|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
77.12.211.159
In Reply to: RE: cMP - the open source high-end Memory Player posted by cics on December 30, 2007 at 05:42:01
dear forum,I think I´ve mentioned this before, but would you please have a look at this battery powered ATX module, and tell us what you think.
http://www.ocean-server.com/smallbattery.html
how viable would it be, to feed a cMP2 system this way ?
looking forward to explore this further.
kind regards
cMP2 Computer (XP minlogon) Intel E7400 cPlay039>Allocator>Lynx Two B. /192kHz throughout. 2x AcousticReality Ref. 202 & 2x AcousticReality Ref. 601´s ICEpower. Magnepan MG3.3R beechwood frames & custom stands. Miller chokes on Ribbons.
Edits: 01/25/11Follow Ups:
Hi Playmate,Reading Gstew’s (and others) reply on your post I asked myself:
Why stick too your Lynx Two B?
Clearly the Lynx Two B is the best available internal soundcard with multichannel analog outputs. I owned the 2 channel anolog out version L22 for a year or so. But your love for this card will put you on high costs when going ‘full linear’. Even ‘only’ going linear/hybride will give you much trouble and costs.
When you add up all the costs (money, time, energy) you need too invest for going ‘full linear’ with the Lynx Two B: is that really worth the trouble and costs?
But why not consider an high quality external 8 channel analog out soundcard like the Lynx Aurora?
The Aurora uses the samen converters as used with the Two B / L22). The Lynx Aurora external soundcard can be connected too your cMP, with use of an firewire - or USB card. (the LT-FW card or LT-USB) too be installed in the Aurora.This way you can go full Linear and/or linear/hybrid without any trouble/hassle and without (much?) extra costs.
The Aurora + FW card will cost: € 2.155,- (ex VAT). The Aurora + USB cards costs 241 euro’s les = € 1.914,- (ex VAT). See link below for pricelist Dutch Lynx distributer. Don't know the German distributor.
I’m still a fan of good old fashion Firewire. Provided you install a Firewire card with an Texas Intrument chipset (!) in your cMP setup. I’m (still not) a big fan of USB connections. But this might come some day.
I know this is al lot more money than your current Lynx Two B.
But I also red, you where considering spending 700 on an extra PSU.
So than it sprang too my mind: if you are willing too spend that much extra (and you than still will have a lot of extra trouble implementing this psu), you better sell your Two B
And buy an Aurora with FW or USB.Looking at the trouble ahead on your road for powering your cMP/TwoB-setup all linear (or even ‘only’ linear/hybrid), the external Aurora option may turn out too be the less costly manouvre in the end.
Just a thought (and calculation exercise) too consider.
Kind regards
Mark
fully optimized cMP2 PC -> ESI Juli@ -> Van den Hul Optocoupler MkII-> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
Edits: 01/27/11
H,
But why not consider an high quality external 8 channel analog out soundcard like the Lynx Aurora?
The Aurora uses the samen converters as used with the Two B / L22). The Lynx Aurora external soundcard can be connected too your cMP, with use of an firewire - or USB card. (the LT-FW card or LT-USB) too be installed in the Aurora.
This way you can go full Linear and/or linear/hybrid without any trouble/hassle and without (much?) extra costs.
I cant quite get my head around this. How can one go full linear with the Aurora? Are you saying that the internal psu IS linear? Otherwise I see no way outside of modding things how to get full linear.
No one here remembers the bending of our minds
Hi Downrazor,
I don't know what type of PSU is inside the Aurora-box.
I ment full linear for the cMP PC.
Mark
fully optimized cMP2 PC -> ESI Juli@ -> Van den Hul Optocoupler MkII-> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
d
No one here remembers the bending of our minds
Playmate,I must agree with Mark.
I would prefer a Prism Sound Orpheus in place of the Aurora though. If you're lucky, you can find it 2nd hand.
By using a firewire DAC, it will be easier to make a linear PSU for your cMP.
Edits: 01/27/11
thanks for you direct words, guys !
that´s by now a decided strategy.
I´m not sure if the financial situation will give away a prism or metric halo yet.
if in that price range, I would go straight for the DEQX, but I guess the aurora or the good old rosetta 800 from apogee could do the trick with a firewire interface.
my "acourate" FIR filter software program has not come to any results yet, as I´m waiting for the up-coming multi-channel ability of this.
new mainboard, processor, new SSD, two new linear power supplies, interface slots and firewire PCIe does all add up....:-)
how easy the world was, when you just had vinyl.....:-)
-anyone up for buying a lynx two-B card ???
cMP2 Computer (XP minlogon) Intel E7400 cPlay039>Allocator>Lynx Two B. /192kHz throughout. 2x AcousticReality Ref. 202 & 2x AcousticReality Ref. 601´s ICEpower. Magnepan MG3.3R beechwood frames & custom stands. Miller chokes on Ribbons.
The second most-important thing I've learned playing with my cMP over the last 2 years is how important the power supplies are in making a good-sounding piece of audio gear. Even on something like a cMP's computer. Even on seemingly un-important pieces like the harddrive, USB port, and monitor power.And the third-most important thing I've learned is that most power supplies for computers and other consumer-grade and business-grade electronics are very poor-sounding supplies indeed. They are just not designed with sound-quality in mind.
I think we should consider ourselves lucky that a PicoPSU with a decent 12v raw supply is a considerable sonic upgrade over a consumer-grade computer power supply. Given what I know now, I would not have expected that.
The sound quality provided by a well-implemented PicoPSU strongly suggests to me that a DC-to-DC converter of the type they use does not pollute the outputs and inputs with as much noise as a normal SMPS does starting from a 120v/240v AC input. So there's hope for the Ocean Server solution. But do remember that people have reported that only the non-WI versions of the PicoPSU result in a sonic improvement, so it's only a limited subset of this type of supply... and no guarantee that this is one of them.
I say all this because while I have no direct evidence that would say the Ocean Server DC123SR is a poor-sounding supply, my experience would suggest that it is likely not a good-sounding one. Might be worth a try, but at about $700 for a starter setup (and I suspect you'll want a larger battery), it's an expensive gamble.
Then on the discussion of battery versus AC powered supplies, I am definitely in the camp of those who feel that while a battery supply can sidestep some of the problems inherent in an AC-powered supply, the battery brings it own set of problems and adds a set of constraints on usability. And in my experience, good design can at least largely if not completely mitigate the AC-powered supply problems. So I don't see battery power as a slam-dunk to a superior-sounding power supply. (And the Swenson supply scheme that Ryelands mentioned is one of the most-reasoned and best solutions to overcoming the deficiencies of an AC-powered supply compared to batteries. But for a fully-linear cMP supply with multiple raw supplies, it will be large, heavy, and expensive. I'd also corresponded with the fellow in Australia who had a Swenson-design-based linear supply built for him and it was somewhat of a compromise with just two raw-DC supplies, one powering the 12v P4 and the other the 12v/5v/3.3v ATX-20/24... and it was about as large and heavy as my setup with five seperate raw-DC supplies for the same solution. But I suspect it was very good... and someday I intend to go to that too! But as Ryelands pointed out, I WILL stick with standard EI-core transformers, preferredly split-bobbin to minimize noise transmission. Toroidal transformers are just not as good in this type of application.)
Raymond in Australia's Swenson-based-design linear supply:
My linear supply:
In any case, bottom line is that there's no guarantee that the Ocean Server ATX-20/24 solution will sound good and having it battery-powered won't necessarily make it sound better.On types of batteries, low output impedance and low noise are important factors in how good any supply sounds, battery or AC sourced. I've heard of others listening tests that strongly agree with Theob's and suggest that LiFePo4 cells sound better than lead-acid or any of the nickel-based cells. See this info from battery-power proponent, Red Wine Audio: http://www.redwineaudio.com/products/lfpv-edition .
I do want to point out that LiFePo4 are a distinctly different cell type from a LiPo battery, which is the soft-packaged cell that most of us have in our cell phones (and that I also use in my high-powered model airplanes). I'm not sure if one of these two types of lithium-based cells are better-sounding than the other... in model airplane use, LiPos are used where the greatest power-to-weight is needed, LiFePo4s where a more economical, sturdier, and safer cell is needed. None of these factors have any bearing on how they sound. Still, the LiFePo4 are a much safer cell type and I'd recommend them to most people over LiPo, even if I saw evidence that the LiPo sounded better (LiPo can violently burst into flame under a lot of conditions, something LiFePo4 don't do as readily). Li-Ion cells (as used in many laptop power supplies) are much lower current cells than either of these two other Lithium types which means a higher output impedance and generally poorer sound.
But this is all in answering your asked question. Your unasked question is 'what's the best bang-for-the-buck real-world solution for powering my cMP setup and getting the best sound I can'. Is this a good summary of your un-asked question?
The two largest improvements I ever heard during my cMP hardware upgrades were 1) upgrading from my modified computer SMPS to the linear P4 supply and the linear/hybrid PicoPSU ATX-20/24 and 2) implementing separate power supplies for my sound card. Even going from the linear-linear/hybrid to a fully linear supply did not produce as large of a sonic improvement (although the fully linear IS a strong improvement over the linear/hybrid, the difference from the modified computer SMPS to the linear/hybrid was greater!)
With this in mind and with your setup where you have a digital crossover and need to use a 'serious' soundcard like the Lynx, I suspect a hybrid-solution might be best. Use a pair of linear supplies with a PicoPSU for the motherboard. And power the Lynx with separate linear supplies, either by using a modified riser-card or modifying the Lynx. The supplies for the Lynx will be pretty simple supplies and can be built-up using a number of available boards and/or kits, such as the Peter Daniel universal power supply board Rick McI mentioned. The key thing is finding a nearby electronics expert to assemble it, modify the riser card or the Lynx, and get it running.
This would not be a 'cheap' solution... but I expect you could do it for no more than and likely less than the Ocean Server solution... and I suspect it would sound much, much better. And if you want to power it via batteries, look at the generic Red Wine Audio supplies... good-sounding batteries plus effective battery management built-in. But this would make it more expensive, of course.
Another alternative would be to use the linear P4 / linear/hybrid PicoPSU ATX-20/24 as above, but add a separate -12v supply to feed the ATX-20/24 that will handle the Lynx current draw. To do this you'd have to bypass the -12v coming out of the PicoPSU and you still might need a local electronics expert, just less of him. Still, you'd be getting your 5v, 12v, and -12v for the Lynx from the motherboard and that's a BAAADDDD source for good sound. It would work, but definitely not sound as good as the separate supplies for the Lynx.
Oh, finally, the most important thing I've learned in the past 2 years? That I honestly know almost nothing about designing good-sounding power supplies and that any sonic improvements I've gotten from all of my power upgrades have been more luck than intent.
My 2 cents!
Greg in Mississippi
P.S. The main driving force for the development and commercialization of LiFePo4 cells is their use in cordless power tools... and for many hobbyists like myself, the least-expensive way to obtain them is to buy an appropriate power-tool spare pack when it's on sale and take it apart!
Everything matters!
Edits: 01/27/11
Hi Greg,
I couldn’t agree more with Theo: a wonderful summation on cMP power supply.
Thank you for taking the time too write that down !
It’s also very informative and useful for me, as it very nicely paints the big picture on PSU improvements.
I was especially happy reading this in your post:
“The two largest improvements I ever heard during my cMP hardware upgrades were:
1) implementing the linear P4 supply and the linear/hybrid PicoPSU ATX-20/24 and
2) implementing separate power supplies for my sound card. Even going to a fully linear supply did not produce as large of a sonic improvement”.
For me this is very informative in 2 ways !
* First.
It describes the diminishing returns on sound quality improvement with each PS-improvement. Which makes me re-consider my aspirations for going fully linear, as I already have implemented:
- the P4 linear;
- the P24 linear/hybrid PicoPSU solution;
- the ‘clean’ and ‘dirty’ PSU concept, through using 2 PicoPSU-150-XT’s. Each driven by there own linear PSU.
- ‘smoothing’ caps on the 3.3v and 5 volt pins on the digital part of the Esi Juli@.
* Second.
Through your post, it’s now possible too (roughly) scale the sound quality improvement, gained by installing a small Kingstone SSD 8 Gb (with the PSU-improvements meantioned above already implemented).
Two weeks ago, I installed a small Kingstone SSD 8 Gb. It made a really big sound quality improvement. Almost as much as the combined P4 / P24 PSU improvements. I’m not making a hyperbole here. The SQ improvement was really far bigger than I ever expected from reading other inmates posts.
Combining the information in your post on sound quality improvements, with the sound quality improvements brought by thiss little 8Gb Kingstone SSD in my setup (music files on a NAS), it (might be) save to say:
- before going fully linear: first implement a (small ?) SSD.
Do you use a SSD in your setup? May be you could comment on this.
Mark
fully optimized cMP2 PC -> ESI Juli@ -> Van den Hul Optocoupler MkII-> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
Thinking more about what I wrote, I realized that one thing I did not highlight was the context of my cMP at the time of the two changes I mentioned... and as you'll see, the context is likely crucial to judging whether the changes will cause a similar impact with your setup.
At the time I implemented the Linear/Hybrid supply (Linear P4, Linear -> modified PicoPSU ATX-20/24), the state of my cMP was:- cMP's Zalman case on compliant footers & some case-top weights & minimal motherboard damping
- HDD, USB, Zalman touchscreen powered via 2 Granite Digital SMPSs
- HDD mounted in the Zalman cage with minimal isolation
- Analog out via an un-modified Juli@Initially:
After linear HDD/USB/Zalman monitor PSUs & some HDD isolation:
When I implemented separate power for my Juli@, the system state had changed significantly:
- cMP's Zalman case on compliant footers w/ some case-top weights & surface damping & slightly increased motherboard damping
- MoBo power via Linear P4, Linear -> modified PicoPSU ATX-20/24
- HDD, USB, Zalman touchscreen powered via hi-test linear supplies
- HDD resting on wooden slab with damping weight
- Analog out via a Juli@ modified with upgraded signal & power supply reservoir caps & 4 key regulators
And when I implemented the fully linear supply, the system state had changed even more:
- cMP's Zalman case on compliant footers w/ some case-top weights & surface damping on the top and sides & much greater increased motherboard damping
- HDD, USB, Zalman touchscreen powered via hi-test linear supplies
- HDD mounted to wooden slab with damping/stray-field-limited weight, ERS paper shielding, & compliant footers with carefully dressed connecting wires
- Analog out via a Juli@ digital w/add-on I2S connected AK4399-based DAC card & separate power
Until I implemented the separate power for the Juli@, any motherboard-connected power supply improvements likely made a larger impact on my cMP's sound quality than they would for a cMP with a separate DAC. This was confirmed when I powered the Juli@ from entirely separate supplies... The change made by upgrading the regulators, but still powered via the motherboard was noticable, but honestly disappointing in magnitude. But adding the separate power made the machine a whole new source (and this was probably magnified by the presence of the 'good' regulators).
How is this relevent to this discussion?
First, Playmate also uses a soundcard for his analog output. So he'll probably get results similar in magnitude to what I experienced if he does something similar. But not everyone will get similar results!
