PDA

View Full Version : Can You Design Me a Crossover That:



toddalin
12-29-2005, 09:18 AM
Can better integrate the 2235 with the LE175 on an HL-91 with an 075.

Changing out the 130As to 2235's left a "hole" in the area between 500 - 1,000Hz. To me, this is perceived as a muffling of peoples voices.

I am currently using an N1200/N8000. I would like to crossover the woofer to the horn at about 800 Hz. This should be high enough to let the horn operate within its range while reducing the size of the "hole." A two-way that retains the N8000 would work fine. All components are 8 ohm.

Thanks

speakerdave
12-29-2005, 10:13 AM
You might be able to build one yourself. You could begin by looking at this thread about the 3133, the crossover for the 4333/L300. Your list of equipment is similar, although there are differences.

http://www.audioheritage.org/vbulletin/showthread.php?t=166&highlight=3133

I'm not sure if there is a free crossover designing service available here. I think there probably would be help available if someone started out on his own and ran into some glitches, but that's just my guess. Our new forum is called DIY, not DIFM.

David

John
12-29-2005, 10:42 AM
Our new forum is called DIY, not DIFM.

David

:applaud: :applaud: :applaud:

toddalin
12-29-2005, 11:28 AM
Thanks for the link. I should be able to adapt the L300 network.

There are limits to DIY and I would still be physically building the crossovers. I just figured that if the design was out there, why reinvent the wheel? :blink: Even with the L300 design, I would probably just use the two-way section and still retain the N8000s just to fill in the hole in the back of the cabinets.

edgewound
12-29-2005, 12:03 PM
Todd...

You should probably look for LE85/2420 or 2425/6 if you want the midrange to be clean and trouble free to work with a network such as for an L300. I'm afraid you'll find the LE175 won't be up to the task...it was meant to work with 130A or D130. Just my opinion...might save some future headaches.

toddalin
12-29-2005, 12:16 PM
Edge, it's certainly a future possibility, but three nice LE85's don't come cheap and and HL92's are scarce. (Wasn't easy to get wife to go for $1,050 just to redo the four woofers.;) )

I assume that I can use higher wattage resistors than indicated (e.g., substitue a 25 for a 20 watt resistor) so long as I retain the proper resistance.

speakerdave
12-29-2005, 12:18 PM
I just figured that if the design was out there, why reinvent the wheel?

Indeed. My wisecrack was a little out of line, but I couldn't resist. I should talk, anyway. The whole crossover thing has given me the hesitation blues.


Even with the L300 design, I would probably just use the two-way section and still retain the N8000s just to fill in the hole in the back of the cabinets.

It seems to me Ian has suggested that cascading crossovers, although apparently JBL-recommended practice in the past with prebuilt units, can involve unnecessary compromises when building from scratch. In the old charts ('70's) JBL treats the 175 and 85 compression drivers as interchangeable vis-a-vis horns and crossovers for HF and the 075 and 077 also when used for UHF. It might be better to go for the whole 3133, either using the charge-coupled or bypassed version. (And then pay for the parts by selling the N8000's on ebay.) Anyway, that would be a specific question I would post for the filter wizards.

David

speakerdave
12-29-2005, 12:24 PM
Todd...

You should probably look for LE85/2420 or 2425/6 if you want the midrange to be clean and trouble free to work with a network such as for an L300. I'm afraid you'll find the LE175 won't be up to the task...it was meant to work with 130A or D130. Just my opinion...might save some future headaches.
The 130A and the D130 both have much higher output than the LE15 or 2235. The sensitivies of the 175 and the 85 are listed as identical. The differences would be in UHF extenstion. Either is adequate when crossed over to a supertweeter.

David

speakerdave
12-29-2005, 12:33 PM
. . . and HL92's are scarce.

I have a pair of H92's sitting in my garage, black, crinkle paint, nice and clean; one of the foilcals has gone missing. I'm afraid I have $175 into them, but they're yours for that plus transportation if you want them.

David

edgewound
12-29-2005, 12:40 PM
The 130A and the D130 both have much higher output than the LE15 or 2235. The sensitivies of the 175 and the 85 are listed as identical. The differences would be in UHF extenstion. Either is adequate when crossed over to a supertweeter.

David

I understand...but my thinking is that the bigger motor will better handle the lower crossover point and the transients that go with home theater sound tracks. Like I said...just my opinion. Also 2425's are still reasonable on ebay, etc.

speakerdave
12-29-2005, 12:47 PM
. . . . my thinking is that the bigger motor will better handle the lower crossover point and the transients that go with home theater sound tracks.
Could be true.

Either driver will be better off with the longer horn at 800 Hz.

David

toddalin
12-30-2005, 02:07 PM
Could be true.

Either driver will be better off with the longer horn at 800 Hz.

David

If I were to change, I would need at least 3.

To be honest, I had the woofers out of the L200 cabinets the other day and with the LE175/HL91, there only appears to be about 1.5" from the back of the driver to the back of the cabinet. (I actually have a piece of styrofoam wedged between the driver and cabinet back.) An HL92 is reported as being 3" longer than an HL91 and is reported to be used in this cabinet..., but I can't see how. Add to that that the LE85 has a bigger diameter and because of its angle, this also affect mount depth, how could it fit???.:blink:

norealtalent
12-30-2005, 02:23 PM
Partial
...An HL92 is reported as being 3" longer than an HL91 and is reported to be used in this cabinet..., but I can't see how. Add to that that the LE85 has a bigger diameter and because of its angle, this also affect mount depth, how could it fit???.:blink:
I believe HL92 is 4" longer. It will not fit a stock L200 cab. :bouncy:

Zilch
12-30-2005, 03:40 PM
No go, Todd. That's the problem.

And punchin' the driver out the back is, ummmm, well, you know....

Upgrade to LE85/2420 later.

edgewound
12-30-2005, 04:44 PM
Maybe a compromise solution would be to crossover the HL91 at 1000hz with 3rd or 4th order Linkwitz-Riley crossover...Maybe Giskard can advise on this. I read another thread about the OASR waveguide being useable down to 750-800hz, so why shouldn't the H91 be ok at 1Khz with steep enough slope?

edgewound
12-30-2005, 06:17 PM
Todd...

Following Giskards recommendation and doing a little searching through my service lit...the L200B uses the HL91 crossed over at 800Hz but uses the LE85 driver with a 136A woofer. The L200A also used an LE85 & LE15B driver crossed at 1200Hz with the same horn. So Giskard seems to know that the LE175 will be fine with the HL91 at 800Hz so go for it. Personally I'd opt to cross the LE175 at 1000Hz just for a margin of safety at higher power levels and transients with home theater sound tracks. Besides that...the 4430 (2235H) crossover is at 1000Hz and that seems to sound pretty good. Theres plenty of software available at Parts Express to design that crossover for you...and they have the parts too. Have at it!!!

Paul Joppa
12-30-2005, 10:37 PM
...
Changing out the 130As to 2235's left a "hole" in the area between 500 - 1,000Hz. To me, this is perceived as a muffling of peoples voices.

I am currently using an N1200/N8000. I would like to crossover the woofer to the horn at about 800 Hz. This should be high enough to let the horn operate within its range while reducing the size of the "hole." A two-way that retains the N8000 would work fine. All components are 8 ohm.

Thanks

First, the "hole" as others have noted may just be the reduced woofer sensitivity. The N1200 won't adjust the horn down far enough to match. It's also possible the different woofer HF rolloff changes its effective time delay, and a phase reversal of the woofer might help fill in.

OK, then here's my perspective on the technical history.

Originally, the LE-15/LE175 (S-7) crossed at 500Hz. My perspective - that was to keep the phase difference small enough to run the drivers in phase. (Plus, the LE-15 crapped out above 500Hz.) The 130A/175 combo at 1200Hz has a phase difference close to a wavelength so the drivers can be in phase at crossover.

The first monitor (4420?) upped it to 800Hz, probably to protect the 175 from damage. But the phase difference at 800Hz is about 1/2 wave length (assuming the woofer phase origin is some 4" behind the front panel). I don't know if they reversed the woofer phase for that or not. The longer H92 at 800Hz came closer to matching the 1200Hz crossover with the H91 horn, again allowing the woofer phase to be the same as the HF.

The crossovers, N500 and N1200, are pretty straightforward and easily adjusted to an 800Hz crossover - or just use the 4420 design, with the reversed phase as necessary. But do the calculations to get the sensitivity matched!

Zilch
12-30-2005, 10:56 PM
Paul: S7 was LE-85, and you mean 4320, not 4420, I would assume....


Can you design me a crossover that can better integrate the 2235 with the LE175 on an HL-91 with an 075?http://www.lansingheritage.org/images/jbl/catalogs/1979-comp/page08.jpg

Paul Joppa
12-31-2005, 10:10 AM
Paul: S7 was LE-85, and you mean 4320, not 4420, I would assume....

http://www.lansingheritage.org/images/jbl/catalogs/1979-comp/page08.jpg

Shouldn't trust memory at my age!

Anyhow, the LX5 was specified for LE175, LE85, and 375 in the earlier versions; in fact before the 4320 with LX13 at 800Hz, the studio monitor used the LX5 at 500Hz. All of these were specified for use with the LE175/HL91 at one time or another, though in later years the minimum crossover for the LE175 drifted up to 800, 100, 1200, and eventually 1500Hz in the "potato masher" horn. I always assumed that was because amps got bigger and people were blowing out the drivers.

Sounds like several classic crossovers could be adapted, and a few were actually meant for the 2235. AFAIK, anything that used the LE85 will work fine with an LE175, losing mostly a litle HF extension. The big problem is always the tapped inductor/autoformer in the crossover, so an original JBL on the used market would be a lot easier than a new scratch-built crossover.

Always wondered if there was much market for suitable tapped inductors. It should not be that hard to get some made up, if one could amprtize the design and tooling costs (and if JBL did not object!)

Paul Joppa
12-31-2005, 11:06 PM
Giskard- sorry, I should probably not have stepped in so fast after being gone so long. (Have no idea what an "LCS" is ...)

Back to occasional lurk mode for me!

Zilch
12-31-2005, 11:15 PM
I should probably not have stepped in so fast after being gone so long.Naw, stick with us here. We'll figure it out. :thmbsup:

LE175 was S6, or something. LCS was JBL "Loudspeaker Component Series," I think. Just found out about that recently in the EN3 thread....

Todd:

If LX200B is N200B, that's what I'd build, 'cause there's no tapped inductor in it, but leaving out the HF boost loop (1uF, 0.16 mH, 5 Ohms,) since you're adding on the N8000 and UHF driver. It's pretty simple. Add bypass caps. Done.

I can measure the inductor DCR's, if that's what you decide to do.... :)

http://manuals.harman.com/JBL/HOM/Technical%20Sheet/L200B%20ts.pdf

toddalin
01-01-2006, 11:54 AM
Naw, stick with us here. We'll figure it out. :thmbsup:

LE175 was S6, or something. LCS was JBL "Loudspeaker Component Series," I think. Just found out about that recently in the EN3 thread....

Todd:

If LX500B is N200B, that's what I'd build, 'cause there's no tapped inductor in it, but leaving out the HF boost loop (1uF, 0.16 mH, 5 Ohms,) since you're adding on the N8000 and UHF driver. It's pretty simple. Add bypass caps. Done.

I can measure the inductor DCR's, if that's what you decide to do.... :)

http://manuals.harman.com/JBL/HOM/Technical%20Sheet/L200B%20ts.pdf


Thanks Zilch.

I assume that the N200B is similar (at least in function) to the LX200B referred to by Giskard. There don't appear to be any tapped inductors in the N200B schematic. I assume that I can get the other parts at Parts Express? Would there be a way to bias this design using the 9 volt battery, and if so, any idea if the difference would even be noticable?

There is currently a pair on N200Bs on epay with a BIN of $185. Seems kind of steep for the value of the parts.

http://cgi.ebay.com/Mint-Pr-Vintage-JBL-N200B-Crossover-800-Hz-from-L200B_W0QQitemZ5846035736QQcategoryZ50597QQssPageN ameZWDVWQQrdZ1QQcmdZViewItem

BTW, in the mean time, I've found that 3 dB of PEQ boost around 1,000 Hz does a nice job of filling in the "hole" thereby restoring vocal intelligability.

Zilch
01-01-2006, 12:04 PM
There is currently a pair on N200Bs on epay with a BIN of $185. Seems kind of steep for the value of the parts.Yeah, and OLD parts, too. The ones you build will be better. We seem to have confirmed it's the one to do.

Parts Express, yes.

Lemme look at biasing here....

EDIT: Yup, biasing's gonna be easy, if we just do it in the HF section. I'll draw it up here, what I propose.

Key to getting it right is DCR of the 0.8 mH inductor.

$20 apiece more will bias the LF, as well, if we change the Zobel per N3134 (bottom). $30 apiece more if not, 'cause the big caps get expensive.

Bypass caps omitted for clarity. Them alone's gonna be $20 to $25 per crossover.

Check me, please:

Ian Mackenzie
01-01-2006, 08:30 PM
I would have thought with Spice and Clio that would be a piece of American Pie (I mean cake for the rest of the world)

A useful and well written article here for those interested to learn the finer points of passive crossover design incl Zobels and their effect.

http://sound.westhost.com/lr-passive.htm

But this is only the beginning.

There ain't no substitute for a good set of ears. The more I learn to appreciate and understand the original vintage JBL crossovers the more obvious the amazing skill of engineers like G.T and others is who painstakingly tuned and voice each system so carefully.

Zilch
01-01-2006, 10:31 PM
SO, Ian, you suggesting to go with the calculated Zobel, or the original design?

Per N3134 is the upper curve, vs calculated (middle) and original design (lower), all with 7.5 Ohms.

'Course, it's how it intereacts with the actual woofer impedance vis-a-vis the total network and system, in the end.

That's an "empirical" determination. :p

O.K., 2 X 47 uF, then.

Them 68's are $18.43 apiece. :(

OR, put in the 68's and try 6 Ohms for the calculated Zobel, 7.5 Ohms for the original spec....

Titanium Dome
01-01-2006, 10:41 PM
(snip)

But this is only the beginning.

There ain't no substitute for a good set of ears. The more I learn to appreciate and understand the original vintage JBL crossovers the more obvious the amazing skill of engineers like G.T and others is who painstakingly tuned and voice each system so carefully.

(With apologies in advance to the thread's contributors for going a bit OT. None of these comments are aimed at the present effort.)

That very thought always leads me to this conundrum:

JBL/Harman have the money, the facilities, the brains, the experience, the ears, and the theory to do just about anything they damn well please, and IMO their better efforts are going to have a very high likelihood of kicking the ass of anything we do; YET one always has the nagging suspicion that what happens in the lab gets mucked up in the translation to production by marketing, accountants, sales decisions, etc., SO we probably get some compromises that result in "it is what it is for a variety of reasons" comments by some of our more experienced hands. HOWEVER we all want to believe we can take a good thing and make it great with just a bit of tweaking, so maybe we can figure out where JBL/Harman cheaped out or took "good" instead of "great" and we can create a better way to do it. MAYBE my very serious effort to improve things will result in my greatest crossover ever which will be better than GT's "it is what it is for a variety of reasons" crossover, but I have to tell you, I seriously doubt it.

I think the biggest improvements for some of us will be in bringing old crossovers back up to spec or maybe hotrodding them a little with better parts than were available with the originals or by bypassing the caps.

I've studied the simple crossovers in the Performance Series for some time, and I've decided anything I do will probably F them up. I guess I'm deliberately placing myself in Giskard's cynically invoked 95% who couldn't do this on his best day, and happy to do it. Now I'll have more time to just enjoy the darn things.

To the other 5% of you, I salute you, and maybe someday I'll pay you a thousand bucks to kick somebody's crossover butt for me. ;)

Zilch
01-01-2006, 11:10 PM
Your observation is relevant, Ti Dome. These old networks are reputedly sound, but the total systems aren't up to today's sonic standards (read "pretty awful," above, and "dreadful" by others). I'd say that's licence for us DIY hacks to "mess" a bit, ourselves. :D

Ian Mackenzie
01-01-2006, 11:11 PM
Not necessarily. You need to determine empirically the desired target acoustic rolloff (taking into account the raw response on the baffle) of the woofer and work from there empirically.

Regards the target acoustic rolloff there is no text book solution for an individual system. Theoretically calculated crossovers often fail to win favour over the empirical method.

Remember the voltage drive and the amplitude response are two different things..you really need to be measuring both to arrive at the rolloff characteristic that best suits the system at hand.

This may end up being a compromise of on and off axis smoothess in the crossover region and a host of other factors.




SO, Ian, you suggesting to go with the calculated Zobel or the original design?

Per N3134 is the upper curve, vs original spec.

'Course, it's how it intereacts with the actual woofer impedance vis-a-vis the total network and system, in the end.

That's an "empirical" determination. :p

O.K., 2 X 47 uF, then, per Giskard.

Them 68's are $18.43 apiece. :(

OR, put in the 68's and try 6 Ohms for the Giskard calculated Zobel, 7.5 Ohms for the original spec....

Ian Mackenzie
01-01-2006, 11:43 PM
I was really thinking more of the consumer JBL L series and corresponding Prof 43xx series and one might wish more finely tuned than the kits.


EDIT: Okay I know see your system:Todd it would be helpful if you could post some close ups of your enclosures in this thread.

Sorry it has taken me a while to catch up with Todd's stuff per the thread below. Todd's drivers do not necessarily mate with any of those above and it appear not to be of the woofer for that original box.:hmm:



http://www.audioheritage.org/vbulletin/showthread.php?t=8591


If Todd prefer to keep with what he's got then its a case of working something out like a poor mans L300 with the ring radiator..nothin wrong with that. But these H 91/92 /93 horn designations confuse the crap out of me. Am I right in thinking the 91 is the 2307 ..a 1200 hertz horn or am I sun struck?

Not meaning to de-rail any noble efforts but if it were me I would use the 175 (I think Rob uses one successfully) with the bum cheek 2344a and it would fit on the front baffle I think. Use the 075 to zap aliens or trade it for a set of 2344's and some change on Ebay

There is no question that that horn works very very well with the 2235 woofer (and with some tweaking others) and will result in a quantum leap in subjective improvement over any reworked crossover for the device on hand (horn) in IMHO






(With apologies in advance to the thread's contributors for going a bit OT. None of these comments are aimed at the present effort.)

That very thought always leads me to this conundrum:

JBL/Harman have the money, the facilities, the brains, the experience, the ears, and the theory to do just about anything they damn well please, and IMO their better efforts are going to have a very high liklihood of kicking the ass of anything we do; YET one always has the nagging suspicion that what happens in the lab gets mucked up in the translation to production by marketing, accountants, sales decisions, etc., SO we probably get some compromises that result in "it is what it is for a variety of reasons" comments by some of our more experienced hands. HOWEVER we all want to believe we can take a good thing and make it great with just a bit of tweaking, so maybe we can figure out where JBL/Harman cheaped out or took "good" instead of "great" and we can create a better way to do it. MAYBE my very serious effort to improve things will result in my greatest crossover ever which will be better than GT's "it is what it is for a variety of reasons" crossover, but I have to tell you, I seriously doubt it.

I think the biggest improvements for some of us will be in bringing old crossovers back up to spec or maybe hotrodding them a little with better parts than were available with the originals or by bypassing the caps.

I've studied the simple crossovers in the Performance Series for some time, and I've decided anything I do will probalby F them up. I guess I'm deliberately placing myself in Giskard's cynically invoked 95% who couldn't do this on his best day, and happy to do it. Now I'll have more time to just enjoy the darn things.

To the other 5% of you, I salute you, and maybe someday I'll pay you a thousand bucks to kick somebody's crossover ass for me. ;)

Zilch
01-02-2006, 12:01 AM
What are we up to now?

1) 1 ea. inductor, 2.2 mH, Erse Super Q - $13.55
2) 4 ea. capacitor, 47 uF, Jantzen CrossCap - $44.20
3) 2 ea. capacitor, 68 uf, Jantzen CrossCap - $33.78
4) 1 ea. N/I resistor, 7.5 Ohms, Mills MR - $3.50
5) 2 ea. capacitor, 33 uF, Jantzen CrossCap - $17
6) 1 ea. inductor, 0.8 mH, Jantzen 18 Ga (tentative) - $3.67
7) 1 ea. N/I resistor, 2.5 Ohms, Mills MR - $3.50
8) 1 ea. N/I resistor, 5.0 Ohms, Mills MR - $3.50
9) 1 ea. L-Pad, 16 Ohms, Dayton 50W - $7.50
10) 1 pr. input terminals, Dayton - $6.88
11) 8 ea. capacitor, 0.1 uF, AudioCap Theta - $25.76
12) 4 ea. resistor, carbon film, 2.2M - $0.20
13) 1 ea. battery holder, 9V - $2.10
14) 1 ea. battery, lithium, 9V, UltraLife - $7.99

Total = $173.13 ea., plus shipping and construction materials.

Add 1 ea. N/I resistor, 6.0 Ohms, Mills MR - $3.50 to tweak the Zobel, as desired....

Mike Caldwell
01-02-2006, 07:24 AM
Hello
Here is a link to crossover design program that works well that
could get you going or at lest to a starting point.
http://www.pvconsultants.com/audio/targetgen/pcdc.htm

Mike Caldwell

spkrman57
01-02-2006, 10:46 AM
I would have thought with Spice and Clio that would be a piece of American Pie (I mean cake for the rest of the world)

A useful and well written article here for those interested to learn the finer points of passive crossover design incl Zobels and their effect.

http://sound.westhost.com/lr-passive.htm

But this is only the beginning.

There ain't no substitute for a good set of ears. The more I learn to appreciate and understand the original vintage JBL crossovers the more obvious the amazing skill of engineers like G.T and others is who painstakingly tuned and voice each system so carefully.


A little time spent reading the "abc's" of crossover design goes a long way for those who have little background in that area.

Ron:p

spkrman57
01-02-2006, 12:14 PM
Does the HF compensation (1 ufd cap - .16mh coil - 5 ohm resistor) operate the same in theory as the 4430 crossover?

I am just curious as to why there would not be a variable resistor to dial in the HF compensation(such as the 4430 does).

Hopefully I am not going off topic with this.

:(

Ron

toddalin
01-02-2006, 01:45 PM
I really didn't mean to start a war here.

Yes, I've heard real 4330s used in studios in the '70s and was impressed at the time.

Yes I attended the old JBL demos in the 70's when they would show off the L300s, L220s, L212s, etc (and still have the "Sessions" LP).

Yes, I've heard real 4430's and 4425's.

All the above sounded very good, but:

No I don't intended to spend several hundreds of dollars for crossovers, nor do I want to pay $175/pr for 30 year old units on epay.

No, at this time it's not economically feasible, nor required, to change the 175s to 85's especially when most of the difference occurs in the range where the 075 takes over. Nor is it economically feasible, nor required to change the 075's to 077s. I'm not trying to build an L300 clone, but am trying to optimize what I already have.

At this point, I'll probably order the parts to make three N200B crossovers, probably leaving out the HF boost circuit if it is unnecessary. Is it?

Maybe then I'll consider biasing the networks. (Thanks for your work Zilch.)

Similarly, I have no intention of paying $20+ each for Kimbercaps or Audiocap Theta's, but have no problem paying $5-6+ each for Dayton's, Jantzen's, or Solens as opposed to cheap non-polarized electrolytics. Is there a recognized preference for one type over the other on this board???

My primary concern lies in the chokes and their dcr. There are a number of chokes listed at PE that have the correct inductance, but I don't know what the dcr (and therefore wire gauge) should be.

And while we're at it:

Is there a sonic difference when combining caps? For example, if a value is specified as 24 uF, is there a sonic difference between using a single 24 uF cap, two 12 uF caps, or three 8 uF in parallel, or mixed values (e.g., a 14 uF cap and a 10 uF cap)?

Is there a sonic difference if a resistor is rated for more current (watts) than is necessary? (Doesn't seem like there should be.)

Thanks to everyone who has contributed to the discussion.

Ian Mackenzie
01-02-2006, 01:47 PM
I am not here to scrutinize anyone but I keep asking myself what is the point of all the majoring in the minor when the simple question and pobably most important question has not been covered with any real depth and support by hard data. Some post have been deleted so how is anyone supposed to follow the discussion in the early stages?

Is the 91 horn and 175 driver a good match for the 2235 reconed driver and is it worth the time and expense of further invesigation? What are the options and trade offs and limitations? Does the current driver line-up suit the intended application?

It would be nice to see some graphs posted and real world experience to support worthy discussion of how the 2235 and 91 horn with the 175 driver will combine.

Zilch
01-02-2006, 02:27 PM
Hi, Todd.

I've built with mostly Daytons, and an occasional Solen here. I grabbed Jantzens for the pricing because the values for the biased version were listed, is all. I've never used them.

Regarding Thetas, I follow forum advice and use them 0.01 uF's for bypass caps instead of the cheaper Daytons. It's just a luxury I've always afforded myself in building crossovers here; I don't pretend to hear a difference. When I've run out of Thetas, I'll use Daytons, but replace them ultimately.

I only use multiple caps to make a non-standard value or avoid an odd huge single on a board. To my knowledge, there's no sonic benefit to using multiples of the same type, or not. Statistically, multiples should get you closer to your target value by averaging, since the tolerances are so broad. I don't pursue that, because every additional component just makes them that much harder to build.

The important thing about the power resistors is to use the non-inductive variety. I use the Mills exclusively, again, possibly, a luxury in comparison to the cheaper 10W ones PE also offers.

PE does not stock the 7.5-Ohm Mills, so you might consider using the 10W variety for that. The other two values are not available in the cheaper line there. Alternatively, combine your resistor order and purchase those from Percy Audio. Nine Mills MR's of the requisite values will meet his $25 minimum, but you'll be paying an additional shipping charge for that.

You absolutely do not need or want the HF boost LCR if you're using these to drive N8000 and a UHF driver. The boost is always there the way it's connected in the circuit. You want that gone.

I'll dig out the crossovers here and measure the DCR of the inductors yet today and update this post accordingly....


EDIT: Had to drill out the pop rivets to get 'em open, of course. :banghead:

Insides shown in top photo below. The three inductors are all laminate core.

Measurements taken with BK Precision 875B handheld LCR meter. The most sensitive scales apparently use a method that doesn't like laminate core inductors. These are taken one scale up, one inductor lead disconnected:

"2.2 mH" - 2.13 mH, 2.11 mH, 0.84 Ohms, 0.81 Ohms

"0.8 mH" - 0.79 mH, 0.82 mH, 0.51 Ohms, 0.50 Ohms

"0.16 mH - 0.152 mH, 0.154 mH, 0.12 Ohms, 0.13 Ohms

So, turning to what's available at PE, it looks like the Jantzen 18 Ga are going to provide the best match for both the 2.2 mH and the 0.8 mH, for starters.

Interestingly, the serial numbers of these two units are 12338 and 12351, but in the second one, shown in the bottom photo, the seals on the capacitors have curled and fallen out, exposing the electrodes and sand inside. The seals appear to be paraffin or paraffin-like in composition. These'd require a rebuild, I'd say.... :p

:hmm: How come I'm seein' a 20-Ohm resistor in that pic? Investigation required....

RE-EDIT: There's 20 Ohms, 10W across the HF output not shown in the schematic. It's in both units here....

Ian Mackenzie
01-02-2006, 03:22 PM
If you go here there is some recent work regards capacitors.

Regardless its been covered in many threads as you will see with the search function.

http://www.audioheritage.org/vbulletin/showthread.php?p=84045#post84045

JBLnsince1959
01-02-2006, 03:32 PM
I really didn't mean to start a war here.



war?.....what war......this has only been the usual "love taps" that happens when discussing something here....no biggie....just the typical "gentlemanly" LH forum discussions .... :D

slxrti
01-02-2006, 07:48 PM
At this point, I'll probably order the parts to make three N200B crossovers, probably leaving out the HF boost circuit if it is unnecessary. Is it?

Maybe then I'll consider biasing the networks. (Thanks for your work Zilch.)

Similarly, I have no intention of paying $20+ each for Kimbercaps or Audiocap Theta's, but have no problem paying $5-6+ each for Dayton's, Jantzen's, or Solens as opposed to cheap non-polarized electrolytics. Is there a recognized preference for one type over the other on this board???

My primary concern lies in the chokes and their dcr. There are a number of chokes listed at PE that have the correct inductance, but I don't know what the dcr (and therefore wire gauge) should be.

And while we're at it:

Is there a sonic difference when combining caps? For example, if a value is specified as 24 uF, is there a sonic difference between using a single 24 uF cap, two 12 uF caps, or three 8 uF in parallel, or mixed values (e.g., a 14 uF cap and a 10 uF cap)?

Is there a sonic difference if a resistor is rated for more current (watts) than is necessary? (Doesn't seem like there should be.)

Thanks to everyone who has contributed to the discussion.

At one time I own LE85's and the HL91 and 92 horn I used this combination with a 2235. The 2235 was mounted in a 4cf enclosure. The crossover freq

was 800hz, after 5 minutes of loud music, I would have to leave the room.

Btw, this never happens with the 4430s.



I believe part of the problem was directivity Index was so different between the two drivers, causing apparitions in the off axis response, which negatively effects what we hear (I read this in an article of the 4430's).



Can the 175 be well integrated with the 2235? The 2235 response extends to 1.2khz and will be 6db down @ 50 degree off axis from the horn center line. (the 4430 uses this design philosophy). This implies that the 2235 can make it to at least 1.2khz.



Can the horn and 175 make it that low without a large reduction in DI?

Take a look at the horn info provided by JBL. If the DI is close in the horizontal and vertical plane then I feel it will work, how well that's difficult to say. But at the very lease the off axis response will be smooth in the crossover region, which is very important.



In regards to the eq, I would leave it in, but optimize it for the 175. Check out the Pi forum. The moderator has plenty of experience in this matter.



As far as caps, the 4430s ripple in the crossover region is due to the ESR of the caps (read this in the 4430 article). Placing caps in parallel reduces the ESR.



Hope this helps, take care slxrti

Zilch
01-02-2006, 09:04 PM
In regards to the eq, I would leave it in, but optimize it for the 175.Hi, slxrti.

Todd's using 075 with N8000 on top, which will cut the VHF to the 175 but pass on erroneous equalization to the UHF driver.

http://www.jblproservice.com/pdf/Network%20Schematics/3106%20Network.pdf

The contour of that eq also varies with the setting of the MF L-Pad in the original N200B circuit.

That's the basis for my assertion that it should be removed....

4313B
01-03-2006, 12:04 AM
The contour of that eq also varies with the setting of the MF L-Pad in the original N200B circuit.

That's the basis for my assertion that it should be removed....I believe you asked me to run this some time ago Zilch.
Orange is bump filter removed. Center frequency for the bump filter is ~ 13 kHz.

Zilch
01-03-2006, 12:38 AM
Yes, Giskard, and thank you.

That illustrates it quite nicely.... :thmbsup:

Ian Mackenzie
01-03-2006, 05:57 AM
This thread has to hold the record for the most consecutive deleted posts by posters.

Pity, if they were edited and reposted it would make great learning for future generations of DIY forumitets.


A bit of history.......

On the 2231A did you ever try one? ...well top this.

I distinctly remember pulling one of a pair out of the carton brand new....

Oh would have been back in the Spring of 79 and Mum saying what are you going to do with that......Ian...Ah make some sub woofers Mum and eventually a big pair of speakers. They sure did sound good in the lounge room on a Saturday night!

Pity kids don't do stuff like that these days...

Still wondering in amazement at just how good they sounded no matter what I did with them. Still got the original cartons and warranty cards.....might post some nice hi resolution pics sometime.....still in mint condition..reconed of course but that driver's sure got some stories to tell!

Never run them up past 300 hertz though, 2121's were the get go back then...just made to work together like nothin else on earth.

4313B
01-03-2006, 06:26 AM
This thread has to hold the record for the most consecutive deleted posts by posters.

Pity, if they were edited and reposted it would make great learning for future generations of DIY forumitets.I really don't want to get into this again. It's best to just let them play and figure it all out themselves. Granted, they will come up with different solutions and invariably they will be wrong the first few dozen times or so but who cares? I've thought about this all night and I suspect they will think they sound just fine and that's really all that matters. Ignorance is bliss. Hell, I thought the 136A and 2231A were the bomb right up until I replaced them with the 2235H. It's the way it is. It's too much effort to try and ramrod the situation so that it's done right. It will be tough enough to keep the Technical Help forum straight and all the wannabees from posting "answers" in it. I was talking to a couple of the guys at JBL about our situation and they felt that it just didn't matter. This stuff is all really old and they certainly don't have any interest in any of it. I was encouraged to let it go and follow suit. I think it's great that we split it off and I think we should let this DIY forum be real open with an "anything goes" attitude. It seems desirable and people seem to have fun with it so whatever.

I see the rules for this DIY forum are pretty short and sweet. I think it will end up being fun for everyone.

On the 2231A did you ever try one? ...well top this.I can't top it. My first set of four 2231A's were hand-me-downs when a buddy upgraded to 2235H's in his 4350's. Before that I used 136A's exclusively. I was real heavy into the Loudspeaker Component Series and knew it all backwards and forwards. I was obsessed. My dad worked at a major University back then so I had access to all their measurement gear. It was fun.

4313B
01-03-2006, 07:34 AM
EDIT: Had to drill out the pop rivets to get 'em open, of course. :banghead: Is that a bad thing? I've drilled tons of them out so am I in trouble? :p
the seals on the capacitors have curled and fallen out, exposing the electrodes and sand inside.There's that "all those old things are worth $25" quote of mine coming into play. ;)
RE-EDIT: There's 20 Ohms, 10W across the HF output not shown in the schematic. It's in both units here....There's that "JBL continually engages in research..." quote coming into play. :)


I guess I'm kind of surprised that the large coil has such a high DCR. I suspect anything you build these days will sound considerably better.

Zilch
01-03-2006, 10:00 AM
I guess I'm kind of surprised that the large coil has such a high DCR.Yes, me too. I'm used to using larger coils there. There'll likely have to be some level balancing once Todd gets them built. If we need a little more bass SPL, it'll be available there.

I suspect anything you build these days will sound considerably better.That's been my experience. It's surprising what pleasing results can be obtained with off-the-shelf parts and a little effort to build them. I'm bettin' they'll make a substantial improvement in Todd's system.

His 2235's are a major upgrade; just have to get it all playing together, now.... :)

toddalin
01-03-2006, 10:18 AM
Yes, me too. I'm used to using larger coils there. There'll likely have to be some level balancing once Todd gets them built. If we need a little more bass SPL, it'll be available there.
That's been my experience. It's surprising what pleasing results can be obtained with off-the-shelf parts and a little effort to build them. I'm bettin' they'll make a substantial improvement in Todd's system.

His 2235's are a major upgrade; just have to get it all playing together, now.... :)

They have got to sound better than the 30-40+ year old N1200s currently in there, which probably used even smaller coils.

Thanks everyone for your help. I'll wait until the end of the week to see if others chime in regarding using one make cap over another before making a parts purchase.

Now's the time to chime in..., anyone?

4313B
01-03-2006, 10:22 AM
I'd cut the DCR in front of the 2235H back to under 0.5 ohms.

toddalin
01-03-2006, 10:49 AM
I'd cut the DCR in front of the 2235H back to under 0.5 ohms.

This one then???

http://www.partsexpress.com/images/255-438m.jpg

The Jantzen 15 gauge air core inductors offer a lower DCR alternative to the other Jantzen lines. They feature the same quality construction including high-purity copper wire wound to a +/- 3% tolerance, and the special Jantzen high-temperature baking process that eliminates the possibility of FM distortion. The perfect layer air core construction improves heat dissipation while eliminating hysteresis and any type of saturation distortion.

Inductance: 2.20 mH
Wire gauge: 15 AWG
DCR: .52 ohms
Power handling: 500 watts RMS
Dimensions: 2.32" Dia x 2.4" H

Zilch
01-03-2006, 11:11 AM
I'd cut the DCR in front of the 2235H back to under 0.5 ohms.The 15 Ga Jantzen is more like what I've typically used.

How about the ERSE Super Q laminated core at 0.154 Ohms for just a few pennies more? I've been using those lately, largely, I confess, because of their convenient and stable mounting.

If my calculation is correct, that'd give 0.87 dB more drive to the 2235H and also improve the damping.

With that, the 20 Ohms added across the HF output of the original may not be needed, recovering that cost. Seems to me the 20 Ohms is just attenuation, no? There'll be some in the following N8000, anyway....

4313B
01-03-2006, 11:17 AM
I'd go with the ERSE Super Q's. I've used them in 4435's and 250Ti's.

toddalin
01-03-2006, 11:53 AM
I'd go with the ERSE Super Q's. I've used them in 4435's and 250Ti's.

This one then. And here, I thought that the air core coils were perferred???? Is there a preferred orientation for the coils?

ERSE SUPER Q 2.2mH 16 GA 500W INDUCTOR
Finally, a sonically superior magnetic core coil! The Super Q line features a specially shaped core that allows for optimal magnetic performance and increases the coil's effective permeability. The result is extremely low DC resistance, reduced distortion and very high saturation levels. The coils are encapsulated in a plastic housing, which eliminates any magneto striction distortion and makes mounting a breeze. The 500 watt power rating makes these coils perfect for high power prosound applications. All coils measure 4-1/2" L x 1-5/8" H and are made in the U.S.A. Recommended for frequencies 0Hz - 3000 Hz.



http://www.partsexpress.com/images/266-900m.jpg

Zilch
01-03-2006, 12:33 PM
This one then. And here, I thought that the air core coils were perferred???? Is there a preferred orientation for the coils?
You'll want to orient the axis of the coil 90 to any others, and keep them 3" away, minimum. Otherwise, there'll be crosstalk between the LF and HF sections via coupling.

Since building this biased (HF only) NL200t3, I've been unwrapping the black tape, bringing the leads out the similar bottom slots for a little neater lead dressing, and then re-wrapping.

Zilch's Wonder Cable feeds the HF, upper right. Yellow caps are the Theta bypassers, and that's two Mills MR's at the lower right. Wiring's on the reverse, not shown. :p

Input terminals on mine go out the back of the board and through a secondary masonite board recessed on the the rear of the cabinet, mounted from inside. You can see the Dayton 091-124x binding posts ($6.88/pair) offer plenty of length and adjustability to work with that. They also take up to 8 Ga wire, for those believers:

Zilch
02-21-2006, 01:14 AM
Project continues here:

http://audioheritage.org/vbulletin/showthread.php?t=9009