[JoshK]

900hz wouldn't work here - B&C DE 250 is specified as 1k-18k with recommended crossover freq of 1.6k

you would need a bass driver capable of going much higher - quite a difficult task - something like 20hz-2k or so.....

Sorry but you are being too literal with the specs.  The driver is rated for pro use, as in 120db.  For home use, you most definitely can drive these down to 800hz.  The poly phragm has a big factor in this too.  Geddes said he crosses around 800hz now.  I don't know if he used to cross at 900hz before.

[anubisgrau]

Sorry but you are being too literal with the specs.  The driver is rated for pro use, as in 120db.  For home use, you most definitely can drive these down to 800hz.  The poly phragm has a big factor in this too.  Geddes said he crosses around 800hz now.  I don't know if he used to cross at 900hz before.

interesting post. i've always thought that the measurements are an exact data which is irrelevant of the way of use - pro or home. how can a driver specified for particular linearity over certain frequency range achive a desired linearity outside of the specified range? can you really improvise here? am i missing something?

[DanTheMan]

interesting post. i've always thought that the measurements are an exact data which is irrelevant of the way of use - pro or home. how can a driver specified for particular linearity over certain frequency range achive a desired linearity outside of the specified range? can you really improvise here? am i missing something?

Measure it in your particular horn and see what it does.  If that's what you mean by improvising than yes it can be and should be, no must be done.

Got any measurements?

Dan 

[JoshK]

interesting post. i've always thought that the measurements are an exact data which is irrelevant of the way of use - pro or home. how can a driver specified for particular linearity over certain frequency range achive a desired linearity outside of the specified range? can you really improvise here? am i missing something?

The manufacturer's recommendation are based on safe operating of the driver, not based on distortion specs.  At 120db rms @ 900hz you run out of xmax and damage is going to be done to the driver.  But at 95db rms (really freaking loud in home use) the driver is loafing. 

Lots of drivers (CDs) are also limited by their resonance frequency which happens to be pretty low for a polymide 'phragm, not always the case with metal 'phragms.

Distortion measurements aren't always linear with increasing SPL either.  Measure the driver xo'd at 90hz @ 85db and the distortion will be minute, measure again at 110db and the same isn't true and it won't just be a parrallel shift up. 

[anubisgrau]

I've had a look into graphs provided by B&C and yes, this driver can be crossed that low, though it's already fast dropping bellow 1k.

Now that I've bought these waveguides, I will have a chance to hear what is all the rave about - especially that I have another pair of possibly one of the better commercially available horns (Oris Swing).

As I have never heard a Geddes speaker, I still can't imagine that such a high crossover frequency (and placed in the most critical range) between a direct radiator and a WG can be without any  impact. I would expect this to be audible - at least around the crossiver frequency, like 600-1200hz range....

Finally - as there are complaints about lack of HF extension of DE250 in OSWG (and that's clear from the Geddes Summa graphs), are there any reasons why a good supertweeter wouldn't be used, like RAAL or similar (except for the price, of course;))

Thanks

[DanTheMan]

If the directivity matches, I'd bet you can't hear it.  I know I can't.  Prior to my current speaker, I've heard every crossover w/ the exception being fullrangers.

Dan

[anubisgrau]

If the directivity matches, I'd bet you can't hear it.  I know I can't.  Prior to my current speaker, I've heard every crossover w/ the exception being fullrangers.

Dan

thanks - do you have summa or something else...?

[DanTheMan]

Something else.   Homemade jobbies, but match criteria fairly well.

This is not a simple project.  Good luck.  There will be plenty to help along the way.

Dan

[anubisgrau]

Something else.   Homemade jobbies, but match criteria fairly well.

This is not a simple project.  Good luck.  There will be plenty to help along the way.

not sure why a summa clone would be a project too difficult - with geddes crossover schematics and his guidewaves.

it's altogether a different story if you want to make something different. for example, i would maybe rather go for a vented than a sealed box. i don't want to make a HT system with subs and would be OK if i can reach -3db 40-45hz response with the room help. but i don't think this changes a basic summa concept - to have a bass driver that matches a waveguide directivity at a crossover frequency.

i'm wondering what data exactly shows that a bass driver has a 90 degrees directivity at 900hz? i may be interested in trying some other woofers that looks better in other areas than B&C, most notably like BD15 or a beyma 115ND/W.... 

[DanTheMan]

Oh, didn't realize you had the correct schematic.   I think I'm mixing up threads from other sites.  It shouldn't be too bad then.  Your probably best off just sticking to the original design but altering the box to ported doesn't seem too complicated if you have the capability to do measure impedance.  There's plenty of free box modeling software.  I guess it depends on how complicated you want to make it.  Memory tells me the Summa only gets down to 80Hz on the low end.  My own experience says that's too high w/o subs to be satisfied with the bass.

Have you looked into AE speakers?  They have some reasonable options with somewhat known off axis performance.  Here's some 15" woofers measured at 0,20, and 40 degrees: http://sites.google.com/site/drivervault/driver-measurements/15

Too bad he didn't go further.  The real problems I'v had are roughly 45-90 degrees--and that's been universal so far.  I don't know that AE would do any better in that department.

Dan

[anubisgrau]

just looked at the 15TBX100 parameters - it has a far too high MMs for my taste (over 160g)... i would much prefer to have this not much bigger than 100g (like beyma 15B100R with 110g http://www.stereo-lab.de/EN/media/docs/15b100r.pdf)....

anyone experimented with different 15" woofers and 15" geddes wgs?

[nullspace]

i'm wondering what data exactly shows that a bass driver has a 90 degrees directivity at 900hz? i may be interested in trying some other woofers that looks better in other areas than B&C, most notably like BD15 or a beyma 115ND/W....

To know any particular driver's directivity @ a specific frequency requires on- and off-axis measurments of that driver. Generally speaking, though, I don't think it varies a great deal from driver to driver of a specific size. Most any 15" driver will be in the ballpark of 90deg. @900Hz.

Also, I haven't seen any graphs for just the 15" Geddes waveguide + B&C DE250 combination, but as horns/waveguides appraoch the cutoff frequency the directivity starts to widen. Earl's waveguides are pretty big so they hold directivity pretty low, but I haven't seen any measurements that show exactly how low they hold 90degrees. Just something to keep in mind...

John

[anubisgrau]

no graphs - there's only earls' writing about the concept that he tunes a similar directivity from the bass driver and his horn at the crossing frequency. i just hope that all the praise you can read about the concept comes from the people who heard enough of really good speakers in their life.

after some search, i think that AE TD15M looks like a good candidate to me, it's extremely low mass cone of 70g. anyone used it?

 

[anubisgrau]

ok, some random thoughts while awaiting for the delivery of the geddes 15" OSWGs.

do any of you, who had a chance to hear summas in better surrounding than geddes' panasonic AVRs, think that the bass driver choice (B&C 15TBX100) might be a weak point of the design. i was looking into critical parameters and a number of things there is really not up to my taste - i'm appalled with the cone mass of over 160g for a driver aimed to play so high. can such a driver really match speed of the sound coming from the waveguide? however i may accept that this is all irrelevant and i'm being too picky with data on paper while in reality it just works different in a given design - never heard summas so i can't have an opinion here.

my choices are now narrowed either to B&C or other european manufacturers of the similar level (beyma) or to more expensive contenders: BD15 by a dutch BD-Design or AE TD15M with or without Apollo mod.

one of the main reasons why i really like TD15M was extremely low Mms and L which makes it ideal for driving it with low powered amps and that's my goal - to drive it passive with a SET amp. BD15 has even lower Mms but L is not so low, although it's a nominal 16ohm driver (instead of 6ohm AE) which gives another opportunity to make an easy load - AFAIK DE250 is also a 16ohm version.

B&C would cost me around $500 a pair while BD15 will be at least $1200. apollo version of TD15M will be in the same range delivered to me - $1100-1200.

i'm really fine with paying more if there's a clear guarantee it will yield sonic benefit in given design - which is a big question for me.

[DanTheMan]

I haven't heard the Summa, but I think you'd be more likely to succeed if you don't mess with the design.  The other woofers may look good on paper, but you'll have to change the design.  It just doesn't seem like a good idea if you are rather new to this stuff.  If you know what you are doing then changing the woofer might be a good idea.  Whether or not the difference would be audible or even truly an audible improvement is questionable.

Dan

[nullspace]

Without having any first-hand knowledge, I would think that you could make an excellent speaker with the AE TD15M plus Earl's waveguide and B&C DE250.

However, you'll be back to square one in terms of the crossover. If you relish a substantial challenge, and we're in The Lab after all, it could be a very rewarding project. Of course, it could also be completely maddening, costing you time and money, and you might never finish the speakers.

I did a 12" midbass + horn/compression driver as my first multiway project, and the crossover between the midbass and horn took >1 year. I'm a bit slow at finishing projects, but an awful lot of that time was learning the basics. And while I'm a proponent of reading the classics on loudspeaker design and measurement, there are some things that you just need to learn first-hand.

John

[DanTheMan]

Without having any first-hand knowledge, I would think that you could make an excellent speaker with the AE TD15M plus Earl's waveguide and B&C DE250.

However, you'll be back to square one in terms of the crossover. If you relish a substantial challenge, and we're in The Lab after all, it could be a very rewarding project. Of course, it could also be completely maddening, costing you time and money, and you might never finish the speakers.

I did a 12" midbass + horn/compression driver as my first multiway project, and the crossover between the midbass and horn took >1 year. I'm a bit slow at finishing projects, but an awful lot of that time was learning the basics. And while I'm a proponent of reading the classics on loudspeaker design and measurement, there are some things that you just need to learn first-hand.

John

Excellent post John!  CD crossovers are beyond the basics. No book I know of covers them.  If it were me, I'd just stick to the original as much as I'd like to do something wild.

Dan

[anubisgrau]

it looks as this will be my scenario. B&C bass drivers for a safe solution and than experiments. luckily i may have a pair of BD15 on disposal for in-room tests if improvements are really audible with this design or not. unfortunately due to rather bad E/$ ratio and extorbitant shipping costs to europe (where i'm based) TD15M will not be an option unless i find them for a private sale at lower prices.

[JoshK]

The way to do it if you are serious about using a different woofer is to buy one of the B&C woofers, fit the stock crossover, measure.   This gives you the transfer function that Dr G is using in his crossover.  Now put in the other woofer in the box and measure its raw response.   Armed with the transfer function goal and the raw response you can now make up a new low pass part of the crossover assuming the sensitivities don't need adjusting (they probably do).  You will have to modifying the padding in the high pass for sensitivity adjustments. 


When you are done, you can sell the B&C woofer.   

[DanTheMan]

That's smart Josh.  I'd add to that, make sure you do both out to 90 degrees--not just one or a couple axis.  That cone break up will impact the off axis performance in currently unpredictable ways unless you have fancy modeling programs and accurate driver parameters including the geometrical stiffness.  Even then, the model will not be as good as just doing the work.

Dan