Eric,
We are in complete agreement. In a conventional magnetic circuit, there is an assymetry in flux induced inductance front and back of the gap. Reducing the inductance in front of the gap seems an illogical direction to go if you want symmetrical conditions of operation.
To all,
Crossover revisions have been made and the new results are posted to the Web site. Preliminary listening results are encouraging but I really feel the need to pause here and do more extensive listening tests.
There may now be a gap in design updates while I am evaluating, but I am still feeling motivated to work on this project and there are still lots of refinements and design extensions to implement.
Good designing and good building,
Mark
We are in complete agreement. In a conventional magnetic circuit, there is an assymetry in flux induced inductance front and back of the gap. Reducing the inductance in front of the gap seems an illogical direction to go if you want symmetrical conditions of operation.
To all,
Crossover revisions have been made and the new results are posted to the Web site. Preliminary listening results are encouraging but I really feel the need to pause here and do more extensive listening tests.
There may now be a gap in design updates while I am evaluating, but I am still feeling motivated to work on this project and there are still lots of refinements and design extensions to implement.
Good designing and good building,
Mark
Greetings All,
Well, the intensive listening tests were so successful they did not have to last very long. I took a pair of the modified buy out polypropylene woofers in .4 cubic foot sealed enclosures with the Vifa tweeter recessed and front loaded on the road this weekend and did listening sessions in various environments, with varied set ups, and associated equipment. The speakers were evaluated as stereo music loudspeakers and as main speakers in home theater set ups.
The general concensus was that the loudspeakers are neutral in sound quality with exceptional detail and imaging.
I specifically asked each evaluator for details about problems with the sound. Respondents indicated that the speakers were bass shy in the extreme bottom end and that they would benefit from a sub woofer. One respondent also stated that he thought there might be a slight problem with the extreme top end (a slight lacking of "edge") but that if there was a problem he adapted to it quickly. He stated that he could no longer hear whatever it was after the second track and couldn't hear whatever it was from then on.
I will add a photo of the loudspeakers in their unfinished MDF enclosures to my Web site very soon. I will also provide a description of the tweeter mounting. The acoustic performance of this combination is already posted to the Web site.
The one woofer, one tweeter combination is a stand alone success. That success allows me to then go on to the design and construction of an array knowing that the sound of the basic components will not be a problem. The design goals for the array are increased bass performance and higher linear range SPL while preserving the acoustical qualities of the simpler loudspeaker design.
I will also try different tweeters in the basic two-way design. Why mate a $28 tweeter to a $5 woofer if you don't have to?
I am also developing a copper plug design for the woofer. I am not doing this because it will cause any change in the performance, but just because it will look cool. If it does no harm, why not?
Good designing and good building,
Mark
Well, the intensive listening tests were so successful they did not have to last very long. I took a pair of the modified buy out polypropylene woofers in .4 cubic foot sealed enclosures with the Vifa tweeter recessed and front loaded on the road this weekend and did listening sessions in various environments, with varied set ups, and associated equipment. The speakers were evaluated as stereo music loudspeakers and as main speakers in home theater set ups.
The general concensus was that the loudspeakers are neutral in sound quality with exceptional detail and imaging.
I specifically asked each evaluator for details about problems with the sound. Respondents indicated that the speakers were bass shy in the extreme bottom end and that they would benefit from a sub woofer. One respondent also stated that he thought there might be a slight problem with the extreme top end (a slight lacking of "edge") but that if there was a problem he adapted to it quickly. He stated that he could no longer hear whatever it was after the second track and couldn't hear whatever it was from then on.
I will add a photo of the loudspeakers in their unfinished MDF enclosures to my Web site very soon. I will also provide a description of the tweeter mounting. The acoustic performance of this combination is already posted to the Web site.
The one woofer, one tweeter combination is a stand alone success. That success allows me to then go on to the design and construction of an array knowing that the sound of the basic components will not be a problem. The design goals for the array are increased bass performance and higher linear range SPL while preserving the acoustical qualities of the simpler loudspeaker design.
I will also try different tweeters in the basic two-way design. Why mate a $28 tweeter to a $5 woofer if you don't have to?
I am also developing a copper plug design for the woofer. I am not doing this because it will cause any change in the performance, but just because it will look cool. If it does no harm, why not?
Good designing and good building,
Mark
Just an FYI update.
I have published crossovers for the 5.25 Poly woofer and two new tweeters.
You can find the information at:
http://madspeaker.com/Projects/projectimages/CAS525/PionandTBtweet/299245Pioneertweet.htm
and
http://madspeaker.com/Projects/projectimages/CAS525/PionandTBtweet/299145TBtweet.htm
Good designing and good building,
Mark
I have published crossovers for the 5.25 Poly woofer and two new tweeters.
You can find the information at:
http://madspeaker.com/Projects/projectimages/CAS525/PionandTBtweet/299245Pioneertweet.htm
and
http://madspeaker.com/Projects/projectimages/CAS525/PionandTBtweet/299145TBtweet.htm
Good designing and good building,
Mark
Small diameter drivers appear to have a problem with low frequency acoustic reproduction independent of T/S characteristics. When Fs is low in frequency, the acoustic output begins to fall off at frequencies higher than predicted. A logical question sourced from this observation is whether multiple drivers closely spaced will acoustically couple and act as if they were a single driver larger in diameter.
By the experiment I devised and conducted, it appears the answer is yes. When closely spaced, small diameter drivers can be made to act as if they were larger in diameter and come closer to the low frequency acoustic reproduction predicted by T/S analysis.
It has not been easy to control for variables in the test execution. Without control of the variables other than acoustical coupling of drivers, experimental conclusions would be suspect. Still, with the construction of specialized test jigs and careful implementation, I believe that the results I post here accurately reflect the performance difference due to acoustical coupling of the two closely spaced drivers.
There is a measured increase in low frequency output such that the combined response comes closer to T/S derived expectations for this driver. In the case of this 5.25 inch advertised diameter driver with Fs in the mid 60s, its performance was already close and simply coupling two drivers achieves maximum response. Adding more drivers to the closely spaced array is unlikely to provide additional performance improvements.
Like all design, there are both positives and negatives to the technique of acoustically coupling multiple drivers. There are indications in these tests of both additional benefits and liabilities caused by using even two drivers in an array. Positively, there is smoothing of individual differences in driver performance. Negatively, there is evidence of accentuation of problems shared by both drivers, even at relatively low frequencies. This is particularly evident in the response just under 200 Hz where the bump and then dip in response is greater than either driver alone. With just two drivers, none of the negative consequences are critical. Whether that will hold true when using more than two drivers in an array will require further testing.
Good designing and good building,
Mark
By the experiment I devised and conducted, it appears the answer is yes. When closely spaced, small diameter drivers can be made to act as if they were larger in diameter and come closer to the low frequency acoustic reproduction predicted by T/S analysis.
It has not been easy to control for variables in the test execution. Without control of the variables other than acoustical coupling of drivers, experimental conclusions would be suspect. Still, with the construction of specialized test jigs and careful implementation, I believe that the results I post here accurately reflect the performance difference due to acoustical coupling of the two closely spaced drivers.
There is a measured increase in low frequency output such that the combined response comes closer to T/S derived expectations for this driver. In the case of this 5.25 inch advertised diameter driver with Fs in the mid 60s, its performance was already close and simply coupling two drivers achieves maximum response. Adding more drivers to the closely spaced array is unlikely to provide additional performance improvements.
Like all design, there are both positives and negatives to the technique of acoustically coupling multiple drivers. There are indications in these tests of both additional benefits and liabilities caused by using even two drivers in an array. Positively, there is smoothing of individual differences in driver performance. Negatively, there is evidence of accentuation of problems shared by both drivers, even at relatively low frequencies. This is particularly evident in the response just under 200 Hz where the bump and then dip in response is greater than either driver alone. With just two drivers, none of the negative consequences are critical. Whether that will hold true when using more than two drivers in an array will require further testing.
Good designing and good building,
Mark
Attachments
Just wanted to post an update about early designs using this driver. First, PE still has them listed on their Web page as in stock.
Second, I have quite a few hours now listening to the 299-145 and the Vifa tweeter. I still like them very very much. In the simple sealed enclosure I used with the Vifa tweeter you can push them right up against a wall and not lose any imaging capability. They are great for small rooms. They are great for bookshelf speakers.
And now the real reason for my posting here. I have been playing around in the full-range forum recently. I have been trying to work with the Fostex FE126 and 127 in a full-range system. The price of the Foxtexs is similar to what you would pay for a 299-145 and a Vifa tweeter combination. The enclosure for the 299-145 and Vifa is about the same internal volume and no more difficult to make. The sound quality is not similar at all. I do not know how to place a numerical value on relative sound quality, but the 299-145 Vifa combination is easily somewhere between 10 to 100 times better than the FE126 or 127. Unlike the Fostex, this speaker has something close to real bass. It does not need BSC. It is clear and detailed across the entire spectrum. It does not sound compressed at any frequency.
We have an excellent library where I live. The last couple of days I have been listening to an Eric Clapton collection. It is so easy to hear the changes as recording capabilities improved and different studios were used. The early Yardbird stuff is very distorted. By the late 1960s and Cream, the vocals have become very good. I think maybe better than they will be later. One of the amazing changes was how much natural ambience was present on vocals during the late 60s. By the late 70s, vocals have really dried.
And it is not just vocals. On the Cream selections, I can just as easily and clearly hear Jack Bruce's bass lines as Clapton's lead.
If you want a small speaker to use as a reference, then this is it. If you want an inexpensive loudspeaker to just listen to for the sheer joy of listening to music, then this is it. If you want to bring the price down 10 USD, then use the Tang Band tweeter. Almost as good.
Good designing and good building,
Mark
Second, I have quite a few hours now listening to the 299-145 and the Vifa tweeter. I still like them very very much. In the simple sealed enclosure I used with the Vifa tweeter you can push them right up against a wall and not lose any imaging capability. They are great for small rooms. They are great for bookshelf speakers.
And now the real reason for my posting here. I have been playing around in the full-range forum recently. I have been trying to work with the Fostex FE126 and 127 in a full-range system. The price of the Foxtexs is similar to what you would pay for a 299-145 and a Vifa tweeter combination. The enclosure for the 299-145 and Vifa is about the same internal volume and no more difficult to make. The sound quality is not similar at all. I do not know how to place a numerical value on relative sound quality, but the 299-145 Vifa combination is easily somewhere between 10 to 100 times better than the FE126 or 127. Unlike the Fostex, this speaker has something close to real bass. It does not need BSC. It is clear and detailed across the entire spectrum. It does not sound compressed at any frequency.
We have an excellent library where I live. The last couple of days I have been listening to an Eric Clapton collection. It is so easy to hear the changes as recording capabilities improved and different studios were used. The early Yardbird stuff is very distorted. By the late 1960s and Cream, the vocals have become very good. I think maybe better than they will be later. One of the amazing changes was how much natural ambience was present on vocals during the late 60s. By the late 70s, vocals have really dried.
And it is not just vocals. On the Cream selections, I can just as easily and clearly hear Jack Bruce's bass lines as Clapton's lead.
If you want a small speaker to use as a reference, then this is it. If you want an inexpensive loudspeaker to just listen to for the sheer joy of listening to music, then this is it. If you want to bring the price down 10 USD, then use the Tang Band tweeter. Almost as good.
Good designing and good building,
Mark
Hi Mark,
Thank you for all the great work with the 299-145 driver. I bought 28 of them a couple of months ago with the intention of building some cheapo arrays to play with. I also made sure the PE person on the phone associated your name with the order.
I have three questions:
1) A 7/8" dowel makes for a very loose fit plug. There's quite a gap. Is this O.K?
2) Are you still planning on publishing the DIY copper phase plug design?
and
3) How about a dimple mod kit to replace the glue bands? It would be a lower mass solution. PE seems to have an endless supply of these drivers so there could be a good demand for mod kits.
Thanks again.
Graeme
Thank you for all the great work with the 299-145 driver. I bought 28 of them a couple of months ago with the intention of building some cheapo arrays to play with. I also made sure the PE person on the phone associated your name with the order.
I have three questions:
1) A 7/8" dowel makes for a very loose fit plug. There's quite a gap. Is this O.K?
2) Are you still planning on publishing the DIY copper phase plug design?
and
3) How about a dimple mod kit to replace the glue bands? It would be a lower mass solution. PE seems to have an endless supply of these drivers so there could be a good demand for mod kits.
Thanks again.
Graeme
Thanks and some ??
My thanks too, Mark, for all your work with these. I just got 14 of these (used the link on your website). I am doing a HT setup and thinking of using 2 per speaker, maybe dipole or whatever, I'm making a bunch of protos to check out.
But to my questions - you say the phase plugs should be 19mm long. I have pulled the dust caps off 4 and the pole piece to end of vc is 20 mm on each one. All the phase plugs I've ever seen project past the vc. So I'm wondering if 19mm is right or if my samples are different - ?
The other question concerns the xo and tweeter. It looks like the crossover is around 14000 Hz. That's about the top of my range (too much Hendrix and Stones back in the 60s) so I'm wondering if I even need a tweeter. If so, I'm looking at the PE 279-234 - another buyout. I know I prefer textile domes, but like you said, a tweeter that costs 5 times as much as the woofer seems like over kill.
I'm looking forward to your next steps with this driver.
Thanks again for all of us who didn't get the engineer gene and can't afford all that testing equipment.
My thanks too, Mark, for all your work with these. I just got 14 of these (used the link on your website). I am doing a HT setup and thinking of using 2 per speaker, maybe dipole or whatever, I'm making a bunch of protos to check out.
But to my questions - you say the phase plugs should be 19mm long. I have pulled the dust caps off 4 and the pole piece to end of vc is 20 mm on each one. All the phase plugs I've ever seen project past the vc. So I'm wondering if 19mm is right or if my samples are different - ?
The other question concerns the xo and tweeter. It looks like the crossover is around 14000 Hz. That's about the top of my range (too much Hendrix and Stones back in the 60s) so I'm wondering if I even need a tweeter. If so, I'm looking at the PE 279-234 - another buyout. I know I prefer textile domes, but like you said, a tweeter that costs 5 times as much as the woofer seems like over kill.
I'm looking forward to your next steps with this driver.
Thanks again for all of us who didn't get the engineer gene and can't afford all that testing equipment.
An offering of copper plugs and a Whisper Kit for the cone are both possible. While PE has not run out of the drivers, they will eventually. Didn't seem to make too much sense to develop a commerical product for a closeout. Dimpling the cones is also much simpler. Doesn't take many dimples. While I have both, most of my listening is done to the glue ring set. Removing the mass of the glue rings makes no improvement in performance. Other factors are dominating by this point in the modification. I will think about a Whisper Kit for the drivers over the next couple of weeks. I will get back to you.
There may also be one more modification in the works for this driver. I have been doing some basic research into surrounds and I have been having some interesting results. That information may apply to this driver. Time will tell (when I find the time to apply it to this driver).
As to other questions. . . The larger the diameter the cone, the less important is the phase plug. It really just fills the void in the center of the voice coil. If it is not filled, the air mass in it resonates and interferes with the desired output. The plug can be just slightly recessed or slightly long. Neither is a problem to the ear or to the test equipment. Again with diameter. A little off is no problem. A lot off is. 95% of the improvement is getting rid of the dust cap and then displacing the air filling the voice coil tube.
Also, you might try listening to the modified driver without a tweeter. The woofer alone will handle all musical fundamentals. The tweeter really just adds a little air. Synthesized effects do benefit from the extra half octave. I can't really testify to how hearing loss may impact one's ability to hear the things a tweeter does above 13 or 14 kHz.
In sum, compared to all the other closeouts deals posted to this forum in the past, this one is the real deal.
Good designing and good building,
Mark
There may also be one more modification in the works for this driver. I have been doing some basic research into surrounds and I have been having some interesting results. That information may apply to this driver. Time will tell (when I find the time to apply it to this driver).
As to other questions. . . The larger the diameter the cone, the less important is the phase plug. It really just fills the void in the center of the voice coil. If it is not filled, the air mass in it resonates and interferes with the desired output. The plug can be just slightly recessed or slightly long. Neither is a problem to the ear or to the test equipment. Again with diameter. A little off is no problem. A lot off is. 95% of the improvement is getting rid of the dust cap and then displacing the air filling the voice coil tube.
Also, you might try listening to the modified driver without a tweeter. The woofer alone will handle all musical fundamentals. The tweeter really just adds a little air. Synthesized effects do benefit from the extra half octave. I can't really testify to how hearing loss may impact one's ability to hear the things a tweeter does above 13 or 14 kHz.
In sum, compared to all the other closeouts deals posted to this forum in the past, this one is the real deal.
Good designing and good building,
Mark
Mark, thanks for clearing that up. I'm going to try pairs of these in several different baffle/box configurations - MMs, I guess.
gl, if you really like the copper look, you might try the local hardware store. Get some end caps for 3/4" copper tubing. I picked up one to check it out and it measured 25 mm OD x 23 mm H. Buff them up and spray with clear lacquer. Don't know if being hollow might cause problems, guess you could stuff it with something.
The ski lift closes the 27th
Might get more done though 
MB
gl, if you really like the copper look, you might try the local hardware store. Get some end caps for 3/4" copper tubing. I picked up one to check it out and it measured 25 mm OD x 23 mm H. Buff them up and spray with clear lacquer. Don't know if being hollow might cause problems, guess you could stuff it with something.
The ski lift closes the 27th
Might get more done though MB
MB,
Great idea on the copper end caps.
gl,
One word of caution, the open ends of the caps are not terminated well. These are all imported from China now and because it is not necessary to their intended use, they do not control length or eveness of the open end. In plumbing and with solder connections this is just not important.
If you are trying to mount one of these inside a tube with close tolerances, the end must be even enough for the the long axis of the cap to be perpendicular to the top of the pole piece. If you have access to a metal lathe, you can turn the end caps to identical lengths and have them be centered and straight in the voice coil.
The narrow lip on the open end does not give you much area to mount the cap to the pole piece. Other than that, being hollow is not a problem. A filled or solid unit with a flat and perpendicular bottom is much easier to work with in this application.
Good designing and good building,
Great idea on the copper end caps.
gl,
One word of caution, the open ends of the caps are not terminated well. These are all imported from China now and because it is not necessary to their intended use, they do not control length or eveness of the open end. In plumbing and with solder connections this is just not important.
If you are trying to mount one of these inside a tube with close tolerances, the end must be even enough for the the long axis of the cap to be perpendicular to the top of the pole piece. If you have access to a metal lathe, you can turn the end caps to identical lengths and have them be centered and straight in the voice coil.
The narrow lip on the open end does not give you much area to mount the cap to the pole piece. Other than that, being hollow is not a problem. A filled or solid unit with a flat and perpendicular bottom is much easier to work with in this application.
Good designing and good building,
Thank you to both Mark and MB for your responses. I really wasn't wanting the copper phase plugs for their looks but for higher performance. For looks I had intended to paint the wood plugs black as a matter of fact. The thing that would improve the looks is having a bullet shape. However, Mark I know, has indicated that this is not important for this driver. It's just that one or two of the madspeaker projects seem to have had some very detailed work done on producing a very precisely shaped phase plug. It's difficult for me to reconcile why so much exacting engineering is valuable for one driver and not for another. I do realize that the diameter of the driver plays a major role in the effectiveness of a phase plug.
Would I loose anything by making the phase plugs bullet shaped and more orthodox looking?
Thanks much,
Graeme
Would I loose anything by making the phase plugs bullet shaped and more orthodox looking?
Thanks much,
Graeme
Graeme,
Thanks for your reply and question. I have addressed this phase plug shape before and one respondent just about took my head off.
As I speak about in the AudioXpress article, the plug isn't about phase. It is about loading the diaphragm. The larger the cone the less impact the plug has. With a two-inch cone, the shape above the front edge of the voice coil is very important. For five-inch and large cones, within certain constraints, it just doesn't matter anymore. Do you remember any of your high school biology? Why is there a limit on the size of a single celled organism? Both area and volume to increasing diameter are non linear relationships.
If, however, you flare out the plug so that it looks like a mushroom, it can still load the cone. This also sets up a reflecting surface within the cone that is never a good thing. It actually becomes a phase plug and does some really bad things.
Somewhere, whether it is on my Web site or elsewhere, I state that copper is used merely for the appearance. I also put a little bevel on the plug so that the tip ends with a simple cone shape. Again, purely for appearance.
Sorry for taking so long to get to your question, but I tend to think about things in complex ways. Anyway, as long as the shape tapers in to a diameter less than the ID of the voice coil, feel free to make it bullet shaped. You will do no harm. Okay, again thinking complexly, almost no harm. Extreme off-axis response will be impacted by having the plug extend beyond the top edge of the voice coil. But it is not very much and unless you orient your speakers in an extremely odd way you will never do any serious listening this far off-axis. So, again, feel free to make the plug bullet shaped and extend it above the top of the voice coil.
I am also glad you clarified your earlier message. I will feel no guilt about not making a Whisper Kit for this driver.
Good designing and good building,
Mark
Thanks for your reply and question. I have addressed this phase plug shape before and one respondent just about took my head off.
As I speak about in the AudioXpress article, the plug isn't about phase. It is about loading the diaphragm. The larger the cone the less impact the plug has. With a two-inch cone, the shape above the front edge of the voice coil is very important. For five-inch and large cones, within certain constraints, it just doesn't matter anymore. Do you remember any of your high school biology? Why is there a limit on the size of a single celled organism? Both area and volume to increasing diameter are non linear relationships.
If, however, you flare out the plug so that it looks like a mushroom, it can still load the cone. This also sets up a reflecting surface within the cone that is never a good thing. It actually becomes a phase plug and does some really bad things.
Somewhere, whether it is on my Web site or elsewhere, I state that copper is used merely for the appearance. I also put a little bevel on the plug so that the tip ends with a simple cone shape. Again, purely for appearance.
Sorry for taking so long to get to your question, but I tend to think about things in complex ways. Anyway, as long as the shape tapers in to a diameter less than the ID of the voice coil, feel free to make it bullet shaped. You will do no harm. Okay, again thinking complexly, almost no harm. Extreme off-axis response will be impacted by having the plug extend beyond the top edge of the voice coil. But it is not very much and unless you orient your speakers in an extremely odd way you will never do any serious listening this far off-axis. So, again, feel free to make the plug bullet shaped and extend it above the top of the voice coil.
I am also glad you clarified your earlier message. I will feel no guilt about not making a Whisper Kit for this driver.
Good designing and good building,
Mark
Just a quick update on the continuing work on this driver. Now, the the glue ring and glue line modification remains valid. This new work, even when completed, will not obsolete the glue ring/line design.
I wanted to look at two things in this new design. One was the influence of the surround and the second was eliminating the parallel notch filter. I want the driver to not need any equalization.
I am getting close, but am not quite there yet. Most of the hump the notch filter was equalizing has been controlled through cone modification. Not a complete success yet because a little peak at 5 kHz remains and the driver still has a slightly rising response. The rising frequency response can be equalized with a .5 mH inductor. This inductor does reduce high frequency extension slightly, but still allows for a bandwidth to 10 kHz.
The driver rolls off at 6th order acoustic rate. It is generally smooth and does not require additional filtering.
A three frequency response graph of the new design is posted below.
Good designing and good building,
Mark
I wanted to look at two things in this new design. One was the influence of the surround and the second was eliminating the parallel notch filter. I want the driver to not need any equalization.
I am getting close, but am not quite there yet. Most of the hump the notch filter was equalizing has been controlled through cone modification. Not a complete success yet because a little peak at 5 kHz remains and the driver still has a slightly rising response. The rising frequency response can be equalized with a .5 mH inductor. This inductor does reduce high frequency extension slightly, but still allows for a bandwidth to 10 kHz.
The driver rolls off at 6th order acoustic rate. It is generally smooth and does not require additional filtering.
A three frequency response graph of the new design is posted below.
Good designing and good building,
Mark
Attachments
Just for fun I thought I would try to present the performance of the new modification by the testing standards of other people. For this posting I used a company generally well thought of as my reference.
I have included frequency response and phase response of the 5.25 mod 3. There is no smoothing of the frequency response. The loss of detail from my typical postings is all done by setting boundaries on the data set for the FFT. No limits were, however, used for the phase response. Vertical resolution (both for magnitude and for phase angle) is identical to that used by this other company.
This other company does not make a 5.25 inch driver. They do make a four-inch driver. While their four-inch driver has an extra octave on the top end, the 5.25 inch driver has an extra octave on the bottom end.
The retail price of the 5.25 inch driver is 12 USD. The other driver costs about 12 times as much. I am certain that that driver is much better than the modified 5.25 driver, but using these testing standards it is not easy to tell that it is.
Good designing and good building,
Mark
I have included frequency response and phase response of the 5.25 mod 3. There is no smoothing of the frequency response. The loss of detail from my typical postings is all done by setting boundaries on the data set for the FFT. No limits were, however, used for the phase response. Vertical resolution (both for magnitude and for phase angle) is identical to that used by this other company.
This other company does not make a 5.25 inch driver. They do make a four-inch driver. While their four-inch driver has an extra octave on the top end, the 5.25 inch driver has an extra octave on the bottom end.
The retail price of the 5.25 inch driver is 12 USD. The other driver costs about 12 times as much. I am certain that that driver is much better than the modified 5.25 driver, but using these testing standards it is not easy to tell that it is.
Good designing and good building,
Mark
Attachments
Keep up the good work, Mark. Some of us need all the help we can get.
On the end caps, a couple of seconds on the belt sander will give you a nice flat bottom and take care of irregularities in length.
I'm glad to see you are continuing to work with these speakers. I'm just starting, but I have several in some test baffles and they sound pretty good.
I don't know what to think about your trying to eliminate the PNF, I'm all for less stuff in the signal path, but then I've already bought all the parts to build 6 of them. Extra crossover parts, I guess.
MB
On the end caps, a couple of seconds on the belt sander will give you a nice flat bottom and take care of irregularities in length.
I'm glad to see you are continuing to work with these speakers. I'm just starting, but I have several in some test baffles and they sound pretty good.
I don't know what to think about your trying to eliminate the PNF, I'm all for less stuff in the signal path, but then I've already bought all the parts to build 6 of them. Extra crossover parts, I guess.
MB
MB,
While I build them, listen to them, and share them with friends, the redesigns I post are simply my design exercises. Even if I am able to meet my standards for driver performance without signal equalization with this driver, it does not mean that the older design is not viable. Indeed, it is very viable.
Also, I am in awe of your talent with a belt sander. You are far beyond my meager skills to flatten and square the bottoms (open ends) of copper cap fittings.
Best,
Mark
While I build them, listen to them, and share them with friends, the redesigns I post are simply my design exercises. Even if I am able to meet my standards for driver performance without signal equalization with this driver, it does not mean that the older design is not viable. Indeed, it is very viable.
Also, I am in awe of your talent with a belt sander. You are far beyond my meager skills to flatten and square the bottoms (open ends) of copper cap fittings.
Best,
Mark
Mark,
All this recent discussion reminds me of a question or two I've had for while.
I am building line arrays with the 299-145 drivers and intend to cross over at 2000 to 2500 Hz to pair of DIY line source ribbons. I will cross over temporarily on the bottom to a pair of elderly Tympany 1D's. The purpose of the whole effort is to provide a test bed to experiment with the ribbon tweeters.
O.K. On your web site you present graphs of the unmodified 299-145 and final graphs of the driver with all the mods. I found myself wishing that you had run a graph for each mod as it added. In particular I am curious to know what the gaphs look like for the driver with the pahase plug mod and the glue ring mod but without the equalizer.
I am doing the physical mods and will be building a low level 4th order LR xover to operate in the 2000 to 2500Hz range. If your equalizer is used to mitigate a 5000Hz peak then it looks like I might be pushing things and may have to add a notch filter. Any advice?
Thanks in advance.
Graeme
P.S. MB - I bought a couple of those copper plumbing caps. They work very nicely indeed. Thank you for the great idea.
All this recent discussion reminds me of a question or two I've had for while.
I am building line arrays with the 299-145 drivers and intend to cross over at 2000 to 2500 Hz to pair of DIY line source ribbons. I will cross over temporarily on the bottom to a pair of elderly Tympany 1D's. The purpose of the whole effort is to provide a test bed to experiment with the ribbon tweeters.
O.K. On your web site you present graphs of the unmodified 299-145 and final graphs of the driver with all the mods. I found myself wishing that you had run a graph for each mod as it added. In particular I am curious to know what the gaphs look like for the driver with the pahase plug mod and the glue ring mod but without the equalizer.
I am doing the physical mods and will be building a low level 4th order LR xover to operate in the 2000 to 2500Hz range. If your equalizer is used to mitigate a 5000Hz peak then it looks like I might be pushing things and may have to add a notch filter. Any advice?
Thanks in advance.
Graeme
P.S. MB - I bought a couple of those copper plumbing caps. They work very nicely indeed. Thank you for the great idea.
Mark, I guess those years as a woodshop teacher got me familar with the belt sander - BTW, the sander is stationary, I hold the caps - you could do as well by gluing some 100C sandpaper to a board and working that way, just slower.
Graeme, glad you liked the idea and it worked for you.
BUT, on to today's adventure. I had 2 identical baffles set up, so I thought I would compare the modified version to the original. Not that I doubt that Mark has improved them or doubt his neasurements, I just wanted to see if I could hear the differance. So I hooked them up and put in my mono test CD. I thought I hadn't connected the modified version. It seemed all the sound was coming from the unmodified side! I got closer and realized that the modified side was playing after all, and sounded like it always had, it was just that the unmodified side was so much louder. I took the PNF out of the modified side and the sound was pretty balanced then. I'm wondering if I somehow assembled the PNF incorrectly, but how could you do that? Is it normal to lose 3-4 db with a PNF?
Anyway, I found that the modified version had a much better vocals, brass, strings and guitar. The unmodified seemed to have slightly more of the lower bass, but mids where not as clear as the mods.
I'm using 2 of them per baffle, eventually to be a MTM if I ever find a tweeter. Another odd thing was that I could tell absolutely no differance in the sound if I turn them with the drivers aligned vertically or horizontaly, a big no-no, from what I read, but I tend to believe what I hear more.
MB
Graeme, glad you liked the idea and it worked for you.
BUT, on to today's adventure. I had 2 identical baffles set up, so I thought I would compare the modified version to the original. Not that I doubt that Mark has improved them or doubt his neasurements, I just wanted to see if I could hear the differance. So I hooked them up and put in my mono test CD. I thought I hadn't connected the modified version. It seemed all the sound was coming from the unmodified side! I got closer and realized that the modified side was playing after all, and sounded like it always had, it was just that the unmodified side was so much louder. I took the PNF out of the modified side and the sound was pretty balanced then. I'm wondering if I somehow assembled the PNF incorrectly, but how could you do that? Is it normal to lose 3-4 db with a PNF?
Anyway, I found that the modified version had a much better vocals, brass, strings and guitar. The unmodified seemed to have slightly more of the lower bass, but mids where not as clear as the mods.
I'm using 2 of them per baffle, eventually to be a MTM if I ever find a tweeter. Another odd thing was that I could tell absolutely no differance in the sound if I turn them with the drivers aligned vertically or horizontaly, a big no-no, from what I read, but I tend to believe what I hear more.
MB
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.