Jump to content


  • Content Count

  • Joined

  • Last visited

1 Follower

About gstew

  • Rank
    Rabid Audio Tweaker & DIY'er

Personal Information

  • Location
    MS, US

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. Gavin, Do understand the actual Ultracap parts used are rated for 2.7V max, which is why 2 are needed for a 3.3V to 5V supply. 12V would take at least 5 UCs. Having followed and used Ian's products for almost as long as he's been active on DIY Audio (and occasionally collaborated with him), my take is that he tends to develop products that he uses in his own DIY builds and when they produce results he likes, shares them with others via his group buys (for which DIY'ers everywhere are eternally grateful!). So while he might take up this challenge, I don't expect it would be a core focus for Ian. Also what power supply is used as the 3.3V / 5V / 12V source is just as important. AND to be used with UC output buffers, you need to be sure the two used together will be stable and safe. Of course, you could use 3 of Ian's 5V UcConditioner boards with 3 appropriate and floating output (non-grounded) 5V power supplies with the outputs connected in series + to - to get a 15V UC-buffered supply. OR 2 5V UcConditioners and a 3.3V UcConditioner for a 13.3V supply. OR 2 3.3V UcConditioners and a 5V UcConditioner for an 11.6V supply. I have some beta-release 5V UcConditioners here that I intend to assemble and try out this weekend/week. Will report on how they work. I expect Ian will have these available for general purchase fairly soon. I have 2 active audio systems and 3 UC pair supplies (2 325F pairs, 1 3000F pair) in each for my digital sources. The improvements I heard moving up to these from my previous reference supplies (which were no slouches) were quite incredible and rewarding. Of course the UC pairs I use are based on Ian's developmental project, the UCPi, where one needs to manually pre-charge the UC pair to the correct target voltage before connecting them for use. Very hair shirt, but as a long-time DIY'er, something I am used to doing in all of my setups. Ian's UcConditioner series is designed to make use of these supplies a bit easier and less user-hostile. Greg in MIssissippi
  2. @Tapatrick, I have not and am sorry to sad I do not expect to do so soon. 3 reasons... first is that I'm REALLY happy with the sonics I'm getting in my 2 setups now. Since I wrote that comment in December 2019, I've: - Tried some alternate Ethernet cables after the EtherRegen and settled on what I liked at first... the lowly Cat6 cable that came with my Baaske isolators (which came with a ferrite core around the cable near one end), a single Baaske isolator (pre-ER I had 2), and a 6" flat Cat6 from Amazon into my USBBridge Sigs. I did try all of the low-cost favs and none bettered this combo either separately or in conjunction with the Baaske isolator in my setups and to my ears. This setup just sounds more like music to me. - Moved to newer model lower-ESR 325F UC pairs on the power feeds to the DACs and associated boards (Katana microprocessor board, Ian Canada FiFoPi) along with using LDOVR.com LT3045 regulator boards as the float-charging regulators when in-use. I still use Ian Canada's LT3042 reg boards pre-use for charge-up as I know they work safely at their native 250mA-300mA native current limit without additional heatsinking... just lazy here! - Moved to 3000F(!) UC pairs on the power feed to the USBBridge Sigs. These of course take a hours to pre-charge at 300mA - 900mA, so when when I power-down the setups for an extended absence or predicted thunderstorm activity, I keep these charged up with a couple of 5V wall warts so I can get the systems back up pretty quickly. - Because I have been so happy with the setups that I was more interested in listening than fooling with them, I moved to Lexar 1Tb SD cards in my Zotac PI320 micro computers configured as LMS servers. In addition to now holding almost 80% of my music files and increasing my music selections, they also brought a small but noticeable improvement over the 512Gb PNY cards there previously. Scratching my head on this one, but happy for the bumps up! As I said above, of this has had me MUCH more interested in listening that building / tweaking / comparing. Second, the 2 setups I planned to use to compare USBBridge Sig to SDTrans384 were my Soekris DAM DAC setups, one of which is tethered to the SDTrans384 and the other is configured for an I2S feed from an RPi-type header (enter USBBridge Sig here). While both are highly modified, neither has been optimized for use with the UC pair power rail final buffers I universally moved to middle 2019 and would need likely a couple weeks of delicate rework to get there. Both of these setups with the power supplies last used are now sonically FAR behind my current setups in most if not all parameters, so I know I need to bring them up to use my lastest power supply scheme if just to have a comparison that I'd want to listen to for the several weeks needed to get a fair listen (and have the Black Gate caps on each board run-in again as they need to do after an extended down time). Third, I've just had less audio workbench time available because I'm shifting my focus over to my model aircraft hobbies. AND it is getting close to flying weather! So likely no comparison until next winter at the earliest. SORRY! Based on what I heard when I switched over to the current setups in early 2019 AND what I heard as I incrementally upgraded them (MANY TIMES!), I suspect that the SDTrans384/Soekris setup will not be far ahead of the USBBridge/FiFoPi/Soekris setup IF AT ALL. A TirNaHiFi-style synchronous I2S reclocker (Acko's old S03 would also work here) after the SDTrans384 might help it pull ahead. BUT likely no work on any of that until late this year earliest. Greg in Mississippi
  3. LOL... capacitors such as these weren't even available 8-10 years ago. Who knows what technology will be available in 8-10 years and if it will even be compatible replacements for the caps in the current Shanti? Maybe Dilithium crystals or zero-point energy? BTW, still loving my Shanti! I have 2 setups and a Shanti in each. Each powerw a USBBridge Signature with a somewhat modified Katana in one setup and a somewhat modified Ian Canada GB FiFoPi/Dual-Mono ES9038Q2M combo with a modified version of his IVStd output stage using a couple of NewClassD Ultimate discrete opamps in the other. Both setups sound very good with the base Shanti, but in the last 6 months I've been using a MUCH larger Ultracap pair as additional output buffers for the Shanti's... originally 350F UC pairs, then a lower-ESR 325F UC pair, and currently a MONSTER 3000F UC pair in each. Those UC pairs have to be carefully pre-charged with a current-limiting setup before connecting them into the setup (they are nearly a dead short for much of the charging time), but the sonics are definitely worth it! For as great at the Shanti are (and they are bargains for the SQ, they compared well with some of my best DIY'd supplies I have here with parts costs up to and over $1000), it is still amazing what you get with roughly 100x the 3.5F UCs in the Shanti. AND even better 1000x! AND except when there's severe storms predicted (such as earlier this week and early next), I leave my Shanti and the rest of the systems on ALL the time too (except for the tube/solid state hybrid amps, which I leave in standby mode when not listening for tube-life purposes)! Greg in Mississippi
  4. Allo has very good regulators on their USBBridge Signature, but some are LDO's with <6V maximums. Feel free to try it with a 6V battery. You may get lucky. More likely you'll have to buy another. Greg in Mississippi
  5. @4est, Nope, it does not. Ian Canada's FiFoPi is a better bet for DSD/DoP source reclocking. I should add that the FiFoPi works VERY well on the USBBridge Sig. See my PM in a couple of minutes. Greg in Mississippi
  6. I've played with using ferrite core chokes (mostly the clamp-on type) since the late 1990's. Where I've found them useful: - On the AC cords of any non-audio-system household electronics and appliances... washers, dryers, microwaves, refrigerators, garage door openers, lamps (ESPECIALLY THOSE WITH DIMMERS!), phone chargers, laptop power supplies, computers & peripherals, basically anything plugged into your AC line in the house that MIGHT make noise! - On the AC cords and DC cables of any audio system pure-digital gear... NASs, USB drives, switches & routers, FMCs, source computers in a 2-computer setup... basically everything up to the endpoint. Also if you have gear