LETS TALK ABOUT TUBE AMPS VS SS

[gartenman]

I remember meeting Saul Marantz (the man himself) after he sold his company to Superscope in Japan for an undisclosed sum, and having a very informative converasation with him. He was the VP of Marketing for Dalqust (remember the DQ-10?) at that time, not liking "retirement" one bit.

He told me that you wanted Tubes for the highs and solid state for the lows because the inherently lower source impedance of solid state had better control of the larger excursions/horspower requirements of the woofers.

thats exactly what i did and it really rules! wish i knew this years ago

[russ69]

DeanG says; "Most tube amps have a warmer or darker sonic signature.........."

That is not my experience. Some tube amps can be dark but they are really the very low end of the market. A good amp will sound very neutral, I have had to go to some extreme measures to make some of my amps sound just a tad warm. I think that blanket statement might be a bit overstated for a newbie.

Thanx, Russ

[Deang]

I've never owned tube gear that didn't sound warmer or darker than just about any solid state I've owned, so I think I'll stand by my statement.

[garyrc]

Don Richard, Hashimoto, DeanG and everyone else,

O.K., if one good approach (Deans?) is to not overdrive amplifiers -- to avoid clipping in the first place (or have it happen a very small percentage of the time and for very short durations) -- then I'm not at all clear on how to predict the multiple of some measured figure --- say RMS power -- at which actual clipping will occur. There is a garble in my head, the field, or both.

I am aware that this came up on the forum before, in 2006, but I don't think that Don Richard and Hashimoto were posting then, and I'm curious to know what they (and others) would make of the following:

• In the old days, before solid state became popular, it was often assumed that "peak power" was twice "rated power," and the better companies used the RMS figure as their "rated power." This rule of the thumb seemed to disappear in the 1960s. Was it well founded?
• Klipsch published an amplifier power rating table, called "Amplifier Power to Drive Klipsch Speakers," showing the SPLs various Klipsch speakers would produce in a 3,000 cubic foot room, with amps of various amplifier power ratings -- specified as continuous average at 8 Ohms." This table went out over Keele's name in 1977, then was revised very slightly and went out a few years later with no author listed. In BOTH versions it was stated that peaks 10 dB above the continuous average rating will pass without clipping. This was well into the Solid State era -- solid state became very popular in the mid '60s, although McIntosh hadn't thrown in the towel at that early time. Did they mean 10 dB above average for both Solid State and Tube amps? Have we ever heard this figure anywhere else?
• We often see "dynamic headroom at 8 Ohms" listed as about 3 dB.
• In various articles I have seen, Tube peak power is listed as 5 to 6.25 times that of Solid State peak power, for amps having the same power rating (RMS? Continouos Average @ 8 Ohms?), always with the admmission that there will be a great amount of distortion at this Tube peak power, but that it will be much more pleasent distortion than Solid State distortion.

How does all this information fit together?

[Don Richard]

In the old days, before solid state became popular, it was often assumed that "peak power" was twice "rated power," and the better companies used the RMS figure as their "rated power." This rule of the thumb seemed to disappear in the 1960s. Was it well founded?

Before the FTC power rating method, various manufacturers used many methods to rate amplifier power. The "better" companies used RMS power at a specific distortion and impedance. The others' ratings were random numbers generated by the marketing department, much like the horsepower ratings given by the auto companies of that era.

Klipsch published an amplifier power rating table, called "Amplifier Power to Drive Klipsch Speakers," showing the SPLs various Klipsch speakers would produce in a 3,000 cubic foot room, with amps of various amplifier power ratings -- specified as continuous average at 8 Ohms." This table went out over Keele's name in 1977, then was revised very slightly and went out a few years later with no author listed. In BOTH versions it was stated that peaks 10 dB above the continuous average rating will pass without clipping. This was well into the Solid State era -- solid state became very popular in the mid '60s, although McIntosh hadn't thrown in the towel at that early time. Did they mean 10 dB above average for both Solid State and Tube amps? Have we ever heard this figure anywhere else?

With music, generally the peaks are about 10 dB higher than the average, irrespective of amplification used. With the VU meters on the recording equipment of that time, peaks were around 13 dB higher than the displayed level.

We often see "dynamic headroom at 8 Ohms" listed as about 3 dB.

That would depend on the design of the amplifier, mostly on the reserve available from the power supply. NAD's first products were rated for low power, 20 - 25 watts or so IIRC, but could output 6 dB more (80 - 100 W) - ie. a 6 dB dynamic headroom rating.

In various articles I have seen, Tube peak power is listed as 5 to 6.25 times that of Solid State peak power, for amps having the same power rating (RMS? Continouos Average @ 8 Ohms?), always with the admmission that there will be a great amount of distortion at this Tube peak power, but that it will be much more pleasent distortion than Solid State distortion.

This is a gross generalization. It is not possible to accurately generalize in that manner. In the case of the NAD's high dynamic power output, it is not likely that any tube gear would output 5 or 6 times that. Running the equipment at high distortion levels would favor tubes over many SS designs because of the more gradual onset of distortion with tube designs, and the type of distortion generated (even order harmonics vs odd order harmonics).

How does all this information fit together?

Simply taking a power number without defining that number completely is meaningless. The FTC power rating method is quite stringent in terms of preconditioning requirements, and must be given with an impedance value and a distortion type percentage. When all amplifiers are rated by that standard, there is no more apples and oranges comparison, and the rated power correlates better with what is heard.

There is also the dynamic power rating, or dynamic headroom. Often this number isn't given in the specs and varies from 0.5 dB to 6 or more dB. In a lot of cases the amplifier cannot pass the preconditioning without overheating, so power output is downrated. The amp will output more for short periods of time and sound louder than it's FTC rating would suggest.

In summary, one must pay attention to all of the specs, graphs, and data on an amp in order to estimate it's audible power output. YMMV

[ClaudeJ1]

With music, generally the peaks are about 10 dB higher than the average, irrespective of amplification used. With the VU meters on the recording equipment of that time, peaks were around 13 dB higher than the displayed level.

PWK himself told to "make it 17 db" when I mentioned the 10 db figure for headroom. This was while he was doing a demo for me at his house back in 1985. He had switched from tubes to solid state. He was using 1/2 of a Crown D60 for his center Belle and a BGW 100 for his R and L channels. He referred to 100 W amps as "stoves" since he thought that "35 Watts is a helluvalot of power.

[Islander]

He referred to 100 W amps as "stoves" since he thought that "35 Watts is a helluvalot of power.

In The Dope From Hope, reference is made to people blowing drivers using what were high-powered amps at that time (the '60s or '70s) and that putting stronger components in the speakers was counter-productive, because they'd have less sensitivity, calling for more power, which caused the problem in the first place.

Has anything changed since then? Lots of people are using much more powerful amps today, Is Klipsch still getting complaints of drivers getting burned out because of them, or do modern amps put out cleaner signals that are less likely to damage speakers? Or are the speakers sturdier than they used to be?

[gartenman]

Most tube amps have a warmer or darker sonic signature, which is a good thing when you jack on the volume -- it helps to prevent the top end of the speaker from running away from the bottom and sounding too forward. The problem is that most good tube amps only give you 60 wpc or so, and unless you're running big Heritage or better -- they're undoubtedly clipping the signal. The argument is always made that the clipping is "soft", "euphonic", etc., and then the discussion goes off into an examination of the different clipping characteristics between ss and tubes. At that point, everyone wants to talk about which kind of amp sounds better when it's being run out of it operating limits and distorting. Does anyone besides me see the absurdity in this?

Finally a voice of reason...

[nola]

JL Sargent:

Right now I'm running active crossovers to tube and SS amps. I'm using
SET amps to the upper mid and tweeter range horns and SS to short bass
horns. I'm really into this type system right now. I have power on the
low stuff and I have the SET voicing I'm after on anything above 4 or
500hz.

tell me / us more about your active crossovers and system. thinking aboout doing this too.

NOLA

[JohnA]

I am seeking some very technically minded people to chat about differences in sound quality, dbs, and power ratings. I have not dived into this much yet but in my past stufy, I find that SS amps are uniquely rated (lie about it). Maybe there is special "sound watt" used but when comparing input fusing and wire size. When looking at tubes, their rating for power is much lower. 50-120wpc seems about right. I run a Rotel RB1080 on my RF-7s at 400wpc. I have to wonder the difference in dbs from a comparable tube amp.

I am not just wanting to compare tube on the Klipsch but all speakers. Sure seems like tube is way more expensive but can really give some nice sound. My experience with tube is only on guitar amps and I would not touch a SS guitar amp now!

Has someone compared tube with SS directly for dbs, accuracy, THD, etc?? Who the heck makes a decent priced tube amp?? They all seem about 3K and up for mono blocks or around 5K for a stereo amp.

Should someone else fail to say it, a watt is a watt, period. It is defined as voltage x amperage. A tube amp capable of 100 watts unclipped is not louder than a SS amp capable of delivering 100 watts unclipped. When over driven, tube amps produce more attractive sounding distortion, so many people routinely over drive them never knowing it. This is the basis of "tube watts" are bigger than SS watts.

Somewhere I have a research paper comparing simple tube circuits to simple transistor circuits. Each circuit was made as near identical as the devices would allow. The tube circuits had less distortion and noise, but less gain. The transistor circuits had more gain, noise and distortion, sometimes a lot more. The take home is that tube circuits can be simpler with less feedback for a given distortion level and thus may well sound better. The increased gain in a transistor circuit allows increased negative feedback for a given gain perhaps increasing accuracy when passing a waveform. Not a clear answer.

In my limited experience with tube amps, a great tube amp is indistinguishable from a great SS amp as long as neither is driven hard. A step down from great to good and the SS amp produces solid, tightly controlled bass and can be a tiny bit brittle or harsh at high frequencies compared to a tube amp.

Equal performance in SS amps is available for less than half of a tube amp and cost is still a big part of the equation for me.

[gartenman]

i know where you can get a just back for a tune up Eico for less than a thousand or a Baldwin which is going for a measley $420. both are really nice amps, lemme know and i can point you in the right direction

" I am seeking some very technically minded people to chat about differences in sound quality, dbs, and power ratings. I have not dived into this much yet but in my past stufy, I find that SS amps are uniquely rated (lie about it). Maybe there is special "sound watt" used but when comparing input fusing and wire size. When looking at tubes, their rating for power is much lower. 50-120wpc seems about right. I run a Rotel RB1080 on my RF-7s at 400wpc. I have to wonder the difference in dbs from a comparable tube amp. I am not just wanting to compare tube on the Klipsch but all speakers. Sure seems like tube is way more expensive but can really give some nice sound. My experience with tube is only on guitar amps and I would not touch a SS guitar amp now! Has someone compared tube with SS directly for dbs, accuracy, THD, etc?? Who the heck makes a decent priced tube amp?? They all seem about 3K and up for mono blocks or around 5K for a stereo amp."

[JL Sargent]

First off let me say that I do not have a golden ear. Just the opposite I'm sure. So take that into consideration when reading my thoughts about sound quality.

With my active setup I have been listening to a compilation of amps. They are an old 5 watts per ch. 6BQ5 SET amp, a 10 watts per ch. PP 6BQ5 amp, old NAD or CREST SS amps. I tend to cross into the squawkers at 400 or 500hz. Above this is tube amp(s) with SS on the bass. I'm swapping a RANE 22B active along with some even less expensive Behringer actives that I picked up here on the forum. I can hear no noise from these crossovers. All my cables are short and I have really have no noise to speak of after a minute or two of tube warmup. I'm using all sorts of configurations with the following speaker elements: The Altec 816a bass horns, Eminence Pro Kappa 15a in ported bass cabs, JBL 2441 drivers on Community horns, Klipsch RF7 horns, T35s, Dukane 5A540 drivers on some pretty big composite horns. I have played with all this ever which way you can think of. I listen and tweak, listen and tweak. I have learned that I can tweak the system for a certain style music or certain artist and its like they are visiting me in person. I can close my eyes and they are standing there. I really have been having a ball with it all.

I can change my crossovers points, invert drivers, play with delay, use sub filters, etc. all with the touch of a finger or turning of a knob. All of the crossovers I have been using are 24db/oct. which has been fine but I would like to try some steeper ones. I've gone 2 way and 3 way with it all. It's been fun and I have really created some great sounding stuff (to me) with all this for not alot of money invested. Hope that answers any questions about it. I'm a mad scientist at heart and these audio explorations are right on target for me.

The idea, the inspiration, and the how to do it I gleamed from Klipsch forum member contributions. Thanks guys for the help.

[Colin]

“a great tube amp is indistinguishable from a great SS amp as long as neither is driven hard.”

This is the famous Gordon Holt argument; that good solid-state and tube amplifiers are indistinguishable when they are NOT clipping! The key however is that they do clip.

The THD rating of an amplifier is at the low point of its THD curve. The amplifier is usually capable of twice that power, although at much higher levels of THD. At its maximum output, the THD of amplifiers screams skyward like a rocket taking off. With tubes as unique stores of energy, a tube amplifier puts out a lot more current than its rating implies. According to a Stereophile article, tube amplifiers may average 5 times the current of its simple RMS rating.

Nonetheless, I hope I remember this right: Let’s say you have a 25-watt amplifier idling along with a typically 85dB sensitive loudspeaker. This is average sensitivity for audiophile quality loudspeakers. That means that it takes about three watts to generate a moderately loud average music volume (85dB) at the listening spot in a typical room.

Every additional three dB louder requires something like a doubling of the amplifier power. To reach a 88dB peak, the amplifier needs to put out 6-watts. Now the typical CD has fast music peaks some 15-16 dB higher than the average music level. So a CD averaging 85dB in music level can have fast peaks at 100dB.

To reach those fast music peaks properly, without clipping, the amplifier with this loudspeaker needs a quick 38-watt burst of energy. You may NOT hear it, but the amplifier limited to just 25-watts is actually clipping, during those micro-second passages, at about 95db, or only 10dB higher than the average music level. As the amplifier reproduces those fast musical peaks, it is putting out maximum power at very high levels of distortion. The result sounds hard, harsh or metallic or edgy, particularly with solid-state amplifiers.

Yet, doubling the power of the amplifier doesn’t make an enormous difference in the volume that it can produce with 85dB loudspeakers. That only makes the sound some 10dB louder, but it does prevent clipping during fast musical passages, so the more powerful amplifier should sound better with these speakers.

(This says nothing about impedance control of the woofer. More powerful amplifiers double their wattage into lower impedance loads. This means the amplifier controls the woofer excursion, not the woofer controls the amplifier. More powerful amplifiers often sound better, not simply because they have more headroom, but because they often have better sounding bass.)

The type of distortion makes a very big difference. In a tube amplifier, some say the type of distortion resembles natural orders of musical instruments, so the heavy distortion when the amplifier is clipping during musical peaks is said to be good. Plus, the weaker horsepower of the tube amplifier rolls off quickly during those peaks, this is called soft clipping.

The heavy distortion of a SOLID-STATE amplifier when it is clipping does NOT roll-off gently, this is called harsh. When your ears are ringing because the music is so loud, it is NOT merely the volume that is making the cochlea cilia of your inner ear vibrate in pain, it is also the tremendous amount of distortion as the amplifier clips continuously.

90 dB/w/m

With super-sensitive 90dB loudspeakers, a 25-watt amplifier clips fast musical peaks at 25-watts, near the musical peak, about 100dB. At the average 85dB playing level, this amplifier and loudspeaker combination should NOT obviously sound like it is clipping fast musical passages.

With an ultra-sensitive 95dB loudspeaker, the little 25-watt amplifier has plenty of power to reproduce the 100dB musical peaks with room to spare. Call this headroom.

Big Ole Horns

With fantastically sensitive big ole horns, at 100dB/w/m or more, things begin to get a little ridiculous compared to the typical solid-state receiver and cone loudspeaker set-up. For one thing, the loudspeaker needs only a hundredth of a watt to reproduce a steady tone at conversation levels of 75dB, as measured two (2) meters away. Peaks on CDs are actually only about 12-15 dB above normal levels, requiring only a third of a watt. Therefore, a flea-powered 3-watt tube amplifier will reproduce the musical highs at 90dB with about a third of a watt, while the 25-watt amplifier could produce an ear-splitting 110dB before clipping!

That is why big ole horns, especially in the hyper-critical mid and high range, need quality watts, not lots of them.

If it wasn’t for the low impedances of the woofers, a good quality (meaning low THD) low-powered solid-state or tube amplifier is all the power anyone needs with fantastically sensitive big ole horns.

[garyrc]

O.K., some of this still is not consistently clear to me.

In one of the Dope From Hope issues, PWK opined that one needs 115 dB at one's ears to simulate a full symphony orchestra's peaks. I have heard other estimates as high as 125 dB, plausible for rim shots and the like.

In another Dope From Hope (January 1977, Keele, revised November 1980), Klipschorns, one of the most efficient speakers in the world (then rated at 104 dB / 8 Ohms / 2.83v / 1 m), in a 3,000 cu. ft, room, were said to emit the following SPLs.

One Klipschorn only, in case a powerful instrument was positioned on the far right or left:

1W would produce 97 dB

out in the room, presumably

in the listening area

(as opposed to @ 1 M,

where we would get about 104 dB)

63 W would produce 115 dB

and ....

The W for 125 dB is off the chart, which ends at 375 W (@122.7 dB).

I presume that even if "dynamic power" or "peak power" that is available for only an instant is what is needed to produce these extreme peaks, we would still need the power in watts specified in the chat, but only for that instant. Wouldn't that mean that for an ittybitty mini-micro-nanosecond we would need > 375 W for one of those 125 dB peaks, and my 150 W amp (really more like 120 W RMS @ , .707 x 171 W clipping point) would clip? Or is it implied that the peak is so brief that the peak would pass without clipping? And that the peak is so brief that the speakers won't blow either? Obviously, this brings me back to my original question of several posts ago. Since I can easily produce the 63 W needed to produce 115 dB, do the Klipsch charts really mean (as they seem to state) that my amp will pass very, very short peaks @ 10 dB higher (125 dB!!) without clipping? Is it that the strain on the power supply, etc., etc. is too brief to cause the amp to run out of steam and the waveform flatten?

[ClaudeJ1]

If it wasn’t for the low impedances of the woofers, a good quality (meaning low THD) low-powered solid-state or tube amplifier is all the power anyone needs with fantastically sensitive big ole horns.

Which is why Roy Delgado said the first place you need a "big ole horn" is in the bass where there is the most cone movement, hence, most distortion, to be tamed...........yet, people do just the opposite, by doing horns in the treble first becasue it's smaller, easier, cheaper, etc.

I know a retired Audio engineer who uses DCM time Windows on top of Khorn bass bins for this same reason.

PS: I'm now working on horns SUBWOOFERS, so I will be selling my excellent DR subs soon.

[ClaudeJ1]

I have heard other estimates as high as 125 dB, plausible for rim shots and the like.

Maybe from the drummer's seat, but not from any house seat, even in the first row. So, the term "I have heard........." is hardly a reason to overpower a system by 10 db (10 times) necessary.

By the original estimation of 115 db peak, which is more realistic to even Stravinsky's "Rite of Spring" going full tilt in the best orchestra hall, a 37.5 watts of total amplifier power should be all one needs in any reasonable sized home. That's less than 20 Watts per channel (R+L). I use milliwatts in my huge treble/midbass horns/super tweeters. I rarely hit 10 Watt peaks in my bi-amped Khorn bass units. Subs are another story and I'm working on horn subs to reduce the wattage. I want realistic reproduction to run on solar panels!!

Long live efficiency and low distortion.