|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
173.200.6.182
In Reply to: RE: LT3080-based regulators & Linear P4 on Dirty AC circuit experiments... posted by GStew on August 16, 2010 at 07:30:15
which took me by surprise.
The power supply he is using now is not regulated at all, just an RC network and lots of capacitance. Of course, what the thing is powering has regulators on board so I would not doubt it does all that needs to be done.
Installing the BELLESON regulator did sharpen up the images and lost a little of the dreaminess I was getting with the choke filetered supply, but on the whole I do think it is better.
One has to think that the LT3080 has to be so much better than what was "there" as to make going to custom supply almost foolish, especially if it is also a series regulator.
I think it is a good idea to have a bleed resistor anyway. It is good the thing forces this on us.
I hope this difference in the treble continues to be a good thing. I was thinking of just making life easy and using the LT1083 for the 5 volts rail and the 3080's for the others, even though I was ALSO thinking I would use the BELLESON on the P4. I can always make a chance to it later if that seems like a good idea.
What are you doing to power the HDD's? I cannot figure out if I have to power them with both 5 and 12 or one or the other. Again, your help, please? How are you doing this?
At this point I have been using the ANTEC for the P24 and the originally recommended from the ART OF COMPUTER TRANSPORTS supply (something Noisetaker?).
I have made so many changes and cannot ascribe what is making the MOST difference but I have been amazed at what this is able to do. It is more spacious than my turntable. (I am going to have to get to work on my phono stage) It is astonishing.
If this can be appreciably improved with the linear supplies I think I might be very happy for a good long while.
I have to think Dave Davenport's new DAC is a very good sounding device. With the large HYNES shunt reg and Dave's HDMI I2S I am hearing as neutral a tonal balance as I get from a good LP. Been a couple of weeks and I am still listening for hours on end.
Set latency to 98, from 128, and I think it is better. What setting are you using?
Thanks for the report.
Rick McInnis
Follow Ups:
P.P.S. Did he say why he dropped the LT3080?
Everything matters!
P.S. what you using for a phono stage?
Everything matters!
Always good to hear from you. Sorry for the slow response... busy week at work. I believe I still owe you some responses from posts you made a week or so ago too.
Interesting news about Davenport and the LT3080. I can see that tho... I've heard that cascaded regulators may not sound as good in most situations as single ones. And since he has local regs on the DAC board, the LT3080 isn't really needed. I do find the larger C curious... don't they say something about 'minimum reactance' in their PS designs?
And of course, that he isn't using it in his raw DC supply does not mean it's a bad reg... just that the choice of setup he's using now (no reg there since he has them downstream) sounds better.
Did you see Thorsten's description of his NOS TDA1541-based DAC in the Tube DIY Asylum here:
http://www.audioasylum.com/cgi/t.mpl?f=tubediy&m=186140
He's gone to some interesting PS configurations including a lot of RC or LC filtering... look at his LCR phono stage or the Cole phono he designed for DIY Audio (I have one of these to finish up someday as I get a phono setup running again). But of course, to prove that there's no absolute truth, it sounds like he has some cascaded regs in the DAC PS.
Very curious about the Belleson regs. Someday I need to try one and also compare it to the Dexa (which are less expensive). But I'm leaning towards either PH ones or no-feedback shunts like the Salas ones from DIYAudio... based on some comments from John Curl & Charles Hansen about not liking feedback regs. I do think either of these will handily beat the standard 317-LT1083-LT3080 regs... just as the Dexa beat them in my Juli@ mods.
Also curious about what you described as 'dreaminess' with the choke-filtered PS for the Juli@. Can you say more about that sound?
The treble difference of the LT3080s in my dirty supplies still seems mostly positive... maybe a bit dry tho. I did not 'sex' the poly filtering caps (Wimas) across the voltage-setting R's, so I'll likely do that this weekend and see if it makes a diff. They also may just need some break-in time. Or it could be something else.
I also plan to put LT3080's into all the single positions (ATX-24 3.3v & 12v, P4 12v) in my linear supplies this weekend and that should tell me a lot about how they sound.
But that slight sense of treble 'dryness' notwithstanding, I'm still very happy with the sound after the LT3080 regs in the dirty supples and the P4 powered from my dirty AC circuit. The 'transparent' range has been extended up and down in the spectrum quite a bit and dynamics are well-improved.
On the bleed resistor, I remember a long time ago reading that the 317/337 worked better when providing about 1/2 the max current and the way to do this was to add ballast R's at the load if it didn't draw enough. A lot of my DIY stuff using these regs from that era ('80s) use these R's, but I can't say I ever heard a diff from this. Worth checking out with this reg, tho.
