Super L100 (or bastard L220)

[dmtp]

I am new to this forum and JBL, but not speaker building. My current system is a biamped slightly modified version of Linkwitz's Pheonix in a custom built (by me) scultpured enclosure. Anyway, I've always wanted some JBL's and seeing items on ebay, I picked up the following:
LE14A
LE5
LE25
My plan is to put them in a cabinet of ~2.5cf. Does that sound about right once you lose volume to the LE5 sub enclosure?
For an XO, I like the minimalist 2caps and L-pads of the L100A and will certainly try that, but I suspect something more like the L220 XO is needed. I am not sure I really understand that XO. Looks like second order to the LE14A, but what is that cap splitting two resistors, and why the 50 ohms across the speaker? The LE5 looks to be third order, but then ? notch filters?? series and parellel??? The HF will need to be different anyway so I'll just ignore that.
Does 800 & 5k seem about right for freq?
Should I follow JBL's lead at least for the LE14A & LE5?
Recommendations for the HF section?
Should I just 'roll my own' using standard formulas?
Recommendations on what order XO?
Thanks for all the help. Finding this forum made me willing to jump in and start building (the speakers arrive this week).

[Robh3606]

Hello

I would scale up the enclosure to something in the 3.5-4.5 range to get the more lowend out of the Le-14. I would set them up like the L240Ti and go from there.

http://www.lansingheritage.org/image...-ti/page07.jpg

You could try the L240 crossover

http://manuals.harman.com/JBL/HOM/Te...240Ti%20ts.pdf

I think you might be better off with the L220. The Le5-2 has a rissing response see this thread

http://www.audioheritage.org/vbullet...&highlight=le5

You want to compare the 104H with the the LE5 resonse curve to see how close they are. Up top I would try to source 035Ti's they are much better than the Le-25's.

Rob

[Earl K]

Hi Mark

- Welcome to the forum !

I am new to this forum and JBL, but not speaker building. My current system is a biamped slightly modified version of Linkwitz's Pheonix in a custom built (by me) scultpured enclosure. Anyway, I've always wanted some JBL's and seeing items on ebay, I picked up the following:
LE14A
LE5
LE25

My plan is to put them in a cabinet of ~2.5cf. Does that sound about right once you lose volume to the LE5 sub enclosure?

- If you like the bass response of the L100 and that's your personal preference / then I fear that you'll find that any le14 variant operating in a 2.5 cu' will be too tight of a box tuning .
- This is an area of personal preference and can only be addressed at that level.
- That's why I always recommend that people make test boxes of various sizes & tunings to determine this (size ) question for themselves .
- 2.5 cu' ( tuned to 30 - 32 hz ) with a bit of electronic EQ would be fine for many here ( including myself ).
- 3 to 4 cu' , represents the net volumes that will offer freer-flowing bass ( at the expense of some actual accuracy or bass articulation ).

For an XO, I like the minimalist 2caps and L-pads of the L100A and will certainly try that, but I suspect something more like the L220 XO is needed.

- I'd study the N220 network as well as the N240ti network and N4412 for some inspiration when pondering your first crossover point .

I am not sure I really understand that XO ( N220 ). Looks like second order to the LE14A, but what is that cap splitting two resistors, and why the 50 ohms across the speaker?

- That cap with 2 resistors in the N220 is a Zobel network ( impedance compensation ). It's usage can be considered optional ( especially after one studies the N240ti network ).
- The conjugate resistors ( 50 to 80 ohms ) across the woofer mitigate ( average down ) some of the resonant peaks that all woofers have at their resonance points. This ( less severe peaking at resonance ) helps the driving source ( amplifier ) deliver a slightly more stable load . The resistors also drop a bit of power / lowering the sensitivety of the woofer / in this case , helping better matchup with the 5" midrange device .

The LE5 looks to be third order, but then ? notch filters?? series and parellel???

- No, it's second order . I count 2 poles per high pass & low pass within the N220s' bandpass section . The N220 does include another Zobel on the le5-9 ( this impedance compensation is formed by the 6uF & 7.5 ohm conjugate ) .
- The notch filter should be left out of your project . It's specific to the needs of the N220 / its' appearance was likely driven by some whacky acoustic resonance created by the mounting of the L94 lense onto a le5-9 .

The HF will need to be different anyway so I'll just ignore that.

- The le25 is a pretty old-school tweeter. FWIW, ( if my project ) I'd rather design a hipass around a 035ti which can be still had on eBay .

Does 800 & 5k seem about right for freq?

- Yes .

Should I follow JBL's lead at least for the LE14A & LE5?

- Most definately , follow JBLs' lead when choosing crossover values .

Recommendations for the HF section?

- Look at the 4412 series of enclorures, if you decide to use a 035ti or 052ti tweeter .

Should I just 'roll my own' using standard formulas?

- Standard formulas ( while fun & instructive ) usually don't work well with JBL components for a bunch of reasons . One reason ; the stock formulas require ruler flat response from the components ( which is rarely the case ).

Recommendations on what order XO?

- Second order ( if my project ).
- Also, use variable Lpads on the midrange & tweeter ( at least to start with ).
- Fixed Lpads are really best left to those with a lot of network design experience . You can swap pad types later, if need be.

Thanks for all the help.

- You're welcome .

[duaneage]

Your going to need a signal generator and a frequency counter/multimeter ( like the Fluke 8060a for example) to do serious design testing, at a minimum. An rta or a sound level meter is a big plus since you can plot the output of the SYSTEM once it is together. The drivers are going to interact on many levels at the crossover frequency, and in the cabinets themselves. Edges can cause peaks and dips which may be corrected with crossover networks or incorporated into the rolloff (sometimes). Driver placement on the baffle board can greatly affect how the system sounds and images.

I have taken differnt drivers from different companies and built decent systems but it is a long process. The bet part is when your done you have a unique sounding system, not a mass produced piece, and you can have it take on any sonic charactor you like. After all, your the customer and the builder.

[dmtp]

I do have a signal generator (old HP) and a SPL meter (Radio Shack) and have done ground field measurements in my driveway for other projects (quiet house in the country). A lot more tedious than an RTA and pink noise, but it at least gives repeatable results. It makes XO a bit tedious, but do-able. That's why I planned to start with a JBL network to start. I have the LE5-5 which is a ribbon coil so not as much high end peak as the even numbered.

At this point, I'm looking at a net volume of just over 3 cf with 1 3/8 thick walls (3/4 MDF + 5/8 particle board with polyurethane construction type glue between - then fancy veneer) I've used this before and get a nice "dead" cabinet. (The standing joke with my friends goes like this: Q: "What is the difference between a speaker Mark built and a factory job?" A: "About 150#")
I assume the best speaker arrangement is crowding everything together like on the L100's with the LE5, vent, LE25 in an arc over the woofer. I'm looking at a baffle of ~16" width.

[Zilch]

Vertical in-line offset from center (mirror-image pair) would be my preferred arrangement of drivers.

It doesn't much matter where the port(s) go(es). It'd be a pair at the bottom with LE14A for me....

[dmtp]

I'd be very interested in more thoughts, opinions, or (best of all) data on the differences between the vertical array and the clustered arrangement. I am familiar with the advantages of the D'Apolito (?sp?) MTM arrangement, but WMT - what is the advantage of lining up the voice coils? Or what about lining up the the edges against one edge of the cabinet?

[Zilch]

Buy Dickason's "Loudspeakers Design Cookbook," now in 7th edition.

Read up....

[dmtp]

I have Vance's book (fifth ed.)along with many others includig Weems and Cohen and I refer to them frequently. I do not, however, find anything on the arrangement of mid/hi drivers which favors the "line array" versus the "clustered" approach. If I have missed it, please point me to the appropriate chapter or page.

[Robh3606]

Look at the polar responses on in line vs. clustered. If you want the same symmetrical radiation pattern on both sides of the cabinet you go in line with offset baffle spacing.

Rob

[Zilch]

It's Chapter 6, wherein:

"As a 'guiding' principle, I think that keeping both vertical and horizontal polar responses as symmetrical as possible results in a better subjective experience...." (6.10A, end)

Also:

"Any time there are more than two radiating sources operating in the same frequency range on a single baffle, the combined output will produce complex interference patterns...." (6.20)

In which regard, consider the L100 crossover frequency range overlaps taken in the context of the intersecting planes of time alignment between woofer to mid, and mid to tweeter.

It's only along the line of their intersection that interference does not occur. Add a second speaker, and that reduces to a single point in space. Combing occurs at all other locations....

*****

See also D'Appolito Testing Loudspeakers 5.3.3.3 "Vertical Polar Response and Slow-Slope Crossovers."

The vertical in-line layout is shown in Figure 5.31, and the performance is documented in 5.4 "Example 5.4: A Three-way, Three Driver System."

Consider crossover upgrades accordingly.

*****

http://manuals.harman.com/JBL/HOM/Te...240Ti%20ts.pdf

[Ian Mackenzie]

If you consider the polar geomtry of 6 db slopes it becomes more understood why certain driver layouts are adopted.

The L100 used 6 db slopes. 6 db slopes are the worst effected by (vertical )off axis variations in (vertical driver layouts) polar response because of the broad overlap in the passband of each driver. By using a horizontal layout or other than vertical layout the listening position on axis may in fact be a smoother response.

[dmtp]

So if you use 6db xo, a clustered arrangement is better, but with 12db the line array is better?