what are the pros and cons of valve amps?

[Æ]

That you have to re-bias tubes frequently, they age and burn out quicker than solid state because they run so hot. (feels good on a cold day) Tubes are fragile and can break and are microphonic.

[jteef]

Grey, when you say that you think tubes "image" better than solid state, do you know of any physical/operational differences between tubes, fets, or bjts that would account for the better imaging? different bias levels, levels of doping, etc?

I think good imaging comes from channel separation and accurate signal reproduction. I think we've been able to reasonably achieve both for quite some time. This makes me wonder how any modern amp could have an effect on imaging unless its being played past its limits.

I, along with many others, would love to know the "why's" of this stuff. You'd think we would have figured them out in 75+ years Do you know if they give masters or PhD's for this kind of research? Journals or papers from the AES maybe?

It almost seems like there is some unknown characteristic(the schwartz?!) that weighs heavily on sound reproduction. On the other hand, it also seems like it might all be in our heads. I don't think i've heard of one person who could consistantly identify one quality amp over another while blindfolded.

The biggest problem I see with imaging is that a lot of speakers/setups strive for the ideal point source model(our current recording methods do cater to it), when a symphony is anything but. Want better imaging?, set up a wall of drivers, each one corresponding to an instrument and its general location on the stage. (more frugal models might include a speaker for each section, instead of each instrument). Start recording 10-50 channels and stick it onto a DVD-A disc. Sure, its got an SAF of about zero, and, it might be expensive and ridiculous, but not nearly as crazy as the clowns who pay for a krell master reference and set of wilson WAMM's to get the "best" sound. There is a market for anything in audio.

jt

[GRollins]

AE,
Aye, stick fragile into my list above. Can't argue that one. Of course, I've seen many a bent heatsink fin, although that doesn't prevent the amp from being used.
Microphonic? Not necessarily. It depends on the tube. I've never had a microphonic 6550, although I'm sure they exist. 12AX7s are sometimes microphonic--depends on the manufacturer as much as anything. 6SN7s are never microphonic here, chez Bear (I use lotsa 6SN7s--love 'em). We can put that in an iffy category.
Bias? That's more a design problem. There's no need to put up with it. I use regulated bias supplies (most people don't go to the trouble, I guess). I check the bias about once a year. Sometimes I find it needs a whisker of adjustment; usually not. If you really don't want to fiddle with bias, there are two excellent options: cathode bias (meaning to use a resistor under the cathode), and current source. I chose fixed bias (an external voltage applied to the grid) because it's more efficient, but either of those other methods are certainly rock-solid as far as bias stability. Sorry, I don't view bias as a problem in a well-designed circuit.
JT,
I've heard people postulate that tube imaging is due to:
1) Distortion products. Bull. There's no distortion product that's going to bring the back wall of a concert hall into focus. Quite the opposite, I'd think.
2) Microphonics. Some think that there's a positive feedback loop going on between sound in the room and the tube. Again, I don't buy it. Anything of that nature would--because of the time/phase lag if nothing else--tend to blur the image, not bring it into clearer focus.
3) Greater linearity. If I say that one out loud, a solid state fan (usually bipolar, for some reason MOS/J/FET folks aren't as defensive--don't know why) will come rushing in complaining. This usually leads to discussions of comparative rates of...
4) Feedback. Generally, lower amounts of negative feedback lead to better imaging, although this is <i>not</i> a hard and fast rule. The idea here is that NFB gives with one hand and takes away with the other. Sure, you get greater bandwidth, damping factor, etc. but you lose imaging and inner detail. Choosing the right rate of feedback is an artform, not a science.
But the long and the short of it is...nobody knows why tubes do the things they do. You can theorize all day, but not get anywhere. The problem is that no one has yet figured out what (regardless of gain device) makes a piece of hardware image well. There are some who go so far as to deny that there is such a thing as depth of image (hopefully nobody argues about image width). Given what they listen to, they're probably right--for their system.
One oddity is why some amps present an image that is further forward than others. The Aleph 2s I built image <i>behind</i> my tube amps by about a foot. Say what? I've heard amps do this before, but never had such an obvious case in my own home where I could play with it. I'm building a second pair of Aleph 2s (up late last night populating the front end boards) so as to try some experiments with this. Nelson and I wrote back and forth about it; his take was,'It's a mystery.' If I'm able to resolve any questions on this front, you guys will be first to know. Don't hold your breath. Nobody's been able to pin this down yet, and I doubt I'll be the one to do so.

Grey

[Æ]

There is one amplifiier that uses a small microprocessor and DAC to set tube bias automatically during each and every turn on/warm up sequence. It adjusts for the aging and different (EL34) brand/date of tubes. Neat huh?
Wolcott Prescence Monoblocks, 220 watts and 75 pounds each. But they cost about 10 grand for a pair. They will drive Sound Lab Electrostatics better than any other amp! Talk about sweet sound!

[GRollins]

AE,
My problem with their strategy is that the tubes are read cold (at least as I understand it), which isn't going to give you an accurate bias setting. They ought to run for an hour or so first.
An even better way to do it if you want to go active bias is to use a sense resistor, monitored by an opamp with a lowpass filter set at, say, 1 Hz. The opamp drives the bias voltage. Voila, real-time bias adjustment, not just when you turn on the amp. I might incorporate something of this nature into the OTL I'm working on.
Supposedly, OTL amps are a wonderful match for electrostats, but I've never heard the combination.
Sound Labs M-1s are one of the two electrostats I've ever given serious thought to buying. (The other was the Stax F-81--a gloriously impractical speaker.)

Grey

[Æ]

So it gives the tubes time to warm up, and you can press the autobias reset switch at any time while playing music, it temporarily mutes the music and starts the autobias cycle all over again. Smart, huh?

[GRollins]

I'd give it an hour or more before trying to check the bias. It changes dramatically during that time (solid state does too, for that matter--this isn't limited to tubes). If you've got to go tell it to check the bias manually every time you listen (after an hour) then you're actually losing, rather than gaining. It'll start not knowing the bias, take a cold read (anything less than fifteen or twenty minutes minimum will be way off), and be wrong. After a bit, assuming that you want optimum performance, you'll have to rebias. Sounds like you're going to have to do this every single time you listen, whereas I only bias once every six months to a year. I dunno, seems like a lot of unnecessary fiddle factor to me.
Incidentally, I haven't heard the amp. I'm not saying that it doesn't sound good. It just seems like a flashy way to bias without much thought towards the practicality.

Grey

[Thatch_Ear]

Put biasing, tube life, power etc all aside the first time I really heard my speakers, the first time I got great imaging and soundstage was when I hooked up a pair of 3.5 watt SE 6BQ5 mono blocks built on the old Mullard circuit to my humble but loveable Shahinian Obelisks which are by no means efficient in my book at 91-92 db. I do get out and listen to "High End" of the latest flavour( all the B&Ws, Martin Logan, Magnepan, Krell, Mac, AR, etc,etc)and recently went to a DIY Speaker Convention in Texas where tubes, SS, great cables and expensive interconnects were all there with about every type speaker represented in some fashion except Elecro Statics and I did not hear the kind of imaging that the SE tube amps provide. There was no SE tubes there. Most of the speakers were less than 90 db and had complicated crossovers and SE wouldn't have worked with them anyway. The point is that I have heard better sound reproduction with $2600 retail (I don't do retail)gear and speakers with DIY cables and interconnects than I have heard on $50K plus systems at dealers. That is the whole point to DIY isn't it? You want what sounds best for the least amount of $$ and the pride of DIY. And by the way I have hooked up Marantz, Carver, Phase Linear and other SS, and Dyna ST-70s, Fisher PP integrated amps and recievers and others to the same speakers without comeing close to the sound of the Mullard 3.3s which are not the best SE out there (well maybe for the money). The only draw back was because of the inefficiancy of my speakers I could easily clip the amps. That level was well above my normal listening
level so the only thing I could not do was really crank it. Bi amping would take care of that problem though. The best imaging of the soundstage I have heard with SS sounded artificial and like it was a thing the amp did and not necesarily something that was on the recording. These are only opinions, but I do want to say that the only way to measure the difference between the different types of amps is to listen to them. And listen to them correctly. Which means with the right type of speaker and cables etc. It really is apples and oranges.

[Æ]

I doesn't take a tube as long to warm up as solid state. Checking bias on solid state is sometimes real tricky, I know how fast it changes with temperature change. Take the cover off of the amplifier and the temperature along with bias changes immediately. Then put the cover back on and watch your adjustments drift again. Frustrating! i once set up a SS amp with external test leads so that I would not have to remove the cover, then adjusted the bias and offset over a two week period.