsoTek 'Qube' to condition the mains. The change is subtle but definitely there (I can easily switch over to not using it so therefore can compare). What I really need to do is get a friend to switch things over randomly without me knowing so as to remove any possible expectation bias. I'll try that soon.

    I'll bow to Jez's deep electrical knowledge but I guess that a clean mains power supply will reduce even further any (perceived) differences between switched and linear PSUs.

    I'm still keen to finish off the part built LPSU I have - need to find someone local who can solder neatly. The guy I got the parts from used the same parts to create a LPSU and he said it made a world of difference. I've yet to be convinced - hence me starting this thread - which certainly has taken an interesting and worthwhile detour in the meantime.
    Current: [P20] Roon/Tidal > Custom PC> Chevron Paradox NDF16 > Phast Pre > Neuro. 686 > Tannoy Berkley (RFC tweaks)


+ Reply to Thread
Page 14 of 32 FirstFirst ... 4121314151624 ... LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •