|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
85.147.82.48
In Reply to: RE: cMP - the open source high-end Memory Player posted by cics on December 30, 2007 at 05:42:01
Hi all,While surfing the internet, I stumbled upon this ‘naked’ linear PSU for industrial use.
What do the more knowledgeable inmates think of this ‘naked’industrial linear PSU. Model NMC 101 S from FG-Electronik?
Van FG-Electronik linear PSU NMC 101 s
http://www.produktinfo.conrad.com/datenblaetter/500000-524999/510259-da-01-en-LINEARE_STROMV_NMC_107.pdf
(see also link provided at the end of this post)And also: http://www.fg-elektronik.de/powersystems/standardgeraete/ac-dc/linearnetzteile.html (unfortunately german language only)
Price: 139 euro (incl VAT)
* Could this be an ‘ideal’ candidate for a linear PSU to power the MoBo 5 Volt and 12 Volt DC rail on the P24?
The product information says:
This linear regulated power supply family with smallest possible residual ripples has been developed for the power supply of measuring and testing systems and other sensitive terminal devices.
There are a maximum of 4 galvanically isolated output voltages available. Three isolated GNDs can be provided in order to minimise faults. As standard, these GNDs are connected (solder points). All outputs are shortcircuit
proof. The transformer is protected by a thermal switch (automatic reset). The maximum output current on the high current branch and the output current during short-circuit is precisely adjustable and thermally stable.
• all outputs are short-circuit proof and thermally protected
• with suppressor OVP protection diode in the 5 V branch
• series regulator system, thus no HF interference
• adjustable current limit characteristic, foldback characteristic for the
high current range
• connections via plug pins and screwless power supply terminal
• European format
• 3 isolated GND
• patented control circuit for the high current branch
• excellent control characteristics
• very small output rippleGeneral technical data:
- Mains input: 230 V + 6 / -14 %
- Load rating: approx. 60 Watts
- Output voltage: high current range ± 10 % adjustable. The short-circuit current is approx. 40 %. i.e. current fold back and thus permanent short-circuit proof guaranteed.
- Static regulating behaviour: for a change in mains voltage of + 6 / - 14%, the output voltage changes by max. 0.02 %, For a load change of 0-100 % by max. 0.02 %
- Dynamic regulating behaviour : for periodic load change between 10 and 90 %, the voltage jumps stays smaller than 200 mV.
- Residual ripple: residual ripple and noise are under 2 mVss from 0 - 1 MHz
- Temperature class: T 40 E
- C-type core transformer: SM 74
- Primary fuse: T 0.5 A
- Dimensions: European plug-in card 100 x 160 mm. Height approx. 85 mm
- Weight approx: 2.0 kg
Some background information on this question:I’m very keen on getting my (cMP) audio PC ALL linear powered (Processor, Mobo and sphericals).
And also every time I read Gstew’s enthousiastic reports on his progress in building his own DIY linear PSU’s, I want to take that jump.
(BTW: Gstew should be elected Audioasylum’s ‘Member of the year’ . Always enthusiast, always willing to share and always with a positive attitude).
But already for one year now, I hesitate for several reasons and I don’t dare too take ‘the jump’.* Obstacles getting there:
-> I’m not very confident in that using 3 or 4 ‘standard’ ‘off the shelve’ linear bench PSU’s will meet the high quality standards that are needed.
(the absence of HF-pollution, stable current supply with rapid changing loads, ect.)
The specs of these ‘standard’ ‘off the shelve’ linear bench PSU’s, simply don’t mention anything about these characteristics.-> If I would press myself really, really hard, I could build myself a linear PSU from recommended circuits like: Gstew, John Svenson, Mihaylov and others.
But somehow I have the feeling that these ‘almost (theoretically) perfect’ circuit designs, heavily depend on the use of also ‘almost (theoretically) perfect’ parts and are thus very costly (Black Gates, etc, ect)
Also: this would be my first time building such a linear PSU and I will probably learn a lot from it.
But….. ordering all the parts, investing all the time to learn to skills needed, ect etc.
It will probably not be a very ‘economical’ project.So may be this ‘industrial’ linear PSU might be a good candidate for my audio (cMP) PC setup.
