Rivendell61

But......they are not tube amps. They are Class D. Good Class D will be as load independant and linear as the best conventional amps. OTOH--Flying Mole are not particularly good Class D amps. Mark

DrWho, The only reason the interconnect I described gets the nod in unbalanced to balanced is because it is more likely to maintain signal integrity (SNR, etc) than the alternatives. Now....if you're getting increased noise using that balanced connection then I would look at your gear. Is the interconnect wired properly? Or, more likely, the ground/circuitry (of the unbalanced or balanced side) may be less than desired (i.e., screwed up). Might be worth looking into. Re power--it should be a non-issue. Again, the lack of a perceived problem is why the 1:1 option gets the nod over the 1:4. The PLX has a gain adjustment for the very reason of matching the input. OTOH.....this is all very gear/situation dependent. While that interconnect is going to be best in most (studio or home) cases there may certainly be situations where the amp is not capable and/or the signal quality issues are not paramount, etc. Mark

I believe this one is cheaper: http://www.musiciansfriend.com/product/ART-CleanBOX-Two-Way-Stereo-Converter?sku=180613 Basically does the same thing. DrWho--I think we have been around this track before? Don't use those boxes. In home audio playback where signal integrity, quality, etc are paramount there is a better way to connect unbalanced to balanced: Create a balanced connection with a shielded twisted pair (ala Belden 1800). Have someone like Blue Jeans Cable wire it with XLR on one end (per QSC input) and an RCA on the other, with shield tied to ground at the RCA end. That's it. You will now have a balanced connection with improved noise rejection/CMRR over the unbalanced or the 'adaptor' method (which wastes the balanced input). Now, that will work in most situations (should always work but....there will always be an exception). The primary requirements are a correctly designed balanced input on the receiving amp, and a fairly low output impedance on the RCA end. And it is really the 'correct' way to go from unbalanced to balanced. Mark

I'm sorry to say, none of my posts were actually related to the cable issue...just the claim re inaudibility of 0.2dB. But roll-off should be vanishingly small in any domestic application. I think elsewhere I quoted for you Benchmark Media to the effect that an unbalanced RCA out of its DAC/pre-amp should be under 1360 feet to maintain a sub 0.1 dB loss at 20 kHz.....which should be sufficient for most houses..... Mark

Exactilioso DrWho re your prior posts.... The 0.1dB was (did I say this already?) ABXed reliably/repeatably in-house by Sony. Interesting the response here. Anyway, there is lots of this sort of testing done--most of which never sees light for various reasons, mainly because it is valuable as proprietary knowledge. Right--no reason to lose sleep over it. Can't imagine why! It is mainly information of importance for controlling variables in listening tests. Your point about larger issues in rooms, etc., is well taken and obviously we do not worry about such tiny anomolies. But it is also worth noting that the ABXing was done via speakers, in a room. My point being that larger distortion mechanisms do not neccesarily mask smaller ones. The paper was saying in part: 'pay attention folks we are underestimating our ears acuity to the detriment of audio quality'. Some other stuff they detected in listening tests were: "Harmonic distortion components at 80dB below signal level, more than 10dB below noise floor", "Limit cycle behaviour of Delta/Sigma DAC converters at 100 dB below max. signal level", etc.... Mark

The data does in fact support 0.1dB. Discrimination of very small artifacts is better understood all the time--in part due to work at accomplishing transparent digital conversion. Sony, Philips, etc., all do very good in-house testing. I have no idea what source you are quoting/reading but things seem to have moved on. Incidentally, headphones (mentioned in your example) have been found to reduce test sensitivity. Mark

Don't mean to pick on you but as I skimmed the thread it jumped out..... 0.1dB (perhaps less) variations in gain or frequency response are audible. So if you undertake to test some of the assumptions in your post regarding cables be sure to match your sources to 0.1dB or better. Mark

I must have had a dyslexic moment..... If you are going balanced to unbalanced just ignore my post above. As I understand it per Whitlock bal to unbal is more problematic because of the potential variety of balanced output circuits.... Mark

As Edgar says here and in the other post.... If you are wired up with the twisted pair, shield tied to ground at the unbalanced end AND you have a very low output impedance going into high....should be fine: having created a balanced connection. If the impedances are less than ideal you could lose some CMRR. That resistor thing will then help. Here is a pic:

Dean, You may already know this--or not be interested but....the stock module (180ST or AD?) tends to get modified a bit in builds for 'audiophile' use. There is some info on the Hypex Application page like this: http://hypex.nl/docs/gainappnote.pdf And the thread 'UcD 180 questions' at diyaudio: http://www.diyaudio.com/forums/showthread.php?s=060891883ae626475ee0493630d6e17b&threadid=38199 The amp designer contributes a lot and while it is a long thread if you skim looking for posts by "Putzeys' and 'Jan-Peter' (and later in the thread maybe Haskins, et al.) it might be useful. There are several other interesting threads there too.... Good luck with the modules. Mark Edit: corrected link

Speakerfritz, my criticism was only of the language/info not your gear. No question: subjective appreciation of the sound is ALL that matters here. Mark

Perhaps it was a mistake to have mentioned a specific DAC. Again, no, as I said in my post above: it has nothing to do with a particular converter--it is inherent in all DAC circuit architecture. No DAC (that is well designed--some poor converters may) will perform better at 192 than 96. In fact, some very good DACs will operate better at even lower rates. Now, this says nothing about what someone may like or not like ala your Onkyo. There are many reasons why a DAC may sound different/better at a less optimal rate--but I won't go there. And subjective preference does not always track signal accuracy. The 192 speed issue is very well discussed. Here is a brief explanation of one of the reasons 192 underperforms 96: "The point about 192kHz being worse than 96kHz centers around practical behavior of real life electronics. The settling time of any sampling circuit is finite. A given sample/hold circuit (or sampled integrator in switch-cap sigmadelta circuits or voltage switch in continuous-time filtered deltasigma circuits) can settle, for example, to 1ppm in 10us but only to 10ppm in 1us. This means that when you're trying to sample faster, you won't get the job done with the same accuracy." But all this is really nothing to someone listening to music--the flaws in a 192 DAC are at the extreme margin and may even be perceived favorably. I was only contradicting you because your original stated a 192 DAC was needed for SOTA performance--quite the contary. Belief in 192 does drive lots of sales--so don't expect main line consumer converter makers to be honest about them the way many pro makers are. Mark

Speakerfritz--my post above was speaking of 192 DACs. The Benchmark incorporates a 192 DAC (Analog Devices) but does not operate at 192 for the reasons I stated above (i.e., signal quality). Mark

I have found that in the cases where it is considered degraded, a 24/96 chip was used upsampled to 192 thru overclocking. Since DAC chips are sold in lots of 1000....it is expected that manufacturers justify their use of older 24bit /96khz or even 16bit /44khz chipsets until they exhaust their bulk supplies. btw, The Lavry DA10 can operate at 192KHz in "Wide" mode. No, it has nothing to do with chip availability, or quality (or 'chips' at all really since the same applies to discrete converters). I should have been more specific: it is inherent in the circuit architecture. Any DAC operating well at 192kHz will perform better (less distortion, noise...) at 96 kHz. 192 sells so we have lots of them and lots of cheerleaders--but a competent converter will give a more transparent signal at a lower speed. (And yes the DA-10 does 192--but that is the cheapest Lavry DAC aimed in part at consumers who think it is better. The top of the line Lavry 'Gold' does not. Nor does the Benchmark for the reasons I stated.) Mark

Lee, There are so many options out there (which is good and bad) that it is hard to give useful suggestions. One way: A cheap Dell is a good base. Add a large second hard drive (internal) ala 300GB Seagate (will fit c. 800+ CDs at FLAC level 0 compression), c.$95. MusicXP web site( http://www.musicxp.net/ ) is good to guide how a PC should be set-up if dedicated to music (Tascam and others offer similar guides). Foobar2000 is an excellent, small footprint, totally customizable player/interface (but needs a day or two dedicated to figuring out/setting up!). Output from PC has many choices but easy and excellent is USB to a DAC. Apogee just lowered the price of the Mini-DAC to sub $800 and it has a USB input. Benchmark is maybe a bit better and another couple hundred (and need a USB/SPDIF converter). For ripping music onto the HD a Plextor 760 (IDE) internal optical drive is a nice choice (or, I think, 716 for external). The Plextor (comes with Plextools software) will rip into FLAC (good lossless compression algorithim) just as secure as EAC (Exact Audio Copy)--with error checking, etc. And you will want some external HD (or two), same size as the internal for back-up copies for when the internal drive self destructs. Wireless ideas: don't really know (I don't do it). Squeezebox is very popular. Mark

Hi Craig! It's been a while because more than 3 minutes of consecutive free time only comes c. 3-4am. I'll e-mail you later today. Mark

Cambridge 640H specs: http://www.cambridgeaudio.com/specifications.php?PID=39&Title=Specifications And FYI: 192kHz DACs are not an indicator of quality, often quite the opposite. Some of the best DACs (Benchmark, Lavry) refuse to output at 192 sample rate because it actually degrades the signal. Mark

I understand this thing can shtomp Max's gizmo (well....maybe not). But it is kind of cool: http://www.altmann.haan.de/turntable/

rofl - is that supposed to be a big deal? [] Yeah--I wasn't sure if he was praising or damning. There are over-the-top subjective reviews of UcD module based amps like the Kharma Matrix: http://www.positive-feedback.com/Issue25/kharma_mp150.htm The specs do clearly indicate some of that--like the flat THD/frequency, and the load independant frequency response. Not 'chip' amps.... Mark

Based on the little knowledge I have of those--if I was looking for a 'convenience' server solution....I'd get the Cambridge Audio 640H. If you were willing to spend a bit more energy you could have a SOTA hard drive system for about the price of the 640 by buying the cheapest Dell (c.$350) and adding/configuring a few bits. Mark

Dean got that pretty near. It is a Hypex UcD 400. He linked to a US dealer--or they can be bought direct from hypex for about 100 Euros (or 130 for the version with the Analog Devices 8620). By the time you get two, power supply, tranformer, etc.....it does get a bit steep just to be curious. Here is the link to the Hypex website: http://hypex.nl/ And here is the link to the module data sheet: http://hypex.nl/docs/UcD400_datasheet.pdf Mark

Pat--Yes they are analogue. Some are pseudo-digital ("power DACs") but most of the better ones are not. DrWho--I hope you get to hear one you like. Class D is actually really difficult to design so right now there are very few good amps out there--and lots of middling dreck. But the good ones really don't have any worry re matching specs or sonics with linear SS. The plot you indicate looks like a TA2024....? FYI: The amp I posted the THD/Power plot of is not a chip but 3"x3" discrete (see below):

DrWho, I wonder if the Power/THD plots for the Tripath amps in Dean's link are the sort distortion you were talking about.....? I'll post a Class D plot here for you. It compares favorably with most linear SS amps--and note the actual THD values at sub-watt power. If you are unconvinced re some 'inherant' Class D problem I can post one with even better numbers later..... Mark

I assume you know this--but in case not: the Monica is an NOS/Filterless DAC. So there are some exchanges being made between signal accuracy and audiophile sonics.....something to consider. Mark

Sounds good DrWho. Dig up those plots and we can have something concrete to discuss. However, my point remains: there is no 'inherant' distortion issue with Class D amps vs linear amps. No large peak as output is reduced, etc..... That is just bad info. "Perfect"? No....but is any amp? Just no less perfect than a Class A/B amp--and maybe a little bit more so....[] Mark