As a side note, all claims about audio gear performance coming from equipment manufacturers should be regarded with extreme caution. Common sense should always be exercised. There is no legal liability for them if they make outright lies to promote their wares. This is not peer-reviewed research, like information in technical publications.
Re. speaker distortion, a lot of good, credible information can be found on John Krutke' s web site, Zaph Audio. He measured distortion of many commercial drivers. John is no longer active with speakers, but the archive is still there.
How about just putting a frequency dip into the amplifier chain, adjustable for freq. and damping to match the speaker peak? Then the resulting flat speaker should be drive-able by either a V or I amplifier to get similar results.
Nowadays this can be done with digital signal processing, and this is definitely a viable approach.
In addition to 2H and 3H, which are relatively benign (if not causing IMD), speakers produce non-harmonic distortion. Probably the most notable kind of NHD is what can be seen in time domain as "waterfall plot". An electric pulse excites resonances at multiple frequencies, all with varying decays. The driving signal gone, a haze of extraneous noise hangs for some time. The more Mms, the longer and the higher in amplitude this cumulative decay. The drivers with least moving mass, such as electrostats, magnetostats, and ribbons have the best waterfall plots.
Moreover, an NFB amplifier exacerbates this kind of distortion. The hangover signal's back EMF is reflected to amplifier's output, and NFB is trying to correct it. Or, in other words, looking at speaker-amplifier as a system, speaker introduces mechanical delays that screw the feedback loop.
The bottom line here is that high Mms in conventional speakers is bad.
Moreover, an NFB amplifier exacerbates this kind of distortion. The hangover signal's back EMF is reflected to amplifier's output, and NFB is trying to correct it. Or, in other words, looking at speaker-amplifier as a system, speaker introduces mechanical delays that screw the feedback loop.
The bottom line here is that high Mms in conventional speakers is bad.
Last edited:
Looking at all this from a differential wave equations viewpoint, zero impedance boundaries just cause the wave to reflect back with inverted phase, while infinite impedance causes the wave to reflect back with the same phase. Neither one absorbs unwanted reflections. N Fdbk just holds the line at the V it is told to. No real problems with N Fdbk unless it is near an instability. What you really want is an impedance matched to the speaker that will absorb any incoming wave energy. ElectroVoice and others provided a critical damping adjustment for this. Not that hard to do.
In spite of all hard evidences against your rants (and your lies here) I presented in post #96, you wan't give up, eh?
Purifi speaker distortion at different SPL levels (up to 99dB/1m) was measured by 3 independent parties: Erin's Audio Corner, Voice Coil/AudioXpress and HIFiCompass:
1. Erin's Audio Corner tested and measured two models of Purifi 6.5".
The fist test ( https://www.erinsaudiocorner.com/loudspeakers/selah_audio_purezza/ ) shows extremely low H3=0.2% distortion at SPL from 79dB/1m to 99dB/1m. Here is, again (post #96), GIF animation/slide show at SPL from 79dB/1m to 99dB/1m:
The second test ( https://www.erinsaudiocorner.com/driveunits/purifi_ptt6.5m-08-nfa-01a/ ) again shows extremely low H3=0.2% distortion measured at SPL=96dB/1m:
2. Voice Coil/AudioXpress magazine test (https://audioxpress.com/article/tes...udio-6-5-midbass-transducer-from-purifi-audio) measured extremely low H3<0.2% distortion at SPL=94dB/1m:
3. HiFiCompass tested two Purifi 6.5" models.
The first test (https://hificompass.com/en/speakers/measurements/purifi/purifi-ptt65w08-01b-ptt65x08-nfa-01) measured extremely low H3<0.1% (!!!) distortion at SPL=97dB/1m (16V at 8 ohms / 1m):
The second test (https://hificompass.com/en/speakers/measurements/purifi/purifi-ptt65m08-nfa-01a) measured extremely low H3<0.1% (!!!) distortion at SPL=102dB/1m (16V at 4 ohms / 1m):
According to manufacturer's specs, all Purifi 6.5" models have 39mm diameter voice coils and appropriate 80W measured power handling with 100 hours signal according to IEC268-5 18.4, and 250W measured power handling according to IEC268-5 18.2. Nothing wrong with that.
Moral of the story: we don't need stinking current drive with H3=0.8% distortion from the non-feedback amplifier alone, when we can have H3=0.2% total (speaker+amp) distortion at SPL>95dB/1m with conventional solid state amp with negative feedback, feeding low-distortion, not-so-expensive loudspeakers.
Last edited:
one of the goals of our PTT drivers is in fact to avoid the need for current drive. it is in fact a challenge to design a current drive amp with low noise and one needs to handle the fs region with care since the electrical damping of the driver is lost with current drive. moreover, current drive only acts to reduce some of the distortion sources (eg BL modulation and suspension distortion are untouched by current drive)
A current drive amplifier shouldn't need to be no-feedback. A small current monitoring resistor (0.1R) in series with the load would make good source for global negative feedback, holding the current on the load constant.
Title of this thread is no-feedback amplifier. Besides, you will have hard time to persuade the OP to include even the smallest amount of negative feedback.
Purifi driver philosophy is very different from what I am proposing here. Purifi is optimized for long stroke. This is important if a single driver is used for bass and mids. Long stroke has advantage for bass, but no advantage for mids.
Long-stroke reproduction of low frequencies by physically small drivers operating close to resonance causes high levels of distortion, as can be seen on all speaker distortion graphs. No mafter how you slice it and dice it, you cannot defy the fundamental law: effective radiator size should be comparable to wavelength.
In defence of his approach, Mr. Putzeys states that harmonic distortion at bass frequencies is not important. This is a "grapes are green" kind of argument.
Gap flux modulation problem, which Purifi claims to solve, is a problem of low efficiency: the lower the efficiency the lower the ratio of VC:gap magnetic fluxes, and the more gap flux modulation by VC. Purifi is a low efficiency driver. High efficiency drivers need small VC magneric field on the background of strong gap field, for the same SPL output, so the modulation is practically unimportant. Nevertheless, for high efficiency FC drivers the flux modulation problem have been addressed 70 years ago - it was countermodulation coil wound on top of FC, in series with VC (Magnavox patent).
Higher noise in a no-feedback amplifier is not a significant problem. In well-designed no-feedback triode amplifies, noise at normal listening levels can only be heard with ear close to speaker. Pentodes may have higher noise than triodes, but in output pentodes this not a problem too.
I believe that NFB introduces more problems than it solves. I know Mr. Putzeys' opinion about NFB. Some say that he "nailed it". He didn't nail anything, it is just one of many arguments in the endless debate. I would like to avoid this thread degenerating into yet another one, so I will not respond to cues about NFB vs. no NFB.
Long-stroke reproduction of low frequencies by physically small drivers operating close to resonance causes high levels of distortion, as can be seen on all speaker distortion graphs. No mafter how you slice it and dice it, you cannot defy the fundamental law: effective radiator size should be comparable to wavelength.
In defence of his approach, Mr. Putzeys states that harmonic distortion at bass frequencies is not important. This is a "grapes are green" kind of argument.
Gap flux modulation problem, which Purifi claims to solve, is a problem of low efficiency: the lower the efficiency the lower the ratio of VC:gap magnetic fluxes, and the more gap flux modulation by VC. Purifi is a low efficiency driver. High efficiency drivers need small VC magneric field on the background of strong gap field, for the same SPL output, so the modulation is practically unimportant. Nevertheless, for high efficiency FC drivers the flux modulation problem have been addressed 70 years ago - it was countermodulation coil wound on top of FC, in series with VC (Magnavox patent).
Higher noise in a no-feedback amplifier is not a significant problem. In well-designed no-feedback triode amplifies, noise at normal listening levels can only be heard with ear close to speaker. Pentodes may have higher noise than triodes, but in output pentodes this not a problem too.
I believe that NFB introduces more problems than it solves. I know Mr. Putzeys' opinion about NFB. Some say that he "nailed it". He didn't nail anything, it is just one of many arguments in the endless debate. I would like to avoid this thread degenerating into yet another one, so I will not respond to cues about NFB vs. no NFB.
Last edited:
For a discussion example:
How many listeners listen to their 2 cubic foot cabinet speaker system with their ears right on the grill cloth, at the speaker driver(s)?
If you hear noise, then move one or two meters away, and relax. No noise.
If you still do hear noise at 1 or 2 meters . . . the output triodes, pentodes, or beam power tubes are not the problem (unless they are bad tubes). In that case, fix the amplifier.
Ask the same question for your own speakers . . . listen with your ear on the grill cloth, or listen 1 or 2 meters away.
And yes, if you are curious, I mostly listen to my amplifiers with the speaker only 2 feet away (0.6 meters away).
I do not hear any noise from the output tubes, even with no negative feedback . . . triodes, pentodes, or beam power tubes.
I must either have a bad ear, or quiet amplifiers.
How many listeners listen to their 2 cubic foot cabinet speaker system with their ears right on the grill cloth, at the speaker driver(s)?
If you hear noise, then move one or two meters away, and relax. No noise.
If you still do hear noise at 1 or 2 meters . . . the output triodes, pentodes, or beam power tubes are not the problem (unless they are bad tubes). In that case, fix the amplifier.
Ask the same question for your own speakers . . . listen with your ear on the grill cloth, or listen 1 or 2 meters away.
And yes, if you are curious, I mostly listen to my amplifiers with the speaker only 2 feet away (0.6 meters away).
I do not hear any noise from the output tubes, even with no negative feedback . . . triodes, pentodes, or beam power tubes.
I must either have a bad ear, or quiet amplifiers.
BTW, speakers under discussion here are kind of universal: equally suitable for for current drive or voltage drive. Low moving mass, high sensitivity, low VC inductance, good mechanical damping, and fast cumulative decay are all signs of quality in a speaker.
Actually, there is one situation when I listen very close to speaker: when I want to hear distortion. Even small amounts of harshness or grain can be heard this way. Distance masks it.
If the Purifi is so revolutionary, I don't see where the revolution is, after reading what is posted on their web site. Proprietary cone material? Others also have proprietary cone materials. Computer-optimized design? Others also did that, like, for example this Scan Speak 6" midbass . Very long stroke? The Scan Speak is also very long stroke. So, in what way the Purifi is better than the Scan Speak?
I tried these Scan Speaks in 2005 and moved on. What replaced them, a bookshelf ADS with metal magnet, had incomparably better midrange. Then moving to field coils was another quantum leap.
What else? Hysteresis? Never heard of this being a problem. All speakers whose iron is close to, or at magnetic saturation, are free from hysteresis. I bet it is same solution in the Purifi.
WOW!! That's nice to know.
Current drive is but one engineering approach and certainly not a panacea. Low Qm drivers help in the fs but region pose their own engineering issues.
Yep - and that could be a good design thread, maybe "Low distortion Pentode amp with high output impedance suitable for current drive experiments"?