DrewP said:
Martin King shows a very similar phenomena with his Lowther DX2/DX3/DX4. Go look under projects on his site for some measurement graphs. http://www.quarter-wave.com/. A teaser below.
dave
Thank you Drew for the very informative run down for the Fostex Sigma series, you sold me to the 108. I might be testing these by Christmas.
The thing is that I want to replace a very shabby "FR" from RS that has a lot of shortcomings which I use without XO, thow I have been trying to add a notch filter for the 2.5Khz peak, but still remains another peak at 7.5K. Its a very lousy speaker to say the least.
What's going on with those funny looking ridged cones from Fostex?
The thing is that I want to replace a very shabby "FR" from RS that has a lot of shortcomings which I use without XO, thow I have been trying to add a notch filter for the 2.5Khz peak, but still remains another peak at 7.5K. Its a very lousy speaker to say the least.
What's going on with those funny looking ridged cones from Fostex?
Dave, those plots look like they are from the same driver, only different samples!
I bet prices arent that near.
I bet prices arent that near.
Hey guys, here is the response I got from the Bert person that Dave mentioned. I have no connection with him, but thought that this might be useful.
Hi Mark,
> What would the cost be (euros is fine) for a pair of 215 RTF Supravox
(single cone) shipped to the USA?
These are 162 Euro each, shipping them to the USA is 49 Euro making the
total for one pair 373 Euro.
> How is the quality on these particular models. I know some Supravox have
had some problems. Are these OK?
I had one problem in the beginning (bad soldered contacts) but that seems to
be okay now.
Ciao,
Bert
-----
BD-Design - info@bd-design.nl
Phone/Fax : +31 341 254500 Mobile: +31 6 51242990
Our Products, only the best! - http://www.bd-design.nl/main.php
Hi Mark,
> What would the cost be (euros is fine) for a pair of 215 RTF Supravox
(single cone) shipped to the USA?
These are 162 Euro each, shipping them to the USA is 49 Euro making the
total for one pair 373 Euro.
> How is the quality on these particular models. I know some Supravox have
had some problems. Are these OK?
I had one problem in the beginning (bad soldered contacts) but that seems to
be okay now.
Ciao,
Bert
-----
BD-Design - info@bd-design.nl
Phone/Fax : +31 341 254500 Mobile: +31 6 51242990
Our Products, only the best! - http://www.bd-design.nl/main.php
____________________________________________________
Thank you Drew for the very informative run down for the Fostex Sigma series, you sold me to the 108.
____________________________________________________
As far as I can see the only downsides to the 108 is a limited bottom end extension a lower sensitivity and a lower power handling. All this adds up to LESS BASS.
Now me, I got funny tastes. I'd prefer fast/tight bass even if there's much less of it. So horn loading a 108 would be fine for me. I also have a loungeroom that's 12ft x 18ft with 8ft ceilings so i don't need mega volume.
You may find the 108's run out of "Ooof" pretty fast. I think BD design used them in a TQWT with 2 in parallel.
Re the "funny ridges" this is a new devolpment in the fostex lineup. I think they break up any standing waves across the cone (or prevent them forming at all in the first place). All the 108 Sigma series have been updated to ESigma. All the whizzer cones are gone which is probably no bad thing. Not precisely sure what's happened with efficiency, power handling etc but reports seem to indicate that the new drivers will be better sorted for horns. Looks like it's trickle down technology from the ultra high end Sigma ES Japan Only units of last year.
Drew.
Thank you Drew for the very informative run down for the Fostex Sigma series, you sold me to the 108.
____________________________________________________
As far as I can see the only downsides to the 108 is a limited bottom end extension a lower sensitivity and a lower power handling. All this adds up to LESS BASS.
Now me, I got funny tastes. I'd prefer fast/tight bass even if there's much less of it. So horn loading a 108 would be fine for me. I also have a loungeroom that's 12ft x 18ft with 8ft ceilings so i don't need mega volume.
You may find the 108's run out of "Ooof" pretty fast. I think BD design used them in a TQWT with 2 in parallel.
Re the "funny ridges" this is a new devolpment in the fostex lineup. I think they break up any standing waves across the cone (or prevent them forming at all in the first place). All the 108 Sigma series have been updated to ESigma. All the whizzer cones are gone which is probably no bad thing. Not precisely sure what's happened with efficiency, power handling etc but reports seem to indicate that the new drivers will be better sorted for horns. Looks like it's trickle down technology from the ultra high end Sigma ES Japan Only units of last year.
Drew.
Lynn,
From your original post in this thread, it seems you're on a search to design a speaker using criteria similar to those that I used a few months ago when working on my speakers.
I posted my findings on this thread: http://www.diyaudio.com/forums/showthread.php?threadid=6644
You might want to look it over.
Regards, Robert
From your original post in this thread, it seems you're on a search to design a speaker using criteria similar to those that I used a few months ago when working on my speakers.
I posted my findings on this thread: http://www.diyaudio.com/forums/showthread.php?threadid=6644
You might want to look it over.
Regards, Robert
DrewP said:
And being discontinued getting harder to find
Not exactly new. This was 1st shown in the Fostex NF1 in 1999, from there into the Japanese limited editions, and now into regular production. Probably the most significant part of this trickle down is the trick surround.
dave
Is this thread dead? Where are you Lynn? Variac, Planet10, DrewP...... all others?
Is the Fostex FX200 a contender?
On axis and dispersion looks good. No whizer cone. Enough low freq to cross at 100 Hz or maybe lower. Only 1 mm Xmax but I guess it's OK for an 8" speaker. Only drawback will be eff quoted at 92dB. Anyone heard these?
Is it a Fostex? No bannana pulp cone, what a disappointment.
Is the Fostex FX200 a contender?
On axis and dispersion looks good. No whizer cone. Enough low freq to cross at 100 Hz or maybe lower. Only 1 mm Xmax but I guess it's OK for an 8" speaker. Only drawback will be eff quoted at 92dB. Anyone heard these?
Is it a Fostex? No bannana pulp cone, what a disappointment.
I see that most on this thread are playing around with 6" FR Fostex's so I had a look at models in this size shown at Madisound, I counted 7 of them.
Strangely enough the most expensive one, the FE168E Sigma, had the worst on axis response and the least expensive one, the FE164, had the best plot.
Have someone heard the FE164?
Strangely enough the most expensive one, the FE168E Sigma, had the worst on axis response and the least expensive one, the FE164, had the best plot.
Have someone heard the FE164?
Those that have heard the FX200 like it... needs some help at the top.
Martin King used the FE164 in his ML-TQWT and is was good enuff for him to go out and get a set of FE208sigmas which has lead to a set of Lowther DX2/3/4s. He liked it. I don't think these are very available any more or at least not for long.
People that have the FE168sigmas are very enthusiastic about them. Hop over to the Full Range Forum and do a search on FE168.
My personal experience is limited to the budget FE103 sold as the Radio Shack 40-1197 and the alnico precursor that i have dubbed the FE103A. The RS is shockingly good, and they get better as you go up. I'd love to have the money to try a quad of FE108s or 168s.
What i'd really like to see is the new banana pulp FE108 with a smaller, more powerful alnico motor.
Almost any of these are the basis of an eventual 3-way, mostly full-range speaker system -- and it can be grown 1 way at a time. The choices and compromises you can make are legion and you can enjoy the music the whole way.
dave
Martin King used the FE164 in his ML-TQWT and is was good enuff for him to go out and get a set of FE208sigmas which has lead to a set of Lowther DX2/3/4s. He liked it. I don't think these are very available any more or at least not for long.
People that have the FE168sigmas are very enthusiastic about them. Hop over to the Full Range Forum and do a search on FE168.
My personal experience is limited to the budget FE103 sold as the Radio Shack 40-1197 and the alnico precursor that i have dubbed the FE103A. The RS is shockingly good, and they get better as you go up. I'd love to have the money to try a quad of FE108s or 168s.
What i'd really like to see is the new banana pulp FE108 with a smaller, more powerful alnico motor.
Almost any of these are the basis of an eventual 3-way, mostly full-range speaker system -- and it can be grown 1 way at a time. The choices and compromises you can make are legion and you can enjoy the music the whole way.
dave
Reality check?
A properly implemented full range driver should be able to reduce these problems. But then, a properly implemented 'electrical' crossover (whether passive or active) should be able to do the same!
Is it really true that fullrange drivers do not introduce phase (or other) distortions similar to that of a crossover? Is a full range driver really better than a properly implemented crossover? Not questioning anyone's wisdom here, just having my moments of doubt. Here are some quotes from the fullrange driver website:surf said:
A properly implemented full range driver should be able to reduce these problems. But then, a properly implemented 'electrical' crossover (whether passive or active) should be able to do the same!
Regarding "Reality check?"
A widerange driver that extends from 80-4khz (nominally flat response within this range) usually means that phase distortion is kept out of the "critical region".
Normally the amplitude and phase responses of a loudspeaker driver are interrelated so that the phase distortion is essentially determined by the amplitude response. By "normally" I mean that the driver has effectively a "single location in acoustic space" (may not be entirely true for some driver designs, but mostly true for most drivers). If a driver falls off at the low end at 80hz with a 12 dB per octave rolloff, then the driver itself is introducing the same phase distortion as a 2nd order crossover at 80hz would. If the amplitude response is flat from 80-4000hz, and crossovers are applied at these frequencies in a manner that accounts for driver falloff and phase response, then the phase distortions are determined by these crossover designs and by the amplitude/phase responses of the lowest and highest frequency drivers at the lower and upper frequency extremes.
In simple terms, we are "mostly OK with our assumptions", but there are exceptions, and there are approximations involved. Certainly driver phase response must be considered as part of crossover design, and many useful crossover desaigns exist that are far from "ideal textbook" but still sound OK.
Hope this helps.
A widerange driver that extends from 80-4khz (nominally flat response within this range) usually means that phase distortion is kept out of the "critical region".
Normally the amplitude and phase responses of a loudspeaker driver are interrelated so that the phase distortion is essentially determined by the amplitude response. By "normally" I mean that the driver has effectively a "single location in acoustic space" (may not be entirely true for some driver designs, but mostly true for most drivers). If a driver falls off at the low end at 80hz with a 12 dB per octave rolloff, then the driver itself is introducing the same phase distortion as a 2nd order crossover at 80hz would. If the amplitude response is flat from 80-4000hz, and crossovers are applied at these frequencies in a manner that accounts for driver falloff and phase response, then the phase distortions are determined by these crossover designs and by the amplitude/phase responses of the lowest and highest frequency drivers at the lower and upper frequency extremes.
In simple terms, we are "mostly OK with our assumptions", but there are exceptions, and there are approximations involved. Certainly driver phase response must be considered as part of crossover design, and many useful crossover desaigns exist that are far from "ideal textbook" but still sound OK.
Hope this helps.
Regarding "Reality check?"
to jag:
Yes, full range drivers are usually not perfect, and usually have response amplitude and phase anomalies. If so, they must be tollerated, corrected, or discarded.
In my previous post I was making an extremely limited statement regarding phase distortion. If the driver has a frequency response anomaly in the intended passband, then for critical designs the response anomalies are "smoothed out" by appropriate equalization using active or passive elements (notches, etc.). If this can be done effectively (to obtain flat response) then the phase distortion is also reduced.
HOWEVER, some response anomalies are not so easily treated, and so some driver/enclosure combinations are a "lost cause". In this case the phase distortion will still be present, but such distortion is typically obscured by the more grotesque audible effects of the non-flat amplitude response. The more difficult to treat response anomalies include uncontrolled cone breakup and diffraction and/or resonance effects that can arise in enclosure/baffle designs that are dramatically poorly designed.
So, for example, a designer who wants to use a driver with significant cone breakup or other resonance will want to place the crossover well away from the breakup point, and possibly place a notch where the peak occurs. This sometimes works.
A designer who wants to avoid crossover distortions within the 80-4khz region (or wherever) will choose a driver that is flat or can be equalized to flat within that range.
to jag:
Yes, full range drivers are usually not perfect, and usually have response amplitude and phase anomalies. If so, they must be tollerated, corrected, or discarded.
In my previous post I was making an extremely limited statement regarding phase distortion. If the driver has a frequency response anomaly in the intended passband, then for critical designs the response anomalies are "smoothed out" by appropriate equalization using active or passive elements (notches, etc.). If this can be done effectively (to obtain flat response) then the phase distortion is also reduced.
HOWEVER, some response anomalies are not so easily treated, and so some driver/enclosure combinations are a "lost cause". In this case the phase distortion will still be present, but such distortion is typically obscured by the more grotesque audible effects of the non-flat amplitude response. The more difficult to treat response anomalies include uncontrolled cone breakup and diffraction and/or resonance effects that can arise in enclosure/baffle designs that are dramatically poorly designed.
So, for example, a designer who wants to use a driver with significant cone breakup or other resonance will want to place the crossover well away from the breakup point, and possibly place a notch where the peak occurs. This sometimes works.
A designer who wants to avoid crossover distortions within the 80-4khz region (or wherever) will choose a driver that is flat or can be equalized to flat within that range.
goudey,
Thanks for your patience in explaining this.
Probably it was tacitly understood by others on this thread (that a full/wide range driver has more problems and you need to be extra careful in dealing with one), but I was getting an impression that using (any) wide range driver automatically resolves the crossover problems in the critical range.
I guess there are compromises involved in any design. And ultimately it boils down to personal preference as to what (compromise) sounds better to you, and which problems would you rather resolve.
Having said that, I am not implying that I am giving up on a full range driver - maybe it will also sound better to me (I have not listened critically to a wide range design yet). But, it is good to know all the issues involved.
Thanks for your patience in explaining this.
Probably it was tacitly understood by others on this thread (that a full/wide range driver has more problems and you need to be extra careful in dealing with one), but I was getting an impression that using (any) wide range driver automatically resolves the crossover problems in the critical range.
I guess there are compromises involved in any design. And ultimately it boils down to personal preference as to what (compromise) sounds better to you, and which problems would you rather resolve.
Having said that, I am not implying that I am giving up on a full range driver - maybe it will also sound better to me (I have not listened critically to a wide range design yet). But, it is good to know all the issues involved.
wideband driver attributes
to jag:
"...a full/wide range driver has more problems and you need to be extra careful in dealing with one..."
Actually, the problems I mentioned are not particularly worse with drivers intended to be operated as very wideband devices. In fact, most "wideband drivers" that have succeeded in attracting the "single driver" crowd are typically far better with regards to cone breakup behavior and other resonance effects than are typical midbass drivers. Its just that when you allow use of a crossover you can "chop off" a problem area in the drivers response, while if you don't allow a crossover, then you need to be more concerned about the typical cone breakup region.
Most people don't place any priority on avoiding crossovers in the 80-4khz region (or wherever) until they hear such a design. But note that we end up with forums that specialize in such designs, and people who search out the ultimate wideband driver. So maybe there is "something to it". I observe that some prefer the wideband approach once they experience it, while others who are exposed to it are not particularly attracted to it.
to jag:
"...a full/wide range driver has more problems and you need to be extra careful in dealing with one..."
Actually, the problems I mentioned are not particularly worse with drivers intended to be operated as very wideband devices. In fact, most "wideband drivers" that have succeeded in attracting the "single driver" crowd are typically far better with regards to cone breakup behavior and other resonance effects than are typical midbass drivers. Its just that when you allow use of a crossover you can "chop off" a problem area in the drivers response, while if you don't allow a crossover, then you need to be more concerned about the typical cone breakup region.
Most people don't place any priority on avoiding crossovers in the 80-4khz region (or wherever) until they hear such a design. But note that we end up with forums that specialize in such designs, and people who search out the ultimate wideband driver. So maybe there is "something to it". I observe that some prefer the wideband approach once they experience it, while others who are exposed to it are not particularly attracted to it.
How to read phase distortion?
Here are some measurements (by NRC) for a 2 way design. This figure shows on-axis, 15 degrees off-axis, and 30 degrees off-axis response. Note the dip at 1500 corresponding to the crossover at 1500.
Post contd. to next thread
goudey said:
Here are some measurements (by NRC) for a 2 way design. This figure shows on-axis, 15 degrees off-axis, and 30 degrees off-axis response. Note the dip at 1500 corresponding to the crossover at 1500.
Post contd. to next thread
Attachments
Contd. from previous thread
Here is the phase response for the same design. Now, I actually see a lot of phase anomalies because of the woofer driver as opposed to the crossover. Or am I reading this right?
I apologize if I am taking this thread offtopic. Pl let me know if I should make this a new thread.
Here is the phase response for the same design. Now, I actually see a lot of phase anomalies because of the woofer driver as opposed to the crossover. Or am I reading this right?
I apologize if I am taking this thread offtopic. Pl let me know if I should make this a new thread.
Attachments
regarding amplitude/phase plots from jag
The amplitude and phase plots are not consistent with one another according to any physical model I am aware of for normal analog circuits. For example, the phase deviation where the woofer falls off stays within +/- 45 degrees, when I would expect phase changes to exceed +/- 90 degrees. In the phase plot there is no sign of the crossover, but phase perfect crossovers do exist. Such crossovers trade the phase distortion problem for problems with total power versus frequency. This type of cure may be worse than the disease. Some will not agree, but constant power versus frequency for multi-way system is a common design objective (all part of the trade space). I find it strange that the phase plot is nice and smooth and the amplitude plot looks like a typical measurement. Are you sure this is not the phase of the impedance of the design, not the phase of the acoustic response?
The amplitude and phase plots are not consistent with one another according to any physical model I am aware of for normal analog circuits. For example, the phase deviation where the woofer falls off stays within +/- 45 degrees, when I would expect phase changes to exceed +/- 90 degrees. In the phase plot there is no sign of the crossover, but phase perfect crossovers do exist. Such crossovers trade the phase distortion problem for problems with total power versus frequency. This type of cure may be worse than the disease. Some will not agree, but constant power versus frequency for multi-way system is a common design objective (all part of the trade space). I find it strange that the phase plot is nice and smooth and the amplitude plot looks like a typical measurement. Are you sure this is not the phase of the impedance of the design, not the phase of the acoustic response?