that have separate AC connections for digital versus analog sections, on the digital side (I do this when I modify gear such as my Sony HAP Z1-ES). - Ethernet cables, especially prior to the connection to your endpoint in a 2-computer setup. I've not found them useful on any analog signal, power, or AC cables. Occasionally on the AC and DC power connections to some endpoints, but as the endpoints get better, less useful... for example I found them useful on a stock RPi endpoint with iFi or other SMPSs or just average linear supplies, not useful on RPis modified with all-linear regulators replacing the DC-DC converters OR with a USBBridge Signature where I'd use them with very good linear power. I treat AC noise control with a 'defense-in-depth'... ferrites on household appliances and other AC-connected items that will produce noise in the house along with local parallel noise filters, audio system computer & network gear on a separate AC circuit with extensive parallel filtering (especially when using a galvanic isolation barrier such as optical networking or an EtherRegen), and good power regeneration (PS Audio Power Plants) along with pre-regeneration parallel filtering. In my setup and to my ears, it all makes a difference! Greg in MIssissippi P.S. Thanks Barrows for your Charles Hanson ferrite story. I'd heard part of that before, but not the part of using a tape demagnitizer on the ferrites to restore them. I'll add that to my regular maintenance regime!
  7. What I posted a week or so ago: On 10V vs 12V, I don't remember seeing any controlled tests that suggested a preference. Greg in Mississippi
  8. @Heckyman and anyone else interested, The SD Card Players such as the SDTrans384 and DIY'd ones based on inexpensive ones from EBay CAN provide a very good I2S source for feeding an I2S-connected DAC, either DIY or the few commercially available ones. In addition, the SDTrans384 includes a wired S/PDIF output and a PS Audio-style I2S-over-HDMI cable output for use with other DACs. AND with a bit of DIY work you can add those to the inexpensive EBay units. Benefits CAN BE a top-notch I2S feed for great sound. Of course that is all dependent on power supplies, clocks, good (and for native I2S, SHORT!) signal routing, and other implementation details. Use of one of these is not by itself a guarantee of good results. Drawbacks are due to the limited source material you can fit on an SD Card... typically the units will operate with 32Gb or 64Gb cards max. AND the SDTrans384 won't read compressed files, which for PCM limits you to non-compressed WAV files. I have a likely mid-sized library of roughly 1.5Tb of music files, most in FLAC, so roughly 2Tb uncompressed. You have to deal with making the music you want to hear available to use on the player, which can be setting up a process for easily copying it to an SD card or pre-loading a bunch of cards with your preferred music and cataloging which card contains what. Definitely back to the stone-age (or vinyl / CD age) for content management. AND they don't support streaming at all, though with some care and advanced DIY'ing, I suspect you can create a setup that supports both. As I've posted here and other places, I implemented an SDTrans384 setup in an effort to get a good reference for my DIY efforts. It has served well in that capacity for several years. In my thirst for the best sound, I'd put up with limited and kludgy access to my music library. Over that time some upgrades to both the SDTrans384 and the attached Soekris DAM DAC along with the attendant power supplies have kept them at the forefront. In Mid-2018 I modified a couple RPi 2B with LDOVR.COM's RPi Mezzanine board to replace the RPi's DC-DC converters with good linear regulation, I felt that setup with the best power supplies I then used put an RPi-based network player at roughly the same level as the SDTrans384 sourced setup when used with basic power supplies. While that configuration didn't challenge the SDT/SDD combo's sonic supremacy in my setups, it was getting into the same range. My recent moves to the USBBridge Signature as an RPi replacement AND Uptone Audio's EtherRegen as my switch have together made such a difference that I suspect once I take the time to compare the SDTrans84-connected Soekris DAC setup to my RPi/USBBridge Sig-connected one, now the SQ difference may be so small that the limitations and headaches of using the SD-Card players may no longer be worthwhile. I'll try that sometime next year and report back. More info: SDTrans384: https://www.diyaudio.com/forums/digital-source/142562-microsd-memory-card-transport-project.html https://www.diyaudio.com/forums/group-buys/260565-gb-sdtrans384-micro-sd-card-digital-transport.html?highlight=sdtrans384 http://www.tirnahifi.org/forum/viewtopic.php?f=15&t=3705 Other SDCard Players: http://www.tirnahifi.org/forum/viewtopic.php?f=15&t=3750&start=20 http://www.tirnahifi.org/forum/viewtopic.php?f=15&t=4541 https://www.diyaudio.com/forums/digital-source/285630-sd-card-memory-flac-wav-192-24-player-ess-crescendo-ii-ak4495-dv20a.html https://audiophilestyle.com/forums/topic/55533-audiophile-sd-card-transport/page/6/ https://audiophilestyle.com/forums/topic/55818-qls-quloos-qa361-dap-as-servertransportplayer/?tab=comments http://www.tirnahifi.org/forum/viewtopic.php?f=9&t=4682 https://www.ecdesigns.nl/en/ (currently offering USB-Stick players, used to offer SD Card Players) Greg in Mississippi
  9. See this post in the Firmware update thread... I felt the author did a good job of describing the experience of the process. Greg in Mississippi
  10. I have 2 Shanti here in daily use, powering a USBBridge Signature in each of my 2 setups. I like them a lot and feel in that appication they compete with much more expensive commercial units and my over the top DIY efforts. See my comments on using the unit in posts 222, 258, and 594 of this thread on DIYAudio.com: https://www.diyaudio.com/forums/vendor-s-bazaar/337859-shanti-dual-lps-5v-3a-5v-1-5a.html I did compare them with some LPS-1.2 setups (LPS-1.2 singly and 2 paralleled via a dual-section MPAudio 3-paralleled LT3045 board) and found it was roughly equivalent in my opinion with some tradeoffs. AND definitely less expensive! To definitively beat the Shanti for powering DAC / digital interface cards and USBBridge Signature's, I have to use supplies that float-charge a pair of 325F Ultracaps as the final output buffer. These sound amazing, but are clearly expert-level supplies that have to be pre-charged before use and require a lot of care in use lest one blow up either the charging power supply or what is being powered. IF you have to have it explained to you, you really don't want to use them. They are THAT challenging! The good thing is that the Shanti, with a 3.3F Ultracap pair as output buffers on each rail give you a good taste of what one can get with a 100x larger buffer! LOL... I agree with @nbpf on the form factor and captive cables, BUT I also don't expect much in and from a $160USD power supply with 2 outputs. In that context, I think the Shanti produces an amazing performance at the price, punches WAY above its weight class, and recognize that a more convenient form factor and better cables would have likely increased the cost and may have compromised performance. I also totally agree with @Tucker on vibration damping for the Shanti. See my comments in post 335 of the above thread and poster @chipsahoy in post 349. Greg in Mississippi
  11. @Heckyman, No I have not. I use the Logitech Squeezebox architecture with the RPi endpoint with the PiCorePlayer distro/Squeezelite SW combo. Playing local files would require running the LMS music server software on the RPi and my experience has been getting the best SQ out of computer audio means the lowest SW load one can achieve on the player. SO I'd expect that to be a step backwards. BUT honestly, I haven't tried it. If you scan the Uptone Audio ER Listening Impressions thread, you'll see comments that the ER brought streamed source up to the level of local file-served source. BUT that is not the same as what you're asking. Sorry, I can't help you here. OTOH, see my comments on TirNaHiFi here: http://www.tirnahifi.org/forum/viewtopic.php?f=15&t=4541&p=89937#p89937 Summary... my reference digital playback source is a modified SDTrans384 SD-card player. In an effort to close the gap between that and my file-served music, I modified a couple of RPi 2B's with all-linear regulated voltage rails from an LDOVR.com Mezzanine board replacing the noisier on-board DC-DC converters. This took the RPi source into the realm of the SD player, though as the lowest