Hello all,
I've got some general questions about FR drivers rattling around in my head that I'm hoping some generous souls will put to rest for me:
First, do I understand correctly that the two enemies of treble performance in FR drivers are voice coil inductance and mass rolloff? Are these both purely first-order phenomena? If so, what is the "Q" and how do you calculate the corner frequecy for mass rolloff? Is it a function of BL?
The reason I'm wondering is that I'm trying to get a handle on how I might create a good 15" FR driver of extremely wide bandwidth. Of course, in drivers of this size, VC inductance and mass rolloff quickly become factors as frequency rises.
I was also wondering if it might be profitable to design a FR driver that centers the inductive and mass rolloff points at the same frequency to create a smooth 2nd-order rolloff and put a 2nd order LC/R shelving filter in series to equalize the response. I know that's gonna make the purists cringe and subtract efficiency but... Wouldn't it extend response? Alternately, you could create a dual voice coil wired in parallel with a high-pass filter in series with one of the coils.
Which brings me to my final question. When you have equal and opposite filters functioning within a single transducer, what's the net result? Does everything average out as long as the filters are 1st order (as Thiel claims on their website) or does other bad stuff happen?
Thanks for setting me straight.
Bill
(PS. I posted this on the full-range driver forum but haven't got any takers yet, ergo the double post)
I've got some general questions about FR drivers rattling around in my head that I'm hoping some generous souls will put to rest for me:
First, do I understand correctly that the two enemies of treble performance in FR drivers are voice coil inductance and mass rolloff? Are these both purely first-order phenomena? If so, what is the "Q" and how do you calculate the corner frequecy for mass rolloff? Is it a function of BL?
The reason I'm wondering is that I'm trying to get a handle on how I might create a good 15" FR driver of extremely wide bandwidth. Of course, in drivers of this size, VC inductance and mass rolloff quickly become factors as frequency rises.
I was also wondering if it might be profitable to design a FR driver that centers the inductive and mass rolloff points at the same frequency to create a smooth 2nd-order rolloff and put a 2nd order LC/R shelving filter in series to equalize the response. I know that's gonna make the purists cringe and subtract efficiency but... Wouldn't it extend response? Alternately, you could create a dual voice coil wired in parallel with a high-pass filter in series with one of the coils.
Which brings me to my final question. When you have equal and opposite filters functioning within a single transducer, what's the net result? Does everything average out as long as the filters are 1st order (as Thiel claims on their website) or does other bad stuff happen?
Thanks for setting me straight.
Bill
(PS. I posted this on the full-range driver forum but haven't got any takers yet, ergo the double post)
There are a lot of reasons I oppose full range drivers-intermodulation distortion prevents even one with a perfectly performing cone from really doing well-but a 15 inch full range driver is really out of the question.
This graph is for a Vifa 12 inch-a 15 inch would be worse. Look at the dispersion curves. Even at 1700 Hz the 60 degrees axis line is way, way over 10 dB down from the on-axis curve.
The only reason this speaker-which is not designed for full range use-is not worse, dispersion-wise, at high frequencies is that the cone has so completely broken up at the high ranges that the high frequencies are being radiated by small sections of the cone acting independently of the larger cone at a whole. A speaker designed for high frequency use will have less cone breakup, therefore worse dispersion.
A large diaphragm, no matter how perfectly manufactured, cannot give good dispersion for the high frequencies. The high frequencies will "beam" forward, and the high frequency sound will be harsh.
This graph is for a Vifa 12 inch-a 15 inch would be worse. Look at the dispersion curves. Even at 1700 Hz the 60 degrees axis line is way, way over 10 dB down from the on-axis curve.
The only reason this speaker-which is not designed for full range use-is not worse, dispersion-wise, at high frequencies is that the cone has so completely broken up at the high ranges that the high frequencies are being radiated by small sections of the cone acting independently of the larger cone at a whole. A speaker designed for high frequency use will have less cone breakup, therefore worse dispersion.
A large diaphragm, no matter how perfectly manufactured, cannot give good dispersion for the high frequencies. The high frequencies will "beam" forward, and the high frequency sound will be harsh.
Attachments
To be honest, Bill, I would not be the best person to answer the rest of your question.
However, if I understand your idea correctly, you want to lower the output of your full range speaker in the lower frequency ranges in order to match the output of the speaker in the high frequency range-in the region where it rolls off. If so, I don't see any reason why that idea would not work.
In an article in the Journal of the Audio Engineering Society, Paul Klipsch went into detail on "Doppler distortion"-the distortion products which occur when a speaker tries to create higher notes when carrying lower notes at the same time. To briefly summarize, the low notes require comparatively large excursions that distort the higher notes greatly-giving out "fuzzy" sound. Frequencies that are not even on the original program material are actually created. And these distortion-generated frequencies-which are not on the same musical scale as the music-can approach the amplitude of the higher notes themselves.
The only reason that electrostatics and Magneplanar type speakers can accomplish full range sound so well is because their surface area is so great, the excursions required to produce low notes are correspondingly tiny. So carrying high notes is a snap for them.
For a regular 6" or 8" speaker, the excursions will be quite a bit higher for the low notes, and the higher notes will be smeared.
Amplitude that the speaker is played at has everything to do with this, so a 6" or 8" speaker played relatively softly might be OK. So, possibly, might an array of full range drivers, to provide more surface area.
I am not trying to shoot down your enthusiasm for a new project. I am just letting you know the principles involved. And in fairness I should mention that knowledgeable audio fans have built single-driver full range units that they swear sound very good and natural.
However, if I understand your idea correctly, you want to lower the output of your full range speaker in the lower frequency ranges in order to match the output of the speaker in the high frequency range-in the region where it rolls off. If so, I don't see any reason why that idea would not work.
In an article in the Journal of the Audio Engineering Society, Paul Klipsch went into detail on "Doppler distortion"-the distortion products which occur when a speaker tries to create higher notes when carrying lower notes at the same time. To briefly summarize, the low notes require comparatively large excursions that distort the higher notes greatly-giving out "fuzzy" sound. Frequencies that are not even on the original program material are actually created. And these distortion-generated frequencies-which are not on the same musical scale as the music-can approach the amplitude of the higher notes themselves.
The only reason that electrostatics and Magneplanar type speakers can accomplish full range sound so well is because their surface area is so great, the excursions required to produce low notes are correspondingly tiny. So carrying high notes is a snap for them.
For a regular 6" or 8" speaker, the excursions will be quite a bit higher for the low notes, and the higher notes will be smeared.
Amplitude that the speaker is played at has everything to do with this, so a 6" or 8" speaker played relatively softly might be OK. So, possibly, might an array of full range drivers, to provide more surface area.
I am not trying to shoot down your enthusiasm for a new project. I am just letting you know the principles involved. And in fairness I should mention that knowledgeable audio fans have built single-driver full range units that they swear sound very good and natural.
Full range driver
Try the Full Range Driver site and forum http://melhuish.org.
One of the best sources for full range driver questions I am aware of.
Cheers,
Try the Full Range Driver site and forum http://melhuish.org.
One of the best sources for full range driver questions I am aware of.
Cheers,
Bill F. said:
A 15" FR seems to me to be a bit on the big side. Large co-axial 2-ways can work -- the Tannoy, and the PHY-HP 12" are two examples. Even then i am amazed the Tannoy 15" cone reachs the 1 kHz XO to the horn (it does sacrifice ultimate bass to do so), and the PHY, as seductive as it sounds is missing something in the range below the entry of the tweeter.
I have come to the feeling that a FR should be less than 8". Smaller gives you better HF response but some still benefit from a super-tweeter. And every good full-range needs a subwoofer.
A FR can be very good but i feel you ultimately loss something at either end -- but boy are the seductive. They make an ideal wide-band mid -- actively XOed to a woofer and a super-tweeter rolled in at the top -- ideally with only minimal active HP filtering on the FR to keep the big excursions out.
dave
Kelticwizard,
I have not read Paul Klipsh's article (can you point me to a link?) However, a couple months ago, I posted my musings/questions about doppler phenomena (I called the thread "doppler encode/decode" if you want to find it).
It has occurred to me that since microphones operate full-range there is must be a certain amount of doppler frequency shift inherent in all recordings. Therefore, even absolutely ideal bandwidth-limited drivers cannot faithfull reproduce the original sound. Rather, they faithfully reproduce the *recording* without fully decoding the doppler information.
Still speaking in ideal terms, only a full-range transducer (operating at a certian volume setting where the ratio of its max amplitude to surface area equals that of the microphone that recorded the piece) will faithfully reproduce the original music. It does this because it applies proportional inverse doppler distortion to the signal, thereby canceling the distortion.
At least this is my understanding. Please correct me if I'm wrong.
In the final analysis, however, I do recognize the weaknesses in current approximations of a full-range driver.
Bill
I have not read Paul Klipsh's article (can you point me to a link?) However, a couple months ago, I posted my musings/questions about doppler phenomena (I called the thread "doppler encode/decode" if you want to find it).
It has occurred to me that since microphones operate full-range there is must be a certain amount of doppler frequency shift inherent in all recordings. Therefore, even absolutely ideal bandwidth-limited drivers cannot faithfull reproduce the original sound. Rather, they faithfully reproduce the *recording* without fully decoding the doppler information.
Still speaking in ideal terms, only a full-range transducer (operating at a certian volume setting where the ratio of its max amplitude to surface area equals that of the microphone that recorded the piece) will faithfully reproduce the original music. It does this because it applies proportional inverse doppler distortion to the signal, thereby canceling the distortion.
At least this is my understanding. Please correct me if I'm wrong.
In the final analysis, however, I do recognize the weaknesses in current approximations of a full-range driver.
Bill
Hello Bill, I had pair of Yamaha NS-20 8" 2way. Threw away the original crossover - connected the 8" woofer direct and fitted RC network across voice coil to null rising impedence characteristic, did the same to the tweeter and fitted 1.1 mF
tweeter series capacitor. The woofer with this impedence compensation network sounded nicely out to 10kHz or so, and the tweeter filled in nicely above. This gave (comparatively)nicely phase linear system, that went clean, and seriously loud, and the amplifier ran barely above warm. (150 Wrms/ch/8ohms)
The downside is that absoloute polarity is clearly apparent, and needs to be inverted accordingly.
This system was interestingly very efficient, stunning depth imaging and immediacy, and never any ear bleeding.
I have found this to work great with correct electrical tuning - Hope this can help.
Regards Eric.
tweeter series capacitor. The woofer with this impedence compensation network sounded nicely out to 10kHz or so, and the tweeter filled in nicely above. This gave (comparatively)nicely phase linear system, that went clean, and seriously loud, and the amplifier ran barely above warm. (150 Wrms/ch/8ohms)
The downside is that absoloute polarity is clearly apparent, and needs to be inverted accordingly.
This system was interestingly very efficient, stunning depth imaging and immediacy, and never any ear bleeding.
I have found this to work great with correct electrical tuning - Hope this can help.
Regards Eric.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.