It ‘only’ needs to be boxed and connected.So any thoughts on the suitability and DC quality output of this linear power supply form FG electronic series: NMC 101 S ?
All thoughts and responses are appreciated.
Mark
N..B. My current power supply situation is:
My audio (cMP) PC is still powered the ‘hybrid way’:
- ‘standard’ linear 12 V PSU on the P4 (processor)
- ‘standard’ 12 V linear PSU powers a picoPSU on the P24 (MoBo)
(this picoPSU model only leaves the 12 V untouched. 3,3 V and 5 V are still regulated)
- a second 12 V linear powered picoPSU powers: USB, HDD, DVD separately.
(this picoPSU regulates all 3 voltages)
fully AOB optimized cMP2 PC -> Lynx AES16 -> XLR AES/EBU -> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
Edits: 09/04/10Follow Ups:
But I'd put gene (first fully-linear supply), Mihalov (super linear supply show-and-tell) and JackWong96 (the most serious cMP ever seen!) ahead of me... and of course, the master cics!
Then of course, I have comments on the power supply...
- First of all, good find and it looks very interesting.
- I'm glad they include a schematic... one can make some informed judgements about it.
- It does use IC regulators for all but the main high-current output. Not a problem, just an observation.
- I am not a trained EE, so I don't quite understand the main h-c output circuit. It LOOKS like a negative rail feedback regulator connected as a positive rail regulator. I've seen this done with positive rail regulators connected as a negative before, but not this way. Not sure what benefits it brings.
- Other than that, the main h-c circuit does not look like anything special, but also not anything bad.
- Looking at the schematic and the different versions they offer, they MIGHT be able to configure one with +5v high-current and +-12v and +3.3v medium-to-low-current. It may take a different transformer wind, but is worth asking. Version NMC 103 looks close to this. From the schematic, it looks like the -5v in that version comes from the bottom-most regulator circuit AND that regulator looks as though it can be configured for either a positive or negative voltage output... so it may not be a significant change.
- It may be a very good platform for later upgrades or rebuilding. You could reuse the transformer and heatsink as long as you got the version with a transformer that provided the right voltages.
- The C-core transformer was mentioned by Thorsten as being the best-sounding option among the different types available... that's a definite plus for this unit.
- It'd be worthwhile seeing what rectifying diodes they use. If not soft-recovery ones, that's a fairly easy and very positive upgrade.
So my suggestion is ask them about a +5v, +-12v, and +3.3v version. Get that if they can do it and if they can't, do the NMC 103.
Then on special parts such as Black Gates, IMHO, there is no substitute for well-selected, good-sounding parts. Build up two supplies with identical circuits, one with well-selected parts and the other with a selection of garden-variety parts from Radio Shack or Farnell or whatever you're local electronic parts outlet and the one with the good parts will sound better. Look inside of a high-end amplifier, preamplifier, or disc-player from Ayre, Audio Research, or other respected high-end equipment manufacturer and you'll see many of those same high-test parts... and a few that you and I can't get because they are custom-made for them. I used these parts when I did my supplies partly because I've learned this lesson over the years and partly because I had some of these parts on-hand and knew where to get the rest. One thing to say here though... while the top tier of these parts are almost always pretty pricey, there is a middle tier where there is good performance available for much less cost, often about the same as the 'garden-variety' parts. Elna Silmec, Panasonic FM & FC, Nichicon Muse electrolytic capacitors and Vishay-Dale CNF55 resistors are all examples of these. But the trick on these is to find the ones that work well for your setup and hearing bias... I personally cannot stand those Panasonics and only rate the Elna Silmec as acceptable, but they have worked well for many others. Thorsten does not like Black Gates at all and they have worked well for me. No part is the best for everyone!
Then this is a good place to make mention of some other options for DIY Linear Supplies. I've not had much time I can devote to this right now, but have been on the lookout. I haven't found any single-unit building blocks such as this one, with the main issue being that many of them are single-voltage out and pretty expensive.
But I have seen some pieces that look very useful for the not-too-advanced DIY enthusiasts:
- AMB Labs has a single-board high-current regulated supply called the Sigma 11. It is all discrete components and with an appropriate heatsink should handle the ATX-20/24 5v high-current rail with no issues. It does not come as a kit, but they can supply the boards, semiconductors, and heatsinks, with you needing to add the transformer, resistors, and capacitors. I suspect it will sound better than an LT1085-based supply too! See it here: http://www.amb.org/audio/sigma11/ . One could also use it for the other voltages, but it would be overkill in current capability, but maybe in supply quality perspective.
- Twisted Pear Audio has a Low-Current Dual Power Supply kit. It uses LM317-type IC regulators. With at least larger heatsinks and possibly upgrades to LT1085 IC regulators, it can handle the ATX-20/24 3.3v & 12v and the P4 12v for an optimized cMP machine. You need only add an appropriate transformer to this kit... see it here: http://www.twistedpearaudio.com/power/lcdps.aspx .
- Peter Daniel at Audio Sector offers a set of bare boards for building up raw-DC supplies. They can be configured for just a raw supply or will also accept either fixed or adjustable IC regulators. These could be used to provide a higher-capacity/higher-current supply for either of the above or to do your entire supply using various IC regulator chips. You can see info on it here: http://www.diyaudio.com/forums/audio-sector/149672-universal-power-supply-pcb.html .
Anyway, I've rambled on enough today... need to go work on trying those LT3080's!
Later!
Greg in Mississippi
Everything matters!
Hi Greg,Thank you for your feed back.
I red the links and also did some more reading and thinking on building or on finishing pre-build linear PSU's and other pre-build power regulators. Like Teddy Pardo’s SuperTeddyReg’s
I also searched the costs for an enclosure / box, wires, terminals, etc etc for the ‘naked’ SMS 101 linear PSU. I also thought about all the hours work needed on putting it in a box ect, ect. I feel it will costs me too much time, effort and money.
So when I stumbled upon a (used) linear PSU this weekend from Thurbly Thandar Instruments (TTi) model: PL330QMT, I bought it (100,- euro’s). It has tree independent outputs. 2 outputs: 0 - 32V at 0 - 3A. Third output: 4 - 6V at 0.1 - 7A
TTi linear PSU’s are considered to be among the best linear PSU’s vailable. On Pinkfishmedia there are even guys who power there Naim NAC pre-amps with TTi linear PSU’s at very good results (they say).
However there is one downside: you need a cupboard or something, to hide these bulky monsters from the eye. They weigh a hefty 15.5 kilo’s (34 Lbs) and are very, very ugly.
If the raw TTi PL330QMT output isn’t good enough, I than consider too use the PL330QMT to feed Teddy Pardo SuperTeddyReg Ultra Low Noise Regulators with it.
But these SuperTeddyReg’s are also ‘naked’ and also have too be boxed. :-(But first: let’s now find out which 12 V, 5 V and 3.3 Volt wires/pin’s on the P24 have to / can be bundled and how to start-up my MoBo.
Thank you for your feedback. I’ll report back (in a few weeks) if I managed too get everything properly connected and if I got the mobo up and working.
Mark
fully AOB optimized cMP2 PC -> Lynx AES16 -> XLR AES/EBU -> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
Edits: 09/13/10
> > What do the more knowledgeable inmates think of this ‘naked’industrial > > linear PSU. Model NMC 101 S from FG-Electronik?
Certainly better than the SOLA/HD that I am using.
short of going for a shunt style supply, this is probably as good as anything with a three pin regulator. At least, it appears to be a discrete regulator which should offer some performance advantage.
The price is quite reasonable. One could not assemble the same thing for much less if ANY less.
You are left coming up with a 3.3 volts rail which would not be difficult to implement.
The packaging appears to be quite compact would allow it to be close to the MB.
The ripple is not THAT low, but I am sure it is plenty low. Of course, the ripple in a dynamic situation is what counts. They claim this to be a notable quality of their patented circuit. One wonders what demands a de-tuned cMP computer has for the supply? One would think it not to vary very much other than (I am guessing) when requesting a new album and that would have little to do with sound quality since the operation is performed before playback.
One wonders if such a thing would make life almost simple. If one wanted to place some BLACK GATES at the P24 and P4 connectors it could not hurt!
I am going to see if the supply is available here. I am lazy and this would have to much better than any switching supply even if it might have a slight disadvantage to a supply built to phono amplifier standards!
Thanks for the notice.
Hi Rick,Thank you for your responce.