level of that unit's range. When I got the USBBridge Sigs in, I had been running my RPi setups using those modified RPi and the USBBridge Sig was a significant step up from them. With the additional jump in SQ from the ER, I now know I need to recompare this source to the SD card player to see where they stand. I'm not expecting the file-served music through the ER on the USBBridge Sig to beat the SD card player, but I suspect it is VERY close to parity now. BUT honestly, the results I'm getting now are so good that I've lost much of my will to even do these comparisons and have been focused more on DAC and power supply upgrades and getting more of my music library available on my LMS server. Greg in Mississippi P.S. I've posted this elsewhere, but I think it bears repeating here. I was not expecting a significant SQ boost out of the ER. My main reason for wanting them in my systems was to eventually make use of hi-quality upsampling with SW like HQPlayer, which will require MUCH more powerful (NOISY!!!) music server computers. My networked music servers setups are very low-power Zotac picocomputers feeding a switch selected for good SQ (ZyXEL) and a optical link to eliminate the possibility of ground noise contamination, again with FMCs selected for SQ (Trendnet sending, Delock receving). ALL of the gear was powered by at least DIY'd linear supplies with crucial pieces (switch and receiving FMC) powered by Uptone Audio LPS-1.2s. I modified the receiving Delock FMCs with linear regulation. AND the music files were stored on an SD card into the picocomputer which I found sounded better than a USB stick in this setup. Given all of that I was expecting a modest SQ boost from the ER. I got much more.
  12. @matthias, Actually I already have. I have used the ER in 2 setups, both using a USBBridge Sig as the player endpoint. One has a somewhat modified Katana stack, the other a somewhat modified Ian GB FiFoPi/Dual-Mono ES90338Q2M/IVStd stack. Both benefited greatly from the ER. I strongly suspect the USBBridge Sig works synergistically with the ER. If you look at my last 12 or so posts on this board, almost all are raves about the ER in these setups. AND my emails to Ioan at Allo about this have been even more enthusiastic! Interestingly, I felt the Katana setup benefited more from the addition of the ER. Before the ER, I would have put my Ian GB setup a little ahead of the Katana. Now I'd say the tables have reversed. Though I'm listening to a different discrete opamp in Ian's IVStd tonight as I type and the tables may have turned again... will see after I get some hours on them. Greg in Mississippi
  13. @ray-dude, Thanks for that very measured, organized, and informative report. Honestly, after getting my ER in and through a couple hundred hours of run-time, things sound so good I've lost my will to try different things out with it and have instead been focused on listening to music and trying various options and tweaks on the downstream side to see where I can take my setup. I did do a couple of quick trials with lesser linear supplies than the 2xLPS-1.2 paralleled via a dual-section MPAudio 3||LT0345 board. They confirmed that the dual LPS-1.2 setup was superior, though I haven't put the stock SMPS back in to see how it compares again. Based on your comments, I'll soon try a single LPS-1.2 soon alone and also with an MPAudio single 3||LT0345 board just for reference. Based on using the dual LPS-1.2 / MPAudio setup powering RPi endpoints (stock, modifed with linear regulators replacing the DC-DC converters, and the very nice Allo USBBridge Signature, basically an audiophile RPi) along with my SDTrans384 SD card player and other various DIY DACs, I expect the dual setup to come out on top. Of course I'm using about $1000 of power supply setup for the ER. AND sorry, living in rural Mississippi, I don't have access to a PH SR4 to throw into the mix. And I'm curious what @kennyb123 finds with the Farad supplies. I suspect they will work very well too. My current top DIY DAC supplies use an LPS-1.2 charging a ultracap pair. BUT part of why I do DIY is that I can do things that are difficult, dangerous, or expensive to build into a commercial product. From what I've read, Farad uses roughly 4F UCs in their supplies, presumably some number in series based on a unit's output voltage due to the typical 2.7V max per UC. In contrast I've been using a pair of 350F(!!!) UCs for each of the rails (typically 2) on my DIY DACs. And another pair for the 5V to the USBBridge Sig. This has to be heard to be appreciated, especially in comparison to other supplies I have available. AND I suspect it will be VERY difficult to turn something like this into a product, as at a safe 1/4 amp charging rate it takes roughly 1/2 hour to get a 350F UC pair charged to 3.3V with 5V taking a little longer. AND I have a pair of 3000F UCs to try on the USBBridge Signature 5V supply based on the recommendation of another over-the-top DIY friend! I'm expecting overnight charging for these! I say this because I DO need to try a 4-350F UC string at 10V into the ER, energized by the dual LPS-1.2 setup I use now. Could be interesting! Honestly, based on what I'm hearing, the ER DOES work very well with the stock SMPS. AND my experience (and why I do this DIY thing), better power always produces better sound! Greg in Mississippi P.S. Since I wrote this, I saw your cable comparison too. THANKS! I haven't experimented much here yet, thanks for a great guide forward @ray-dude!
  14. Not a single one that I know of. See John Swenson's comments on the various LPS-1/1.2 threads to understand that topic better if you don't already. AND read the monster 'A novel way to massively....' thread and others to see reports on various power supplies that suggest that leakage matters but is not everything. Jud, You hit the nail on the head. I still need to do some comparisons with much more humble supplies and see how they do. BUT if I can't improve over a $50 or so supply with a roughly $1000 setup, I need to turn in my DIY'er / Tweaker secret decoder ring! Greg in Mississippi
  15. An ER Firmware Upgrade experience… Alex sent it to me as I had an interesting issue with my somewhat unique setup when I first put in the ER a couple of weeks ago. I have a very simple setup with no streaming and not even a connection to the Internet. I have a Zotac picocomputer I use as an file server (Logitech LMS server) and an old 2.4G WiFi router I use as an access point for connection via a cellphone to control the system. They were both connected to a ZyXEL switch which fed a sending Trendnet FMC (sounded better than the TP-Link ones I had here) and that connected to fed a slightly-modified DeLock receiving FMC (added a 3.3V regulator on board so I could feed it with a good linear supply instead of the included SMPS). When I got the ER, I pulled the 2 FMCs and the ZyXEL switch out and connected everything to the ER. I could not get a connection to the file server via the WiFi even though the lights on the jacks were blinking. I played around with it a little and found that bringing the ZyXEL switch back, connecting it to the ER, and connecting EITHER the Zotac file server OR the WiFi Router through the switch to the ER A side worked ok. Weird! I’d been in touch with Alex about the issue (it is such an un-common system that I didn’t want to trouble-shoot in the forum) and he sent me the firmware update to try. First, Alex’s instructions were flawless and it worked like a charm as described with no issues. Second, when I pulled the ER to do the update, I just connected everything back into the ZyXEL. Instant SQ drop. YUCK! I’m GUESSING that was due to the increased noise fed via the Ethernet connection into my endpoint even though the bits being played were ones still in the buffer that had been transmitted via the ER (my setup will play several minutes after disconnecting the input to the endpoint). I did the firmware upgrade and went back to the system. It might have been my imagination, but the SQ drop seemed larger. Possibly because now the bits being played came through the ZyXEL with no ER regeneration. DOUBLE YUCK!!!! I put the ER back in with the setup configured just like before and things worked as before. Then I tried reconnecting everything through the ER and now it worked ok without the ZyXEL switch! Happy camper. AND my impression was that the overall SQ went up a notch post firmware update. Most notable was better definition and dynamics at very low levels (while my wife is home and I can’t play it as loud as I like). A winner got even better! Greg in Mississippi
  • Create New...