During the transition from vacuum tube to solid state, the large full-range speakers were replaced by small midrange and squeaky tweeters. So there was nothing more that could produce a bass.
That's phase reversal, some early opamps showed it too, but it rapidly became de-rigeur for new opamps to be immune to phase reversal due to its unpleasant nature (compared even to hard clipping).
Not only were the speakers too damn small, but they quit taking advantage of bass boost that came for ”free” from the speaker/amp synergy. Sounded better to many than something that started to roll off at 200 Hz and just keep right on falling.
The reversal itself may not be that bad, but the reversed signal gets amplified more than the other one and quickly clips on the other side.
I think that would even be done on purpose. The output impedance of the tube amp was about the same as the speaker impedance and could even be greater. That dampened the resonance and what was left, as you say, was the bass boost.
Transistor amplifiers, on the other hand, have a very small output impedance due to the global negative feedback loop, and they can therefore destroy the loudspeaker in an instant in the event of resonance.
And now: My amplifier theoretically has an infinite output impedance (current source). Practically it is 100..300 ohms. The resonance is almost unsupported. This doesn't result in any additional bass boost, but the speaker gets just as loud at the resonant frequency as it does at 1kHz ("clean bass"). This means there is no lowering of the volume in the bass range as with standard solid-state amplifiers (voltage source).
So being more highly damped leads to more resonance - is that what you are claiming? I think the point is a low impedance amplifier can deliver more current when asked to, for instance when the speaker impedance is at a minimum. However speakers have a peak of high impedance at resonances, not low, so I don't get it.
Sold state amps mostly destroy speakers when under powered, leading to severe clipping which puts tones of energy from low frequencies into the tweeter, creates big DC offsets, and is generally nasty. And being over-powered isn't great either if abused. Perhaps an underpowered amplifier is more likely to clip badly if it has a high damping factor - don't use an underpowered amplifier!
In real life, demonstrations at exhibitions for home audio/hifi and in hifi studios usually give the impression that tube amplifiers have significantly better sound characteristics than transistor amplifiers. And this impression is right.
But the reason therefore isn't the fact, that tubes are provide better audio than power BjT's and MOSFET's.
The reason therefore is the fact, that most tube power amplifier output stages runs in Class-A or nearly in Class-A and most transistor resp. solid state amps runs in Class-AB with 20-50mA idle current through the output in case of BjT's and 100-150mA in case of MOSFET's.
If solid state amps runs in class-A or nearly class-A (e. g. Pass "Aleph", "First Watt" or "XA" series so as Musical Fidelity "A1"), sound is very close to those of tube amps - independent, whether BjT's or MOSFET's are in use.
Maybe Rod Elliott's approach is, what you need:
https://sound-au.com/project70.htm
https://sound-au.com/project36.htm
or the wide range of Nelson Pass' class A projects - maybe even Andrea Ciuffoli's Power Follower projects - both single ended and push pull (Cool Follower).
That sounds paradoxical... but if you have an amplifier with an operating voltage of +/- 40V..100V, what is to prevent it from mechanically destroying a loudspeaker at the resonant frequency? The speaker's resonant impedance becomes large, but not infinite. There will always be a nudge.
The main reason why amplifiers with high output impedance sound different is that the speaker's crossover malfunctions. Stereophile regularly measures the effects of an amplifier's output impedance on a simulated loudspeaker load. Tube amplifiers and no-global-feedback amplifiers do poorly on that test.
Ed
Ed
The resonant frequency may not have any more total woofer excursion than anywhere else. Excursion goes up below box tuning, given a constant voltage input. At that high impedance resonance, the amp (tube or transistor) simply runs out of voltage and clips at less power. You only get “more bass” for a while, and then it limits.
If the speaker can’t handle the excursion produced by 40, 100 or whatever volts, you can rip the suspension out. When you do that, it’s usually well below the resonant frequency.
Tubes sound different because of the output transformer design.
If the transformer is designed to 30 Hz, it would be bigger than the power transformer.
Guitar amps have 80 Hz (or 60 Hz for the bass amps)
Generally, If you hear low bass, that would be the higher harmonics of the root note dominating.
Less bass, the amp will sound very clean.
Regards.
If the transformer is designed to 30 Hz, it would be bigger than the power transformer.
Guitar amps have 80 Hz (or 60 Hz for the bass amps)
Generally, If you hear low bass, that would be the higher harmonics of the root note dominating.
Less bass, the amp will sound very clean.
Regards.
Actually, I didn't want to imitate the tube sound. This happened by itself. I wanted a robust amplifier and I managed to achieve this with two tricks: 1. class A and 2. no global negative feedback loop.
Then the topology from model ES200 (Audiolabor) could be of interest:
https://www.audiolabor.de/images/audiolabor/info/es200-info.pdf
https://old-fidelity-forum.de/thread-23346-post-932704.html
https://www.hifiengine.com/manual_library/audiolabor/es200.shtml
no global NFB loop is one of the features.