Indeed it would make the job more easy.
But the linear PSU model NMC 101 S from FG-ELECTRONIK also has my special interest because FG-ELECTRONIK not only stresses the absence of HF-pollution but they also stress the importance of (dynamic) load regulation.
I have no education or background in electronics, so I don’t know if it is really important, but I’m under the impression that a very good load regulation is also critical for sound quality coming from digital resource. I don’t see any in dept articles on the importance of (dynamic) load regulation for digital resources, but I do see more and more hints that point in this direction.
* Some producers of PSU’s for audiophile digital resources stress the importance of there PSU’s designs having a very low internal resistance.
* On this forum I think it was Bertel (if remember correctly) who stressed the importance of low resistance in order to maintain a stable voltage and steady current flow too the mobo.
* And also in this (sales talk!) article on upgrading the squeezebox PSU the importance of dynamic load regulation is mentioned.
http://www.squeeze-upgrade.com/philosophy-behind-sbooster-concept-pm-30.html#A%20linear%20regulated%20power%20supply%20improves%20the%20sound%20quality%20only%20for%20X%
(Although I seriously doubt if the series resistance of the DC power line is a real problem in that matter).I agree, it looks like if a de-tuned cMP PC is not a difficult load for a PSU. But on the other hand: some inmates on this forum do suggest that all the sound quality differences that can be heard on a cMP-setup through under-volting, under-clocking, using less RAM, an SSD, ect etc are because off less stress and demand on the PSU.
I’ll wait (ad hope) for some more response from other inmates, but since the PSU is only 139,- euro’s (so not super expensive) it might be worth to give it a try.
Mark
fully AOB optimized cMP2 PC -> Lynx AES16 -> XLR AES/EBU -> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
Edits: 09/04/10
for these supplies. It is not listed and when it comes to regulators it is a bit of smoke and mirrors. One can get great static measurements but when the load changes, like in a class AB amplifier, it becomes hard to be sure what it is. It is an indicator of good dynamic performance and the only way for us to find out is to use it and even then a shroud of mystery surrounds since - what are we to compare it to other than the ANTEC switcher which it damn well BETTER be better than since it cost about five times as much! But better is better - the question remains - how much better?
How much better than the ANTEC and then how much better than the PICO?
What one needs to know is: is a MB more like a class A amp or a class AB/B amplifier? Does an MB in cMP mode, more specifically, mimic a class A amp? My instinct is that it does but who knows what that is worth (my instinct that is)?
If the MB is more like a class A amp then the power supplies duties become much less difficult since it is not being asked to respond to changes - the definition of class A implies a steady state.
The supply does not seem to be easily available in the US - one would have to import one since none of the majors seem to represent the company, which is a shame since it does appear to be a step up from the average over the counter linear supply. I like the fact that it uses a discreet regulator instead of the ubiquitous three pin chip. One would like to assume this assure much better high frequency performance, important with computers playing music. I cannot imagine the average switcher having good performance in the high frequencies but there is another guess.
The trick with MB and CPU generated noise is how to sink it? The big advantage of the shunt regulator is that is does just that as part of its mode of operation. How any series regulator can reduce noise, other that its own noise, is a bit of a mystery. One would expect the power supply to be "quiet" - how it handles the garbage being returned to it via the ground is another thing. Maybe these clever Germans have found a way. If they have we will be very lucky music listeners.
If we could get a triple output including the 3.3 volts it would be very desirable. I have not found an over the counter supply with 5, 12 and 3.3 volts. We are the odd ones wanting to power am MB with a linear supply! I still like the idea of powering P4 with a separate supply. I continue to wonder if it makes any difference what powers the HDD and the USB. I just can't see how it makes any difference to use anything but the GRANITE supply.
This does seem to be an above average over the counter supply.
Are you considering ordering one?
Bye,
Rick McInnis
Hi Rick,
To me it’s totally unclear what a digital audio circuit exactly ‘needs’ from a PSU. I also don’t know what kind of digital audio circuits designs are more sensitive than others, too the quality of the power that is feeding them.
As also mentioned by other forum members: swithing PSU’s can be used with great success in digital audio equipment. For instance: my Lavry Black DA10 DAC uses a switching PSU inside, but it’s considered to be among the best sounding mid-level-quality DAC’s. May the digital circuit designs of the Lavry are not so sensitive for the power quality that is feeding them. Or: the switching PSU design is optimized for feeding digital audio circuits in the Lavry. I don’t know. But I guess: probably both.
I also don’t know / have (strong) reservations on ‘just’ copying what is ‘good practice’ in the analogue audio domain too the digital audio domain, assuming that the same practice will also be ‘good practice’ in the digital domain. Some times it is, because essentially all ‘digital’ electronics work analogue too. But also a whole set of different rules apply when transporting and processing block pulses at mega Hertz speeds instead of processing ‘old fashion’ sine waves. Inmate Ryelands sometimes stresses this by pointing at Ivor Catt’s (a.o.) book Digital Hardware Design. http://www.ivorcatt.org/digital-hardware-design.htm#contents. I think Ryelands stresses a very good point here.
So that’s what do I do. I just wonder around in ‘the land of PSU improvements’ for my cMP setup while trying to avoid getting lost. I read how other inmates try too improve the sound of there cMP setup through optimizing the power supply. When it works (other inmates report sound quality improvements too) I try to copy that improvement. But only when it’s not too complicated or difficult because my knowledge and skills are (very) limited in this matter.
So the common sense around here on PSU improvements is to apply a combination of:
- Using a PSU that can deliver currents that vastly vary while keeping Voltage rock steady.
- avoid noisy PSU’s by using quit (mostly linear) ones and or battery’s
- avoid or remove noisy components
- ‘sink’ or ‘filter’ noise generated by noisy components on MoBo, processor, sphericals, etc, ect (too a REAL earth or to GND ?)
- using separate power supply’s in ‘dirty’ and ‘clean’ sections.
But I suspect that the ‘dirt’ from the dirty sections somehow travels around and still manage to reach the clean sections through various ways.
As a hobbyist with limited knowledge and skills I concentrate on the use of (pre-build) linear PSU’s and the sinking / filtering of ‘noise’.
Since using linear PSU’s is relatively easy, most inmates concentrate on using / building linear PSU’s. Unfortunately I don’t read so much on ‘sinking’ or ‘filtering’ noise that ‘dirty’ components are throwing into the power supply lines.
Although I sometimes post long posts (like this one) I’m not the kind of guy who likes lengthy exchanges on ‘philosophically’ or ‘scientifically’ arguments if something might work or not. My motto is: if you think (after good, sound and thoughtful consideration) something might work: than just try it.
So yes I’m considering too buy one. Gstew already provided me with a lots of information on the subject and also some guidance on which model too choose. But he also provided so much more other info and alternatives, that I first have too think and read about that.
Meanwhile I’m going too ask FMAK how he is doing his LF and HF filtering. It seems he belongs to the few inmates who not only concentrate on providing clean power but also on keeping it clean (filtering). I’m very interested in how he does it.
fully AOB optimized cMP2 PC -> Lynx AES16 -> XLR AES/EBU -> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
as opposed to linear supplies which are pretty easy to get ninety percent right.
No one is interested, it seems, in building a LINN or NAGRA quality switcher for the computer so we are stuck doing it the old fashioned way.
The task is easier for a cMP machine versus a gaming computer since our demands on the machine are pretty steady. I cannot imagine there being large jumps in current demand while playing music.
I think fmak is using the PICO but I do not believe he has said what he uses to power that. He is using an ATOM board which probably makes the powering easier in exchange for not having as much control on the MB's BIOS. All is compromise.
It would be interesting for someone with the gear to test the PICO with a high quality 12 volts supply for ripple. It is probably unfair of me to compare the SILENT PC test results when they used the supplied power brick. It very well is almost as good as the Herculean approach. One can still power the P4 with a separate linear supply and get almost to where the fully linear supplies can get us.
Sorry for the surmising. I hope you do not feel I am waxing philosophically.
Embarrassed that I missed the three pin regulators on the German supply. Just looked down there too quickly and assumed it was just rectifiers. That does seem to make 2/3's of the supply rather pedestrian and something you could do for yourself. It is not that hard to assemble a three pin regulator supply.
Looking forward to hearing what you decide to do.
You are right in nominating GStew for member of the year!! He has help me tremendously.
RayBan
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: