JBL 4333a/Sansui EC138

[Oyster150]

Hi All,

Sometime ago I had bought a Sansui EC138 which is essentially a reworked JBL4333a/L300
Components used are:

1) JBL 2235H
2) JBL 2425H (8ohms)
4) JBL 2405H
5) Horn: JBL 2307
6) Crossover : JBL 3110a + JBL 3106

While the woofer and midrange sound balanced, the 2405H tweeter has an extra sizzle in upper midrange
The 3106 crossover point is 8khz per schematics below:


Question:
I believe I need to move the crossover point to above 8khz to eliminate the 2405 sizzle
How can I move the crossover point to something like 9khz-10khz in the above network ?

I tried copying the HF circuit from post #170 https://www.audioheritage.org/vbulletin/showthread.php?1812-Building-the-4345-4344&p=51639#post51639 :


However, this resulted in too scooped midrange

 

[toddalin]

Replace the 1.5 mfd with a 1 mfd.

 

[Oyster150]

@toddalin
Thanks, that worked out great!! The 2405 now sounds much smoother
There is some void audible in the upper midrange ; 7khz-8khz
Anything I can do in the 3106 LF circuit to lift it ?

 

[Oyster150]

Another question:

The Nelson Pass published crossover for JBL300 here https://www.firstwatt.com/pdf/art_l300.pdf

L300 has 16ohm midrange driver. I have an 8ohm variant of it. Do I still need 8ohm lpad in midrange circuit ?

 

[SaturationPt]

Another question:

The Nelson Pass published crossover for JBL300 here https://www.firstwatt.com/pdf/art_l300.pdf
View attachment 2964443
L300 has 16ohm midrange driver. I have an 8ohm variant of it. Do I still need 8ohm lpad in midrange circuit ?

The L300/L300A/4333 all have the same midrange and HF impedance & VC, but for some reason they have different impedance specs. They are not different and the DCR has been measured many times. The HF does have a blue vs gold diaphragm, but the same coil (this has been discussed ad-naseum on the Lansing Heritage site as well as here on the Lansing (JBL) forum).

You have the 2235H, which would be in the A models of both speakers (136H/2235H), go with the A model specs (spoiler: the crossover did not change but specs did). My L300As had been re-coned when the surround deteriorated (by the JBL dealer) with the correct C8R2235H kit, measured in-room response was down into the low-30Hz range +0/-6dBA and with the stock crossovers measured very flat, phase linearity was also very good.

Nelson Pass developed his crossover on a self re-foamed L300 Summit (early 136A woofer) so there is a difference there, and if you read his (published) account of the experience he stated that it was very hard to improve on JBL’s network, and he wasn’t sure that all of the effort was worth the small (observed) improvement. It was also a pair of old speakers at the time, so not comparing new-component original crossover to new-component modified, he started with an old/original crossover for comparison.

 

[Oyster150]

@SaturationPt Do you have your drivers all in negative polarity ?
While the woofer and tweeter have reversed polarity, I think 2425/2420 midrange are positive drivers.

 

[Oyster150]

@toddalin
In the other thread I read about your views on the alnico vs ferrite jbl drivers and the Nelson pass crossover for L300
Do you have a crossover design in mind that I can follow for the drivers I have?
Peaks and nulls , I haven't been able to get rid of these reasonably
The stock crossover has tad too much sizzle, and there is the horn honk somewhere (Can't point out where in the FR range)

While I can make the Nelson pass version, I am worried if it'll still have any peaks/voids anywhere
The cabinet design also is different vs L300. Single bass port vs dual port. I can block one port but it sounds wonky

 

[toddalin]

I made the crossovers for my speakers (Erse inductors, Solen caps with Theta by-pass caps) and sent one to Zilch labs for testing. It got a thumbs-up and I installed them, but was never happy with the performance of the 2235. I did several mods to get the horns to play a bit lower, just as Nelson did, but still never took a cotton to them after redoing the woofers to the 2235s.

Then I started to concentrate my efforts on the Mermans, which just blow them away.

Your best bet is to get an RTA or program that plot the response and try the horn and tweeter with alternate phasing and see which is smoothest in the sweet spot.

 

[SaturationPt]

I made the crossovers for my speakers (Erse inductors, Solen caps with Theta by-pass caps) and sent one to Zilch labs for testing. It got a thumbs-up and I installed them, but was never happy with the performance of the 2235. I did several mods to get the horns to play a bit lower, just as Nelson did, but still never took a cotton to them after redoing the woofers to the 2235s.

Then I started to concentrate my efforts on the Mermans, which just blow them away.

Your best bet is to get an RTA or program that plot the response and try the horn and tweeter with alternate phasing and see which is smoothest in the sweet spot.

IIRC, your L300s weren’t L300 cabinets but modified L200, and the 2235s were not 2235s (unless this is a different set than you’ve discussed in the past) but re-coned L200a drivers with 2235 cones (not sure the manufacturer of your cone kit either).

My recollection is that you’ve never owned actual L300s or L300As. Without searching for old posts, is this correct? And you are still basing your opinion of the L300 series and the 2235H on your L200 mod project?

I think you’re talented and knowledgeable, but I also feel that your repeated talking down the L300s and 2235H based on your modified drivers and projects is a disservice to those of us who have experienced the real article.

 

[toddalin]

They are enlarged L200 cabinets. The two ports have been extended and flaired. The horns are the long L300 horns. The woofers were 2205s reconed as 2235s, which use the same basket and magnet assembly, only a different cone and the addition of the mass ring. They were also recharged, twice. Grumpy ran WT-2 on them and all parameters came out as spec for 2235s with a BL of over 21 after the second recharge.

Anyone with an L300 who has had the 136A woofer reconed (not re-edged) is dealing with the same beast I am.

 

[toddalin]

They are enlarged L200 cabinets. The two ports have been extended and flaired. The horns are the long L300 horns. The woofers were 2205s reconed as 2235s, which use the same basket and magnet assembly, only a different cone and the addition of the mass ring. They were also recharged, twice. Grumpy ran WT-2 on them and all parameters came out as spec for 2235s with a BL of over 21 after the second recharge.

Anyone with an L300 who has had the 136A woofer reconed (not re-edged) is dealing with the same beast I am.

BTW, nothing about the size and porting of the cabinet affects the top end roll off of the 2235. Cabinets and porting work with the bottom end of the spectrum and the 2235 never reaches the horn.

And..., I also had a pair of stock 4430s (but refoamed) and didn't think the 2235 met the horn on those all that well either.

I should also note that I listened to several LE85/2420s before I gave up my LE175s because the LE175s go lower to better match the woofer even if their top end is not as extended. But eventually, I found a set of 2420s with original tangential diaphragms (like used in the LE175s), and these were a bit sweeter on their bottom end than other LE85s/2420s I had auditioned and these replaced the LE175s.

 

[Oyster150]

Woofer: (Could get only a partial shot sliding phone cam into the port)


Midrange:
Those crumbs are from the original damping disc that had rotten. Replaced with new unit from JBL. Had tried some diy discs earlier, but all screwed the FR


Tweeter:


Horn:


Stock damping disc:


Replacement disc : (JBL Part number 61286)


"BTW, nothing about the size and porting of the cabinet affects the top end roll off of the 2235. Cabinets and porting work with the bottom end of the spectrum and the 2235 never reaches the horn."

This sounds about right. Midrange lacks the force/push at some places while the 2405 has thin/sizzly upper midrange making the overall balance even worse.

 

[GordonW]

Trying to use a 2235 in a 4333/L200 design, is one of those cases where using an aftermarket recone kit may well work better than factory,..

With an aftermarket kit- you have control of how much Aquaplas and how much total mass ring weight is added to the cone. I am of the impression, that subtracting about 5 or so grams of mass from the 2235 cone assembly, may well give it just enough extra HF extension, to work better with the horn in these speakers...

It's funny- to replace the LE15 in the earlier L300s- JBL recommends the 2234, instead of the 2235. That's at least partly because it can much better go up to the 1200Hz crossover point in those speakers, than the 2235, due to the 2234 not using a mass ring...

Though nobody who's a "purist" would ever agree to this- I would wonder what a hybrid woofer- using the 2235 basket and voice coil, the softest spider that could be gotten, and the 2226 cone, could sound like in an L300. That 2226 cone EASILY goes up to 1200Hz... flat.

Or, starting with a complete 2226- put in the softest spider possible, and add mass ring to tune it to the L300 enclosure. That also could work well, potentially. Probably less distortion than a 2235 basket- based driver, due to the better magnet geometry.

Regards,
Gordon.

 

[toddalin]

Trying to use a 2235 in a 4333/L200 design, is one of those cases where using an aftermarket recone kit may well work better than factory,..

With an aftermarket kit- you have control of how much Aquaplas and how much total mass ring weight is added to the cone. I am of the impression, that subtracting about 5 or so grams of mass from the 2235 cone assembly, may well give it just enough extra HF extension, to work better with the horn in these speakers...

It's funny- to replace the LE15 in the earlier L300s- JBL recommends the 2234, instead of the 2235. That's at least partly because it can much better go up to the 1200Hz crossover point in those speakers, than the 2235, due to the 2234 not using a mass ring...

Though nobody who's a "purist" would ever agree to this- I would wonder what a hybrid woofer- using the 2235 basket and voice coil, the softest spider that could be gotten, and the 2226 cone, could sound like in an L300. That 2226 cone EASILY goes up to 1200Hz... flat.

Or, starting with a complete 2226- put in the softest spider possible, and add mass ring to tune it to the L300 enclosure. That also could work well, potentially. Probably less distortion than a 2235 basket- based driver, due to the better magnet geometry.

Regards,
Gordon.

I would think that the 2205 may be the better starting point.

And here's a beautiful pair for just three bills, though it would be a shame to rework a perfectly working pair. These are what I have in my enlarged L200 cabinets as well as my center channel except that they were all reconed as 2235s. Had I known the outcome before hand..., I coulda, shoulda, woulda had them redone as 2205Bs again.

https://orangecounty.craigslist.org/ele/d/garden-grove-jbl-speakers-2205b/7657805671.html

 

[GordonW]

I don't get to convert many 2205 alnico frames to anything but D130s. People want those alnico magnet frames for guitar speakers.

Regards,
Gordon.

 

[Oyster150]

IIRC, your L300s weren’t L300 cabinets but modified L200, and the 2235s were not 2235s (unless this is a different set than you’ve discussed in the past) but re-coned L200a drivers with 2235 cones (not sure the manufacturer of your cone kit either).

My recollection is that you’ve never owned actual L300s or L300As. Without searching for old posts, is this correct? And you are still basing your opinion of the L300 series and the 2235H on your L200 mod project?

I think you’re talented and knowledgeable, but I also feel that your repeated talking down the L300s and 2235H based on your modified drivers and projects is a disservice to those of us who have experienced the real article.

I think , at least in the build I have, the 2235 are fantastic. No mid bass bump, very explosive and gradual bottom roll off

 

[Oyster150]

try the horn and tweeter with alternate phasing and see which is smoothest in the sweet spot.

This actually worked out to be very effective. Oddly , I am finding this combination to be most agreeable (most blend in that is): woofer and midrange positive and HF negative. This blunted the HF response a bit and sounds cohesive

 

[Oyster150]

Dunno if this has been discussed anywhere, but
merely changing the tightness of centre phase plug on the 2405 attenuates audibly
Too tight and there is a huge 12k peak
Too lose, where the plug moves almost freely results in a non peaky but very splashy/unfocused treble
Need to stop turning screw as soon as it starts to resist. There is still a peak >10khz, but very bearable
Of course, hold the plug firmly while doing this to avoid diaphragm twist

 

[tarior]

Trying to use a 2235 in a 4333/L200 design, is one of those cases where using an aftermarket recone kit may well work better than factory,..

With an aftermarket kit- you have control of how much Aquaplas and how much total mass ring weight is added to the cone. I am of the impression, that subtracting about 5 or so grams of mass from the 2235 cone assembly, may well give it just enough extra HF extension, to work better with the horn in these speakers...

It's funny- to replace the LE15 in the earlier L300s- JBL recommends the 2234, instead of the 2235. That's at least partly because it can much better go up to the 1200Hz crossover point in those speakers, than the 2235, due to the 2234 not using a mass ring...

Though nobody who's a "purist" would ever agree to this- I would wonder what a hybrid woofer- using the 2235 basket and voice coil, the softest spider that could be gotten, and the 2226 cone, could sound like in an L300. That 2226 cone EASILY goes up to 1200Hz... flat.

Or, starting with a complete 2226- put in the softest spider possible, and add mass ring to tune it to the L300 enclosure. That also could work well, potentially. Probably less distortion than a 2235 basket- based driver, due to the better magnet geometry.

Regards,
Gordon.

Didn't you do a write up on a "super woofer" that you built using something like a 2245 coil, 2226 cone, and a B spider?

*Edit*
Nvm, I found it. I have a pair of 2226 baskets now, too. ;-)

 

[Oyster150]

So, been fooling around with the crossover modifications for the last several months
Some observations:
Giskard's 3145 HF section - JBL 2405 sounds the most behaved with Giskard's 3145 HF section (The one in OP). However, with the HF crossover point now shifted over to 10k, there is a massive void in lower treble. I don't know how to make the JBL 2425 fill in that 8-10k region, hence given up on this network

JBL 3107 HF section - Way too much of that 9-10k sizzle. Upper treble also sounds elevated. It's not peaky, but a very broad HF lift. I added 20ohm resister but couldn't eliminate all of that sizzle. Lpad also didn't help much. I wish I knew how to pad down the 8-10k region because this one sounds the most resolving/vibrant of all.

Stock JBL 3106 - Sounds dirty, uneven, dead and veiled

Nelsen pass L300 HF section - This actually sounded very nice, despite internet complaining of 6db/octave. It's not as sizzle free as the 3145 HF section, but tolerable. Because I did not copy 2425 section from it, upper midrange is slightly muted. And this is something I need to fix

So question to the crossover wizards of AK:

Following is the Nelsen Pass L300 crossover schematic and I want to copy the midrange section of it. This one is for a 16ohm mid horn . The one I have is an 8ohm 2425. What changes do I need to make to make this work with 8ohms ?