Another example is this amp from LC Audio - go to
http://www.octave-electronics.com/lcaudio/temil.shtml
https://www.manualslib.com/manual/1267874/Lc-Audio-Millennium.html
P.S.: global NFB is from my view only in cases a disadvantage as long as more than one gain stages in the NFB loop.
Unfortunately the case in most power amplifiers (usually two voltage gain stages in the NFB loop - 1) input differential-amp resp. LTP - 2) VAS with Cdom)
If there only one gain stage like a folded cascode in the NFB-loop, global NFB isn't a disadvantage maybe because Cdom isn't longer necessary - check out the schematic from the developer Günter Mania (AVM) as an example or the simplified schematic of the op amp AD817 under
https://www.analog.com/media/en/technical-documentation/data-sheets/ad817.pdf
a good choice for a front end for power amplifiers with medium output power due only +/-15VDC
https://www.audiolabor.de/images/audiolabor/info/es200-info.pdf
https://old-fidelity-forum.de/thread-23346-post-932704.html
https://www.hifiengine.com/manual_library/audiolabor/es200.shtml
no global NFB loop is one of the features.
Another example is this amp from LC Audio - go to
http://www.octave-electronics.com/lcaudio/temil.shtml
https://www.manualslib.com/manual/1267874/Lc-Audio-Millennium.html
P.S.: global NFB is from my view only in cases a disadvantage as long as more than one gain stages in the NFB loop.
Unfortunately the case in most power amplifiers (usually two voltage gain stages in the NFB loop - 1) input differential-amp resp. LTP - 2) VAS with Cdom)
If there only one gain stage like a folded cascode in the NFB-loop, global NFB isn't a disadvantage maybe because Cdom isn't longer necessary - check out the schematic from the developer Günter Mania (AVM) as an example or the simplified schematic of the op amp AD817 under
https://www.analog.com/media/en/technical-documentation/data-sheets/ad817.pdf
a good choice for a front end for power amplifiers with medium output power due only +/-15VDC
That is largely incorrect, even if GNF may play a significant role. Common collector has a relatively low output impedance without any GNF loop.
On the other hand, common base and emitter have comparatively high output impedances,
It's the other way round: Lower source impedance increases the damping.
The electrical equivalent to the speaker resonance is a parallel resonant circuit, where damping increases with lower parallel resistance.
An amp having a large output impedance acts a a current source.
With a current source the output voltage is proportional to the load impedance.
At the bass resonance the speaker impedance (often strongly) increases and so does the output voltage of a high impedance source (amp).
Consequently the power delivered to the speaker increases at the resonant frequency resulting in a bass boost.
I recommend to actually measure the amp's frequency response with a speaker load.
It will no longer be flat as with a constant impedance.
To my knowledge published frequency response of speakers or headphones is always measured on a low impedance amp.
Last edited:
The “bass boost“ is just a higher small signal GAIN at low frequency, as gm times the parallel combinion of the reflected speaker impedance and the anode resistance is higher in that narrow band. It does run out of voltage sooner, so the boost doesn’t happen at full power/volume. But at normal listening levels it is there.
Most SS amps would have an output impedance lower than any tube amp (triode, pentode, with or without feedback) even with with low or no global feedback. It’s not universal, though, because not all SS amps have common collector output stages, and they vary in how much current gain those output stages have. Ones that have a lot of current gain tend to retain their high damping factors when clipping. When you are clipping, there is no feedback and the output impedance is what intrinsically is. For hi-if use where clipping is avoided at all costs it’s debatable whether there is any value in keeping Zout low in clipping. But for DJ, concert, and rave use it is the single most important characteristic, aside from reliability. If you can drive an amp 6-10dB into clip and it still sounds clean it is a major advantage.
Tube amplifiers don‘t have have sky high damping factors to begin with, so when you clip there isn’t as much an abrupt change in characteristic so it tends to be less objectionable than a cheap SS amp running out of steam.
Most SS amps would have an output impedance lower than any tube amp (triode, pentode, with or without feedback) even with with low or no global feedback. It’s not universal, though, because not all SS amps have common collector output stages, and they vary in how much current gain those output stages have. Ones that have a lot of current gain tend to retain their high damping factors when clipping. When you are clipping, there is no feedback and the output impedance is what intrinsically is. For hi-if use where clipping is avoided at all costs it’s debatable whether there is any value in keeping Zout low in clipping. But for DJ, concert, and rave use it is the single most important characteristic, aside from reliability. If you can drive an amp 6-10dB into clip and it still sounds clean it is a major advantage.
Tube amplifiers don‘t have have sky high damping factors to begin with, so when you clip there isn’t as much an abrupt change in characteristic so it tends to be less objectionable than a cheap SS amp running out of steam.
And that means artificially increased SPL at the resonance and high frequencies.