Understanding Tube Power Ratings

What do I need to know about the power ratings for tube amplifiers?

With solid state models, I looked to see if the power output doubled from 8 ohm to 4 ohm as one sign.

What about tubes? Are 250 tube watts comparable to 250 solid state in terms of driving power? Please, I understand equal measurements should mean equal results, but am I missing something?

I'm looking to move from the Bryston 4B ST currently driving my Thiel CS 2.3s. I have a Rogue 99 magnum preamp and have the new Rogue Zeus on the short list. The Thiel need for high current had me shying away from tubes, but I am rethinking this.

Don't be shy! Go for the tubes. I had Thiel 3.6's and ran them with older VTL 225 DeLuxe monoblocks, 225wpc, and they sounded great . . . better than I when I ran them with a Levinson 333, 300wpc. Tube power gives you "more," watt for watt, and tubes sound better.
Tubes watts are reputed to be louder than solid state watts. Of course this is not technicaly true. Tubes tend to overolad more gracefully than transistors. Jim Thiel told me that the recommended power given for his speakers were given with quality solid state amps, which could double down, in mind. He further recommended doubling those recommendations when using tubes, and to use the 4 Ohm taps on tube amps. Make sure the amps are up to the job at hand.
No problem. The Rogue 99 has a 23db gain and 30V p-p which is great for any power tube amp with the Thiels. I use a 3-watt 2A3 and a 26db/50Vpp CAT with my Thiels - and it works. And, no, you're not missing anything. Whether you have a tube or a SS at 250 watts, they both can, at full volume, blow out drivers which can only handle, say, 100 watts maximum.

As to why some will say that a "tube watt" is more than a "solid state watt" is not contraindicating the laws of physics. Tube amps have higher dynamic head room than SS amps. Headroom is the ability of an amp to push out more power than rated for a brief time, which comes in handy during transients. Each 3db of headroom doubles the peak power delivery. Tube amps have headroom ratings as high as 5 to 10 db whereas SS amps can have HR less than 1 db and 2 db is typical. The best SS have very low values (0.5 db and less) due to better power supplies.

If you look to SS amps for power doubling as load impedance is halved, then you run the risk of being baited by the sales brochure. Very few SS amps can double down to 4 ohms, much less to 2 ohms. While Ohm's law says this is inevitable, Ohms law also states that the voltage decreases with an increase in power (current). As the load is halved, the PS voltage drops and lessens its ability to keep up the current flow. The PS voltage and current regulators have to be of top quality and the transformer has to have plenty of reserve in order to double down. That's why a single $20,000 100 watt Class A monoblock will have a power supply transformer thats more than twice the size of a $3,999 250 Wpc and more elaborate voltage and current regulation stages. That's where all the money goes.
1) Gs: I've never seen anything in Ohm's law that states the following:

"While Ohm's law says this is inevitable, Ohms law also states that the voltage decreases with an increase in power (current). As the load is halved, the PS voltage drops and lessens its ability to keep up the current flow."

The only time voltage will sag or drop is if there is not enough current to be had within the circuit. The "pressure" ( voltage ) within the circuit will remain consistent so long as the "volume of flow" ( amperage ) is maintained. As such, an amp can act as a voltage source ( doubles down in power output as impedance is halved ) so long as the power supply and output stages are capable of passing enough current to meet the demands placed upon them.

When you find an amp like this, you will notice that the clipping power at 4 ohms is almost exactly twice that of what it is at 8 ohms. If it is really a "high current" design, it will produce twice the power output at clipping into a 2 ohm load that it does at 4 ohms.

2) Since there is SOOOO much "fudging" of power output specs and manipulation of numbers for marketing purposes, forget about "rated power" and start looking at what the amp does at clipping at various impedances. This will tell you a LOT more than what they rate the unit at. On top of that, forget about frequency response on a power amp. Look at the power bandwidth. Frequency response is spec'd at 1 watt of output ( typically ) whereas power bandwidth is the frequency response as measured at rated power.

3) As to tubes vs SS, tubes appear to be more powerful than an equivalently rated SS amp for multiple reasons. SS amps tend to clip much harder and sound much worse when running out of steam. Tube amps can be pushed into clipping without sounding nearly as harsh. In effect, this makes them sound like they actually have more headroom since they aren't getting "nasty" sounding yet they are being run at or above rated power output.

4) As to tube amps having a much higher headroom rating than SS amps, that is a pretty broad generalization that i can't agree with. Either design can have a lot of headroom. While most of this will have to do with the power supply of either unit, the transistors would obviously have to be able to pass enough power to do this. While the same is true for tubes, some tubes will do this far easier than others. Tubes with a high Mu are probably more suitable to "peak power" or dynamic headroom than those with a lower Mu with the same power supply / circuitry supporting them. At least that's what we've come to find out with RF circuits. 8417's are a perfect example of a tube that works excellently in an RF circuit and was also used in audio power amps. Too bad these tubes aren't available anymore. 6L6's are another example, but they are nowhere near the tube that an 8417 is.

5) Regardless of all of the above, it is harder to match a bunch of output tubes than it is to match output transistors. This is not to mention that tubes are "born to die" and replacing them at random as needed will surely create greater imbalances / poorer performance over time. As such, most "experts" agree that all of the tubes within one section of a circuit should be replaced at one time. Matched parts should be used when doing this if possible.

6) While i have nothing against tube amps or those that favour them, you should know what you are getting into should you choose to go that route. Tube power amps are SURELY higher maintenance than an equivalently rated SS amp and i don't think that anybody ( tube-head or not ) would argue that point. If you like the sonics of a tube amp enough to put up with the maintenance and expense that is required to keep it running at optimum performance levels, go for it. Sean
I had more success running my 2.2's on 45 tube watts (C-J) than on 70 (Classe) or 100 (NAD) SS watts. Now I have them on 200 tube watts (VTL) and they're even better (this isn't a parable though - my SS amps weren't of comparable cost or quality to yours, or to my tube amps). The 2.3's are a little more demanding than my speakers, but for average listening levels in an average-sized room, I would imagine that good 100-watt tube amps could contend. Stick with beefy power supplies and output transformers, and leave maximum rated power to be influenced by how loudly you like to listen and your room size. (For an amusingly perverse take on the conventional wisdom concerning 'doubling-down' and SS vs. tube power, check out Ralph Karsten's white paper on the Atma-Sphere website.) Bottom line: Sound quality is difficult to directly correlate with power ratings (and that goes in both directions!) or choice of active devices (tubes or transistors) - better to trust your ears.

Sean: Since tube amps, as a rule, don't 'double-down' as some SS amps can, but are certainly capable of sounding just as good, why do people focus on this aspect of SS amp performance and not of tubes? Is the fact that this capability doesn't seem to reliably correlate with sonic quality indicative that in the real world, most good-quality and properly-matched amps are not being called upon by actual speakers and listening styles to test their ultimate limits in the current delivery department, thus obviating the distinction? (In other words, at listening levels comfortably below the onset of clipping, even tubed amps are putting out all the current being demanded of them into varying speaker impedances when playing music?) If true, pure voltage-source behavior into static loads using typical test signals at maximum rated power would seem to be a red herring of a specification for anything other than offering bench confirmation how over-built any particular big SS bruiser may be - not a bad thing, but maybe not the most important or telling of qualifications either.
Z: Tubes and transistors are required to do the same thing in an amplification circuit. The advantage that tubes have is that they have ( typically ) a much greater voltage potential than that of most conventional SS amp.

As mentioned, so long as one can provide enough current to drive the load as needed, the amp with the highest voltage potential will typically play louder and sound cleaner doing it. As such, a tube amp is ahead of the game so long as the power supply, tubes and output transformer are relatively "beefy". The only times that you might run into severe problems with tubes on the output are if the speaker is very low in impedance and the output transformers are puny with impedance taps that are not well matched to the speaker. Otherwise, you will have a harder time controlling the driver and you'll eat tubes up like there is no tomorrow.

My personal preferences in SS amps are for those that run very high rail voltages and can provide gobs of current. Most SS amps will never have the voltage potential of a tube, but for all practical purposes, a tube will never have the same amount of current. As such, tubes will work best with higher impedance loads that are non-reactive in nature whereas a well built SS amp will offer greater control on low impedance / high reactance speakers. These are both very "broad" generalizations at best, especially since there are so many varied types of tubes / ss output devices and the support circuitry and designs that go along with them.

Having said that, almost all of my speakers are low impedance and make use of multiple woofers. The one system that i do have that has a tube amp in it is of a higher impedance that is of very high efficiency and a single woofer. As such, the single woofer is not making long excursions to produce high volume levels due to the very high efficiency, so there is a minimal amount of reflected EMF to deal with. As such, the need for "muscle" or "high current" is not very critical.

As to what is required of most amplifiers during normal listening sessions, bare in mind that speaker impedance, reactance, loading characteristics, etc... all vary as spl is raised. Whether or not one needs "brute force" for their system will deal with their individual speakers, room and type of music listened to.

Besides that, it all boils down to personal preference and system matching. Does anybody REALLY need the power, handling, braking, etc.... of a Porsche or Ferrari ? Even if they don't "need" such things, you can bet that they are running the vehicle on high octane fuel, have good tires on the vehicle and go in for regular maintenance. That is, if they want the vehicle to perform at peak potential should they ever really "need" it. : ) Sean
Tubes rule!
I have to put in my newbe .02.

I think sound should be your starting point. Forget about power. First decide whether or not you like tubes or SS
I am new to the tube life and let me tell you,I feel that tubes sound much better.
I a\b'd my Jolida 202 against a high end, high powered SS amp and like my cheapo Jolida heads and shoulders above the other.
This was on box speakers as well as Maggies.
It made me realize that there really is a difference and a large one at that.
Let me remind you that I'm really new to all of this. I haven't gotten into any of the tweaks other than tube rollin'. So the difference was not taken lightly when I heard it.
Agreed Sean, I wouldn't think tubes first if I had a low-impedance, multiple-woofer, highly reactive load to drive. I'm more just curious about the audiophile psychology when it comes to this type of spec (and anyone can feel free to chime in on the subject).

Surely we've all noticed how an amp's ability to 'double-down' - or even the illusory capability of it, engendered by fudged specsmanship (where the output capabilities of a SS amp into a 2-ohm load is presumably taken for the baseline figure and then successively halved into each doubled nominal load, thus underreporting the amp's 8-ohm capabilities in order to make the amp appear completely load-indifferent) - is trumpeted for SS amps and ignored for tubed amps.

Yet no mention is ever made of the obvious fact that this supposed attribute or deficiency, depending on which direction you're coming from, apparently makes very little to no actual sonic difference in most 'normal' set-ups. I mean, it's not like in most systems, a good tubed amp of appropriate power rating is going to fail to transmit the bass frequencies or something. Still, many audiophiles will schizophrenically look down their noses at SS amps which admittedly can't 'double-down', while at the same time holding tubed amps to a completely different standard where this ability isn't even a consideration. I am proposing that there may actually be a pretty good reason for the latter phenomenon: 'doubling-down' is apparently an ability that normally won't make a predictable sonic difference in the real world.

Presumably, this would be because most good tube amps of appropriately-matched power ratings, used with most typical speakers at common volume levels in average-sized rooms, will still be able to supply all the current demanded of them throughout the spectrum, regardless of the speakers' varying loads. Ralph Karsten (see my previous post) takes glee in proposing a divergent view of the situation as it applies to his own OTL tubed amps: he chooses to fault SS amps for wimping out and providing less power as impedance rises - instead praising tubed amps for putting out relatively constant power (discounting the effects of matching the output transformer to the load, since of course his amps don't have output transformers) across a broad plateau of load impedances. He also says, in effect, watts are watts, and most speakers won't even come close to demanding all the current theoretically available from the big SS arc-welders: power draw at normal volumes, as determined by common speaker impedance and sensitivity ranges, and as supplied where Watts = Amps x Volts from any type of amp, will virtually never exceed the current-delivery capabilities of a sufficiently good tubed amplifier. He winkingly asks, Should we truly want varying power into varying loads, or should we really want constant power into varying loads?

Actually, for myself, I suspect his argument is at least as much sophistry as the 'doubling-down' one. To me, the real issue is likely that how any amp responds at its maximum rated average continuous power limits is largely beside the point, since if it is correctly-matched with the system, room, and desired volume level, it will be running well below its maximum average continuous power limit, and thus won't be called upon to deliver more power than it can supply when actually playing music. In other words, whether an amps responds by supplying 200w at 8 ohms and 400w at 4 ohms - or 200w at both (or 400w at both) - at its maximum average continuous output limits when playing a static test signal, is largely immaterial when the question in the real world is: Can such amps supply a few watts of averaged continuous output power with occasional brief peaks an order or two of magnitude higher when playing music into a load that varies? For most systems and listeners, the answer in both cases, tube or SS, will probably be yes, thus rendering any amp's abilities with continuous power demands not found in the real world academic - and thereby explaining why we don't actually hear the maximum rated continuous power-into-load behavior of either type of amp as being a sonic liability as long as it's up to the job at hand.
Thanks all!

I am preparing for an audition soonest.