Thanks for the feedback guys. It look like two differant experiences and possible direction.
This sounds like with the picoPSU 150 I should have some headroom.
From Peter's original post it looks like he is using his supply to power a 2.5" laptop drive too. I assumed he was using a monitor and mouse too?
Brad
HaLo6 said:
This sounds like with the picoPSU 150 I should have some headroom.
Greg Stewart said:
From Peter's original post it looks like he is using his supply to power a 2.5" laptop drive too. I assumed he was using a monitor and mouse too?
This would be the end all solution that I would like to achieve but right now I would like to just get it up and running if possible without wasting money on the earthwatts and work on separate power supplies at a later date.Greg Stewart said:
Brad
ecir38 said:
Thanks, I'd forgotten that Peter also powered the HDD via the PicoPSU-150. I don't think he did any other separate power supplies.
So you should be ok... I'm just trying to be conservative, I don't want you to burn up anything because of my advice!
I understand what you mean about not spending extra money... I have two Antec Earthwatt supplies, my original one that I've heavily modified and a NIB spare
and now I'm almost exclusely running the PicoPSUs.
OTOH, I have a Pico that is acting a bit flaky right now and it was nice to have an Antec around to help troubleshoot it.
Greg in Mississippi
The Juli@ Analog Section...
With the part mods and separate power supply work that I did is very, very good. If one can do the scratch-building needed, it is also a very cost-effective option. I spent about $350-$400 (edit, I forgot about the high-bux regulators) for the parts I used (I used VERY high-quality parts). I can publish a parts list if requested.
The sound I'm getting from the modified Juli@ analog section is very detailed, dynamic, musical, and fun to listen to with a lot of PRAT.
One could easily do a lot worse. For example, I'm playing with some very simple, but sophisticated I2S-connected DAC cards with zoned high-quality regulation and a very simple, but effective output stage. These are mounted right on top of the Juli@ digital stage in a manner similar to that of the Juli@ analog stage. The Juli@ analog section loses a small amount of detailing in comparison, but still wins from a musicality perspective.
Over the next few months, I'm working to try separate I2S DACs based on AKM's AK4397 & AK4399, Crystal's CS4397 & CS4398, and the fabled Buffalo ESS 9018 card, all compared against the AK4358-based setup on the Juli@ analog stage.
Hopefully I'll be able to decisively better the modified Juli@ analog section and take my system up a level or two.
Greg in Mississippi
P.S.
I've seen a couple go for under $100 USD on Ebay, but most vendors have them at about $165 USD shipped there.
With the part mods and separate power supply work that I did is very, very good. If one can do the scratch-building needed, it is also a very cost-effective option. I spent about $350-$400 (edit, I forgot about the high-bux regulators) for the parts I used (I used VERY high-quality parts). I can publish a parts list if requested.
The sound I'm getting from the modified Juli@ analog section is very detailed, dynamic, musical, and fun to listen to with a lot of PRAT.
One could easily do a lot worse. For example, I'm playing with some very simple, but sophisticated I2S-connected DAC cards with zoned high-quality regulation and a very simple, but effective output stage. These are mounted right on top of the Juli@ digital stage in a manner similar to that of the Juli@ analog stage. The Juli@ analog section loses a small amount of detailing in comparison, but still wins from a musicality perspective.
Over the next few months, I'm working to try separate I2S DACs based on AKM's AK4397 & AK4399, Crystal's CS4397 & CS4398, and the fabled Buffalo ESS 9018 card, all compared against the AK4358-based setup on the Juli@ analog stage.
Hopefully I'll be able to decisively better the modified Juli@ analog section and take my system up a level or two.
Greg in Mississippi
P.S.
HaLo6 said:
I've seen a couple go for under $100 USD on Ebay, but most vendors have them at about $165 USD shipped there.
I just picked up my Juli@... I plan on using a AMB Labs Sigma PSU's for the Pico, Juli@, and Hard Drives...
I am debating on whether or not to keep a 200GB Laptop HD in there as well so the CD rips are stored locally rather than on a network... I think I can set up a "volume shadow copy" to ensure the drives are identical across the network... since the RAID 1 on the network stores all the CD rips redundantly... then I'll just have the Music PC clone itself every night to ensure the two volumes are indentical...
