PDA

View Full Version : L-100 crossover



Too Tall clist
04-27-2007, 09:47 AM
The L-100 and by many other names is a curiosity to me.
The specific models I am referring to are 3-way with a 12" woofer, a 5" mid (about) and a hard cone tweeter with a center dome.
More important is the crossover.
The entire xover is a series cap with an L-pad on both tweet and mid and nothing on the woofer.

If you measure each on axis -
The 12” goes up to about 5kHz flat. (Rather amazing really)
The mid from 400Hz to 6kHz.
The tweet from 1kHz to 15kHz

With casual inspection you have ALL THREE drivers running in the 1kHz to 5kHz region. The capacitor on the tweeter cuts it back some, but not enough to stop overlap.
Obviously the cap on the mid does nothing to its high frequency extension.

When you model the speaker with the simple xover it uses it does not take to long to figure out that if you turn off the mid driver there is still an overlap of the woofer and tweeter (though the 12” does beam sharply as it goes up in frequency).
This causes a hump in the response that no amount of twiddling on the tweeter L-pad can eliminate (without losing the highs).

So what to do?
Well we do have a mid driver. What happens IF we flip the polarity on the mid and turn it up just a bit?
Will it cancel the hump?
Yes!
It does just that nicely.

And unless all of the owners changed the wiring that is how the drivers are wired stock.
So turning the mid up causes a “hole” to counteract the hump from all the drivers working in the same region.

I have had three similar boxes come through here from local people. It had been my intention to publish a simple 3-way 2nd order xover for free on the net since there are literally thousands of these still in use. But I ran into problems.

Many of the drivers measure different. Some no doubt came from different runs at the factory since these drivers were in production for decades.
Much of it because of use. When one mid or tweet was toasted they would order just one and the years of use on the old one that still worked made it respond very differently.

Also some models have the tweet and mid in line and others the mid is off to the side.
All these changed things enough to mean I could not count on a design working for most owners.

Anyway if you want to do some DIY fun a simple 2nd order at 300hz and another at 3kHz can make these speakers shine as never before. Keeping the high frequencies from coming from more then on driver is a design goal that vastly over shadows any foam you might paste on the tweeter or anywhere else on the baffle.
With the highs coming from three different places in space I don’t see how foam can help at all with how they sound since arrival times from the three drivers scramble things beyond what foam can do. Unless you hang it in front of the woofer.
OTOH for visuals it makes it easier to eat Mac & Chesse in the same room. That somewhat rotten orange foam always made me a little queasy.
Good luck,
Too Tall

PS- I have included a measurement of the raw responses. Please note you have to add 9dB to each curve to see true 1 meter sensitivity.

4313B
04-27-2007, 10:05 AM
Neato!

http://www.audioheritage.org/vbulletin/showthread.php?t=13105

Earl K
04-27-2007, 10:17 AM
Hi TTc,

Swerd has nicely walked through a L100 network solution ( for those interested ) here ! (http://audioheritage.csdco.com/vbulletin/showthread.php?t=13105&highlight=L100+Crossover)

EDIT: Ahh, I see Giskard is a much faster typer than I .


:)

4313B
04-27-2007, 10:36 AM
EDIT: Ahh, I see Giskard is a much faster typer than I .It had been bookmarked and sent to JBL as an example of a really nice project so it was easy to find.

Earl K
04-27-2007, 10:40 AM
It had been bookmarked and sent to JBL as an example of a really nice project so it was easy to find.

Ahhh, that explains it .

For some reason it wasn't in my "bookmarked" folder / so I had to search it out.


:)

Swerd
04-27-2007, 01:47 PM
Too Tall
Greetings from another L-100 looney.

If you look on my post for the unfiltered LE5-2 midrange trace, you'll see a big difference between it and what you measured. I guess that is an example of the differences in drivers you mentioned. My midrange did not go low enough for a 2nd order crossover at 300 Hz. As a result, it was crossed at 1,000 Hz. The woofer is quite smooth there, at least on axis.

Do you have a schematic of your crossover?


Giskard & Earl K
Gentleman, thanks for remembering my post. It makes me feel good this Friday afternoon. Did you really submit it to JBL? I'm flattered. However, since Harmon Intl. was recently sold, no one has yet contacted me about becoming the new CEO. What gives?


Zilch
I finally filed down those plumber's test plugs so they fit into the cabinet ports. They do very nicely.

Too Tall clist
04-28-2007, 08:54 AM
Neato!

http://www.audioheritage.org/vbulletin/showthread.php?t=13105

Thanks for the link.
I know of Dennis from Liberty Instruments (LAUD and Praxis), LspCAD forum and the Bass List Listserv (Now the DIY Speakers). He has done a great deal of quality work.
I have done only a very small fraction of the work he has done in home audio. I have spent most of my time on small Pro audio boxes.

From reading through the thread where Swerd describes how he set up the mic for measurement I find some of it surprising (Since we use the same tools).
I think some of it may be because he moved into a new space where he measures the drivers. It might be significantly larger and he is trying to take advantage of it.

For a microphone he is most likely using the Kim Giardin mic that can be purchase with the Liberty Instruments analyzers back to LAUD (DOS MLS)

Too Tall clist
04-28-2007, 09:22 AM
Hi Swerd,


Too Tall
Greetings from another L-100 looney.

If you look on my post for the unfiltered LE5-2 midrange trace, you'll see a big difference between it and what you measured. I guess that is an example of the differences in drivers you mentioned. My midrange did not go low enough for a 2nd order crossover at 300 Hz. As a result, it was crossed at 1,000 Hz. The woofer is quite smooth there, at least on axis.

Do you have a schematic of your crossover?


I did three pairs of boxes and have three different xovers. 
I will see if I can figure out where everything is and put something coherent together.
Part of the problem is I switched computers and Ingemar updated LspCAD.
As a result in older filter designs when the program goes to look for the driver measurements for the model it can’t find them.
In some cases it was because I was testing a beta version and things just got scrambled. During some of it LspCAD version 6 was replacing the old V5.25 Pro.

One of the toughest jobs to learn when you do measurement and modeling is being a disciplined secretary so you label the data with an appropriate name that makes sense to you three years later and then file it correctly.



Giskard & Earl K
Gentleman, thanks for remembering my post. It makes me feel good this Friday afternoon. Did you really submit it to JBL? I'm flattered. However, since Harmon Intl. was recently sold, no one has yet contacted me about becoming the new CEO. What gives?


Zilch
I finally filed down those plumber's test plugs so they fit into the cabinet ports. They do very nicely.

Zilch
04-28-2007, 09:36 AM
I finally filed down those plumber's test plugs so they fit into the cabinet ports. They do very nicely.I spun the rear disc on a hand drill, as I recall, and it just took a minute or two to remove a bit of the outside diameter.

Tell a bit about the result, please. The boom is gone, no, but the bass is still balanced? I'm interested in how people perceive the result in comparison to how the change measures. :thmbsup:

Swerd
04-30-2007, 07:43 AM
I spun the rear disc on a hand drill, as I recall, and it just took a minute or two to remove a bit of the outside diameter.I tried to mount it on my drill, but the carriage bolt was too large to fit in the 3/8" chuck, and nothing else I had would keep that disc from spinning. So I clamped it in a vise and did it the hard way. I spent more time running up and down the stairs from the basement to see if it fit than I actually spent filing. :banghead:

Tell a bit about the result, please. The boom is gone, no, but the bass is still balanced? I'm interested in how people perceive the result in comparison to how the change measures. :thmbsup:I only did the filing about 2 weeks ago! Yes, I get the lazy man's award! Right away, I could tell the boom seemed like it was gone. Although, I'm still listening to various kinds of music, my initial feeling is that the balance is much improved. I have not done any measurements to see how level the bass has become. But so far nothing has sounded like there was not enough bass. It doesn't have that "dry" sound that I hear with sealed speakers that have Q in the 0.5 to 0.7 range. "Dry" is my poor attempt to describe this sound. What do vented high-Q speakers sound like? I don't think "wet" is right :blink:.

With classical music (Beethoven's 7th symphony) the lack of balance, since I put in the new crossovers, with the port open, was obvious and sounded wrong. With the plug, the balance sounds much better and seems to make things more articulate. The bass fiddles sound more distinct and less muddy. A clear improvement.

With various rock and jazz, the JBL boom was less objectionable to my ears, but remember, I had listened to these speakers, unmodified, for a long time! So the difference may be less obvious to me. It depends on the music itself. On some bass heavy rock, plugging the ports has removed the overemphasized bass. On one recording, I noticed that I heard the drummer more distinctly. I guess the bass hump had previously masked that. I'm still listening, and so far the results are pleasing.

Now that I am thinking about this, I have a recording of a solo piano (Linus & Lucy by George Winston) that I should pull out and listen to. With the new crossover and the port open, the grand piano sounded way too bassy. I'll check tonight - that may be a good demo of the Zilchplugz™.

Have you ever measured T/S parameters of the 123A woofer to see what size cabinet was best?

Swerd
04-30-2007, 08:18 AM
From reading through the thread where Swerd describes how he set up the mic for measurement I find some of it surprising (Since we use the same tools).
I think some of it may be because he moved into a new space where he measures the drivers. It might be significantly larger and he is trying to take advantage of it.

For a microphone he is most likely using the Kim Giardin mic that can be purchase with the Liberty Instruments analyzers back to LAUD (DOS MLS)Some of those details, especially the microphone placement, were from my incomplete, and perhaps faulty, memory. Dennis also did a lot of work when I was not there, and I'm not at all sure where else he moved his mic. Please don't read too much significance into those particular comments.

You may have read references to the so called Zilchplugs in this thread. I realized that you man not have known what we were talking about. Zilch, a frequent poster here, pointed out that the L-100 speakers used a vented cabinet that was deliberatly too small, and as a result, there was a prominent bass peak in the 50-80 Hz range. Zilch used a simple method to seal the port vent and tame the bass peak. He bought a 3" diameter plumbers's test plug. I found them at Home Depot for less than $4 each. Here is a link (http://images.google.com/imgres?imgurl=http://www.hardwarestore.com/media/product/195651_front200.jpg&imgrefurl=http://plumbing.hardwarestore.com/52-300-test-caps-and-plugs.aspx&h=200&w=200&sz=13&hl=en&start=4&tbnid=Vv0lyBL0hHrmuM:&tbnh=104&tbnw=104&prev=/images%3Fq%3Dplumbers%2Btest%2Bplug%26gbv%3D2%26hl %3Den%26sa%3DG) with photos. Because the vent in the L-100s are a little larger than 2¾", some filing is necessary.

Zilch
04-30-2007, 11:37 AM
Have you ever measured T/S parameters of the 123A woofer to see what size cabinet was best?I've cranked them through Bass Box 6 Pro a couple of times, which produces "Suggested" alignments for vented (Hi-Fi and Extended Bass) and closed boxes. There's four versions of that driver, actually:

1) 123A:

CB = 3.34 cuft (gross internal volume)
HF = 12.68 cuft, 19.3 Hz
EB = 17.06 cuft, 18.07 Hz

2) 123A-1 (2212) Not in the BB6P database, alas.

3) 123A-3 (2213):

CB = 2.354 cuft
HF = 8.71 cuft, 19.3 Hz
EB = 11.77 cuft, 18.07 Hz

4) 2213H

CB = 2.365 cuft
HF = 8.76 cuft, 19.3 Hz
EB = 11.83 cuft, 18.07 Hz

They're closed-box drivers - EBP = 48

The calculated alignments for maximally flat vented boxes are impractical, not only because of size, but also really low power-handling capabilities. That's not to say that some intermediate tunings aren't workable, tho.

Gordon W did some fun stuff with bandpass subwoofer designs using these drivers. There's territory to explore there, as well.

I've put them (123A-1) in L100t3 cabinets and closed the ports, 3.3 cuft. Very nice; a bit of boominess remains, but I think I can fix that with more fill. :thmbsup:

Zilch
04-30-2007, 11:48 AM
If you measure each on axis -
The 12” goes up to about 5kHz flat. (Rather amazing really).Well, no, I don't believe.

You need to measure the woofer (and port) nearfield or both together ground plane to see what's really up.... :yes:

4313B
04-30-2007, 11:50 AM
They are what they are.

See posts by Don McRitchie and Harvey Gerst.


I've put them (123A-1) in L100t3 cabinets and closed the ports, 3.3 cuft. Very nice; a bit of boominess remains, but I think I can fix that with more fill.

Naturally the first thing to do would be to measure the system Q so you understand where you are at.

Zilch
04-30-2007, 12:04 PM
Naturally the first thing to do would be to measure the system Q so you understand where you are at.Yup, still have to study them more.

I have a brand new pair of 2213Hs to try in them, as well. It'll be interesting to document the differences....

4313B
04-30-2007, 12:10 PM
Yup, still have to study them more.Ick! Sorry :( Just the thought activates the gag reflexes.
I have a brand new pair of 2213Hs to try in them, as well. It'll be interesting to document the differences....Maybe to you and a few others but I'd personally be spending my time with "better" transducers to begin with. Try Greg's 128H-1 for instance if you like vintage. Anyway, I know you won't be disuaded so carry on. We will wait as patiently as we can for the results.

boputnam
04-30-2007, 12:42 PM
Ick! Sorry :( Just the thought activates the gag reflexes.Maybe to you and a few others but I'd personally be spending my time with "better" transducers to begin with. Try Greg's 128H-1 for instance if you like vintage. Yea, I myself have posted a couple times on the merits of that transducer over the 2213H. There is no point in comparing...

Zilch
04-30-2007, 01:14 PM
Well, nobody's stepped forward with experience using L100/4311 woofers in their "correct" closed box alignment, so I'm doin' it.

[There's certainly plenty available.... :p ]

4313B
04-30-2007, 01:29 PM
Well, nobody's stepped forward with experience using L100/4311 woofers in their "correct" closed box alignment.Well, the 1st clue might be that nobody's stepped forward with experience using the L100/4311 woofers in their "correct" closed box alignment...
so I'm doin' itImpervious to embarrassment. :)

Zilch
04-30-2007, 01:34 PM
Impervious to embarrassment. :)


:D

[I'll make a point of playin' them for Bo when he comes over.... :yes: ]

4313B
04-30-2007, 01:58 PM
:D


[I'll make a point of playin' them for Bo when he comes over.... :yes: ]Outstanding! :p