Printable View
Hi,
if I would build a Yuichi A290 horn for my JBL 2441 drivers, is there a passive crossover I could use?
It would be great to have a highpass around 650hz-700hz for the JBL 2441.
Above 10khz would be a 2405, I think this is easier as I could copy the 3155 network part for that.
Below 650hz is a TAD 1603 with a 650hz passive crossover.
It´s the 2441 highpass that gives me headaches.
Best regards,
Olaf
I just know I will catch hell for this, but,, on mine I used a single cap, 3uF or 4.5uF.
That alone got me +90% there. I will see if I can find the measurements. If I can you will be surprised.
My Yuichi’s are out on loan at the moment so I can’t re shoot them easily.
Barry.
Hi Barry,
appreciate your answer! :)
I would have thought a much bigger capacitor would be necessary... Something around 15uf and maybe 3mh for the inductor...
I haven’t found my Yuichi tests but use this for food for thought. If I remember correctly the results were similar.
Here is a JBL 4338 horn with a 2451SL with a generic LR24dB at 500Hz. With only a 3uF for the white trace or a 4.5uF for the yellow trace, one cap for protection and signal shaping. Measured on axis at just under 2M in my house garage with little regard for reflections as can be seen in the measurement, those fairly symmetrical ripples are reflective interference.
In DSP this takes 15dB of cut or boost and 4 points of PEQ to match what one proper cap in series will do.
Is it perfect? Nope. This was an experiment in simplicity. Ironically it is better than a lot of the complete systems I measure.
I will hopefully find the Yiuchi tests this weekend. The Yiuchi is a well behaved horn. It doesn't take much to make it delightful.
One day I will combine all my data on one computer, one day.
Barry.
Thanks for sharing your data, very inspiring! :)
Olaf,
Don't let the absence of a ready-to-go HiPass deter you from building a pair of Yuichi 290's .
Once built, you'll get the help you need ( though you'll want to buy a test mic and a decent 2-channel sound card to facilitate others efforts > @ 200 Euro ) .
:)
Here's a plot Barry did years ago on a Woody Banks Yuichi semi-clone (I think it's just a bit smaller than the A290 Yuichi). I'm not sure if this is the test plot Barry was thinking of or not, I bet this is just driver/horn. 2450 w/Be diaphragm
http://www.audioheritage.org/vbullet...l=1#post311655
Giskard did quite a comprehensive study of different 4" diaphragms placed in different drivers on a few horns ( + most importantly included the 2 different phase plug types ).
HERE!
The following 2 pics ( of the many found in that thread ) are the most applicable to Olafs project.
Important here is that the 2445 & 2441 share the same ( old style ) phase plug topology.
http://www.audioheritage.org/vbullet...1&d=1269133138
http://www.audioheritage.org/vbullet...1&d=1269133147
Giskard had mentioned ( somewhere ) that he usually displays with 1/2 octave smoothing ( I use 1/6 octave for filter design work ) but these are still very good ( and look a lot sharper than most other programs 1/3 octave smoothing > which is all a bit weird ).
Olaf will need to take his own readings ( if he wants custom networks done up for his Yuichi horns ) .
The 2446 phase plug has a couple db more midrange output than the older type phase plug, see below .
http://www.audioheritage.org/vbullet...1&d=1269131225
Here's one possible filter ( for 650hz ) that's usable only if Olafs horn/driver combo matches Giskards nominal measurements ( here I've worked with Giskard's traces to show an example filter ) .
Attachment 85615
:)
A shot of Woody's handy-work for inspiration .
http://www.audioheritage.org/vbullet...1&d=1303429942
:)
Quote:
Originally Posted by JeffW
I wasn't able to directly compare Barry's work with Giskards > it seems to be comparable ( after making allowances for a few obvious differences ) .
http://www.audioheritage.org/vbullet...1&d=1303438006
http://www.audioheritage.org/vbullet...1&d=1269133138
The larger Yuichi shows a bit of a suckout at 2300hz while Woody's smaller version has a similar dip a bit lower .
:)
Hi Guys;
Thank you Jeff. That shot is a D16R2441 at 2 meters.
These files are not on my laptop. My home shop PC is packed in a corner but I can check the two at work to see what if any filtering was used if those measurements reside there.
I have LMS also but i can get 8000+ data points in less than three seconds on an HF measurement with TEF. LMS is great but painfully slow, also limited to 400 data points if I remember right.
Barry.
Thanks all of you for the effort you have put into this thread!
I wasn´t aware of the difference in the freq.-response using the 2445 vs. 2446 throat-driver. Thanks a lot!!
I´m currently using 2446 bodies with 2441 diaphragms... So my horndrivers are allready a mismatch with the existing crossover-networks :blink:
If I would use Truextent BE-diaphragms in the future, the freq-response will change even more...
In the beginning I thought about using the highpass of the JBL M9500 e.g. S9500 crossover, as it is a 650hz network for a 2"-driver.
My plan was to use the 12uf capacitor and 3mh inductor of this network within the 3155 equivalent network.
The 3155 originally uses a 8uf capacitor and 1mh inductor....
Could this be a reasonable start-point or is it a total mess up?
You're welcome!Quote:
Originally Posted by Dr.db
No, your mash-up approach to a crossover is not reasonable, it'll surely be a mess up ( OTOH , there's no predicting what others are content to call good ;) ) .
IMHO, you need to learn how to take frequency response measurements ( plus get impedance measurements ) so that a proper HiPass can be designed for your unique ( driver/horn ) combination .
:)
Hi Earl,
thanks for this very clear statement! :)
The post I quoted had the 2450Be on the Woody horn, but I didn't post the graph:Quote:
Originally Posted by 1audiohack
http://www.audioheritage.org/vbullet...1&d=1303363834Quote:
Originally Posted by 1audiohack
http://www.audioheritage.org/vbullet...1&d=1303363834