The dirty supply for my HDD is a single transformer providing about 11v or so raw DC that I regulate down to 5v. These laptop drives only need the 5v line (I'm pretty sure that's true of all SATA drives) and I measured about 900MaH peaks with my drive, so the LT3080 works ok. I'm using a standard-rotation drive (5400 RPM instead of 7200 RPM) and this helps keep the current down.
Your setup sounds like it is really satisfying now. I suspect at least a good bit of it is your 'alternative' casing... and of course the K&K DAC, which is one that would be on my list if I were not sorta-scratching DACs. And IMHO, you will be even happier with the improvement made by linear supplies. And I have to finish getting a phono setup running again so I have a good point of comparison.
On PCI latency, I'd have to go and check, but I believe the GA-G31M-S2L board I'm using only offers 128 as an option, so I'm pretty sure that's what I'm using. Will let you know.
Later dude!
Greg in Mississippi
Everything matters!
I just thought it was interesting. CERTAINLY NOT a bad regulator. I think it is a big step up from the older generation.
The main reason he gave was the fact with his new DAC he needed more voltage rails and this was easier to do without all of those regulators.
It was not said, but implied, that he must have thought it sounded at lease as good without it.
I went with the BELLESON because it was small and easy to place right at JULI@. I am sure there are some better approaches but they are bigger(I am guessing!). The fellow at BELLESON told me the bad 'scope pictures were made using the regulator I had first intended to use. YIKES!
The dreaminess meant that the edges were softer, very euphonic. Like I said the images sharpened with the reg but there was a toll in that liquidity which was a distortion, though an attractive one. Not to imply the BELLESON adds no distortion of its own ...
I am going to use the LT3080 everywhere but the 5 volts rail. Only because I already have the 1083 and can't imagine it making a big difference there. My thinking (not saying experience) is that getting the noise out of the supplies to pure digital devices is worthwhile for our purposes but going to extremes might only result in the tiniest of improvements where more might be gained with efforts elsewhere. If I had an unlimited budget I would say, "what the hell", but I do not.
I thought you were using the "new" board, which is why I asked about the latency. I remember on the old board that was a BIOS choice. The new board gives you no choice, hence the PCI LATENCY tool.
My phono amp, which was a temporary solution while I build a John Broskie AIKIDO phono, is the ALTMANN. I used to think it sounded pretty good. It did sound really good compared to my old cMP set-up. Now the new digital system makes it sound really closed in and lifeless. That is how happy I am with this new set-up.
At least I can listen to music while I finish this absurdly long running project.
Thanks for the advice on powering the HDD's. I am using WD CAVIAR BLACK and I think they do spin at the higher speed. I am thinking I will use fixed voltage three pin regs for this. LAZY!
I worry that Thorsten is like Allen Wright (I greatly respect them both) I think what they show you is not enough to get the results they are speaking about so I do not pay too much attention. I am not going to try making a DAC. I do have a pair of those TDA1541 chips, though. Single crwons, when they were being made so I know they are not fakes.
I am completely satisfied with Dave Davenport's line stage and DAC. That is another area I am not going to delve into.
You can see the Dave's new schematic for the power supply at his website: raleighaudio.com. It is not ridiculous amounts of capacitance!
I wish I had thought to give him a hard time about the minimal reactance aspect of the new supply! That is funny.
I am jumping back and forth between your note and mine so forgive the disjointed nature of this.
By the way, have you figured out what jackwong was doing with his memory? I cannot for the life of me make any sense of it at all!
Work does get in the way of all of this, doesn't it?
Thanks for your help, as always,
Rick McInnis
Hi Rick,
Regarding distortion, can we presume the Belleson Superpower innocent until proven guilty? :-)
More seriously, thanks for your comments. Sometimes I'm upset at the engineer after I listen to a recording on a really good system. The other day my wife said "I think something is wrong with the stereo, it doesn't sound right." I checked everything out and it was playing as normal.
"What were you listening to?"
"Journey."
"That explains it. Great music, bad recordings."
Brian Lowe
Belleson
but, I figure I will try the cheaper route first, at least, to get things going and then substitute a superior device.
I am afraid I have no choice but to use the LT1083 for the 5 volts rail which seems to be the most important or the hardest working!
cics has suggested that the 12 volts P4 supply for the CPU is not quality critical. I am not, at this point, looking for minute improvements. I still have my doubts about how much of an improvement a fully linear supply will make but the challenge is too interesting to not give it a try.
I have not lost interest in your regulators.
I hesitate to take a chance of going overboard in praise of anything though some would say I have in my appreciation of cics's software. But, one can let their emotions fly when dealing with another's labor of love which they have decided to give away. Not to say I do not sincerely love what his work has allowed - for me, enjoying digital music. Not to imply I have heard much of what else is out there, not in my own system.
You should lead the way and come up with an easy to implement set of regulators for linear power supplies for the music computer!
Bye,
Rick McInnis
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: