L-200 woofer changes

[mbottz]

I have a set of empty 200 cabinets which I am looking to fit with components. I have aquired a set of 2225H's which are quite nice and was wondering if these would be a good fit? Looking at the specs they say 30 to 2K with a recomended crossover of 12K. This seems to fit the criteria of the original crossover parameters. I need some expert opinions here as I do not wish to spend a lot of time tweaking these cabinets. I would rather enjoy listening to them. If they are a reasonable fit, should I plug one port or leave them as is?

Feel free to chime in here Zilch as I know you have lots of experience in this area.

Thanks in advance for your help

Mike

[spkrman57]

But since you have all the parts, give it a shot and report back to us.

I would not expect a lot of bass response, but should sound decent enough.

Mostly, it comes down to what you need it to do!;)

Ron

[jbl]

Try it with the 2225. I have the 2225 in the 4507, 5 cubic foot cabinets. About the same size cabinet as yours. Sounds great. Produces plenty of bass.

[Zilch]

For 2225H, don't plug a port, which would tune it too low. Just swap in the new drivers and enjoy.

They will not have the extended bass that 2235H and retuning the cabinets would provide, but they'll have plenty of "kick."

What're you doing for the high end then?

[Mr. Widget]

I have a set of empty 200 cabinets which I am looking to fit with components.How empty? Meaning which crossover will you use, one of your own design or an off the shelf model?

What compression driver/horn will you use?

A 2225 will work depending on the answers to the other two questions... I would recone them with 2235 kits myself, but if you are more into head banging and loud than hi-fi the 2225 is a better choice than a 2235. Conversely...


Widget

[mbottz]

I can either use the LE85's or 2441's I have the short horns for the 2441's and the h91 for the le85. I wish to have more hi-Fidelity, as I listen to artists with talent, as well as some vintage Rock and roll. I like solid Bass but not boomy. If I can do something simple to better tune the enclosure like building a box of some type to reduce the cabinet size I can do that easily. I have the stock crossover for 12K but am not apposed to building a custom unit if the benefit warrants it. Keep the suggestions comming as I wish to only do this once if possible.

Thanks all for the great Advice

Mike

[Zilch]

The standard tuning of L200 is fine for 2225H. If I get time this weekend, I'll load one and measure it.

Assuming the efficiency of 2225H and LE15B are similar, the L200's LX16 crossover (1.2 kHz, actually,) may work. I'd keep LE85 with H91.

BUT, your real problem for Hi-Fi is the VHF. You need to add a tweeter, perhaps 2404H or 2405H (077), and an additional crossover, 3105 (N7000, N8000). The "Keeper" crossover developed in these forums is also an option, if you want to go three-way, and Giskard posted a version of the L300 crossover that might be adapted, though I haven't tried that one myself.

L200 has poor high frequency response, and attempting to boost it as in L200B runs up against a drastically narrowing vertical beamwidth with H/L91. I can try that, too, and see. It may work well enough.

I don't know if 2441 on the short horn can be "pushed" to play VHF, or how it behaves. Adding the tweeter is the best answer, unless you're willing to get radical using a different horn in a quasi-4430 two-way conversion....

[Mr. Widget]

I have the stock crossover for 12K...I assume you mean the 1200Hz LX16. This crossover was designed for the 4 ohm LE15B... I absolutely wouldn't use it. I also wouldn't use ANY 15" woofer above 800Hz in a Hi-Fi application.


Widget

[mbottz]

Is there a fairly simple modification I can do to the LX 16 to bring it down to 800hz and mate with the 8ohm driver. Then I can start looking for a 2405 / 077 with 7000 crossover to add to it.

Keep it simple for us novices!!

Thanks

Mike

[Zilch]

I'd use 3110 for a clean shift to 800 Hz.

Widget's right. The frequency will only go higher with a higher impedance driver.

I'll measure it, anyway, to confirm....

[Zilch]

LE15B Fs = 27 Hz, Zmin = 4.10 Ohms
2225H Fs = 47 Hz, Zmin = 7.09 Ohms

2225H in L200 Cabinet:

Fm = 34.24 Hz
Flo = 24.16 Hz
Fhi = 63.94 Hz
Fc = 60.25 Hz
Fb = 32.28 Hz, the box tuning with 2225H in it

I can only approximate the LF response with a nearfield measurement between the ports and the woofer (top photo).

Red = no filter
Green = with 3110

Then, normal 1M measurements:

Cyan = add LE85 on HL91. The dip at 650 Hz is not corrected by reversing phase. Time alignment in the box may be better. Woofer and horn are kinda far apart in my test setup, as well.

And finally, add 3105 and 2405H (courtesy Johnaec) = Blue

It sounded quite "horney" until I added the 3105 and UHF driver. 3105 has a lowpass filter to shut down the LE85 above the 7 kHz crossover frequency.

Bass is good and strong, but you'll want a sub or switch to 2235H for the extended bass this box can produce....

[jbl4ever]

Check to see if you have a LX 16 or 16A. The 16A uses a .7mh coil and a
6uf cap part #10296 if you can see the number on the cap. To rebuild one
with better parts cost about $50 with Solens,Audiocap and Mills. Does help the Xover.

[Zilch]

O.K., here's what BB6P says we've got (black) as opposed to 2235H with one port closed (red).

The difference is major.

JBL's recommended 40 Hz tuning for 2225H (blue) would require shortening the ducts.

BB6P suggests 35 Hz tuning for 2225H in this size box, intermediate between the black and the blue.

[Zilch]

So, can I confirm the BB6P sims via measurement?

Using sinusoidal rather than MLS, I can read down to 10 Hz, and it does better measuring vented box nearfield, ~8" centered between port(s) and woofer.

2225H, both ports open, as above, Red.

With both ports open, 2235H tuning is 34.43 Hz, roughly equivalent to 4430, Orchid.

One port closed, it's 24.66 Hz, extended bass, Blue.

For comparison, LE15B, Green, both ports open per stock L200(A), has the extension of 4430 plus 3 - 4 dB of "boom" from 60 - 200 Hz. :D

[Zilch]

Can LE15B be tamed by retuning the box?

Only minimally. It can achieve greater bass extension, with one or both ports closed. There are likely other limitations, such as cone excursion, in these alignments, tho.

All done with this, now. 2235H is back in, and one port closed.... :thmbsup:

[mbottz]

I appreciate your hard work and testing, I am not sure exactly what all this means, but I will tell you what it looks like to me and you can tell me if I am close.

Looks like the 2225H will work fairly well, but will not go as low in the bass area as the 2235. I should use the 3110 crossover or other crossover that crosses at about 800hz. Tuning will be about 35 htz with stock config of 2 ports. Closing 1 port with the 2225 would push the tuning down below the usable response of the 2225. I should install the LE85 over the 2441 and add an 077 for good highs. Am I anywhere close?

Thanks

Mike

[Zilch]

Exactly, and 3105 is your UHF crossover. It's important to roll the LE85 off with low pass; something simpler for the tweeter like a series capacitor won't get it.

If you're up to building the crossovers yourself, I can do some more work with the "Keeper" design for your system. It needs some tweaks to move 077 up to 10 kHz, and perhaps optimization of the Zobel for 2225H.

Your box tuning with both ports open is more like 32 Hz than the 35 Hz BBB6P recommends as "optimally flat," but that (32 Hz) gets a bit more bass.

Between the lines, I'm advocating 2235H, but you have a nice pair of 2225Hs to work with. Use them and add a sub, if desired, or watch for a nasty pair that you won't feel bad about reconing to 2235H in the future.

Thank you for affording me the incentive to look closer at L200 this weekend. I learned a buncha stuff.... :thmbsup:

[mbottz]

Sure if you have some extra time I have always wanted to make a custom crossover. If you can tweak a design, I will attempt to build it.

BTW if I choose to look for a new set of woofers, which is better, the 2235 or the LE15 with the foam surrounds.

Thanks again

Mike

[jbl]

Zilch was good enough to run some tests for me a few months ago regarding the 2225H vs 2235H in my 4507 5 ft3 cabinets. Thanks Zilch! The result was with the 2235H, I'd be loosing 4 dB in the upper bass region. As a result, I would also need to pad down the 2425J. I didn't think this was a good idea since I'm using my system since 1985 and am very satisfied with the sound.
Why not try the 2225's in your cabinets before changing woofers? This way you can tell what you like/dislike of the system. Graphs don't always tell the whole story. You need to factor your room into the equation. In my audio room, I have a good bass boost in the right side and a steep loss in the left side. If I turn the balance full left and stand in the right corner, the bass is as good as it is with the right channel on while standing in the right corner. No sub needed.
BTW Zilch, the fs of the 2225H is 40 HZ, not 47 Hz.
Hope this helps.

Ron

[Zilch]

BTW Zilch, the fs of the 2225H is 40 HZ, not 47 Hz.Hi, Ron,

Thanks for your comments. :thmbsup:

I posted the Fs as 47 Hz from the CLIO impedance curve.

I'll run both 2225s on WT2 today and see.

I'm STILL looking for a pair of 116A or H for Minis that actually are Fs = 28 Hz, the published spec. They seem to all be off by about 10 Hz.... :p


BTW if I choose to look for a new set of woofers, which is better, the 2235 or the LE15 with the foam surrounds.Many design advances are incorporated into 2235H, making it the better choice by far.

Finding some tired or compromised 2225s and reconing with 2235H kits will produce "as new" 2235Hs. The motors and baskets are the same, so that's a good way to get there....

[mbottz]

Why not try the 2225's in your cabinets before changing woofers? This way you can tell what you like/dislike of the system. Graphs don't always tell the whole story. You need to factor your room into the equation. In my audio room, I have a good bass boost in the right side and a steep loss in the left side. If I turn the balance full left and stand in the right corner, the bass is as good as it is with the right channel on while standing in the right corner. No sub needed.
Ron

Great advise Ron,

Since my financials are not unlimited I will be trying this first. I am hopeful all will sound good to my tin ear.

Thanks all

Mike

[Zilch]

The 2225H pair here, per WT2:

#1 - Revc = 5.7411, Fs = 47.4519

#2 - Revc = 5.8826, Fs = 47.2248

So, by two different methods, now, they're Fs = 47 Hz.

The specs are a good starting point, but, especially when dealing with vintage used gear, it's useful to know the actual values.... :yes:


If you can tweak a design, I will attempt to build it.I'll try to do it this week, then. It's gonna take a bit of "research.":p

Mac needs it for his project, too:

http://audioheritage.org/vbulletin/showthread.php?t=12550

Here's the basic schematic. The HF's gonna be different, of course:

http://audioheritage.org/vbulletin/showthread.php?p=93836#post93836

Gonna cost about $250 the pair for parts, now. I doubt you can acquire a pair of decent 3110s and 3105s for that.

And here's how I built it:

http://audioheritage.org/vbulletin/showthread.php?p=105370#post105370

Here it is running 2405:

http://audioheritage.org/vbulletin/showthread.php?p=94077#post94077

Reviewing some of that, I find I did a direct comparison of this L200B-derived crossover against 3110/3105, and found it to be "smoother" sounding, whatever that meant at the time. ;)

In the course, I'll hunt up Giskards L300 DIY version, as well, so you'll have some options to consider....

[toddalin]

The 2225H pair here, per WT2:

#1 - Revc = 5.7411, Fs = 47.4519

#2 - Revc = 5.8826, Fs = 47.2248

So, by two different methods, now, they're Fs = 47 Hz.

The specs are a good starting point, but, especially when dealing with vintage used gear, it's useful to know the actual values.... :yes:

I'll try to do it this week, then. It's gonna take a bit of "research.":p

Mac needs it for his project, too:

http://audioheritage.org/vbulletin/showthread.php?t=12550

Here's the basic schematic. The HF's gonna be different, of course:

http://audioheritage.org/vbulletin/showthread.php?p=93836#post93836

Gonna cost about $250 the pair for parts, now. I doubt you can acquire a pair of decent 3110s and 3105s for that.

And here's how I built it:

http://audioheritage.org/vbulletin/showthread.php?p=105370#post105370

Here it is running 2405:

http://audioheritage.org/vbulletin/showthread.php?p=94077#post94077

Reviewing some of that, I find I did a direct comparison of this L200B-derived crossover against 3110/3105, and found it to be "smoother" sounding, whatever that meant at the time. ;)

In the course, I'll hunt up Giskards L300 DIY version, as well, so you'll have some options to consider....

And here's how I built (and prefer) the originals. (Three for L/C/R.) The schematic does not show the 0.01 Theta by-pass capacitors that are in parallel with all shown caps. Also, in reality, the resistors are 12 watt Mills.


http://www.largescaleonline.com/eimages/lsolpics/Team_Member_Pics/toddalin/Crossover.jpg
http://www.largescaleonline.com/eimages/lsolpics/Team_Member_Pics/toddalin/Crossover3.jpg

[jbl]

I'm using the 3110A with my setup. Transition between 2225 & 2425/2370 is seamless. If you can get the 3110A, that would be a good choice.

Zilch is the master on graphics. He'll steer you in the right direction.
BTW Zilch, JBL refused to recone my 2225 to 2235 when I asked eariler this year. Said,"Won't work". I left it at that with them. IIRC, same basket, motor, etc.

I've always felt that getting a sub is easier than building a system that has a great dynamic range (punch). The 2225 offers a very realistic performance. No one woofer will give you everything. The 2225 and 2235 are both great JBL 15" woofers. Different results can be obtained, depends on what you want. The 2225 will offer about 4 dB more output up to 500 Hz. The 2235 will offer about 6 - 8 Hz more in the low end along with greater output given the same cabinet.

Zilch can help with the technical aspect. I'll chip in with the 20 year experience with the 2225H. :D

Hope this helps.

Ron

[Zilch]

JBL refused to recone my 2225 to 2235 when I asked eariler this year. Said,"Won't work". I left it at that with them. IIRC, same basket, motor, etc.They were wrong, of course. Well, that PERSON was. :(

Same basket, same motor, etc., like you say.

I'm running a pair of them right now in 4507 cabinets.... :thmbsup:

[jbl]

They were wrong, of course. Well, that PERSON was. :(

Same basket, same motor, etc., like you say.

I'm running a pair of them right now in 4507 cabinets.... :thmbsup:
I know. Everything is the same. I'm sure you measured the reconed 2225H as 2235H against the original 2235H.:D
I got the impression that JBL is egar to brush off us users of the older gear. His response was too fast.

Ron

[Zilch]

I'm sure you measured the reconed 2225H as 2235H against the original 2235H.Nope, never did.

Giskard said they are the same.

[Don't hafta measure.... :p ]

[jbl]

Zilch, do you need to change the spider as well? They should be different for the 2225 and 2235.

[Zilch]

New spider comes on the 2235H cone kit.

The progressive spider is a major feature of 2235 "upgrade...."

[jbl]

Thanks Zilch. So the end result is a 2235H. On e-bay, some reconed 2225 to 2235 were considered 'Frankenstein' woofers.

Ron

[Zilch]

If they got the factory kit, they're 2235H, pure and simple....

[mbottz]

I aquired a set of L220 Crossovers very reasonably the other day and was wondering since these cross at 800hz and 5000hz, can I use these with any success until I get the custom units completed? Since the 220 used the cateye version of the 077 what is the drawback to adding a 077 and crossing at the lower frequency in the 200 cabinet using the drivers discussed?

I appreciate all the good advise. Keep it coming

Mike

[Zilch]

I wouldn't use it, no.

The impedances are different, and the crossover frequencies will shift....

[Mr. Widget]

...what is the drawback to adding a 077 and crossing at the lower frequency...Bleeding ears.... and pained looks from your friends.


Widget

[scott fitlin]

The 077 will sound rough and ragged crossed over at 5K. Personally, I wouldnt use the 077/2405 below 8K.

[Zilch]

Sure if you have some extra time I have always wanted to make a custom crossover. If you can tweak a design, I will attempt to build it.O.K., here 'tis, using L300's 8.5 kHz. A major change in addition to raising the crossover frequency is that the 077/2405 is now running with a 3-pole highpass filter which attenuates frequencies below that at 18 dB/octave, cleaning up its sound considerably.

The circuit shown is the "cascade" configuration. The UHF tracks the MF L-Pad adjustment. If you want the UHF and MF independently adjustable, connect the UHF circuit before the MF L-Pad rather than after. I tried it both ways, and both work fine. It's a matter of preference - you can dial in lot more unnatural UHF "sizzle" (and sibilance) at lower MF settings with the independent UHF connection.

If you separate the UHF entirely, as Toddalin suggested above, it's necessary to add fixed attenuation to bring the UHF L-Pad into a resonable adjustment range.

Giskard's L300 circuit is here for comparison:

http://audioheritage.csdco.com/vbulletin/showthread.php?p=1042#post1042

[toddalin]

The circuit shown is the "cascade" configuration. The UHF tracks the MF L-Pad adjustment. If you want the UHF and MF independently adjustable, connect the UHF circuit before the MF L-Pad rather than after. I tried it both ways, and both work fine. It's a matter of preference - you can dial in lot more unnatural UHF "sizzle" at lower MF settings with the independent UHF connection.

If you separate the UHF entirely, as Toddalin suggested above, it's necessary to add fixed attenuation to bring the UHF L-Pad into a resonable adjustment range.



I don't agree that it is necessary to add fixed attenuation. Using the L-pad alone, the UHF can be adjusted over the full range, from nada to too bright.

I actually prefer it up toward the brigher side and that's the way my RTA says to run it (though I don't put a lot of stake in my RTA up in that range as I think the mic rolls off).

As the 077s are even less efficient than the 075s I used, and crossing it over about 8.5K (with smaller chokes and using an additional cap), as opposed to the ~7.25K I crossed it over at would further reduce its volume and any additional attenuation, beyond the L-pad, is really unnecessary.

It seems to me that the only impact of adding fixed resistance would be to change the position of the L-pad, and add additional components into the signal path. My thought on crossover design, and purity of electronics in general, was that "less is more." It seems to me that the fewer components that a signal has to go through, the less chance there is for each component to leave its own sonic mark.

[Zilch]

Using the L-pad alone, the UHF can be adjusted over the full range, from nada to too bright.Yes, of course. HOWEVER, it's difficult to get a precise or reproduceable adjustment in the lower half of that range.

Plot the attenuation curve of an adjustable L-Pad vs. rotation for us, please.

I'd suggest the "signature" of a fixed L-Pad attenuation, is, uhmmm, nil.

In any case, here's the 2405H adjustment range with the UHF filter direct-connected to the input, LE85/HL91 midrange balanced with 2235H woofer. It needs about 4 dB additional padding, just about what the fixed pad in the MF circuit provides. Blue is at 9:00, i.e., ~20% "on."

That's why I favor the "cascade" configuration. Balance the UHF with the MF, then adjust MF to balance with whatever woofer, and they will track. It may be necessary to alter the fixed L-Pad (R2, R3) to balance both with a substantially more or less sensitive woofer.

Having just heard it, I can assure you there's decidedly scant merit in playing 2405H 6 dB "hot" (red, max).... :eek:

[Zilch]

To complete the picture, here's the midrange adjustability, below.

How do these sound?

Quite nice, in fact.

HF is clean and natural, unlike my previous experience with 2405H crossed lower. Playing the slot alone, the vertical beaminess is obvious. Gotta be in the "window" to hear them at all.

MF is very dynamic, a hint of horniness remains. I don't have any H92s to compare. Some resonance when Patricia Barber wails, but that may be from horn and driver not being mounted fixedly to a baffle.

LF? Well, it's 2235H in 5 cuft. @ 34 Hz. End of THAT story.... ;)

[Earl K]

Zilch,

- Up in post #36 there's an @ 7db hole centered around about 725 hz . It's quite wide.

- What's going on there ?

- Does that dip really exist / or / is it just a funny artifact of the way CLIO pastes together 3 separate traces ?
- ie ; if you were to put your Behinger 1/3 octave RTA on it / would you still "see" that hole ?

:)

[Zilch]

Zilch,

- Up in post #36 there's an @ 7db hole centered around about 725 hz . It's quite wide.

- What's going on there ?
Hi, Earl!

I've fought that one before.

Look at the cyan there -- the two (LF & MF) are not summing acoustically. I believe it's because the woofer and horn are nearly two feet apart, and there's a 4" lip at the top of the cabinet.

I can't get back far enough for them to sum, apparently, under these conditions.

3110 XO did the same thing earlier in this thread, so I'm ignoring it. Flipping the phase on the MF makes it worse.

I'll move over onto a 4507, which has the woofer up near the top, and test that hypothesis....

[Earl K]

Look at the cyan there. The two (LF & MF) are not summing. I believe it's because the woofer and horn are nearly two feet apart, and there's a 4" lip at the top of the cabinet.

So does this cyan trace represent a "true" warts & all, fully summed response in CLIO ?


I can't get back far enough for them to sum, apparently, under these conditions.

You shouldn't need more than 3 or 4 ft to obtain the effects of proper summing.


3110 XO did the same thing earlier in this thread, so I'm ignoring it. Flipping the phase on the MF makes it worse.

Please, it's called "flipping the polarity" / you've got Bo's' disease . ;)

:)


ps ;
- I think you need to raise the lowpass point for these crossovers. I see, quite consistently, the same "hole" showing up across the board ( also with dmtp(s)' le14s ) .

- As an exercise, calculate the F3 point for each of your individual passive components ( in the lowpass section ) . ( Your Zobel sets the impedance to 7 ohms / so use that as your load impedance .) Then multiply those two frequencies together . Then take the sqrt of their product / & / then let's talk . ;)

result ;

- For 7 ohms / I get ; 506.4hz * 947.35hz = 479738.04
- the sqrt of ( 479738.04 ) = 692.63 hz
- This is not the F3 point since the "poles" are less than 1 octave apart. It is the intersection of the 2 reactive impedances. In this case, the F3 will be slightly higher / though I don't yet know the formula to derive the real F3 .

[Zilch]

I DID hafta flip the polarity of the MF, tho.... :hmm:

All I'm doin' is turning off the various drivers to get the individual curves.

Orchid is full range, i.e., all of 'em running.

Area of interest, bottom.

I COULD dial it to 800 Hz, if I cranked up the MF a bit.

Cursor calls that "X" point 886 Hz.

Raising the MF 2 dB makes it 827 Hz.

This LE85 shows a fairly noxious peak at 1350 Hz, tho.

Original circuit calls for 7.5 Ohms in the Zobel. I spec'd 7.0 Ohms because PE don't stock 7.5.

[It's not critical, according to Elliot, but ideally, Zobel'd be tweaked for Le of the specific woofer used and the desired response characteristic....]

[Earl K]

- Thanks for doing that ! It looks much better now .

- Ah yes, polarity . That almost perfect "V" ( of a suckout ) was a good hint .

- The nasty peak at 1350 hz ? Well, if it was my circuit / before I went and designed a series LCR ( strapped in parallel ) to attenuate it / I would explore just what percentage, the present hipass "bump-filter" is contributing to that "nasty" .

- I calculate the reactive center point ( Fcr ) of that 3 pole bump-filter to be @ 1265 hz ( at 6.3 ohms ? ) . That's not that far from your measured 1350 peak .

- For laughs, I'd swap in a "text-book derived", 850 hz, 3 pole Butterworth hipass and then remeasure. The comparison will show the effect of the present filter . This empirical study is ( perhaps ) still easier/cheaper to implement ( parts-wise ) than building up a "LCR fix". The text-book design does require some fairly large value caps .

- You may find that a new bump-filter that is "broader , with less gain and designed around a slightly lower center frequency" / offers enough levelling to be considered a decent compromise ( without resorting to LCRs ) . I have one waiting in the wings for you to try ( if you're so inclined ). It uses all of the exisitng passives values / you'll just need to add in a 10uF cap to the exising 24uF / and lower the final load impedance to 5.5 ohms ( change the buildout resistor in the fixed Lpad ) to make it work properly . This filters "Fcr" is around 1177 hz / so / it's "peaking" only slightly lower .

:)

[Zilch]

Ah yes, polarity . That almost perfect "V" ( of a suckout ) was a good hint.I SWEAR I tried that in the original setup, as well, but I wouldn't put it past myself to have merely reinserted the leads into the samer terminals. It's all back on the L200s again, so I'll get another chance.... :thmbsup:


The nasty peak at 1350 hz ? Well, if it was my circuit / before I went and designed a series LCR ( strapped in parallel ) to attenuate it / I would explore just what percentage, the present hipass "bump-filter" is contributing to that "nasty".I think the first thing I'll do is try a buncha different LE85s to get a statistical "feel" for whether the 1350 Hz peak is coming from the filter or the driver/horn/lens. I'm reluctant to start tweaking the crossover circuit until I establish it as the source. I've made plenty of crossovers that work beautifully on just one driver here. :p

Also want to take a more detailed look at whether cascading the UHF on the MF yanks the main MF highpass around any. First look said not, but I'm not 100% confident of that.

In the course, I'll try your mod and show what happens. It's clear I can't put this away just yet.... ;)

[Earl K]

I think the first thing I'll do is try a buncha different LE85s to get a statistical "feel" for whether the 1350 Hz peak is coming from the filter or the driver/horn/lens. I'm reluctant to start tweaking the crossover circuit until I establish it as the source. I've made plenty of crossovers that work beautifully on just one driver here.

- That's a wise approach. This forum is lucky to have someone such as you ( with enough LE85s ) and the test gear ( and the willingness to use it ) to get to the bottom of this question.


Also want to take a more detailed look at whether cascading the UHF on the MF yanks the main MF highpass around any. First look said not, but I'm not 100% confident of that.

- It's worth the look. The cascade approach / should offer more overall impedance isolation / than the "traditional" topography . I'd suggest running WT2 impedance studies, for both wiring scenarios.


In the course, I'll try your mod and show what happens. It's clear I can't put this away just yet....

- Well, LCRs do eat up resources / its better to know if they are truly needed .
- OTOH, NP electrolytics bypassed with 1 uF polyfilm caps are a cheap enough alternative . Coils can be "right-sized" towards the correct impedance /further driving down their cost ( ie 22 gauge wire types or smaller ). I'm not sure if you remember / but Giskard showed ( in a nice graphical comparison ) that there's no effective difference ( in the notch ) at how one arrives at the correct total value for "R" ( in the LCR ) .



:)

[toddalin]

Also want to take a more detailed look at whether cascading the UHF on the MF yanks the main MF highpass around any. First look said not, but I'm not 100% confident of that.

In the course, I'll try your mod and show what happens. It's clear I can't put this away just yet.... ;)

When I tried the three wiring schemes (i.e., cascaded after L-pad, cascaded before L-pad, and non-cascaded, I found some minor variation in drive levels and crossover freqs for the mids and highs. If I remember correctly, the variation was less than 20 Hz.

Readings were obtained driving the crossover with a function generator and a/c voltage readings obtained with a digital multi-meter.

[Zilch]

Thanks, Todd. I confirm your findings:

First study: Does cascade versus direct drive yank the slopes or crossover points around?

Nope (top).

Not at UHF XO, either (middle).

There are some minor variations in the bandpass area, tho (bottom).

I'm back on L200 with behavior similar to previously. Outta phase summing (subtraction, actually,) is shown in black.

Dip in measured 2235H response slips the XO frequency down to 754 Hz. Some of that may be room stuff. I can't get decent windowed quasi-anechoic MLS in the 700 - 800 Hz range....

[Zilch]

Second study: The 1350 Hz peak.

Four LE85s with the same horn and lens.

Orange is the red one on a different HL91, same vintage.

There's a peak at 2 kHz, too. Which is bigger is driver dependent.

I'm callin' it "Ripple."

Compare to the original measurement here:

http://audioheritage.org/vbulletin/showpost.php?p=130246&postcount=36

A broad, shallow notch from 4 - 8 kHz would tame any perceived over-brightness.

I don't hear any.

Ship it? ;)

[I'm building up a pair of DMTP XOs. We'll see more over there when they're done.... :thmbsup:]

[Earl K]

Zilch ;

re; post #48,

- It's all very interesting . Nice work !

- It still appears to me that the lowpass section of this network needs work . - IMO, the effective F3 needs to go upwards ..
- Moving the lowpasses F3 upwards, will worsen the "bump" that exists in the 2235H stopband ( down around 950 hz ). This should make worse the existing ripple of the composite response in the 1K area. One solution would be a series LCR ( as a conjugate ) set to 950 hz and inserted between the existing passives ( in the lowpass ) and the woofer . This is a tricky area to work in since it's so close to F3 .

re; post #49,

- The ripples ?
- I see spots for 4 Series LCRs within that horn circuit ( at the present magnification ) .
- What does a lower-rez shot look like, such as one from your Behringer ?
- Sometimes it pays to squint ones' eyes ( to figure out if this stuff is worth going after ) .


Ship it? ;)

- Sure, ( if you stop showing us these high-rez shots of all the "hills & dales" ;) )

:p

[Zilch]

- What does a lower-rez shot look like?Squint away.... :D

[Zilch]

mbottz: I think we're done, Mike. :p

Build ON!

L3 is 18 Ga, PE #255-216

L4 is 20 Ga, PE #255-024 unwound to 0.16 mH

I'm running with UHF connected before the MF L-Pad right now.

PM me if you have any questions.

Thanks again to Johnaec for loaning me the 2405Hs.... :thmbsup:

[mbottz]

mbottz: I think we're done, Mike. :p

Build ON!

L3 is 18 Ga, PE #255-216

L4 is 20 Ga, PE #255-024 unwound to 0.16 mH

I'm running with UHF connected before the MF L-Pad right now.

PM me if you have any questions.

Thanks again to Johnaec for loaning me the 2405Hs.... :thmbsup:

Thanks for all of your help on this Zilch. Let me see if I have the recipe right. Attached is the schematic to build with the modifications for L3 and L4 as posted above. This will work well with my 2225H's, LE85's, L91horns, and an added 077 tweeter. Am I understanding you correctly?

Thanks

MB

[Zilch]

Nope, this be the one: