|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
174.51.211.143
In Reply to: RE: MacMini vs Bryston BDP-1 Compared-It's posted by Ryelands on April 12, 2012 at 01:56:47
"It's less that a given PSU is linear rather than SMPS and more that it is well designed regardless of technology. One significant issue is that typical stock PSUs are notorious for noise polluting their mains supply (thereby degrading the rest of the system) whereas better units tend not to."
And you've measured this and have screen shots from your oscilloscope and spectrum analyzer to backup this audiophile perpetuated "fact" that stock power supplies are junk and need to be replaced? You know the noise spectrum that needs to be dealt with, the frequency range, amplitude, harmonics, and byproducts, for the Mac Mini as well as most PC power supplies?
And not just on the AC mains, but such noise conducted into and through the quality power supplies built into many of the audio components we use? You've seen this noise appear on the DC output side from the power supplies built by these manufacturers? Don't you think these guys have given thought to implementing noise mitigating measures at the input side of their power supplies as well as the output side where it matters to their gear?
If you are going to attack the switching power supply at the Mac or PC, you had better check every other electronic gadget and appliance in your home and replace their noisy power supplies, too. Based on audiophile paranoia I bet they are all bad and need to be replaced.
Not to diminish the quality of Swenson's power supply design but it is basically just a pi network (been around for several decades) in front of a 3-terminal voltage regulator. And as you know, it won't power a Mac or PC which is what we are talking about.
Follow Ups:
Hi Abe,
I DID do a lot of checking of all kinds of equipment with scopes, spectrum analyzers etc. In particular how the power supply in one piece of equipment can travel through the mains and affect the power supplies in other components, and yes get through the supplies in other boxes.
The frequency range that was most problematic was 50KHz to 300KHz or so. A lot of the boxes I looked at do NOT attenuate these frequencies well. Many have networks that work well for 1MHz and up, and many have networks that work well in blocking noise in the audio range, it's the range between these that seems to be "wide open" in many boxes.
Unfortunately this is the range that many switchers run at and it's also the range where diode "switching noise" occurs. MANY boxes, both linear and switching have resonances in this range which can be excited from noise in other devices. Unfortunately very few voltage regulators have any attenuation in these frequency ranges either. So if it gets through the main PS, the local regulators are not going to be much help.
I did find that there is a definite distance affect here, noise generated in this region injected into the mains at the other end of the house does seem to get attenuated quite a bit so noise generated by devices outside of the audio system does need to be fairly close by to have an affect on the audio system. Devices "in the rack" are going to cause more problems than devices in a different room or on the opposite side of the house.
Of course I did not test every piece of audio equipment on the planet, just what I had in my house, which ranged from inexepnsive "consumer" gear to very expensive audiophile equipment. It wasn't just audio but things like TVs, satelite tuners, microwave ovens, routers etc. Every single one of these devices injected noise in this range into the AC mains, how much vaired a lot from box to box. ALL of the audio boxes let some of this noise through, none of them did a very good job of blocking it.
The issue of internal switching converters is an interesting one. I did several experiments building one of my liner supplies to drive different digital devices (small computers, routers, switches etc) and found that the amount of noise in this region injected into the mains went WAY down, even though the device in question had internal switching DC-DC converters.
Overall conclusion: power supplies which inject very little noise in this region into the mains are very useful, IF they are near your audio system. Doing this to stuff a room away or more may have a slight affect, but it's probably not worth it.
For example in my main system the squeezebox Touch and the netgear switch are run off linear supplies, and the DAC has one (well actually 5) built in. The preamp and amp both have supplies using these techniques (I can do that because I build it all myself). The result is the AC in my rack is very clean in this region. All this equipment has damped supplies that do not resonate in this region so noise from outside sources which do make it into the system has relatively small effect. I haven't bothered with any of these techniques for computers in other rooms, TVs etc. The listening room does not have a TV, cable box etc and there is no computer in the room other than the Touch.
It's obvious I have not testsed everything in existance, there certainly may be some companies that do produce boxes with low noise injection in this region and supplies that do a good job of attenuating what is on the line. If you have a system that's composed of these, great, you are lucky. But there is a LOT of equipment out there that is not.
On the comment that it is just a PI netowrk, which have been around for a long time, why is that important? I'm not claiming I have created anything new or mysterious, any new physics or anything. The issue is that a properly designed PI network, properly tuned with the right value caps etc will deliver very low noise into the AC mains and will deliver DC to the regulator whose residual ripple has almost no high frequencies. Sure the concept has been around for a long time, but how much equipment uses it? The probability is pretty high that not a single piece of electronics in your house has a PI filter in it, let alone one that has been designed to meet the criteria of mine.
So if a PI network is so good, how come everybody doesn't use them? Price and tradition. It costs more money, space and wieght to do it. The other issue is tradition. PI networks were designed for high voltage tube circuits, and most electronics engineers today if they even know of the existance of such a circuit, assume its only good for high voltage supplies. Since they have never seen one used for low voltage circuits they never even think about using one, it doesn't even occur to them as a possibility.
John S.
Yeah, Pi networks. Described in the ARRL Radio Amateur's Handbook , 1957 edition, as I recall. :-)
Did you try low pass filters at the power supply input of affected audio components to protect them from offensive radiators?
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
And you've measured this and have screen shots from your oscilloscope and spectrum analyzer . . .
Of course not - I no more have a spectrum analyser to illustrate my point than you do to disprove it. If you had, you'd have shown us all your measurements and screenshots. Would you not?
As it happens, I do have a 'scope but it's broken so I have to look for answers to the many reports, screenshots etc etc to be found on the net for those who trouble to look. I also use my ears - you might try the same though I doubt you will.
. . . stock power supplies are junk and need to be replaced
I said no such thing. The SMPS in a typical desktop computer is well engineered and, as anyone who remembers linear PSUs from the mini-computer days well knows, far better suited to general computing tasks than any linear PSU on weight, cost, efficiency and cooling grounds. And they're getting better all the time.
But, in some situations, the picture is less clear cut. Any digital machine performing precision tasks in realtime is sensitive to noise, not just audio boxes.
Someone earlier in the thread made the scornful remark that "I've heard that different computers print to the same printer much differently . . .".
Very droll - but silly with it. I worked for years with high-resolution imagesetters and platemakers (which are essentially large-format laserprinters) in the graphics arts industry where output resolutions are typically 2540 dpi and tolerances much tighter than are called for by the average office printer where ink and toner helpfully cover a multitude of sins. An error of a few microns in the placing of a single line of pixels on a printing plate (and there can be over 200,000 lines on a large plate) can render it useless because the error is readily apparent to the eye.
Manufacturers of top-end imaging kit must therefore pay attention to issues of noise, ground integrity and so on and ensure that their power supplies and the mains power lines that feed them are up to the mark. In short, different computers can and do print differently to the same printer whatever the engineering philistine may think. I'm sure similar points apply in dozens of engineering fields outwith my experience.
. . . you had better check every other electronic gadget in your home and replace their noisy power supplies, too.
As do others, I power my audio kit from a couple of dedicated spurs (one for analogue, one for digital) with no "gadgets" on them, an approach I first encountered in the early 1970s, those halycon days of "clean lines" that suppliers would inspect before delivering kit.
As a parlour trick to illustrate the point, try connecting a commodity-level TV set first to the workaday mains supply then to a proper "clean" line. The difference in picture quality is very marked. I didn't bother installing spurs for years but took the opportunity of other work being done to install them. Yes, they make a difference and, yes, I wish I'd done it earlier.
Not to diminish the quality of Swenson's power supply design but it is basically just a pi network (been around for several decades) in front of a 3-terminal regulator.
Thank you but I do know how it works. Well enough in fact to note that you miss the point that it's a well implemented design, not just any old rework from text books.
I seem to recall from the original discussion that you struggled with some of its basics. Whatever, as JS noted at the time, "I never stated that this was a new, unique, never-seen-before power supply design, just that it was a carefully optimized design targeting parameters I have found to be important for audio purposes".
My experience with it (I've built a couple - have you?) is that John seems to have got it about right.
And as you know, it won't power a Mac or PC which is what we are talking about.
It's not easy to get them to power a conventional motherboard but more than one list member has done so and reports good results. Others report good results from comparable though (IMHO at least) less refined circuits.
I accept that the Mac does not readily lend itself to that sort of approach. In fact, it's one of the reasons I find it hard to take Macs very seriously as devices for decent domestic audio reproduction despite their undoubted merits in other fields.
Not surprising at all...And, to tell the truth, most likely wouldn't make any difference in your case if you did - you do KNOW everything, right? Meaning, only the small part of it your could wrap you head around...
Edits: 04/12/12
I've listened enough to know that many so-called power supply and power conditioning upgrades have little effect on gear that have decent power supplies to begin with.
That's certainly possible - in one of two cases:
- resolution of a system as a whole, or of particular piece of gear, is insufficient to demonstrate the differences power conditioning makes. BTW, the difference can be negative, like for instance with most active conditioners on power amps - but it's ALWAYS there;
- hearing of a person is not discriminate enough, to hear these differences.
Being somewhat familiar with your posts about the (fine) gear you're using - I'll have to go with the reason #2 in your case.
You conveniently misquoted me by omitting the rest of what I said in my previous post. And you ignored or simply cannot comprehend the other possibility that I have already mentioned.... but I've seen you do this many times before so I am not surprised.
The power supplies in some audio gear may already be sufficiently well designed such that additional power conditioning may provide little or no benefit.
But in true audiophool fashion you and many folks here jump to the immediate conclusion that every piece of gear will benefit from a modded or aftermarket power supply and power conditioning. That's fine as it helps to keep those vendors in business.
"every piece of gear will benefit from a modded or aftermarket power supply and power conditioning."
Line cords, don't forget line cords and plugs and sockets and outlet covers and the metallurgy of the screw coupling the latter two and the hanging of tikis on the fuse box and...
On the other hand, have you looked at your powerline with a scope or spectrum lately? Not a pretty picture by any means. And if your stuff is like mine it's all unbalanced and probably designed by an audio guy whose whole RF experience is listening to the game on a transistor radio and doesn't know what EMC stands for.
So what we have are systems that are out of control at the system level and thus almost any alteration to them has the potential to change the performance for good, ill or just different. That's just currently part of the rich pageantry of audiophilism. It comes with the territory. And I'd it didn't, we wouldn't be chatting about it! It's part of the charm of the hobby.
Naturally, believing the above, I believe that what's good for the goose may not actually be good for the gander so he needs to sort out his own nest. It all depends on what the nest is made of, how it's assembled and the environment it's in. Enough metaphores?
Folks who generalize their experience rather than just sharing it bug me. It takes a lot of gall (and a high level of ignorance) to state that "anyone who doesn't find that hanging a rabbit's-foot keychain on their power plug gives their system more hop must either have a 'low resolution' system or bad hearing".
Maybe my system doesn't need it?
Rick
"and the hanging of tikis on the fuse box"
Most likely it's the tiki's. I've provided a link for those who aren't familiar with the concept. (Unless your fuse box is huge, I believe the proper term is Hei-tiki. My Uncle served in the Pacific in WW II and he brought back a tiki that was larger than a pendant but suitable for an inside table.)
Another possibility could be the tassels on oriental rugs in the house. These ward off evil-eye, according to rug experts and merchants. Tassels are undoubtedly why my system sounds good, even though it does not "resolve" differences created by many audiophile tweaks.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
"That's fine as it helps to keep those vendors in business."
It's fine only if those vendors are selling products that actually improve their customer's sound. Otherwise, it's not fine. If people waste money on tweaks that don't work they have less to spend on equipment that actually works, and as a result there is less incentive for competent products to be developed and marketed. No one benefits from useless tweaks except the scam artists peddling them. (The fools may think they benefit, but then they are fools.)
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
You were asked before what those are - and pretty much refused to name them.
So, once again - what are the tweaks that you tried, that didn't work in your system?
Of course, one has to be judicious - in other words, I wouldn't spend any money on, for instance, Coconut Audio products, however I'd be willing to try them for free.
The only tweak that I personally didn't hear any difference with, when tried in my system, was Shunyata cable lifters. Left them in the system anyway, since they didn't cost a lot, and look kinda nice.
Damned if I recall. There are a quasi-infinite number of failed tweaks one might wish to avoid, so unless they fit into some kind of pattern then they tend to evaporate from my mind, just like I tend to forget the bad things that friends have done (but not enemies). You have to understand. Most tweaks do something. At the very least they provide an excuse for one to admit that things are now different. But if this is all they do the effect doesn't last and that's the only thing that is relevant. There are many tweaks that do matter and I remember these. These include all matter of minor changes in speaker positioning and adjustment, adjustment of shades and furniture in the listening room, positioning of tube traps, minute (0.5 dB) volume changes, reversals of polarity, switching from foobar2000 to cPlay, etc., resampling from 44/16 to 176/24 including various settings of Sox or iZotope SRC.
Sometimes changes that I did not expect to make much difference had a rather dramatic effect. These include replacing the large power supply capacitors in a MOSFET AMP with slightly larger valued units, as well discovering that two "identical" files sounded different and then inspecting them and discovering that, indeed they weren't at all identical, differing in the dithering algorithm used to create them. There are some things that I simply do not hear despite my best efforts. These include offset errors in CD rips, and bit-identical files that were made by ripping the same CD using different optical drives. These differences I consider to be "fringe" but just barely. What would be "beyond the fringe" would be a claim that HDtracks.com downloads are (good, bad, your choice) because of the color of the Ethernet cable used to connect the server to the hosting provider's Internet router. I suspect there are some audiophiles that subscribe to the "everything matters" dogma, in which case I am definitely not with them.
I've heard too many obvious changes over the years that in the cold sober light of day proved to be mistakes to naively trust my ears. Yes, I trust my ears, but only after quadruple checking and carefully using my mind. And yes, I believe some "authorities" but relatively few of them because most of them don't know what they are doing, unless it is liberating money from marks. If I am going to pursue theories based on non-standard science, I'm not going to waste my time on things like high-end audio. There are much more important subjects, such as cold fusion, zero point energy, psycho kinetics and remote viewing that are about equally "out there" with some high end tweaks.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: