Seas W18EX Active Lowpass

[paulhfx]

I'm looking for suggestions for an active crossover lowpass for the Seas W18EX - something that does a very effective job of killing upper resonance when crossover frequency is approx 1800Hz. A basic 4th order isn't doing it well (no surprise).

I thought I'd post this as someone has likely done this before with notch filters or higher order filters? What sounded best?

Thanks,

Paul

 

[Paul W]

Paul,
If you are using something like a DCX2496, try using one of the variable-Q bandpass filters as a notch filter. That works well for me...though with a 600Hz LR4. 1800 may be a stretch for the Excels.
Paul

An externally hosted image should be here but it was not working when we last tested it.

 

[angel]

PaulW - what ribbons are those?

I'm guessing it's a dipole speaker, since the image is called "wings".

My GF actually liked their appearance a lot, which may be my only shot at getting to build a cool set of speakers in the near future

 

[paulhfx]

Paul,
If you are using something like a DCX2496, try using one of the variable-Q bandpass filters as a notch filter. That works well for me...though with a 600Hz LR4. 1800 may be a stretch for the Excels.
Paul

No, I should have mentioned - this is a diy analog filter. 1800Hz may be a slight stretch, and I'll likely try lower, but I've got a dome tweeter to crossover/protect also.

Paul

 

[catapult]

When in doubt, consult the filter bible.....

http://www.linkwitzlab.com/filters.htm#7

 

[Jonasz]

Wich tweeter are you using?

Jonas

 

[Henckel]

Using the DCX 2496 and the SEAS Excell 18WEX001- i am doing:

48dB/oct at 2,2 Khz and two notch filters - 15dB, Q= 10 at 4,88 Kh and respectivly 4,98KHz

My experinces is that it needs at lot to kill the break up resonace when you go actice.

As a consequences of this experiences i am now looking for a "tweeter" that than can do 1,5-1,8 KHz as cross over

 

[Paul W]

angel...ribbon is DIY. here's a link...

WTW ribbon

 

[Jonasz]

I would recommend the RS28 wich easily can go down that low. I'm designing an MTM with them right now as a matter of fact. The RS28 is basically a bettered SS 9800 aludome and is not supposed to have any of the zing found in older alu-domes.

Here's a link to the RS28: http://www.partsexpress.com/pe/showdetl.cfm?&DID=7&Partnumber=275-130

Here you can find loads of info about it: http://www.htguide.com/forum/forumdisplay.php4?f=6

Here's a subjective comparison between
the RS28 and the Neo 3: http://www.pesupport.com/cgi-bin/config.pl?read=234053


Other possibilities is oc Seas Milennium (at a higher cost), the Seas 27TDFC, SS 95- 97- 9900.

Jonas

An externally hosted image should be here but it was not working when we last tested it.

 

[paulhfx]

I'm using a Scan Speak 9700, and planning to use a 4th order (electrical) highpass with it.

Paul

 

[Paul W]

"4th order (electrical)"

Paul,
For best results with the active analog xo, I think catapult's suggestion to dive into SL's bag of tricks for EQ, delay, etc. is a good idea.

I tried a commercial textbook LR4 electrical -once- and, because of driver response, ended up with passive traps, all-pass filters, etc. hung on as band-aids.

Though not necessarily a good permanent solution, you might try a passive parallel LCR trap in series with the 18EX...just to get a flavor.
Regards,
Paul

 

[paulhfx]

"4th order (electrical)"

Paul,
For best results with the active analog xo, I think catapult's suggestion to dive into SL's bag of tricks for EQ, delay, etc. is a good idea.

I tried a commercial textbook LR4 electrical -once- and, because of driver response, ended up with passive traps, all-pass filters, etc. hung on as band-aids.

Though not necessarily a good permanent solution, you might try a passive parallel LCR trap in series with the 18EX...just to get a flavor.
Regards,
Paul

I'm using LSPCad to model response,, and using SL's circuit topology and formulas to get initial values - no textbook xovers here

Thanks for the suggestions - I've put in a notch filter around the primary resonance and it's certainly an improvement. I'll try lowering the crossover frequency next.

Any other suggestions and comments appreciated,

Paul

 

[Paul W]

aaaah...I interpreted "electrical" as "textbook". The only other things that come to mind are a very steep slope (LR8?) and frequecy...maybe cauer but haven't tried one. I did have good success with the 9900 at 1500.

 

[paulhfx]

The 'electrical' was as opposed to 'net acoustic slope'.

I briefly hooked up an LR8, but got some muddiness and ringing (I'm being very critical and picky - it wasn't terrible).

While I built a passive cauer and could design one, I'm not sure how to do an active cauer - got a schematic/layout?

Paul

 

[Paul W]

Try Jon Marsh. If it exists (and I think it does) he'll have it.

 

[Jonasz]

Hi Paul W,
I'm veeery interested in you cauer-elliptic filter for the 9900. Wich mid did you cross it too? I have a 3-way dipole topsystem (yet without a crossover). I would like to try a passive crossover (between mid and tweet) before I try active. With your crossover I might not even need to try an active... That is, if you're willing to share it with me oc.

If you are, please mail me the details to jonaslfc@spray.se

Regards

Jonas

An externally hosted image should be here but it was not working when we last tested it.

 

[paulhfx]

I believe PaulW said he hadn't used one.

I used one with an XT25 tweeter, courtesy of a Jon Marsh design, and was extremely pleased with the results. Much info on cauer elliptic filters may be found on htguide - just search for 'cauer'.

I have not seen a published cauer filter specifically for the 9900.

Paul

 

[Jonasz]

Oh sorry....

Misread his message. Well anyway, if he did a good crossover I would still be intersted in the design.

Jonas