Bottleneck's crossover experiment

Chris, if you want more accurate results, try a program called SPL Trace to trace the response and impedance curves in to FRD and ZMA formats respectively, then use Speaker Workshop to simulate. Both bits of software are fee, and using the help / tutorials available for Speaker Workshop, it shouldn't take more than an evening or two to get somewhere.

 

HI Tenson

presumably I'd need a measurement mic for this also?

 

Not if you use the data from the manufacturer and trace it in with SPL Trace. Measurements are better though as they include baffle step and phase shifts due to relative driver positions. On the other hand it can be exceedingly difficult to get accurate measurement especially if it's a three way and you want a low xover point. High xover points are easier to measure because you can gate the measurement sooner to remove reflections.

If you send me all your datasheets and details of the driver positions and offsets then I might find the time to do a simulations for you. Probably not in the next couple of days though.

 

Hi Simon,

Rob gratefully did the measurements for me with his laptop and mic.

Really though - with an in-room measured response of +-5db all the way from 30hz to 20khz (with no significant bumps) it's pretty much bang on..

 

It should be possible to improve further by making sure the drivers phase is matched at the crossover points. This usually gives more clean and immediate sound, and it also affects the dispersion of the speaker as the phase difference between drivers can steer the sound in different directions.

We are still waiting for those pictures!

 

I would agree with Simon that proper dirver measuremtn prior to crossover design is best. In this instance we are also designing according to nominal impedance, which pays no heed to actual impedance around the crossover point.
The slight problem here is that the mid and tweeter drivers are free-standing, so while a highly accurate crossover is possible, Chris will need to decide precisely where the drivers are to be placed - and stick to it.

 

http://www.photobox.co.uk/album/444075726

speakers, front view



tad 4002, JBL smith horn



tad 1601b



jbl 2405, tad 4002, smith horn, 1601b in bass cab



crocodile clips!

 

we're still waiting mescalito !

let's have the photo's of your kit!

 

They look like real fun. The mid horn looks exceptionally narrow top to bottom, what's the thinking behind that? I'd expect it to beam height-wise to a large degree, i.e. stand up and the snare drum vanishes!

Tony.

 

Hi Tony

The 2397 (or Smith horn as it's sometimes known)

Details:

http://www.lansingheritage.org/images/jbl/specs/pro-comp/2397/page1.jpg

http://www.lansingheritage.org/images/jbl/specs/pro-comp/2397/page2.jpg

I bought the 2397 specifically because of it's shape. With a mid-horn, I want wide horizontal dispersion, and limited vertical dispersion to prevent slap-echo from the ceiling and floor at the listening position.

The 2397 was designed to control this in mastering studios, and it works well to do that also in domestic environments.

Of course, in a free-space environment I would prefer another dispersion pattern, but sadly that isn't the listening environment of the homes I've been in so far.

At the listening position, you get about 2-3 feet vertically before treble roll-off, so standing is fine, as long as you dont move around too much.

A variation of the 2397 is also used in the following -

 

That makes complete sense. My Heresys do a similar thing, except they fire the mid at the listener's knee caps! Being a bit of a 'hot seat' listener I do like the directionality of horns, I'm convinced it brings a real clarity, which I assume is largely down to the lack of room reflection. My Tannoys beam the mid and top too, I suspect more than most as they are back-mounted in a rather thick cab.

Tony.

 

The 2397 is a diffraction horn-the thicker lips on the Westy version smoothes the response ripples around 750 hz and increases vertical dispersion.
Up to getting the SAM1 horns I thought the 2397/Smith/Westlake/2440 to have the 'nicest, smoothest, horn none-sound I'd ever heard.
The only issue I had when using my TM1's was that very wide dispersion of the 2397 in my 4.5m wide room meant I couldn't get get a tight enough stereo image without room treatment.
All I'd say Chris is be careful with the xover point/slope on the TADs-I'd defo slap a cap across them to stop any switch on pops taking out £3k diaphragms.
Cooky

 

Hi Cooky

They are on a bandpass, so it's cap then coil.

Would this stop the switch on pops? (sounds like a kellogs cereal)

 

Hi Chris,
It's just a precaution that prevents dc switch on/incorrect active connection damage by running a cap of sufficient high value to be outside the low slope xover passband.
In most cases you could wing it but at £3k for TAD replacement dias I'd not feel comfortable not to at least mention it as it is sort of a given practice over on the Lansing forum;-)
I'd talk to Rob- he's more clued up with networks than I.
Cooky

 

Hi Cooky,

We've got capacitance in series with the TAD tweeter so no dc issues.

 

Hi Rob,
As I understand it the inclusion of this larger cap also prevents any unloaded dia excursion below horn cut off too-more an issue in high level PA situations I know but even with 12dB slopes you can still get significant input outside(below in this case)the passband right in an area where the dias close proximity to the phase plug can become an issue as excursion increases with decreasing frequency..I'm talking bi-amped here-but if it is recommended in active systems I don't know what the equivalent passive protection would be.
Found this among my JBL stuff;
bugger-'tiz a 4001 yaar..
An relevant thread on LH from a while back http://www.audioheritage.org/vbulle...513-1st-Order-High-Pass-on-Compression-Driver

 

Thanks Cooky, just had a read of the linked thread.

I think the issue discussed only relates to active drive as in that instance the power amp is connected directly to the driver. Any DC or switch on thump from the power amp goes straight into the driver. In that scenario a large cap in series would certainly afford protection while leaving the active crossover slope intact.

With purely passive connection we have series capacitance which will protect against d c and LF thumps.

 

Rob,
As a user of rather expensive Be comp drivers myself, in my own case with very high power available it's just too easy to slip up with the volume control so purely for the reasons of out of band diaphragm excursion/proximity to phase plug I mentioned earlier, I wouldn't risk first order passive either-despite some possible sonic benefits.

 

Good points. Chris seems set on 1st order but as the crossover will be mounted externally in a box it would be easy to swap in something with steeper slopes.

 

Oh cool.
It can certainly be a whole lotta fun.