Compression driver selection & Revision
I recently completed my second pro class speaker build--Version 1.0, anyway.
The drivers I selected for this "two-way" are the Eminence Beta-12LTA and the Dayton Audio D250P-8. I chose the Denovo SEOS-12 Waveguide.
I was attracted to the Beta-12LTA for it's point-source nature, and consequentially extended high frequency response. My expectation of the driver has been surpassed--I am extremely pleased with my selection. My dissatisfaction and subject of this thread comes from the compression driver.
My design goal was to create a two-way speaker with an extended range or full range piston driver to keep the crossover out of the sensitive vocal range as much as possible, 3k or above. The Beta-12LTA promised this, and fulfilled it's promise.
The only active crossover I own right now is an iNuke1000DSP amp, which I realize is not the most ideal thing to use. Combined with considerable time listening to different crossover points and slopes, I concluded these speakers sound best when the Beta-12LTA is run full range, implementing a first order high pass on the tweeter at 4kHz. There is a dip in this range on the LTA according to the manufacturers stated frequency response. I have not done any measurements myself. I plan to in the future, I have not found time to clean out the basement for suitable set up room.
The iNuke amp is not audiophile grade by any stretch of the imagination, so I had to get away from it. Because the best crossover I could determine was so simple, I switched to passive implementing Jantzen silver z-caps and moved to a much better sounding amp. These upgrades improved the sound phenomenally. It was actually at this point I realized my suspicion was true--The D250P-8 is not cutting it.
I can not get the Dayton D250P-8 to sound acceptably enough. I don't want to mislead anyone considering the purchase of a D250P-8. Don't get me wrong, these speakers rock (literally) the way they are now. However, according to what is my opinion, the D250P-8 can not keep up with the LTA in terms of fidelity.
In my dissatisfaction, I turned to my pair of Bohlender Graebener Neo3-PDRW isodynamic planer tweeters, which are currently open-back. I understand an open-back planer tweeter is pretty far removed from a horn loaded compression driver, but I knew the Neo3 had the clarity I was looking for, and I wanted to confirm my crossover decision was good. (Most notably wanting to confirm a passive xo could sound good.) Turning the cabs upside-down for minimum driver distance, I sat the Neo3s on top and wired them passively with Jantzen Superior-Z caps. (Different Z so I had to use different capacitors, I already owned the Superior-Z and they coincidentally fell exactly on the 4kHz mark! I should note I used a simple online crossover calculator for this and not an actual speaker designer spreadsheet. I felt this method would suffice due to the simplicity.)
The level of the tweeter needed attenuation, but otherwise perfect match! The LTA actually blended with the Neo3s so well I haven't yet put the effort towards attenuating the Neo3s. It's so close, and it sounds sooo very good. So this is where I am now, I would like to find a compression driver with great clarity. I don't want to spend $800 per driver, however. I found the B&C DE250-8 which looks interesting.
Attached is an image of the two speakers sitting on their accompanying sub. The baffles have 45 degree faceplates to reduce diffraction. This, combined with a high frequency crossover, allows the Beta-12LTA to be spectacular in it's job of reproducing vocals, which was my goal. Vocal clarity is phenomenal!
Any advice? Thanks for reading!
I've used the B&C DE120 extensively and it's very open and detailed sounding. The only problem I see with your design is that when choosing a crossover point you need to pick a spot where the LF and HF have approximately the same coverage pattern. The Beta 12LTA will beam considerably because it's such a large diameter driver. I don't see any polar charts on the eminence website. I've attached a typical polar plot of at 12" woofer. You can see it seriously starts to beam at 2kHz and is down -6dB only 15 degrees off axis. The SEOS-12 horn appears to be a nominal horizontal coverage pattern of 90 degrees creating a large mismatch in coverage.
Also, I always try to avoid running the LF past where the driver starts having modes. You can tell this driver is starting to breakup as low as 600hz by looking at the irregularities in the impedance curve. (see also attached).
Cone breakup will add to the beaming effect, and also limit it's power handling.
My advice would be eliminate the horn from your cabinet and place a larger horn on top of the cabinet, or build a second enclosure for the horn and crossover.
For the Beta 12LTA I'd be looking at cutting it off at 1.5 kHz and using a B&C DE250 with the ME20 horn. The midrange/vocal clarity of the B&C will exceed that of the Beta 12LTA so don't worry about that.
Hope this helps.
It's good to hear positive feedback about the B&C brand, I will consider the DE120 as well. The DE250 seems to have more power handling, which is important. Which B&C driver do you own, conanski?
My bad for leaving out intended usage. These actually will be used as PA speakers. I DJ weddings, so I wanted to create something simple and sleek looking. I have yet to use them on the job, but I have taken them to a bridal show. I received positive feedback from brides about their look. I think they look better in person than in that picture. My smartphone camera could not capture the black levels perfectly. The sub looks slightly purple in that picture even, which it is not. Lol.
I measured the impedance curve on one of my Beta-12LTA drivers with my DATS. I did not measure in free air, rather left it in the enclosure. To my surprise, the entire impedance and phase lines were jagged throughout. They were not consistently jagged, however. I measured a dozen times in a row, and the jaggedness of the lines were almost never close to identical. Sometimes the impedance would do crazy things near 20kHz like spike higher than normal, or jump erratically. Does this mean it has break up modes throughout the entire frequency spectrum? Does this mean I don't have to worry about below 600Hz anymore--Rather every frequency?
Thank you for the observation of the importance of coverage pattern. I had never previously been instructed to pay attention to it, but I suppose most assume as conanski pointed out, it doesn't matter as much with home listening. I have a few observations, but I am still learning, so please bear with me.
You referenced a normal 12" driver, but I would like to point out this Beta-LTA is not a normal 12" driver. It has a whizzer cone. To my understanding, this smaller cone improves high frequency dispersion , although I will admit to not knowing how well.
I have to say, I trust my ears a lot. I spent a lot of time playing around with crossover configurations. As you stated, a lower point should sound clearer. I did observe this; I found a lower point with a steeper slope on both the LTA and the CD produced "clearer" sound. However, it did not sound better that way. Strangely, vocal inteligability was actually lower. When I finally settled on running the LTA full range and the CD xo at 4kHz with a first order slope, it was the best compromise of everything. It sounds the best that way. Also, because I am using such a shallow slope, it sounds to my ears as if the tweeters are still filling in their share in the 2kHz region. But heck, I'm still learning, and I'm happy to receive correction.
Building something on top is out of the question. These speakers are built to look sleek; their looks are more important than their sound because of this application. I do however, have the ability to replace the horn by modifying the cabinet. I'm not yet convinced I need to, however. In my mind, whizzer cone + shallow xo = close enough. Is that so totally wrong?
A few additional notes:
I found the speakers sounded more pleasant listening with the horns and CD at close to a 90 degree angle from the back wall perpendicular to the listening position. With the Neo3s, the same listening angle achieved similar results, but the speakers could be turned to approximately 43 degrees with no penalty to "pleasantness" and a substantial increase in sound stage, which otherwise was never as strong with the horn at any angle.
The enclosure walls are all an inch and a quarter thick, consisting of 3/4 birch plywood on the outside, and 1/2 MDF on the inside. There is oak bracing running from the front of the baffle to the back corner. All of the corners as well as gaps created by the steel handles are filled in and sealed with Bondo. This thing is a tank.
Waited too long to edit. Here is a measured in-box example.
It's good that you used Baltic Birch in your design, and even better having the MDF inside as well, and to boot you've made it trapezoidal to reduce internal standing waves. I recommend getting the Dayton UMM-6 reference measurement mic. It connects directly to a PC via USB. I use a free program called Holm Impulse to do Gated frequency response plots (quasi anechoic). It's helped me tremendously to design my own passive crossovers. I even used it to time align my drivers.
Another thing you might try is to tilt the upper frequencies by building the following passive crossover.
It will give you the curve shown below, with the highs tilted upwards, this compensates for the weakening highs found in most compression drivers.
Experiment with different values. Increase the difference between the to resistors to give a more aggressive tilt. To add attenuation increase R1 and decrease R2.
This may not be 'by the book' but is simply something I discovered using a program called 'Passive Crossover Designer 7'.
Also, I would add a passive filter on the BETA-12LTA to shelf down the highs.
Give that a try. I did this and it substantially increased the amount of bass (which I like!)
Your system efficiency will drop, but to me it provided a much more rounded sound that is much fuller.
I'm getting more exited about this project! I used Passive Crossover Designer 7 extensively in my first pro class speaker build. Back then I didn't have the measurement and calibration tools I do now, so I could only do so much.
I own the Dayton Audio OmniMic V2 Precision Measurement System. The microphone looks identical to the UMM-6, save color. Is it correct to assume it can perform the same functions? I recently purchased the OmniMic V2 for this very project, but have actually yet to use it due to aforementioned basement obstacles. What kind of environment do you test in Joseph? I was going to go all out and create wooden frame panels to hang from the ceiling made from fiberglass and studio foam.
I think I'll go ahead and place an order for the parts in the values you suggested, probably throwing in a dozen extra resistors or so. Thanks!
I had already made previous arrangements to use these speakers gigging a small band for the first time later today. We will see how that goes, I'll post my observations.
I measure in my basement and I move everything away as far as possible and place the speaker in the middle on a stand. Vertically it is positioned approximately in the middle between the floor and ceiling. The mic distance has to be two times the largest dimension of your speaker. If your speaker is 20" tall for example then your mic will be 40" away from the front baffle.
Be sure to download the correction file for your particular mic.
Also get a good mic stand, as this will effect your results.
Also make the speaker stand as low profle as possible, a pole is ideal.
Shown in the pic is a Harbeth speaker being measured at a town hall. If Harbeth uses this method then it's good enough for me or you!!
|All times are GMT. The time now is 09:30 AM.|
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2015 DragonByte Technologies Ltd.
Copyright ©1999-2015 diyAudio