Gold wrote on Fri, 17 July 2009 20:16 |
Andrew Hamilton wrote on Fri, 17 July 2009 17:29 |
But for most contemporary equipment, with a balanced input in the range of 10-20kOhms and output impedance much less than 100 Ohms, the "mismatch" caused by connecting the "low" side of the source directly to ground (as versus the driven output) may very well be less significant than the difference in the input impedance due to normal tolerances in the input resistors of the balanced amp. It probably depends on whether the parts are super-high accuracy (like 0.1 percent tolerance) or are hand-selected/hand- adjusted.
|
I can't see how normal 1% resistors found in most pro audio gear would be a worse common mode match than grounding a leg.
|
Well, Sir, he did say that you need 0.1% tolerance in the network to be spot-on... I should think that four [mere] 1.0% tolerance, non-cherry-picked, resistors are indeed capable of ending up way off the mark compared to "...only a little worse than 0.1%..." shouldn't I? Maybe his 0.1% is too stringent to begin with.
Regardless, my complaint with the system has never been CMRR, or at least not specifically line noise, itself. It's a generally brighter sound, when balanced, and possibly a slight "brilliance" is added to the transients. Is this what happens when CMRR degrades? This sounds more like a loading effect. Or maybe there's something to the audibility of differential amp hiss, on a psychoacoustic level, at least... Or maybe my balanced input circuit is broken in some other way. It was bought second-hand as a factory refurbished. They might have not even tested the XLR's thinking most people would only use the RCA's who bought it second-hand. My Dunlavy's are second-hand, too (or third!). They had broken crossovers. One of them had apparently been improperly assembled at DAL (missing a resistor). The other had a resistor with a broken leg, rendering it, essentially AWOL from the network. I'm open to even spooky action at a distance as the culprit. (or not);
Can't do this double A/B/X blind. Only 3rd-eye blind.
Alas, I can not implement Mr. Putzeys' ideal balanced approach on either my amp or my console, ironically, because the designers of the gear that I happen to have* didn't trust us to know that we could brake the rules responsibly - that we could lift pin 3 on output as long as we remember to shunt (is it?) the inverting leg (say) of the load to ground by the amount of resistance determined by the driving amp's output Z. And/Or/Instead, they came up with a proprietary scheme (at least one which is not publicly discussed) that allows any dummy (such as I) to use an off the shelf XLR interconnect, and use it to send either a traditionally (fully) balanced or an unbalanced one with excellent noise rejection qualities, either way. At least this is what it seems like.
Mr. Putzey's words are "Science" (as Ron Burgundy would put it), and no one is disputing that. But there's a locomotive analogy to Science. Pure Research is the locomotive engine, which is 50 or more years ahead of any likely application. I'm hopefully somewhere near a dining car, but I realize that it's way in the rear cars of the train.
This audio mastering has so many debates. Like, do I want to risk using a pinch roller in order to get rotating lifters? Or do I want to suffer with stationary lifters and a slow wind and risk servo-snap in order not to have to deal with a pinch roller?
ITB or signal jack? Just when the domain converters are getting good, so is dsp (getting good).
Do I want more songs on my iPod at low rez, or do I want them to have better digital resolution, even if there will be forcibly fewer songs, due to the increased demand on storage capacity?
Do I want to go for bottom or for top (end)? etc... (i.e., 15 or 30 ips). Choice is nice. But sometimes there is no satisfaction. Sometimes, both choices are wrong, and there is no third way. ):
Spend / Save
I'm just glad he didn't say unbalanced is for turkeys. I'd have to introduce him to my distinguished e-Mentors on the Blackwood list. (;
* M•DA-824; MEA-2; MLA-2; MPL-2... But also, Bob's Trakkers, and possibly other transformerless active balanced stuff? Just a caveat lector...
Andrew