I am debating on whether or not to keep a 200GB Laptop HD in there as well so the CD rips are stored locally rather than on a network... I think I can set up a "volume shadow copy" to ensure the drives are identical across the network... since the RAID 1 on the network stores all the CD rips redundantly... then I'll just have the Music PC clone itself every night to ensure the two volumes are indentical...
After researching SATA power specs it looks like 3.3V isnt really used... is the 12V used? Some cables only offer 5V (6Pin SATA Slimline)
What voltage is actually needed for hard drives , or does it vary depending on manufacturer?
Greg, have you swapped the two XO's yet? if so where did you source them from?
What voltage is actually needed for hard drives , or does it vary depending on manufacturer?
Greg, have you swapped the two XO's yet? if so where did you source them from?
I probably won't use cMP intself just really using the mantra of the system build fundamentals and a reduced windows footprint...I use Albumplayer as its high enough software quality (i.e. works with ASIO) and has a high WAF and ease of use/ great looking...
I downloaded a trial of XPe to give a whirl but it looks somewhat involved to set up intially so I may have to devote a tad bit more time to setting it up...
I probably will get a low power 200GB HD though to reduce the network hassle...
EDIT...
I may make the SSD dual boot with cMP/cPlay as a second partion though!!! I just thought of that!
I downloaded a trial of XPe to give a whirl but it looks somewhat involved to set up intially so I may have to devote a tad bit more time to setting it up...
I probably will get a low power 200GB HD though to reduce the network hassle...
EDIT...
I may make the SSD dual boot with cMP/cPlay as a second partion though!!! I just thought of that!
Hi Greg,
I noticed that you removed the four filter caps on the analog board next to the J1 pins. When I did my Juli@ mod, I replaced those four caps with Elna Silmic IIs instead of simply removing them.
I want to take your approach by adding Black Gates/Os-Con directly to the voltage rails. If I am to do so, should I remove the Silmics, or is it okay to keep them there?
Thank you in advance!
I noticed that you removed the four filter caps on the analog board next to the J1 pins. When I did my Juli@ mod, I replaced those four caps with Elna Silmic IIs instead of simply removing them.
I want to take your approach by adding Black Gates/Os-Con directly to the voltage rails. If I am to do so, should I remove the Silmics, or is it okay to keep them there?
Thank you in advance!
Hi Greg,
I noticed that you removed the four filter caps on the analog board next to the J1 pins. When I did my Juli@ mod, I replaced those four caps with Elna Silmic IIs instead of simply removing them.
I want to take your approach by adding Black Gates/Os-Con directly to the voltage rails. If I am to do so, should I remove the Silmics, or is it okay to keep them there?
Thank you in advance!
It depends, from my experience SILMICs are great, they make the resulting sound quite euphonic, "musical", or "warm"... They may not be the best in decoupling digital rails but are nevertheless high quality caps. Best used for analogue. So you have to take the rest of your system into account, if you need more punch, or brighter sound you can swap them to OSCONs.. Blackgate... i don't know, never used them. With OScon, be sure to get the SP or SH series type, or the newest and lowest ESR SEPC.
It depends, from my experience SILMICs are great, they make the resulting sound quite euphonic, "musical", or "warm"... They may not be the best in decoupling digital rails but are nevertheless high quality caps. Best used for analogue. So you have to take the rest of your system into account, if you need more punch, or brighter sound you can swap them to OSCONs.. Blackgate... i don't know, never used them. With OScon, be sure to get the SP or SH series type, or the newest and lowest ESR SEPC.
Thank you for the advice.
I agree that the SILMICs are very good caps, and much better than OSCONs for analog applications. However, the only reason I consider replacing them, is because Greg's solution of directly adding "appropriate" caps to each of the supply rails seems better than cramming filter caps on the PCB.
Sorry for not making my question clear the first time. Let me try again:
If I were to add Black Gates and OSCONs (for analog and digital, respectively) directly on the power pins, would they go well with the SILMICs on the PCB? How would SQ be affected? In this case, is it okay to have a bit of "excessive" filtering?
I noticed that you removed the four filter caps on the analog board next to the J1 pins. When I did my Juli@ mod, I replaced those four caps with Elna Silmic IIs instead of simply removing them.
In the picture of my modified Juli@'s analog section in post #1, those caps shown around J5 are all additions, not replacements. The four caps on the analog section near the J1/J5 connection (C76, C79, C81, C83) are coupling caps for the balanced outputs, not power supply coupling caps.
