Trachorns done need xover advice

[3dzapper]

I finished my trac horns this morning and am enjoying them as I type.

For you xover gurus a question: the driver I want to try falls off at 4500Hz pretty sharply, what values of components do i need to lower the HF point of my ALKs to cross at that point? Is it economically feasable? Am I nuts? please leave the answer to the last question unamswered.

Rick

[charlieboy]

Rick,

I am thinking of trying the wooden horns with the JBL's also. What are your impressions vs. the stock setup? Thanks.

Charlie

[3dzapper]

Remembering that this is just a trial set-up with the horns just placed atop the Khorns with no bracing:

The artifact that I was looking to get rid of, a jazz grand piano sounding somewhat like an upright is mostly gone. The remaining tad is , I'm sure, a function of the K-55. Remember I am picking nits here.) I had to reverse the polarity of the K-55 to get it to blend better with the bass bin to my ears and subsequently the tweeters to keep that phase relationship equal. That may well be from the placement further from the bass horn. There maybe, again nit picking, a small resonance or beaming artifact at the extreme high end of the squaker or a crossover peak I'm not sure yet. havn't had time to do any listening to sine waves. (My wife hates it!)

They are subtly better than the K-400s, female voices are more natural, less nasal, vibraphones are more smooth and as I said pianos more realistic. Still not quite acoustic piano in the room real, rather miked and amplified real. Of course that is what I am listening to, DUH!

I will give another review when I get them firmly installed at the proper, lower, level. So far they are worth the time it took to build them most of which was waiting for glue to dry. It sure gets boring watching that stuff.-O

Oh yeah, I know that Al Ks are better than mine. I can't approach Martinelli's level of quality.

Rick

[Deang]

"For you xover gurus a question: the driver I want to try falls off at 4500Hz pretty sharply..."

The driver you're using now falls off at 4500Hz pretty sharply.

[ajsons]

For you xover gurus a question: the driver I want to try falls off at 4500Hz pretty sharply, what values of components do i need to lower the HF point of my ALKs to cross at that point? Is it economically feasable? Am I nuts? please leave the answer to the last question unamswered.

Rick,

I'm not an xover guru, but I'm doing the same thing. I'm building my second set of ALK's rossovers, and I'm crossing at 3500 for an EVM 1824 midrange, based on Al K's tips.

Here's what Al K said:

"Frequency scaling is easy to do. You simply scale the parts by the ratio of the change.

For example: if you have a 2.2 uF cap at 6000 Hz and you want to move it to 7500 Hz you multiply 2.2 X (6000 /7500) = 1.76 uF.

The inductos work the same way. The problem you will run into is that you can't buy a 1.76 uF cap. You will need to parallel several that add up to what you need. To impedance scale you make scale inductor values proportional to the change in impedance and the caps inversely proportional. I think you will want to keep 8 Ohms though."

----------------------------------------------------------------------------------------------------------------------------------------------

Good luck!

Armando

[ajsons]

But if your driver rolls off faster than the crossover, you need the "sensei" for that!

Your horn looks pretty cool though!

[Deang]

But if your driver rolls off faster than the crossover, you need the "sensei" for that!

That's not such an uncommon practice actually -- PK did it with the bass bin of the Khorn for almost 40 years. The acoustic roll-off is 18dB/octave, yet the low pass section is 6dB/octave. That same 6dB/octave low pass is used in both the ALK and the SAA. In fact, the situation with the squawker is much the same: Al hits it with a 12dB/octave filter, but acoustically it's falling off much faster. At any rate, my point with the plot was to show that the original K-55-V with spring loaded terminals is pretty much finished at 4500Hz, and though the ALK doesn't bring the tweeter fully in until 5800Hz -- there is plenty of tweeter energy at 4500Hz to compensate for the K-55's fast roll off. This is no accident -- as the ALK was designed using the K-55-V with the spring loaded terminals (single port phasing plug). I suspect Rick should have to do nothing -- plug and play I believe.

[ajsons]

I knew one of the "senseis" will explain that!

In my situation, am I right to say that my squawker's roll off is too far off (3500hz ) from 6000hz that I really need to change my crossover values? The EV 1824 rolls off rapidly like the K-55, only at 3500. The T-35 is good down to that frequency. Some EV products, like the Sentry IV (horn version) cross the mid and the tweeter at 3500.

[Al Klappenberger]

If you can figure out the computer schematic, here's the 4500 Hz crossover network values.

AL K.

[Daddy Dee]

Rick,

That horn looks very cool. Could you say more about the construction process?

[3dzapper]

Dean, this is the plot of driver Y. It seems to me in my limited experience with it more "musical". The plot by Al is in a K-400 and as you can plainly see, it needs the lower xover.

Al and others thank you. I was able to figure out the numbers AL, I think I have enough pieces around the house to breadboard the change.

Rick

[3dzapper]

Dee, Building one is not difficult if you realize that you will not get the perfection Al offers.

