Been following this thread with great interest, having been introduced to single, wide-range drivers several years ago, and having built a few pairs since then, [thanks to all who's input made the "Frugel-Horns" possible!] believe I've found my niche.
Are there any Western Canadian, Pacific Northwest Feastrex owners around? Would love to hear a pair!
Don
Are there any Western Canadian, Pacific Northwest Feastrex owners around? Would love to hear a pair!
Don
rjbond3rd said:
Indeed. And it would have seemed to me to make a lot of sense to have drivers made available to the designers to facilitate development -- not (by far) that I know everything that's going on, my endless posts notwithstanding . . . perhaps there is an understanding between folks somewhere that I don't know about. But it sure seems like an odd way to bring an enclosure into the world, and my hat goes off to the people who have made it happen despite being (apparently) handicapped by less than ideal circumstances.
Cal Weldon said:No Don, we're too poor. What with our exorbitant mortgages, skyrocketing food bills, outrageous fuel prices and taxes so high they fall over, we're lucky just to afford a beer from time to time.
Cal: Maybe we can put our $100 Gov't cheques toward this years beverage supply at Dave's DIY Fest!
Don
If I were The Simpsons' Mr. Burns trying to market a high-end driver, I'd fiendishly leave a pair on Scottmoose's front porch, then anonymously mail some to Planet10, and scatter a few more in certain select cities (you know which ones), then I'd sit back and laugh, mwah-ha-ha-ha-hah, as my company took over the ultra-high-end fullranger world! The world I tell you!
Seriously. Advertising budget: zero. Probability of success: 99.999%. rubbing hands together Mmmm, excellent...
Seriously. Advertising budget: zero. Probability of success: 99.999%. rubbing hands together Mmmm, excellent...
doorman said:
You know Don, this'll be the first road trip ever where the travel costs outway the beverage costs. Yikes, I'm getting old(er)!
I'll bring the cooler as usual and leave a little room for contributions, especially those dark manly beers. You know, the kind that scare everyone out of the tent later on?
Oops, that's way OT. Back to regularly scheduled programming
That's one of the interesting things about audio though isn't it -the difference between theory & practice. Designing a theoretically optimal cabinet, either via T/S math, or earlier types of alignments is fine as far as it goes, but actually refining & voicing something is often a more complicated issue. Speaker design is neither art nor science -personally, I regard it as the artistic implementation of science. Remove the former and you spell instant tedium. Remove the latter, and you won't have a speaker at all.
Anyway, without hearing a unit, you're partially flying blind in this respect, although the raw data & personal experience does count for a lot, so you can usually get in the general zone, with the final tweaking needing to be done by the end user. Far from ideal, as you say Chris, compared to designing & being able to physically test & refine -but we do what we can, with whatever is available to us.
Best
Scott
well, along that route...
So on the science part, as one of my first duties at Feastrex, I am to measure all the production drivers for accurate TS parameters using our Woofer Tester Pro.
So far, I have been an enclosure guy, and not a particularly scientific one at that. I did cut and try. All by trusting my ear and building on traditional designs. This method works great for making drivers too apparently, as it is what Mr. Teramoto has been doing since before I was born.
Prettymuch I know enough to know I don't know a darned thing about actual electrical driver measurements so I am going to play dumb here. I have the instruction manual of WTPro and am plugging through it. Basically, I am asking for advice on how to properly test these Feastrex drivers from people well versed in driver measurement (perhaps), but probably more importantly, I am asking for advice from folks who have graciously designed enclosures for these drivers so far, and have been hampered by the limited specification information on them.
Since these drivers, like most drivers change in response characteristics beyond hearing perception upon the application of higher power, I would like to test both with the 1w standard testing, and also with a crown amp we have laying around. To perhaps give a few different points of view.
So far, I have played with the FE87 (old version (and this is important for many reasons)). From the WT2 measurement of QTS, with the driver having sat in a box for a few years no play I got values that were all whacked out. This was to simulate a new or "stiff" driver. The QTS came out to around 2.2 After running juice through it (50hz at a decent excursion for 24 hrs) and then sticking them on the bench immediately, I got a QTS of .80. After leaving them for a few hours to "cool", QTS was 1.0, and then after a several more hours it was 1.05, then finally 1.08 (which is very close to Fostex's specs). After repeated measurments, this value remained roughly the same.
So it seems that run in is important, but that also there is some "cooling" which must happen after "burn in." Huh?
On to VAS. This is a technical aspect of driver measurement that evaded me. I tried the Delta mass method with the FE87, and even remotely plausible values evaded me. I threw as many different kinds of pocket change Mr. Teramoto had laying in his desk drawer. all different weights and nationalities. nothing even came close to working. Reading around, it seems that the Delta mass method is fhooey anyways for getting accurate measurements. Temporary gluing things to washi? Aitch naw. Moving on...
We have resources to do closed box measurements instead. How should these go? Are the Feastrex drivers, using a phase plug going to leak too much for a closed box measurement? Are there other ways of measuring Vas, including other programs or equipment that I should be aware of? (I am pretty sure Feastrex would be willing to do whatever it takes to get a system setup to be doing these measurements right every time.)
I am not afraid to do a little math here. It is the technical aspects of all this that I wish not to screw up.
Any advice would be a great help. I would like to help provide some incredibly accurate specs for both customers and enclosure designers in the near future. Also, if things like the high power testing or any other tests are useful, please let me know.
Running the risk of getting chewed out by any die hard objectivists, let me say that the Feastrex drivers are indeed "special" and might require some special measurements in addition to the traditional measurements to see what is actually going on with them.
Finding these "special" measurements might be more of a long term pursuit. something for dedicated acousticians (of which I am not) to develop in grad school or beyond. Seeing that these drivers are at the leading edge of cone driver technology (and especially magnetic circuit quality and raw evenly distributed power at VC gap) the proper way of measuring them might not have even been invented yet.
It is like the man made element saga of the last century (continuing to this cent.) the old "laws" of physics break down upon newer observations of heavier particles and even rarer subatomic particles.
I have absolutely no idea what those special processes might be, but I do understand that it is generally held belief that a powerful magnetic circuit mated to a lightweight paper cone yields a lack of bass response. These drivers are simply impossible according to that traditional logic.
I am willing to beg that question. I sincerely hope others are too. Naturally, the same would go for many other drivers out there, somewhat shortchanged by a 1970s tiny Fe magnet heavy cone mass production paradigm. (my own feelings here).
Thanks in advance,
-Clark
So on the science part, as one of my first duties at Feastrex, I am to measure all the production drivers for accurate TS parameters using our Woofer Tester Pro.
So far, I have been an enclosure guy, and not a particularly scientific one at that. I did cut and try. All by trusting my ear and building on traditional designs. This method works great for making drivers too apparently, as it is what Mr. Teramoto has been doing since before I was born.
Prettymuch I know enough to know I don't know a darned thing about actual electrical driver measurements so I am going to play dumb here. I have the instruction manual of WTPro and am plugging through it. Basically, I am asking for advice on how to properly test these Feastrex drivers from people well versed in driver measurement (perhaps), but probably more importantly, I am asking for advice from folks who have graciously designed enclosures for these drivers so far, and have been hampered by the limited specification information on them.
Since these drivers, like most drivers change in response characteristics beyond hearing perception upon the application of higher power, I would like to test both with the 1w standard testing, and also with a crown amp we have laying around. To perhaps give a few different points of view.
So far, I have played with the FE87 (old version (and this is important for many reasons)). From the WT2 measurement of QTS, with the driver having sat in a box for a few years no play I got values that were all whacked out. This was to simulate a new or "stiff" driver. The QTS came out to around 2.2 After running juice through it (50hz at a decent excursion for 24 hrs) and then sticking them on the bench immediately, I got a QTS of .80. After leaving them for a few hours to "cool", QTS was 1.0, and then after a several more hours it was 1.05, then finally 1.08 (which is very close to Fostex's specs). After repeated measurments, this value remained roughly the same.
So it seems that run in is important, but that also there is some "cooling" which must happen after "burn in." Huh?
On to VAS. This is a technical aspect of driver measurement that evaded me. I tried the Delta mass method with the FE87, and even remotely plausible values evaded me. I threw as many different kinds of pocket change Mr. Teramoto had laying in his desk drawer. all different weights and nationalities. nothing even came close to working. Reading around, it seems that the Delta mass method is fhooey anyways for getting accurate measurements. Temporary gluing things to washi? Aitch naw. Moving on...
We have resources to do closed box measurements instead. How should these go? Are the Feastrex drivers, using a phase plug going to leak too much for a closed box measurement? Are there other ways of measuring Vas, including other programs or equipment that I should be aware of? (I am pretty sure Feastrex would be willing to do whatever it takes to get a system setup to be doing these measurements right every time.)
I am not afraid to do a little math here. It is the technical aspects of all this that I wish not to screw up.
Any advice would be a great help. I would like to help provide some incredibly accurate specs for both customers and enclosure designers in the near future. Also, if things like the high power testing or any other tests are useful, please let me know.
Running the risk of getting chewed out by any die hard objectivists, let me say that the Feastrex drivers are indeed "special" and might require some special measurements in addition to the traditional measurements to see what is actually going on with them.
Finding these "special" measurements might be more of a long term pursuit. something for dedicated acousticians (of which I am not) to develop in grad school or beyond. Seeing that these drivers are at the leading edge of cone driver technology (and especially magnetic circuit quality and raw evenly distributed power at VC gap) the proper way of measuring them might not have even been invented yet.
It is like the man made element saga of the last century (continuing to this cent.) the old "laws" of physics break down upon newer observations of heavier particles and even rarer subatomic particles.
I have absolutely no idea what those special processes might be, but I do understand that it is generally held belief that a powerful magnetic circuit mated to a lightweight paper cone yields a lack of bass response. These drivers are simply impossible according to that traditional logic.
I am willing to beg that question. I sincerely hope others are too. Naturally, the same would go for many other drivers out there, somewhat shortchanged by a 1970s tiny Fe magnet heavy cone mass production paradigm. (my own feelings here).
Thanks in advance,
-Clark
Hi Clark,
Glad you're over at Feastrex. FWIW, I'd be using closed box method to calc. Vas. The thought of sticking weights onto very delicate cones is enough to make me reach for the box of nerve pills. You'll probably get some leakage, but I doubt it would being so severe as to screw things up badly. Won't be perfect, but there's always variation between drivers anyway. Good enough is good enough IMO. Other people know way more than me about this though.
WRT the trad. notion that you never get any LF from light cones & high power motors, it's too simple for its own good. You can. Nothing particularly mysterious about it though. OK, so we sometimes curse them, but that's where Thiele & Small's work makes life so much easier.
Looking at the T/S parameters of the D5nf & D9nf that Chris took a while back, certainly, their Bl factors certainly indicate high motor strength, and relatively light cones. However, the electrical & mechanical damping / control of a drive unit also plays a significant role in its LF & lower midrange performance. Here we see they have a middling total Q: roughly 0.59 for the D5nf & 0.4 for the D9nf. So taking their -3db roll off point as Fhm = 2Fs / Qe then we get a relatively moderate ~331Hz & ~299Hz for the D5nf & D9nf respectively. Contrast that with, say, the FE166E (454Hz) & FE206E (433Hz) which, being inherently better damped, are classic examples of what people expect from FR units, i.e. weak LF & either need major Eq, or to be mated to a horn or XO to a big fat bass unit.
Cheers
Scott
edit: Qt is often substituted for Qe in the above, but the basic trend remains.
Glad you're over at Feastrex.
FWIW, I'd be using closed box method to calc. Vas. The thought of sticking weights onto very delicate cones is enough to make me reach for the box of nerve pills. You'll probably get some leakage, but I doubt it would being so severe as to screw things up badly. Won't be perfect, but there's always variation between drivers anyway. Good enough is good enough IMO. Other people know way more than me about this though.WRT the trad. notion that you never get any LF from light cones & high power motors, it's too simple for its own good. You can. Nothing particularly mysterious about it though. OK, so we sometimes curse them, but that's where Thiele & Small's work makes life so much easier.
Looking at the T/S parameters of the D5nf & D9nf that Chris took a while back, certainly, their Bl factors certainly indicate high motor strength, and relatively light cones. However, the electrical & mechanical damping / control of a drive unit also plays a significant role in its LF & lower midrange performance. Here we see they have a middling total Q: roughly 0.59 for the D5nf & 0.4 for the D9nf. So taking their -3db roll off point as Fhm = 2Fs / Qe then we get a relatively moderate ~331Hz & ~299Hz for the D5nf & D9nf respectively. Contrast that with, say, the FE166E (454Hz) & FE206E (433Hz) which, being inherently better damped, are classic examples of what people expect from FR units, i.e. weak LF & either need major Eq, or to be mated to a horn or XO to a big fat bass unit.
Cheers
Scott
edit: Qt is often substituted for Qe in the above, but the basic trend remains.
I totally disagree. I use the added mass method exclusively and have never had problems with accuracy. I had more trouble with the test box method. If you recognize that T/S parameters are small signal measurements and characterize the driver's resonant behavior you should be fine using the added mass method. The added mass is known exactly (or at least to a number of significant digits that is acceptable) with no requirement to calculate an air volume that includes the volume added in the cone region or displaced by the magnet structure. I have used this method on Lowther, Radio Shack, and Fostex drivers with excellent consistent results.
I use US nickels (5 grams each) and just rest them on the cone to produce the required frequency shift, I might make two measurements with different numbers of nickels to check the results for consistency. You must use a low input signal that just barely moves the cone, the coins should not vibrate. This will provide an accurate measure of the driver's T/S properties in the linear range of motions. The key is small applied signals.
You will not damage a delicate full range driver using the added mass method if you are careful and know what you are doing. The results are repeatable, you won't get exactly the same results for each measurement since there is always some level of inaccuracy in a scientific measurement, from measurement to measurement and day to day. There is no magic or voodoo, you are characterizing a simple single degree of freedom system.
Hello,
I am prety amazed about Mr. Teramoto's work. I mean... from what Mr. Witmer said, these drivers seem to be made almost exclusively by ear and just based on experience.
Looking into driver design, it's seems a lot of trouble. Compliance, mass, cone profile, dampening, motor strenght, uniform magnetic gap... puting this together seems a task exclusively for R&D designers with all the know-how and equipement of the large companies.
I can say just.. wow!
I am prety amazed about Mr. Teramoto's work. I mean... from what Mr. Witmer said, these drivers seem to be made almost exclusively by ear and just based on experience.
Looking into driver design, it's seems a lot of trouble. Compliance, mass, cone profile, dampening, motor strenght, uniform magnetic gap... puting this together seems a task exclusively for R&D designers with all the know-how and equipement of the large companies.
I can say just.. wow!
Clark, I would like to put in one effort to see that this is no more complicated than it has to be. My pal Mike (Panomaniac) has WT3, and the last time I went to his house, we tested all the oddball no-specs drivers I had in the house. It seems to go like this:
Mount the WT3 'ware on your machine.
Read the directions.
Plug the tester tail into your USB port.
Bring up the 'ware, and select your options from the menus.
Hook the tester up to your speaker.
Run the first sweep.
Add mass to the cone (we used 2 nickels and he put 10g in the window).
Run the second sweep. If the speak is sitting on its magnet on the table, there's no need to attach the weight; the test signal is very small, and the 'ware can tell if it's bouncing.
Ta-dah; all your TS params come up in the window.
I learned how to do all this in about five minutes, so it will probably take you about three. Pano is coming over this weekend, probably tomorrow, and he's offered to let me borrow the thing (he's tested every durn thing he has in the place, took him about fifteen minutes), so I'll tell him to bring it with him. When I have had a better look at it, I'll post back here.
I mean, the thing is supposed to be easy.
Aloha,
Poinz
Mount the WT3 'ware on your machine.
Read the directions.
Plug the tester tail into your USB port.
Bring up the 'ware, and select your options from the menus.
Hook the tester up to your speaker.
Run the first sweep.
Add mass to the cone (we used 2 nickels and he put 10g in the window).
Run the second sweep. If the speak is sitting on its magnet on the table, there's no need to attach the weight; the test signal is very small, and the 'ware can tell if it's bouncing.
Ta-dah; all your TS params come up in the window.
I learned how to do all this in about five minutes, so it will probably take you about three. Pano is coming over this weekend, probably tomorrow, and he's offered to let me borrow the thing (he's tested every durn thing he has in the place, took him about fifteen minutes), so I'll tell him to bring it with him. When I have had a better look at it, I'll post back here.
I mean, the thing is supposed to be easy.
Aloha,
Poinz
Exactly, Martin. After years of not knowing what we had in our hand, of manufacturers and vendors either not publishing or 'optimizing' the parameters, the thing is a Godsend.
The part about putting a pair of nickels on (and off) the rice paper cone of a four thousand dollar driver might still make me a little nervous, though.
Aloha,
Poinz
The part about putting a pair of nickels on (and off) the rice paper cone of a four thousand dollar driver might still make me a little nervous, though.
Aloha,
Poinz
SunRa said:
No mean feat indeed. But I wonder how better it would be if there was some science in the mix on top of good taste.
I would like to get a full test sheet with serial number against a reference driver if buying a $4K unit. Especially with the versions and tunings going on, I would like to know the spec and match of a given pair.
No matter how much guru someone is, not digging the tech at hand is a handicap.
With WT2, I used the tiny weights that came with my precision beam balance, but any small object of known weight that will fit on the cone could work. The weights that I used had a happy proprensity to wedge themselves between the main cone and subcone, and the small signals input by WT2 would not easily dislodge them. (A pair of tweezers came in handy.)
I think that as far as basic T/S parameter measurements are concerned, the Woofer Tester Pro will probably be about the same as the WT2 and the WT3 in the general approach taken.
Once you have it set up to work properly and know what you are doing, the Woofer Tester Pro will make itself most useful in allowing you to make various measurements of speaker sound output with a microphone.
Also, the ability to analyze the drivers' behavior with more powerful signal inputs will probably be useful as you go about developing and tweaking enclosures.
-- Chris
I think that as far as basic T/S parameter measurements are concerned, the Woofer Tester Pro will probably be about the same as the WT2 and the WT3 in the general approach taken.
Once you have it set up to work properly and know what you are doing, the Woofer Tester Pro will make itself most useful in allowing you to make various measurements of speaker sound output with a microphone.
Also, the ability to analyze the drivers' behavior with more powerful signal inputs will probably be useful as you go about developing and tweaking enclosures.
-- Chris
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.