Generally for power supply bypassing I suggest using caps with similar sonic signatures together on the same rail.
My current Juli@ setup follows that for the most part... I have BG standards on the 5v rail on the digital section and an additional one on the 3.3v rail on the analog section side of J5.
Then I use an Oscon on the output of the 3.3v regulator with a couple local ones near a couple of the key chips (as recommended by 'marcello'). I also have an Oscon on the 3.3v rail on the analog section side of J5.
So I've got Blackgates on the 5v all the way through and Oscons on the 3.3v all the way through. But if I did it again, I'd likely use Oscons for both of these positions as these are all digital chip supplies.
On the analog rails (+-12v) since I currently power my Juli@ from separate linear supplies, I don't have any caps at +-12v rails on the analog side of J5, those caps are at the output of the diodes in the linear supplies. I do have BG N series 'bypass' caps at the inputs of the +-9v regulators and 5v DAC/ADC analog regulators. But if I was powering the Juli@ from the computer's +-12v rails, I'd have BG Stds on those rails on the analog section side of J5 as you see in post #1 above.
So to directly answer your question, I would not mix the Black Gates at the power supply pins with the Silmecs at J1/J5 for power supply filtering. But I do recommend having that local filtering on the J5 analog section side of the power supply rails, especially if you are powering the Juli@ from the computer rails. In that case I'd suggest Black Gate Stds at the J5 on the +-12v, Black Gate Ns before and after the +- 9v output stage regulators and 5v DAC/ADC analog regulator (which is powered from the +12v), and small Black Gate Ns or NXs at the chip rail input pins. Then I'd have Oscons at the 5v input to the 3.3v regulator on the digital side, Oscons at the output of that regulator and at the 3.3v rail pins at the chips, and an Oscon each on the 3.3v and 5v rails on J5.
I hope this helps!
Greg in Mississippi
Last edited:
SunRa,
Sorry for the slow response. My work has been crazy the last month and I've had little or no time for audio tweaking.
I am still working towards the shoot-out. I am currently listening to an AKM4399 DAC card. I have AKM4397 and AKM4396 setups that are close to being ready to try along with a Buffalo I am setting up for a friend.
But I am currently looking at alternative output stages for the AKM DACs... the stage I'm using currently is based on the OPA827 and is very good, clean, clear, extended, and detailed, but not as musical as I expected. As I indicated above, the modified Juli@ analog stage still wins from a musical perspective. I'm setting up some alternate output stages... one based on the AD815 (per CarlosFM) and another based on the 6SN7 stage that Lukasz Fikus used with the Behringer Ultramatch Pro. Trying these and getting the Buffalo going are the highest priorities now and the DAC shootout will follow after them.
I've also traced out the signal paths in the Juli@ and figured out how to intercept the analog outputs from the DAC before it goes to the mixer stages and output stages. Once I've settled on an output stage, I'll try feeding the AKM4358 output directly to that stage to hear what the Juli@ can really sound like.
Many things to do, just very little time now.
Greg in Mississippi
Sorry for the slow response. My work has been crazy the last month and I've had little or no time for audio tweaking.
I am still working towards the shoot-out. I am currently listening to an AKM4399 DAC card. I have AKM4397 and AKM4396 setups that are close to being ready to try along with a Buffalo I am setting up for a friend.
But I am currently looking at alternative output stages for the AKM DACs... the stage I'm using currently is based on the OPA827 and is very good, clean, clear, extended, and detailed, but not as musical as I expected. As I indicated above, the modified Juli@ analog stage still wins from a musical perspective. I'm setting up some alternate output stages... one based on the AD815 (per CarlosFM) and another based on the 6SN7 stage that Lukasz Fikus used with the Behringer Ultramatch Pro. Trying these and getting the Buffalo going are the highest priorities now and the DAC shootout will follow after them.
I've also traced out the signal paths in the Juli@ and figured out how to intercept the analog outputs from the DAC before it goes to the mixer stages and output stages. Once I've settled on an output stage, I'll try feeding the AKM4358 output directly to that stage to hear what the Juli@ can really sound like.
Many things to do, just very little time now.
Greg in Mississippi
hey greg,
what mods that you made improved the digital out(spdif or more interestingly i2s from envy) if any? i just injected 3.3v from a clean supply and it did absolutely nothing to the sound, maybe my buffalo is not as sensible to incoming digital input as other dacs? what you think?
what mods that you made improved the digital out(spdif or more interestingly i2s from envy) if any? i just injected 3.3v from a clean supply and it did absolutely nothing to the sound, maybe my buffalo is not as sensible to incoming digital input as other dacs? what you think?
