Incidentally, just to clarify, he just said that he has heard two drivers with the same T/S data sound very different. For all I know, he might have been talking about drivers made by some other company! Or maybe even two drivers made by two different companies.
So, I don't think we can assume too much from this.
Mike
So, I don't think we can assume too much from this.
Mike
cdwitmer said:
That makes perfect sense and is also very wise imo, just like you should never only repair the brakes on one wheel but rather do them in pairs!! I would imagine the same approach should also be taken in the boob job industry, otherwise one could assume very lop sided results
Dave
planet10 said:
Well, admittedly my info dates from Altec, JBL back when vendors/employees usually took some pride in their work. I've seen the +/- 10% bandied about in a few threads, but nothing to validate it. WRT individual specs being far off, as long as the calc'd eff. is reasonably close to spec it doesn't matter from a box design POV.
GM
Hmm, there seems to be some similarity in response of these and EnABLed drivers. Although I have not listened to any of these. But the multiple SPL peaks seem somehow related with people hearing more detail.BTW said:Here's the graph that Nelson Pass posted on the Feastrex Fun Thread....
jogi59 said:
1. Always build speakers in pairs.
2. Pay careful attention to consistency of all components.
3. Follow exact rituals for all aspects of assembly.
4. Take in-house electronic measurements
5. Extended listening to and confirmation of operation of drivers, including repetition of step 4 ("extended" is relative; days as opposed to a few seconds for most mass-produced drivers)
Remember, the voice coil former and cone assemblies, including surrounds, are all made by one person, who is also in charge of assembling all the drivers from beginning to end.
The consistency of matching does not seem to be inferior to what people get through mass production techniques.
There is more than one way to skin a cat, as they say . . .
-- Chris
t/s
When I spoke with Mr. Larson, he enlightened me to the incredible usefulness of the woofertester pro, (and the wilyness of getting useful data for box design from T/S alone)
For instance: essentially, many of the driver's parameters can change quite a bit with drive level. this is one aspect of wilyness that is well documented on WT website.
Wow. So boxes that are tuned specifically to T/S might "lock in" at a certain music power level (whether or not that is actually the volume you want to be hearing), and then get a little out of whack when driven harder or softer...
This is why PA drivers are so different from computer speakers, from hi fi speakers and studio monitors. (that is, IF they are being well designed to their application).
Since learning that notion (and a few other curious tidbits from Mr. Larson), I have started drifting more and more into the "cut and try" ethic. I am in a woodshop. We can easily make a mockup of something in no time. That is the quickest and cheapest way for us to design. Always has been. This is not the case for everyone else out there though.
But even for my friends who enjoy speaker making as a (basic level) hobby, I am starting to encourage them to use laminated cardboard, particle board. anything really. just mock it up. Its like legos instead of algebra. Not that I have a problem with algebra. I use math regularly, I come from a family of mathematicians, and at the age of 16, I was even paid as a statistical consultant who prepared data for a court case! But at heart, back to my roots, I am a lego kid. As my grandpa used to say: "just put the saw in the wood..."
I think that t/s data will provide useful data to the people who truely know how to interpret it in relation to basic box design. I am only partially one of those people, self admitted. I had a real problem "streamlining" the measurement process of Feastrex drivers to a satisfactory level while I was there. It was not for a lack of motivation, not for a lack of people helping me. Just a simple learning curve and an ethic that I would not publish data that was taken through methods which were not totally ready for "prime time."
Watching Mr. Larson describe the process of measuring, tradeoffs of different measurement methods, and take the measurements of the NF5ex at RMAF was enlightening for us all. We all looked at Mr. Teramoto, Mr. Tanaka, acknowledged them for the mastery of their craft. Looked at Mr. Larson, acknowledged him for the mastery of his craft. It would have taken me a long time for me to have felt ethically OK (scientific method speaking) with publishing data from my taking. Certain things are best left to professionals or to people who have been through the learning curve.
I will certainly hold myself guilty of having been an "armchair critic" too many times in my earlier days as a hobbyist, trying to speculate what a driver would sound like from t/s data alone. How easy of a trap, and how false of a notion it is... Terry cured me of that right quick though. Saw in the wood... everybody has their own approach to problem solving and it takes a foolish one indeed to knock or knock up a particular method. The best projects I am involved in are almost always team efforts.
Eventually, I will get a WTpro to play around with. its a remarkable tool.
-Clark
When I spoke with Mr. Larson, he enlightened me to the incredible usefulness of the woofertester pro, (and the wilyness of getting useful data for box design from T/S alone)
For instance: essentially, many of the driver's parameters can change quite a bit with drive level. this is one aspect of wilyness that is well documented on WT website.
Wow. So boxes that are tuned specifically to T/S might "lock in" at a certain music power level (whether or not that is actually the volume you want to be hearing), and then get a little out of whack when driven harder or softer...
This is why PA drivers are so different from computer speakers, from hi fi speakers and studio monitors. (that is, IF they are being well designed to their application).
Since learning that notion (and a few other curious tidbits from Mr. Larson), I have started drifting more and more into the "cut and try" ethic. I am in a woodshop. We can easily make a mockup of something in no time. That is the quickest and cheapest way for us to design. Always has been. This is not the case for everyone else out there though.
But even for my friends who enjoy speaker making as a (basic level) hobby, I am starting to encourage them to use laminated cardboard, particle board. anything really. just mock it up. Its like legos instead of algebra. Not that I have a problem with algebra. I use math regularly, I come from a family of mathematicians, and at the age of 16, I was even paid as a statistical consultant who prepared data for a court case! But at heart, back to my roots, I am a lego kid. As my grandpa used to say: "just put the saw in the wood..."
I think that t/s data will provide useful data to the people who truely know how to interpret it in relation to basic box design. I am only partially one of those people, self admitted. I had a real problem "streamlining" the measurement process of Feastrex drivers to a satisfactory level while I was there. It was not for a lack of motivation, not for a lack of people helping me. Just a simple learning curve and an ethic that I would not publish data that was taken through methods which were not totally ready for "prime time."
Watching Mr. Larson describe the process of measuring, tradeoffs of different measurement methods, and take the measurements of the NF5ex at RMAF was enlightening for us all. We all looked at Mr. Teramoto, Mr. Tanaka, acknowledged them for the mastery of their craft. Looked at Mr. Larson, acknowledged him for the mastery of his craft. It would have taken me a long time for me to have felt ethically OK (scientific method speaking) with publishing data from my taking. Certain things are best left to professionals or to people who have been through the learning curve.
I will certainly hold myself guilty of having been an "armchair critic" too many times in my earlier days as a hobbyist, trying to speculate what a driver would sound like from t/s data alone. How easy of a trap, and how false of a notion it is... Terry cured me of that right quick though. Saw in the wood... everybody has their own approach to problem solving and it takes a foolish one indeed to knock or knock up a particular method. The best projects I am involved in are almost always team efforts.
Eventually, I will get a WTpro to play around with. its a remarkable tool.
-Clark
Thanks for that very enjoyable post, Clark. Although there may be folks out there who would simply plug the numbers into an enclosure design program, build whatever comes out and assume they had thereby obtained the best enclosure for the task at hand, and blame the driver if the results were less than stellar, I get the impression from most of the posters here who have been enquiring after the T/S parameters that they expect those numbers to put them only in the general neighborhood of where they are trying to go.
That T/S parameters change with signal level was also a surprising revelation for me when I first learned that fact about two years ago, but it is probably old news to most of the guys who are itching to get to work on their own Feastrex enclosure designs. Since music signals are constantly varying in strength, it means the appropriateness of the enclosure is constantly changing . . . ?
Anyway . . . a man's gotta do what a man's gotta do, even if the T/S parameters are only good for horseshoes and hand grenades.
In Japan, a few builders who have come up with real humdinger enclosures (or so I'm told, not having heard them myself) report that T/S parameter information was of "surprisingly little use" and that they had to build multiple prototypes relying largely on their intuition and past experience in order to get the drivers really singing at their best. What kept them from getting discouraged after initial setbacks was their recollections of the sound of the drivers that they had heard in Feastrex's listening room. I don't know the details of what formulae they may have been using -- not every enclosure design program or formula is necessarily going to be using T/S parameters in the same way -- but later on when I have a little more time I hope to translate their comments and post them here, or perhaps our bilingual Japanese friend Bunpei will beat me to it. Anyway, you can expect some translated comments from Japanese enclosure designers sometime in the next day or so.
-- Chris
That T/S parameters change with signal level was also a surprising revelation for me when I first learned that fact about two years ago, but it is probably old news to most of the guys who are itching to get to work on their own Feastrex enclosure designs. Since music signals are constantly varying in strength, it means the appropriateness of the enclosure is constantly changing . . . ?
Anyway . . . a man's gotta do what a man's gotta do, even if the T/S parameters are only good for horseshoes and hand grenades.
In Japan, a few builders who have come up with real humdinger enclosures (or so I'm told, not having heard them myself) report that T/S parameter information was of "surprisingly little use" and that they had to build multiple prototypes relying largely on their intuition and past experience in order to get the drivers really singing at their best. What kept them from getting discouraged after initial setbacks was their recollections of the sound of the drivers that they had heard in Feastrex's listening room. I don't know the details of what formulae they may have been using -- not every enclosure design program or formula is necessarily going to be using T/S parameters in the same way -- but later on when I have a little more time I hope to translate their comments and post them here, or perhaps our bilingual Japanese friend Bunpei will beat me to it. Anyway, you can expect some translated comments from Japanese enclosure designers sometime in the next day or so.
-- Chris
Yeah, T/S really only calcs the most space efficient, maximally flat alignment, though of course it can be manipulated to suit, but it's up to the designer to measure the right specs for the app, so a prosound alignment will be somewhat under-damped in a typical HIFI/HT app due to being designed based on a much higher VC resistance which raises Qts and suspension non-linearity that affects Vas.
I learned by cut n' try and a basic understanding of resonant systems before T/S and for all the T/S derived designs I've done over the decades, none were for my listening pleasure.
GM
I learned by cut n' try and a basic understanding of resonant systems before T/S and for all the T/S derived designs I've done over the decades, none were for my listening pleasure.
GM
Very interesting responses over the past couple days on this thread. Granted, no small task on the T/S measurements due to the huge variables and such. I know these are basically committed to by Feastrex and it's just a matter of time. I'm starting to think that providing some tried and true enclosure designs (on the Feastrex web site) would probably go a long way to improve customer perception and increase sales. Please note that I am a huge fan of Feastrex and would love to be in a position to purchase another pair of drivers in the near future.
Again, I'll suggest that providing useful information for all drivers is key, but focusing on the D5nf makes more sense as it's the least expensive driver and has the best chances of being purchased by the DIYers.
To date, I'm only aware of 4 enclosures for the D5nf... they are:
1- Moose BVR (aka Freddie Chang and there are two versions)
2- Ron Thackery's baby-TL (which I discount as a true TL - more below)
3- Feastrex 30-liter dual-ported enclosure
4- Feastrex 50-liter ported enclosure (two versions that I know of)
As of now, I'm using the BVR enclosure. I like the looks... clean and simple. It has limited bass extension and is on the short side, albeit the latest version has the driver mounted higher which should improve things for normal (seated) listening position.
Ron's baby-TL is not a true TL in my view as the large cross-section of the spherical magnet assembly consumes such a large amount of the line cross-section, that you effectively end up with a small chamber feeding a short tapered line thru a very small cross-section. Note that this has nothing to do with how it sounds, but I don't think of it as a TL due to this aspect.
The Feastrex 30-liter enclosure has only been shown in some photos... to date I've not seen a single detailed drawing so being able to build one is not possible without guessing.
The Feastrex 50-liter enclosure is still a bit iffy for me... I have two versions which differ in the port size and location and they show as being for the D5e, not the D5nf. The details are minimal which means the builder gets to calculate the rest of the detail.
So, perhaps Feastrex (aka Clark on this part) could update the website and include some downloadable PDFs with driver specifications and some (more than one) recommended enclosures with sufficient detail to ease construction. The first two enclosures are not Feastrex designs, so perhaps either links to them or getting permission to make them available would work. For the 30- and 50-liter enclosures, there's no reason why they shouldn't make these available. Last, I would like to suggest that Feastrex include a PDF spec sheet and some recommended enclosures with each driver and a mounting template and gasket. It can't possibly cost much to do this.
Regards, KM
Again, I'll suggest that providing useful information for all drivers is key, but focusing on the D5nf makes more sense as it's the least expensive driver and has the best chances of being purchased by the DIYers.
To date, I'm only aware of 4 enclosures for the D5nf... they are:
1- Moose BVR (aka Freddie Chang and there are two versions)
2- Ron Thackery's baby-TL (which I discount as a true TL - more below)
3- Feastrex 30-liter dual-ported enclosure
4- Feastrex 50-liter ported enclosure (two versions that I know of)
As of now, I'm using the BVR enclosure. I like the looks... clean and simple. It has limited bass extension and is on the short side, albeit the latest version has the driver mounted higher which should improve things for normal (seated) listening position.
Ron's baby-TL is not a true TL in my view as the large cross-section of the spherical magnet assembly consumes such a large amount of the line cross-section, that you effectively end up with a small chamber feeding a short tapered line thru a very small cross-section. Note that this has nothing to do with how it sounds, but I don't think of it as a TL due to this aspect.
The Feastrex 30-liter enclosure has only been shown in some photos... to date I've not seen a single detailed drawing so being able to build one is not possible without guessing.
The Feastrex 50-liter enclosure is still a bit iffy for me... I have two versions which differ in the port size and location and they show as being for the D5e, not the D5nf. The details are minimal which means the builder gets to calculate the rest of the detail.
So, perhaps Feastrex (aka Clark on this part) could update the website and include some downloadable PDFs with driver specifications and some (more than one) recommended enclosures with sufficient detail to ease construction. The first two enclosures are not Feastrex designs, so perhaps either links to them or getting permission to make them available would work. For the 30- and 50-liter enclosures, there's no reason why they shouldn't make these available. Last, I would like to suggest that Feastrex include a PDF spec sheet and some recommended enclosures with each driver and a mounting template and gasket. It can't possibly cost much to do this.
Regards, KM
Wince. That is the problem with T/S parameters. Far too many people take them, and the basic mathematical alignments, as being gospel. They're not. They can get you in the ~ballpark, but as Dave & GM point out, like anything, you have to know how to use them, what they do & do not show you / provide. Goes hand-in hand with software: again, a very useful / powerful design tool, but you still need to know what you (and it) are doing.
...
maybe i'm too naive but i just can't believe what you post here, on a DIY forum, guys
on what basis should a potential buyer decide what kind of enclosure to build for feastrex drivers? are there any other known and reliable techniques apart from those we have been using for last decades that are based on T/S parameters or does someone expect that if you pay $$$$$$$$ you can built 5-6 boxes to see if the drivers can cope or not....
i wouldn't be so open about this if i don't know someone who exprienced a real nightmare with these in a top class TQWT box with helmholz resonators that sounds excellent both with lowthers (any) or AER or fostex in it, and if i didn't hear them mounted in one of the best (again TQWT) ready made cabinets that perfectly fits (again) lowthers and fostex. the results were not satisfactory and the drivers were sold, not only because of the performance but also because of such an attitude.
IMHO, before you sort the proper data, you should not be on a DIY forum.
maybe i'm too naive but i just can't believe what you post here, on a DIY forum, guys
on what basis should a potential buyer decide what kind of enclosure to build for feastrex drivers? are there any other known and reliable techniques apart from those we have been using for last decades that are based on T/S parameters or does someone expect that if you pay $$$$$$$$ you can built 5-6 boxes to see if the drivers can cope or not....
i wouldn't be so open about this if i don't know someone who exprienced a real nightmare with these in a top class TQWT box with helmholz resonators that sounds excellent both with lowthers (any) or AER or fostex in it, and if i didn't hear them mounted in one of the best (again TQWT) ready made cabinets that perfectly fits (again) lowthers and fostex. the results were not satisfactory and the drivers were sold, not only because of the performance but also because of such an attitude.
IMHO, before you sort the proper data, you should not be on a DIY forum.
In fairness, that will rather depend on the driver. For e.g. the D5nf & D9nf have rather different parameters to most Fostex / Lowther / AER units, according to the measures Chris has taken. Ergo, like any other driver, they would be unlikely to work especially well in a cabinet designed for something with different requirements. That huge motor housing won't do them many favours in a QW based cabinet (say, a tapped, mass-loaded conical horn) either, as it will interfere with the behaviour of the standing waves present in the line. And so on & so forth.
All of which makes me agree with your other point though -people do want the basic parameters as a guide for developing their own cabinets, and they would have done themselves a favour if they'd measured the drivers before, or had someone do it for them. I believe that's something they've learned the hard way though, so hopefully they'll be coming very soon. T/S parameters are not perfect, nor are the be-all & end all, and as GM, Dave, Clark, myself & a few others have indicated, like anything, you have to know how best to use them. But if you do, they'll get you in the general zone quite quickly, allowing you to make detail changes / modifications / tweaks in the physical domain. Cut, shut & measure has many advantages too, but as we all know, not everyone has the time, money or facilities to build numerous prototypes, so a combination is typically the most practicable solution.
All of which makes me agree with your other point though -people do want the basic parameters as a guide for developing their own cabinets, and they would have done themselves a favour if they'd measured the drivers before, or had someone do it for them. I believe that's something they've learned the hard way though, so hopefully they'll be coming very soon. T/S parameters are not perfect, nor are the be-all & end all, and as GM, Dave, Clark, myself & a few others have indicated, like anything, you have to know how best to use them. But if you do, they'll get you in the general zone quite quickly, allowing you to make detail changes / modifications / tweaks in the physical domain. Cut, shut & measure has many advantages too, but as we all know, not everyone has the time, money or facilities to build numerous prototypes, so a combination is typically the most practicable solution.
I don't understand this obsession with T/S parameters.
What did people use before Mr Thiele and Small published their paper in iirc 1961?
The same driver is different enclosures will sound different. The same driver in the same enclosure with different damping material will sound different. The same driver in the same enclsure with the same internal volume and damping but different size ratios will sound different, different enclosure materials and treatments sound different, you get my drift.
Sims are great (not just for speaker design) but they are not the end all of anything.
I build several enclosures a year. I have just built another enclosure for the (my fave) FE103 Fostex, I have done well over a dozen enclosures for this little speaker and am still cutting and trying and listening, they don't all sound different but many of them do.
As Clark said, you have to put saw to wood and build and experiment.
There is no substitute for experience.
What did people use before Mr Thiele and Small published their paper in iirc 1961?
The same driver is different enclosures will sound different. The same driver in the same enclosure with different damping material will sound different. The same driver in the same enclsure with the same internal volume and damping but different size ratios will sound different, different enclosure materials and treatments sound different, you get my drift.
Sims are great (not just for speaker design) but they are not the end all of anything.
I build several enclosures a year. I have just built another enclosure for the (my fave) FE103 Fostex, I have done well over a dozen enclosures for this little speaker and am still cutting and trying and listening, they don't all sound different but many of them do.
As Clark said, you have to put saw to wood and build and experiment.
There is no substitute for experience.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.