The basics, especially building the required jig can be found in the reprint of Dr Edgar's "Speaker Builder" article here: http://www.volvotreter.de/dl-section.htm

I built my jig from 1/2" construction plywood, doubling for the upright. ( For the next ones, if ever, I will add more thickness at least at the apex. After the jig is built and the flat sides are cut and screwed to the jig applying the thin plywood is not hard if you have three hands. I cut the forms with a jig saw after plotting a 1"sq grid on the "pattern" in pencil on the first piece of 1/2" birch ply. I was able then to trace the pattern onto other places in the sheet and cutting them out. The resulting patterns, when applied to the jig will have slanted edges. For that reason, I used Liquid Nails for my adhesive. It has the ability to adhere tightly, fill gaps and remain flexible. The first two points are self explainitory, the third less so. The flexable adhesive when applied heavily between the plates along the horn acts as both a filler and a dampening agent. Win,win.

To bend the 3/16" plywood horn sides I used both hands to first draw the center in. Of course I could not reach my drill driver to insert the first screws( #8 or 10 x1/2"). I then placed my knee in the center of the bend freeing one hand to drive a preloaded screw. Once the first few screws are in the other two dozen or so are easy.)

When both sides are layed up thusly, tt is time to get out your old rocker and watch the glue dry for about 24 hours. Might as well cue up some tunes.

After the glue sets remove the screws and lay up the second layer with plenty of adhesive. I used an expanding ess pattern down the center and a bead on both edges. Repeat the process using hands and knee to apply the panel.

One thing, make the side panels a bit oversized. It is more difficult, to impossible to put on a precisely cut panel. After all the glue is dry it is easy to trim the overhang with a jig saw and a belt sander.

Rick

[3dzapper]

After typing and editing for 1/2 hour I didn't want to loose that last post.

The next step is termination. In his article Dr Edgar suggests using the scraps from constructing the jig for the flat side braces. Although that would work fine, I used one for a pattern and made mine from 3/4" pine. The cut using a table saw, miter guide and a block of wood made a better glue surface than the hand cut jig scrap IMO. I glued these blocks to the horn sides after drilling two screw holes each. I used some lomg sheet rock screws just biting the horn to prevent the blocks from riding up under clamping pressure. The braces were applied with wood glue with the ends extending ever so slightly over the horn.

At the side of the horn I glued pieces of 1x6x1" pine using the same method and a slight overhang of the horn. After all that glue dried, I was tired.

The next step is to square off the mounting area. I used a bench belt sander to square and even all the edges and mounts.

For the end plate I uses a 1x6x6" piece of pine. after cutting the ecact center is marked and drilled. Then I put the piece on the horn centering the hole and marking the location on the braces and the location of the braces on the plate. Holes were drilled and countersunk for the plate to attach to the horn. I did not use adhesive on the plate to allow easy removal for experimentation.

Edit: I forgot to put in that I there will be some small gaps inside the pattern due largly to the hand cutting process of the flat sides and the resulting angle. I ran a bead of Liquid Nails down the corners and pushed it in while smoothing with a latex gloved finger.

Walla! You have a horn!! Put on an adapter from Parts Express for your driver and have some fun!!!!

Rick

[Daddy Dee]

Rick,

Thank you for detailing your process. I'm impressed. It takes a fairly handy guy to turn out those things nicely.

I don't think it is a very large club of those folks who can do the woodworking and also sling solder with excellence.

Congrats on the project.

[Deang]

Nice posts Rick. Yeesh, too much work!

The plot: I'm confused. I see 3725Hz in the bottom left hand corner, and the corresponding marker on the plot -- apparently where the driver begins its rapid descent. Where did 4500 Hz come from?

[DrWho]

Nice work...and what "more musical" driver are you moving to? Is its impedance response similar to that of your K55v? Because a change in impedance will also move the crossover point around (by as much as 1 or 2 octaves if you're not careful).

[3dzapper]

4500 is where I figure the falling mid and the rising tweeter could meet without a large drop off or spike. Like the K-55/k-77 combo crossed at 6K. I would also leave in the P-trap to dampen that strong return.

The JBL could be crossed as low as the 3KHz recommended crossover frequency. The 4500 point would be very near the 3xresonant crossover point recommended by some. Looking at the brochure, the resonance looks to be centered at +/-1800Hz.

DR, If Al gives me numbers, I listen. [] I just wish he would give my the coming PowerBall number.[<)]

Rick

2404.pdf

[sfogg]

Rick,

Nice work on the horns!

"The JBL could be crossed as low as the 3KHz recommended crossover

frequency. The 4500 point would be very near the 3xresonant crossover

point recommended by some. Looking at the brochure, the resonance looks

to be centered at +/-1800Hz."

Be careful there though as the 2404H isn't flat the lower it goes per

Al's measurements. On the JBL forum I've read several recommendations

to cross it higher, not lower for best sound quality.

Shawn

[3dzapper]

Here is Al's plot. I knew I had it someplace. Smoothed it would look like JBL's within about a DB.

Don't ask me how it would sound, that is what I want to find out.[8-|]

Rick

[3dzapper]

Well, Scratch plan Y. Don't match the wood horns at all. Onto Plan A now.[] 808-8 that is. louder than the K-55s with good sounding mids but there seems to be a hole at th bottom of the range. The 808-8 is a 500Hz rated driver and it sounds it.[]

So Ole Dean is right, the K-55 is a fine driver even if slightly colored. Hmmm, maybe if I steal the solder terminal units from the Belles?[6]

Rick