View Full Version : "New" JBL Design project

07-07-2004, 08:52 PM
Well, I'm embarking into uncharted territory, in a very small way.

A customer of mine at work, asked me to build him a center channel to match his home theatre system, which currently has JBL L100T mains, JBL L20T side speakers, and JBL L20T rear speakers. However, he's been using some other center, due to the fact that JBL really never made a center channel for this series (naturally, as there WAS no home-theatre, or at least nothing that used a center channel, in the mid 1980s!)...

I suggested that we take the "guts" from some L20Ts, and make an MTM out of it. He brought me the parts from a pair of L20Ts he found on Ebay (damaged cabinets, but perfect drivers), and I got started tonight.

I first took a baseline measurement in LMS, then I tried a few things, with the woofers and a test cabinet... which led to the following measurement, with two of the 115H-1 woofers and one 035Ti tweeter in an MTM arrangement, woofers wired in series. Wow... this thing worked better, sooner than I thought it would. With the stock tweeter crossover (which acts like a third-order acoustic highpass at about 3KHz), and a modified version of the woofer crossover (adjusting the values for a 12 ohm woofer system, instead of the stock 6 ohm setup), it started to draw curves in LMS, that I swear, I doubt, by listening, ANYONE could tell from a stock L20T, without some serious concentration.

Here's a snapshot of the last graph we took- the dark color (purple) line is the stock L20T, and the bright green line (despite saying "tweeter only", we forgot to change the descriptor from an earlier test) is the prototype MTM arrangement as described above. I'm stoked- it's FLATTER than stock, while having almost IDENTICAL "effective tonal balance" contours. Unless something radical happens when I transfer the system to the final cabinet (which I plan to build, to match the cosmetics of his oak L100Ts), it should kick some serious butt!

Now, I'm planning, if possible, to try attaching some bucking magnets to the drivers (for a modicum of video shielding)... I know those will change the frequency response a tiny bit, but I don't expect to see anything, I can't compensate for. That said, I think it's remarkable, that this result happened SO QUICKLY. As someone said... sometimes it's better to be lucky than good! :D

07-08-2004, 06:47 AM
I took the mid and tweet horns to be the basis for a center to match my Klipsch Heresy based HT. In my case I used a stock Heresy Xover that matched the horns and modified it to work with two Fostex wired in parallel to serve as the mid woof with most stuff below 80hz going to the sub. Much like you I got suprisingly outstanding perfomance on the first go round.

What I found (and cannot quickly recover the information) was that I had an off-axis issue. Applying some calcs on baffle design (the part I can't find - that project was several years ago) I found that I had to do something that was counter intuitive - angle the baffle so that the mid-woofs were angled slightly toward the middle. The horizonal off-axis improved dramatically.

Enjoy - this will be a dramatic improvement for your friend...

07-08-2004, 08:04 PM
Well, I listened to the combo... and yes, there is a midrange dip, when you're FAR off axis. I mean, like 50 degrees away from straight on. Even then, it's just a dip, with no peaking.

The nice thing, that I think allowed this process to go so seamlessly, was that the acoustic rolloff of both the woofer plus lowpass filter, and the tweeter plus highpass filter, measured at an effective 18 dB/oct... or, a third-order acoustic slope. This is IDEAL for a D'Appolito-type MTM. So, no real topology issues to deal with... just the transformation of the lowpass from operating with a 6 ohm load, to operating with a 12 ohm load. The spacing of the drivers, with the 3000 Hz crossover, is certainly within the range of plausible workability of a D'Appolito type MTM as well, from an acoustic wavelength/cancellation/lobing point of view...

And also, if I want to, I can easily ameliorate some of the dip, electrically... a bit of it occurs, in both stock and modified form, right at the same place the dip off-axis occurs. By changing the lowpass filter to one with a higher Q (smaller inductor, larger cap, which will "bump up" the response at the resonance/crossover frequency a bit), I can probably eke another 1 to 1.5 dB out of the midrange dip. That is, if it turns out to be necessary... which from listening and measuring so far, doesn't seem to be the case.

The really nice thing, is how easy LMS makes it to make changes and re-test/re-measure things, and see immediately what the change does. This certainly takes a LOT of the time, and much of the trial-and-error aspect, out of the process.


07-08-2004, 09:10 PM
Hello Gordon

Wow I just would have went for another L20. Thats really cool it worked out that way. That could make a nice MTM with a sub under for mains. Can you post a picture when you have the center all fleshed out?? I have 5 L20t's use one set for rears and have another for mains with a spair. I really like them. They work great with the L80T's.


07-09-2004, 07:56 AM
Nice idea, I'm sure it'll work smoothly.

We actually used the guts (and baffle) of a JBL 4311 (2213, LE 5-2 and LE 25) to build a centre, it now sits long ways with the JBL badge and attenuators along the top, with a new grille.

All I can say is it's the best centre I've ever heard, mainly because its got "balls", low voices and effects in the centre channel are amazing. I think thats where small commercial ones, with the odd tweeter and 2 4" mids lack.

So good luck with the rebuild, it will yeild great results!