OK, so Neo-8 will have no benefit from the current drive whatsoever! That is because of flat, purely resistive impedance nature of the driver. Which drew us to another question (which I already asked): why you don't want to use Zobel network compensation to get flat impedance of the 12" mid driver (in the 100Hz-1kHz range)?
You didn't answer - which is that miraculous 12" mid driver which "sounds much better than 6.5" Scan Speak midbass". Have you actually compared that 12" driver with a 6.5" ScanSpeak? How you compared them, did you actually build optimized crossovers for each of them, to be able to compare apples to apples? Which model of 6.5" ScanSpeak? You know, not all 6.5" ScanSpeak drivers are very good.
But you do subscribe to the belief that low speaker THD equals good sound, don't you? After all, lower speaker distortion is the main reason for introducing current drive amp, isn't it? Then, how on earth you don't care if the 12" driver (you are intending to use) has higher distortion (with the current drive, of course) than the Purifi or ScanSpeak driver???
Last edited:
I think I have detected multiple 'Hot Voice Coils' on this thread.
We can do better than that.
We can do better than that.
Current drive: the main benefit is that it reduces distortion due to modulation of the voice coil inductance. This modulation is from a combination of position dependency and magnetic hysteresis (inductance being a function of the past signal and exhibiting instantaneous jumps in inductance).
The flatness of the impedance curve is not an indicator of inductance modulation hence cannot be used to judge suitability for current drive. However, absence of the usual fs impedance peak means that the driver Q is unaffected by current drive which is an advantage (no boomineas of the bass when going from voltage to current drive). However, that is rather rare. maybe a heavily stuffed TL comes close
The flatness of the impedance curve is not an indicator of inductance modulation hence cannot be used to judge suitability for current drive. However, absence of the usual fs impedance peak means that the driver Q is unaffected by current drive which is an advantage (no boomineas of the bass when going from voltage to current drive). However, that is rather rare. maybe a heavily stuffed TL comes close
Driver impedance as series resistor and inductor is overly simplistic. It is static. Dynamic impedance has another component associated with voice coil acceleration. Instantaneous acceleration component could be positive or negative; averaged over one period it is zero. With voltage drive, but not with current drive, this dynamic impedance component will cause intermodulation distortion. Moreover, the dynamic component should increase with moving mass. Thus a driver, especially one with high moving mass, is a poor (non-linear) V-I converter. This is in Mr. Merilainen's book.
It may be one of the reasons why a driver with relatively low level of harmonic distortion, such as the mentioned 6" Scan Speak ( a highly reputed driver), may sound below the expectations.
It may be one of the reasons why a driver with relatively low level of harmonic distortion, such as the mentioned 6" Scan Speak ( a highly reputed driver), may sound below the expectations.
Disagree. With small radiator size, the radiator must have long stroke. Driver's suspension acts as a spring whose force increases with deflection. That means at different distances of cone travel, deflection force is different. One can design a driver with as linear deflection/force characteristic a possible, but it will do nothing do reduce the third order distortion due to spring characteristic. If a spring works unequally back and forth, then there will be also second order distortion.
The only way to reduce this spring-related distortion is to lower cone's excursion. But that would require increasing driver size.
Also disagree. Magnet, permanent or FC, can be matched to iron yoke in such a way that the iron (not only magnet) is fully saturated. If coercive force is well beyond saturation, then there will be no hysteresis: the iron of the pole piece will remain saturated under all working conditions.
Curious minds want to know how strong suppression of flux modulation is achieved in the Purifi drivers. Without knowing the mechanism, it is impossible to analyze the trade-offs of the proposed approach and judge on its merit. I understand it could be a trade secret, but in the absence of detail it sounds like something absolutely good, which is a telltale sign of a sales pitch.
Yes, you can regard it like this, but only below magnetic saturation of the pole piece. At saturation (which is the preferred way of running a field coil), the ferromagnetic coupling between FC and VC is lost.
You contradict yourself here.
Your disagreement can not erase this simple fact: despite small radiator size and long stroke, 6" drivers from Purifi, ScanSpeak, Seas (W18NX003), Satori (MW16P-8 and MW19TX-4) and Wavecor (WF182BD03) have third order distortion less than 0.2% at SPL=96dB/1m!
Can you, please, post measurements of third order distortion of a 12" mid driver (with current drive, of course) capable of lower (or equal) third order distortion at SPL=96dB/1m ? Or you will choose to not answer my questions because they shatter in pieces your firm believing in the fairy tale that such current driven pink unicorn exist?
That's not necessarily convincing. John Krutke (for me the ultimate authority on speaker testing) says in introduction to his tests that comparisons can be only made within a group of drivers measured on the same rig and under the same conditions. No comparisons should be attempted between tests performed by different investigators.
Hifi compass is funny. They post data for different drivers that were obtained using different tests. For some drivers it is distortion across the spectrum, but for others it is harmonic spectra at different frequencies. For some drivers there is no distortion data. Seems like they don't have the uniform approach to measurements.
I must admit though that I could have been wrong about denying the possibility of HD in fractions of percentage points. Some drivers in Mr. Krutke's measurements (e.g. Satori) indeed have such low distortion in some parts of their spectra.
Critical evaluation of the validity of measurements require knowledge of detail of equipment and procedures used, which Mr. Krutke provided. Others just mention (if they do) the software used.
Mr. Krutke's departure created huge vacuum in independent speaker testing. Whatever labs are there now are not up to his standards. Adequate comparisons between Purifi and similar drivers from other manufacturers is yet to be done.
4x2 cm voice coil cannot continuously dissipate 250 W of heat. If it doesn't open or melt in a few minutes under such power, it will get glowing red hot.
These IEC tests are gentlemens' agreement between manufacturers about how they can dupe their customers without slitting each other's throats.
I repaired a number of SS amplifiers with blown output transistors. The problem was that users believed in manufacturer's continuous output specifications. An amplifier rated 50 W continuous had hardly enough heat sinking for 5 W. Sadly, such hugely exaggerated power ratings of amplifiers and speakers are a norm today, and Mr. Putzeys is one of the many perpetrators.
OK, can you, please, post here measurements of 12" mid driver (with current drive, of course) with lower intermodulation distortion than 6" Purifi? Pretty please?
Moreover, please explain how it is possible 12" driver to have less moving mass than 6" Purifi driver? Any real-world example?
Such measurements would be impossible to do across the spectra of frequencies. Measurements with two fixed frequencies are inadequate to characterize a broadband transducer.
Later on I will discuss the specific 12" driver for this project and post some of its parameters. Please stay tuned.
Oh really? What if different investigators have the same measuring rig, and perform measurements under the same conditions (temperature, humidity, pressure)?
Utter nonsense. You cherry-picked some old measurements. Also, no distortion data for some drivers as a proof of lack of uniformity? Do you believe in your own words?
No, Mr. Krutke didn't say that!
Quote from Mr. Krutke himself: "Do Not compare these harmonic distortion plots to others anywhere else, including on this site. The test method was different for this particular set and the level was a bit lower. Make comparisons within this test group only. If you want to compare results with the 5.5" grouping, the 6.5" group will need approximately 4dB added to level and harmonics."
Are you deceiving yourself on purpose?
Mr. Krutke was using different measuring levels for his own test (side note - do you remember what you wrote minute ago about HiFiCompass tests?)
Utter nonsense, the most (XXXcensoredXXX) thing I have ever heard or read in my whole life!
For your information, both Erin's Audio Corner and Audio Science Review (which measured Purifi drivers) are using Klippel Nearfield Scanner - the most sophisticated speaker measuring equipment on the Earth!
All distortion test in Voice Coil/Audioxpress are performed by industry-standard Listen, Inc. SoundCheck AudioConnect analyzer.
Nobody is saying it will continuously dissipate 250 W of heat. You made it up, which is really sad. Specification of 250W power handling is according to IEC268-5 18.2 with appropriate signal which you may don't like, but it is accepted standard.
You choose not to see the 80W power handling specification for the same driver (measured with 100 hours signal according to IEC268-5 18.4). That is very, very sad.
Your disgusting of Mr. Putzeys and Purifi drivers definitely is worrisome.
Last edited:
???
You can compare two different drivers with any intermodulation measurement method, including two fixed frequencies! If one of the drivers have lower intermodulaton distortion measured with two fixed frequencies, it will have lower distortion with any other measurements method! Do you agree?
Can you, please, describe what is the adequate intermodulation test for drivers?
I am holding my breath.
Impossible? Is the multitone measurements of intermodulation distortion adequate enough for you? For example, simultaneous 32 signals across the 20Hz-20kHz:
Unfortunately, this is indeed inadequate. The arbitrarily chosen test frequencies may fall on speaker's local resonances (peaks or dips), which will make measurements unrepresentative. Measurments should be continuous over infinite number of frequencies. Hence my opinion that it would be impossible to do.
With my limited DIY goals, I have no need for such measurements. Following general principles of what a good current drive speaker should be, and how the lowest distortion in a no-NFB pentode amplifier can be achieved, I will complete my project and decide whether I like the result or not. If, after doing all I can do to optimize it, I don't like the result, I will conclude that current drive is not worth it. But not before it. I feel excited exploring the unknown.
OK, if 32 signals for multitone distortion measurement are not enough, what about 1024 signals? Or 4096 signals? (Both are possible with modern measurements equipment/software.) Are you aware that graphs in distortion measurements we are looking on our computers now have smaller number of pixels horizontally? Or that your computer monitor has smaller horizontal resolution than 4096 pixels? Or that speaker local resonances are way wider than that resolution?
So, multitone testing is more than adequate, and totally possible!
Last edited:
Psychoacoustics of low frequency harmonic distortion is an interesting subject. Do we need to expend effort in reducing subwoofer HD, or we are OK where we are?
Indeed, if a simple tune like "Twinkle Little Star" is transposed for an octave between 30 and 60 Hz, a few people, even those with good musical ear, will be able to pick up the tune. For most listeners, large amounts of 2H distortion in bass-heavy music is not noticeable.
Yet differences between pure harmonic tones, for example at 90, 60, and 30 Hz are very apparent. I hear 90 as a music tone and 60 as bass tone, but 30 gives a qualitatively different sensation. I actually don't hear it like other sounds with my ears, I feel it with my whole body. Does it matter for music? I don't know. At the beginning of my favorite recording of Mahler's 5th symphony, there are few strokes of big orchestral drum. Not powerful strokes, more like gentle touches. If the lowest frequencies are not reproduced well, these strokes dont sound right. I believe the effrct was meant like this by the composer.
So, should we, or shouldn't, compromise on the fidelity of bass reproduction?
Indeed, if a simple tune like "Twinkle Little Star" is transposed for an octave between 30 and 60 Hz, a few people, even those with good musical ear, will be able to pick up the tune. For most listeners, large amounts of 2H distortion in bass-heavy music is not noticeable.
Yet differences between pure harmonic tones, for example at 90, 60, and 30 Hz are very apparent. I hear 90 as a music tone and 60 as bass tone, but 30 gives a qualitatively different sensation. I actually don't hear it like other sounds with my ears, I feel it with my whole body. Does it matter for music? I don't know. At the beginning of my favorite recording of Mahler's 5th symphony, there are few strokes of big orchestral drum. Not powerful strokes, more like gentle touches. If the lowest frequencies are not reproduced well, these strokes dont sound right. I believe the effrct was meant like this by the composer.
So, should we, or shouldn't, compromise on the fidelity of bass reproduction?
I agree that such high resolution measurement is possible, but is it necessary?
Listening is and will remain the ultimate test. Complex sounds like orchestral strings, chorus, or full organ are very revealing of non-harmonic distortion, which makes the sound homogenized, as opposed to being able to hear the distinct multitudes. If an amplifier/speaker system passes such test, is there anything else to be desired?
Last edited:
Effort of manufacturers for reducing distortion in drivers, including subwoofers is evident.
So, we should't compromise on the fidelity of bass reproduction... like compromising perfectly good subwoofer drivers with current-drive amplifiers.