Last edited:
Most of the mods I've done were measured by the improvement in the overall sound of the card, not improvement in the sound from a SPDIF or I2S-connected DAC.
I did do a listening comparison of the basic digital card's mods (upgraded PS bypassing) effect on the SPDIF out... it was a noticable, but not huge improvement. Replacing the on-board 3.3v regulator with an aftermarket non-monolithic board (Dexa) while still powered from the computer buss was a larger improvement via the SPDIF out.
Then I installed a BNC connector to replace the breakout cable and that was a very large improvement. It'd be interesting to do it the other way around... I think the breakout cable and RCA's are so bad here that might mask most of the improvements from the basic digital card mods.
Sound through my Peter Daniel NOS DAC improved in dynamics, liveliness, frequency extension, cleanliness, and clarity via the BNC.
I did not test the SPDIF again after I powered the Juli@ from linear supplies separate from the computer supplies, so I can't comment there. I also have not done any testing on the effects of any of these mods on an I2S-connected DAC... still working to get the best sound out of them.
And what I understand is that the Buffalo DACs are less sensitive to the quality of the incoming signal than most DACs. But people have reported significant improvements connecting it via I2S as opposed to SPDIF.
Also, I'm assuming you injected the 3.3v in, bypassing the on-board regulator? What was the supply you used and how did you connect it? I'd guess (only a guess) that a good-quality regulator right on-board is better than a better-quality regulator via a 3' cable, so I'm curious about exactly what you did.
Let us know more about what you did, that it didn't produce an improvement is still a very useful data point!
Greg in Mississippi
I did do a listening comparison of the basic digital card's mods (upgraded PS bypassing) effect on the SPDIF out... it was a noticable, but not huge improvement. Replacing the on-board 3.3v regulator with an aftermarket non-monolithic board (Dexa) while still powered from the computer buss was a larger improvement via the SPDIF out.
Then I installed a BNC connector to replace the breakout cable and that was a very large improvement. It'd be interesting to do it the other way around... I think the breakout cable and RCA's are so bad here that might mask most of the improvements from the basic digital card mods.
Sound through my Peter Daniel NOS DAC improved in dynamics, liveliness, frequency extension, cleanliness, and clarity via the BNC.
I did not test the SPDIF again after I powered the Juli@ from linear supplies separate from the computer supplies, so I can't comment there. I also have not done any testing on the effects of any of these mods on an I2S-connected DAC... still working to get the best sound out of them.
And what I understand is that the Buffalo DACs are less sensitive to the quality of the incoming signal than most DACs. But people have reported significant improvements connecting it via I2S as opposed to SPDIF.
Also, I'm assuming you injected the 3.3v in, bypassing the on-board regulator? What was the supply you used and how did you connect it? I'd guess (only a guess) that a good-quality regulator right on-board is better than a better-quality regulator via a 3' cable, so I'm curious about exactly what you did.
Let us know more about what you did, that it didn't produce an improvement is still a very useful data point!
Greg in Mississippi
yes, i removed the regulator and took a sigma22 psu with about a feet of wiring,L to p3v3D1 N to dgnd1. the psu is not supposed to go that low(3.3V) and regulation probably suffers(as the designer pointed out). BUT still this psu blows out of the water all el cheapo regs or psus that it replaced (3.3V). but what makes me think that the buffalo is kind of unsensitive to input signal is that whether i use mu x-fi spdif or julia spdif and i2s, i get pretty much the same sound. so i abandonned my buffalo for now and am going to try the mods you suggested with my zapfiltered zhaolu, which is psu modded a bit.
by the way that bnc mod might be REALLY cost effective as upgrading the rca connector on my zhaolu made a very noticeable difference.
edit: forgot to say: with the buffalo i added some blackgates 47uf and nichicon muse 100uf at the place recommended at page 1 and it also did nothing to the sound.
by the way that bnc mod might be REALLY cost effective as upgrading the rca connector on my zhaolu made a very noticeable difference.
edit: forgot to say: with the buffalo i added some blackgates 47uf and nichicon muse 100uf at the place recommended at page 1 and it also did nothing to the sound.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.