Second, to Mark's comments about HDD's & SSD's, first I have to ask Mark what is the current state of his cMP (sorry, I just don't remember or know all of it). I really can't comment with any authority until I know more (and likely not absolutely until I try an SSD), but I suspect that if Mark does not have as sophisticated a level of vibration control & power for his HDD/SSD, he may have gotten a greater improvement than I'll hear if I implement the same drive. Again, my trying it is the only way to know... And I may just have to do that. But an important data point is that the HDD isolation I've done makes a subtle, but very important difference in my system's SQ... if the HDD is touching a nearby structure or if the cables are not dressed correctly, the system will sound very dead, digital-ish, and not very tuneful.
***EDIT***
Mark, I see that you do describe a great deal about your system and how its powered in your SSD-question subthread. So the main thing I'm interested in knowing is how was your HDD and now how is your SSD mounted? Did you do anything to reduce the impact of vibrations from the HDD?
Third, I REALLY have to do a better job of detailing the context as I discuss changes I make to my cMP and their sonic impact... And ask that others do that too. I am pretty sure that someone using a separate DAC, especially one with galvanic isolation, would hear smaller differences if they implemented a linear/hybrid or fully linear supply than I experienced.Finally, regarding the smaller difference I experienced when I implemented my fully-linear supply than when I implemented the linear/hybrid one... By the time I put in the fully linear, I had tweaked many other aspects of my system that would have been affected by that supply... Separate sound card power with very directly-powered DAC primarily, but not exclusively. If I took my system back to the state where it was when I first implemented the linear/hybrid supply & compared the two supplies then, I suspect I might hear greater or lesser differences in sound quality... Honestly, I am not sure which way it would go, but I'm pretty confident it would be different.
So where I end up is with greater clarity about how much less clarity we have in comparing sound quality impacts across our variously-different cMP systems. GACK, not what I expected to learn going down this path!
Greg in Mississippi
P.S. First, I won't be trying that experiment to take my system back to an earlier level and compare the two types of supplies... Sorry, just not enough time for that, at least not now. I will try to compare my modified Antec (additional filter caps, fan and extraneous cables removed) to the fully-linear and my system's current state as I'm getting ready to lend it out to a fellow inmate... I'm hoping that will be instructive.
Everything matters!
Edits: 01/27/11 01/27/11
Hi Greg,Thank you for yet another very informative post. I think it's very helpful and informative too read about how other inmates did there tweaks and optimizations. I learned a lot from that.
So here’s how my cMP setup evolved overtime when climbing the sound quality improvement stepladder.
Right from the beginning I used my cMP setup with these hardware tweaks:
* step 1.
- bituminous felt pads on case.
I put adhesive bituminous felt pads on the bottom plate and the lid of the zahlman case. You can buy them at DIY speaker shops or at DIY car stores.
I placed these bituminous felt pads because I don’t liked the cheap tin-can-like rattling sounds that comes from metal computer cases when you open them. Also without the damping of these bituminous felt pads, one can hear how the HDD makes the metal computer case resonate and humm.
Only afterwards I red on this forum that this is also good for sound quality of the cMP.
Van on filtering, demping, ect - HDD loosely placed on layers of foam / cardboard paper.
Van on filtering, demping, ect - PSU: ATX on P4 + P24
- Fan in Antec ATX PSU and fan on PSU disconnected.
- using outboard switching PSU to power: HDD + USB bracket.
I didn’t like the idea of two 230 volt powered Granite’s inside my case so much, so I used an outboard desktop switching PSU to power HDD + USB bracket (and dvd. dvd only connected for installing the OS, or making / restoring image backups with UBCD for Windows).- sound card: ESI Juli@ -> RCA coax -> Lavry Black DA10
* Step 2:
- smooting caps on ATX 12v line to P4.
- smooting caps on ATX P24 lines.
* Step 3:
- Linear PSU on P4 (whow this taste like more !)
- implemented: Epcos mains filters and started using two separate 230 V AC spurs.
In this stage I also experimented a lot with different sound cards used as digital interfaces only. Too name a few: Emu 1212M, EMU 0404 USB, Lynx L22, RME 9632.
* Step 4:
- PicoPSU 150 XT on P24.
- implementation of ‘dirty’ and ‘clean’ PSU–concept (with smoothings caps.
Because I first bought the wrong PicoPSU model, I now had 2 Pico’s.
So I also implemented the ‘dirty’ and ‘clean’ PSU–concept (like Cics suggests) with these 2 pico’s.
On both ‘dirty’ and ‘clean’ pico’s I placed good quality Panasonic 4700 smoothing caps (paralelled with 4,7 + 0,047 high quality polymer caps) on the 5 DC volt rails.
Just too be sure, I also placed smoothing caps on the 12V DC outlets of both linear PSU’s., Because linear PSU also may have some HF pollution on there 12V DC out.I finally settled on the: Lynx AES16 card -> XLR AES/EBU -> Lavry Black DA10
* Step 5. Going all linear (failed)
Trying too implement an all linear setup with help of a bulky TTi Linear PSU which can output 3 voltages. It failed and I also fried my GA-G31M-ES2L MoBo. I also found so much linear PSU’s casings around my cMP case very annoying.* Step 6.
- ‘total galvanic’ isolation with use of a good quality toslink.
Looking for more improvements: I started too read about filtering HF pollution ect, coming from PSU’s, the effect of HF pollution on DAC’s etc.
This led too going back to the ESI Juli@ but now with a good quality toslink: ESI Juli@ -> Van den Hul Optocoupler MkII-> Lavry Black DA10.
This was really rewarding!
I already had used an optical cable before: the one that came with the RME 9632.
When using that (cheap) toslink cable, I had (falsely) concluded that there wasn’t any SQ too gain. But the sound quality improvement coming from using a good quality toslink cable was a real significant one! Something like the fourth step: going from the ATX P24 too the picoPSU 150 XT on the P24.* Step 7.
Cleaning power ESI Juli@ soundcard like Promethk suggests.
I first put some heatshrink tubing on the pins which are not going to be used. This way one nicely prevents short circuit too the neighbour pins, that are quite close.
Once the heat-shrink tubing is in place, it’s very easy too fit the caps with use of metal screw-in type power connectors. Just as Promethk shows on the photo in his post.
It’s a very easy tweak which can be done in 30 minutes. Provided you isolate the un-used pins with heat-shrink-tube. Otherwise one will almost surely fry the ESI Juli@ through short-circuit.
It gave a small but nice Q improvement. Like one of the best improvements we sometimes got through upgrading cPlay or when upgrading the firmware/drivers of the Lynx EAS16.* Step 8.
- Removing the 500 Gb Samsung HDD and installing a small 8 Gb Kingstone SSD.
A totally unexpected and a real big improvement in sound quality !
Much (!) more than the third step: adding the linear PSU on the P4.
A whole new layer of mico-details ‘emerged from the deep’.
Also the sound-stage and stereo-imaging improved very much.
The whole sound character also changed.
At first I wasn’t sure if I really liked it.
But I really like it now.
For only 38 euro’s, I never expected such a big SQ improvement.So I don’t know at which point on the improvement stepladder implementing a SSD best should be done. Should it be done already at an early stage? Or better later on when other improvements are already done?
Since it's is only 38 euro’s, I would suggest: one should do it anyhow. No matter at what step one is at the improvement stepladder.
Now something on the use of a Networked Attached Storage. I use a NAS too store my music files on (instead of an external USB or Firewire HDD). I read some other inmates use networked storage too. But I also read that they use a very complicated approach too get the network running, while running in cMP mode.* How I use a NAS in my cMP setup.
Since the SSD I installed is only 8 gb, the music files must be stored somewhere else. I store them on a NAS.
Too access the music files I must have the network function enabled.
Only one extra service is needed for that: the 'workstation' service.
So I use a total of 3 services: ‘PnP’, ‘Remote Procedure Call’ and ‘workstation’.
The LAN-functionality also works in cMP-mode under Minlogon.The use of a wired LAN connection has too be planned in advance when optimizing/tweaking the Win XP sp2 operation system. Before one starts disabling the windows services, a few network configuration steps have to be done first, which can’t be done afterwards when the windows services are already disabled.
Before you start with disabling services, you have too do these two steps:
1. make your cMP computer part of the same workgroup as too which the NAS also belongs.
2. Don’t use DHCP. Instead: manually assign a fixed IP address too the network card.
That’s it.Applying Minlogon later on, doesn’t reverse the fixed IP address nor does it reverse the workgroup assignment. Both stay intact when applying Minlogon.
The last 2 years I did a lot of intensive listening if wired-LAN enabled would degrade SQ. I can’t hear any loss in sound quality when using a wired LAN connection.
Probably the water will still be muddy after this post, but at least now it’s clear which steps I made when trying too improve sound quality and at which point on the improvement stepladder, I installed the small Kingstone 8 Gb SSD.
As always: any suggestions, tips, comments, ideas are welcomed very much.
Mark
fully optimized cMP2 PC -> ESI Juli@ -> Van den Hul Optocoupler MkII-> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
Edits: 01/30/11 01/30/11
Wonderful summation of the state of the art on PS. Nice not only on what's out there but on what matters for sound.
I'm glad it was useful.
BTW, I just made some edits to clarify and add some info.
Later!
Greg in Mississippi
P.S. BTW, my next iteration of a modified Juli@ will have a plug-in for the 3.3v so I can compare several different regulators and a LiFePo4 battery setup. Curious which will win... that or the Paul Hynes reg.
Just need some time to finish it!
Everything matters!
One big battery powering a bunch of dc-dc converters would make sense when one wants to maximize the between charges times of the battery but next to useless for a truly upgraded power supply for a music computer.
No question, good linear regulators for each voltage rail is the only true improvement available.
If the switching supply is the problem going from an AC input to a CD input is not going to make a difference, you will end up right where you began, at best.
AS with any audio component, the more isolation, through separate power supplies, you can effect to the systems within the total circuit the better one's result. Of course, there is the point of diminishing returns but that is up to each set of ears.
Batteries are not a panacea - I wished as hard as anyone out there they were the answer to power supply problems. I have used them in many applications. But experience has taught me they offer no advantage over a good AC power supply and a whole lot of trouble.
They are not the simple answer we would all wish to exist. And DC-DC converters are not the answer at ALL.
Bye,
Rick McInnis
thanks for your straight words, rick
I do acknowledge that the "Ocean" solution is similar to a PICO in design and ultimately a compromise.
maybe it´s just the dispointment that the TX150 PICO module cannot run my cMP computer, but it´s of course significant that I get a solution. -especially because my 6 channel DAC´s are integrated in the computer PSU.
we still have to find a linear supply for the ATX...and as I can read between the lines, it seems simply not done with a battery based supply.
okay, back to square one....
kind regards
cMP2 Computer (XP minlogon) Intel E7400 cPlay039>Allocator>Lynx Two B. /192kHz throughout. 2x AcousticReality Ref. 202 & 2x AcousticReality Ref. 601´s ICEpower. Magnepan MG3.3R beechwood frames & custom stands. Miller chokes on Ribbons.
It is VERY interesting, but without measuring it is hard to say how ripple and tension stability are through battery.
Battery power can even be worse. Just to give you an example, I recently measured a Fit-PC2 on its USB port, which would be the bus for a possible DAC.
This is tension stability with supplied AC-12V and battery respectively:
This is ripple, again with supplied AC-12V and battery respectively:
Edits: 01/25/11
While getting off ac mains is always desireable I believe LiFePo4 batteries are better...much lower output impedance. I have listened to Li ion batteries, lead acid, and LiFepo4 batteries and LiPo's beat them all, lead acid is 2nd then Li ion.
While getting off ac mains is always desireable . . .
Designs such as John Swenson's choke-based PSUs (there are others) should give batteries a pretty good run for their money without the hassle. I forget who it was who said he was planning to try them on a cMP^2 rig.
I did hear from an inmate from, IIRC, Australia who built some and was delighted with them but, sadly, I don't think he's still on the list.
Batteries aren't necessarily as quiet as some think (though I can't speak about LiPo's - whatever they are). See e.g. link and also:
http://www.diyaudio.com/forums/chip-amps/13121-battery-powered-supplies-21.html#post187556
I'm sure the Swenson PSU is a very good one. My whole point was that there are better batteries than Lithion Ion. The LiFePo4 (lithium iron phosphate) batteries are a relative new discovery coming out of MIT in Boston. Variations of them are used in some hybrid vehicles now a days. My whole thing about batteries is that when used in critical digital circuits they protect the rest of the ac mains (not dumping digital noise into it) and are protected from noise on the mains best of both worlds. Yes they are a pita but Lipo's I use last 5-7 days on Juli@ and 7-9 days on the Buf32s before needing a charge. I run 3 3.3v 2.0amp rated batteries in parallel to get below 10 millohm output resistance (very fast response).
I'm sure there are excellent alternate solutions but if one is going to use batteries Lipo's are the ones.
Thanks for your responses guys,
My recent implementation did unfortunately cost the life of my motherboard (and maybe the lynx two-B card too. not confirmed yet).
the pico XT150 can certainly not be used in conjunction with a power hungry PCI card like the lynx. it seems that the lynx requires a -12V 150mA supply which the Pico does not provide....
so that´s my main issue here : can this battery system be a serious alternative ?
more comments welcome !
cMP2 Computer (XP minlogon) Intel E7400 cPlay039>Allocator>Lynx Two B. /192kHz throughout. 2x AcousticReality Ref. 202 & 2x AcousticReality Ref. 601´s ICEpower. Magnepan MG3.3R beechwood frames & custom stands. Miller chokes on Ribbons.
I don't know the Lynx card but as you may know many of us here use the Juli@. The originator of the LiFePo4 battery idea (also used a TP Buffalo dac like I do) experimented with batteries on the digital circuit and the analogue circuit of both and said there was much more bang for the buck on using these batteries for the digital circuits. Its like the Linn philosophy: get it right at the source and pay less attention downstream. My experience has been exactly that with these LiFePo4 batteries. I would not use Lithium ion batteries though.
dear theob,
before we continue the discussion about which battery type is the better one (I take your statements on LiFePo4 for granted), could you explain how these batteries have been implemented into the power regime of the computer in question ?
I think we are a bunch of cMP fans who would like to explore all the ways of making high quality power supply possible.
apart form the few tweaks, the pico way and mikhailov´s linear supply it´s very much it....
since the lynx two-B card I use is a 6 analog-out DAC, it´s obviously requiring not only ripplefree power, but also quite a lot of it ....:-)
(sadly more than the PICO can handle)
I think the device that Ocean Server is offering could be a solution to a lot of us....sort of a step up from the PICO, and I would like we could discuss this offering and it´s eventual advantages and drawbacks.
the question is : does it seem a viable solution for cMP?
cMP2 Computer (XP minlogon) Intel E7400 cPlay039>Allocator>Lynx Two B. /192kHz throughout. 2x AcousticReality Ref. 202 & 2x AcousticReality Ref. 601´s ICEpower. Magnepan MG3.3R beechwood frames & custom stands. Miller chokes on Ribbons.
I'm just telling you my experience with Lithium ion batteries there is no obligation to listen to me. Try it and let us know how it worked out. Mikhailov´s linear supply is not a trivial tweak it is a serious attempt.The way I use these batteries are only for the 3.3v digital section of the juli@ soundcard period. I also use batteries on the 3.3 v digital section of my Buf32s DAC.
In my cmp I use 2 ATX power supplies for my pc (one for p24 and one for all else). I tried a linear supply for p4 and that was a pretty good mod but I don't feel the need to go back there after the 2 ATX supplies. I use the Ryelands p4 capacitor bank mod plus the juli@ 5v cap mod of Promethk.
Edits: 01/26/11
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: