Tube Amps Watts vs SS amp

The tubes vs SS discussion should focus more on the type of music one prefers rather than the technical merits of each amp. I have multiple tube and SS amps that I interchange frequently. If you primarily listen to Jazz, voice, acoustic music of any kind there is nothing better than a quality tube amp.

The only time I would caution someone about moving toward tubes is if they primarily / exclusively listen to rock & roll at higher volumes. i don’t feel the benefits of tubes are realized in that case.

This is one of the more common myths in audio- that somehow, tubes or transistors (or loudspeakers) have some sort of taste or are better somehow at a certain genre of music. This idea is entirely false- what makes an amplifier or speaker good at one genre is precisely what will make it good at an entirely different genre. IOW an amplifier that is 'best' at rock will also be 'best' at everything else (whatever 'best' means...).

My understanding is that the concept of “bandwidth” is very important. Perhaps more important for tube amps than solid state. But I have a First Watt SIT-3 which is a low watt solid state amp. It apparently has some bandwidth to it because it drives my Magico A3s well. They have a sensitivity of 88dB, an impedance of 4 ohms and recommended minimum power of 50 watts. The SIT-3 drives them just as well as my Bryston 4B3, which has a lot more power. But what I would like to know from the group is: how is bandwidth measured? How do you calculate bandwidth?

Bandwidth is important to keep phase shift at a minimum. To this end, phase shift is minimized if bandwidth is 10x the maximum frequency to be amplified (20Hz, so 200KHz required) and also 1/10th the lowest frequency to be amplified (20Hz, so 2Hz response required). Bandwidth is measured by either a sine wave or square wave; with a sine wave the signal is applied to the circuit and the output observed to be within usually + or - 1/2dB to be considered ’flat’; with a squarewave rounding of the edges can be seen to show a rolloff at high freqencies and tilt on the top of the squarewave shows a rolloff at low frequencies. This is fairly easy for transistor amps, and there are tube amps that meet the ’2Hz-200KHz’ requirement too, but to my knowledge they are all OTLs (Output TransformerLess).


Keeping phase shift linear has to benefits: more accurate presentation of the soundstage and more accurate presentation of tonality. As an example of the latter, if there is a rolloff at 10Hz, phase shift will cause a lack of impact up to about 100Hz despite the amp measuring flat to 20Hz on the bench. This is why if there is a problem at 50KHz it can often be heard as well, since phase shift artifacts will exist down to 5KHz. Again, this will be interpreted by the ear as a tonality.


So one takeaway: three things affect tonality: actual frequency response (which is different from bandwidth), distortion and phase shift. 

In this particular case the 87 db spec is stated to be based on an input of 1 watt, rather than on an input of 2.83 volts. Also, as I mentioned in my previous post the brochure for the speaker states that an input of 8 watts results in an SPL of 96 db at 1 meter, which is consistent with 87 db/1 watt/1 meter. So if those numbers are accurate the efficiency is 87 db, rather than 84 db.

@almarg

Apparently it matters where you look. The specs I found stated 4 ohms and 87 dB Sensitivity; since the sensitivity spec is a voltage spec and not power, this translates to an efficiency as I stated. I’ve seen speaker manufacturers confuse efficiency and sensitivity before; is this one of those cases?

Under the power paradigm you want speakers of steady impedance and relatively high efficiency. How does Soundlabs achieve this with an electrostatic loudspeaker. My Acoustats drop perilously low at high frequencies.

@mijostyn Actually a steady impedance is not required of either paradigm. What is required is best put as the designer’s intention- and the behavior of the technology. In the case of ESLs, the impedance curve is based on a capacitance rather than a driver in a box (with its consequential resonance). On this account, ESLs have an impedance curve that essentially decreases as frequency increases. The Sound Lab for example is about 32 ohms in the bass but only 1.5-3.0 ohms (depending on the position of the Brilliance control) at 20KHz. The impedance curve of a box speaker maps out its efficiency with respect to frequency; with an ESL the efficiency remains constant despite its impedance.


One sign of a speaker that is a Power Paradigm device is that it will have controls in the crossover- like you see in the Sound Lab, but also like you see in vintage speakers like Altec, JBL, KLH and so on. These controls are not there to adjust the speaker to the room (although they are often used that way) they are there to adjust the speaker to the voltage response of the amplifier, which is an unknown. In the Voltage Paradigm the voltage response is a given; the idea was pioneered by MacIntosh and ElectroVoice in the late 1950s to assist with plug and play. The problem is to do so audible distortion in the form of brightness is usually introduced by the operation of the feedback loop in the amplifier, and one has to come to terms with the simple fact that no loudspeaker is actually flat in frequency response- so getting ’flat response’ is a bit of a red herring.


Add to that the fact that the ear/brain system interprets distortion as tonality (which is why we perceive the distortion added by feedback as brightness), often favoring it over actual frequency response, and you have a situation where you can often achieve greater neutrality without feedback. Crazy world, but we really can’t change how our ears perceive sound except by damaging them. So it seems pragmatic to come up with engineering solutions that take the human ear/brain hearing perceptual rules into account rather than ignoring them!

@bestbaker  You might want to read this thread- it deals with exactly the same issues you are dealing with.
https://forum.audiogon.com/discussions/recommendation-for-higher-power-tube-ampIn a nutshell since your speakers are 4 ohms the efficiency is 3dB less than the stated 87dB.


Efficiency is a much more useful spec when dealing with tube amplifiers as tube amps don't double power as impedance is halved.


The bottom line is if you want satisfying sound pressure *and* you want tubes, you are far better off getting a speaker that is not so inefficient!

Sanders biases his arguments towards his own amps which are admittedly great values in high powered amps. Yes, high powered tube amps waste a lot more energy as heat than an AB SS amp. He is also expressing his opinion with a bias toward driving his electrostatic speakers. He dismisses OTL amps off hand. He is right that you need to use a lot of tubes to reduce the impedance of the output section but once you do so they are every bit as effective at driving electrostats as SS amps.

@mijostyn
ESLs don't obey the same rules of loudspeaker operation as conventional box speakers; no doubt because a box isn't involved for starters :)  As a result, to make them more compatible with solid state (as otherwise ESLs are *far* more compatible with tubes), Sanders made his speakers fairly low impedance, just like Martin Logan. This results in impedances that are south of 1 ohm at the upper range of the speaker. For more on how this works seehttp://www.atma-sphere.com/Resources/Paradigms_in_Amplifier_Design.php