Data Comparison: Stock vs. Aftermarket Recone 2245H Parameters

[GordonW]

I had an interesting experience yesterday.

I have been helping a friend repair some derelict JBL woofers- mostly loose magnets which we've been re-attaching, and then reconing said drivers. I took one of the test subjects- a 2245H- and tested it on the LMS/LEAP system.

Well, as I said, it was VERY interesting. As in, how CLOSE the specs MATCHED original!

Now, keep this in mind when looking at the specs- I DELIBERATELY altered the cone mass of the reconed unit, a bit. I treated the cone with a lacquer stiffening agent, which added about 11 grams to the "stock" aftermarket cone. See how that plays into the specs... really, all it did, was to marginally lower the broadband efficiency about 1/2 dB, while reducing the need for bass EQ on the lowest bottom end... which was what I was intending.

OK, here's the data:

Stock 2245H (from specs on file in various locations), from a broken-in unit:

Fo=20 Hz
Vas = 820 liters = ~ 28.9 cubic feet
Qes= .31
Qms= 2.2
Sd = 1300 sq.cm =~ 201 square inches
Xmax = 9.5mm
Re= 5.8 ohms
Mms= 182.7 grams
Bl= 20.73
Qts= .2717
Sensitivity @ 1w= 95.3 dB

OK, here's the reconed unit data, from LMS:

* Loudspeaker Measurement System
* LMS(TM) 4.5.0.342 May/19/2004
* (C)opyright 1993-2000 by LinearX Systems Inc
* Electro Mechanical Parameters
* Oct 27, 2004 Wed 7:34 pm
* Library=gordon-jbl.lib
* Reference Curve=2245h
* Delta M/C Curve=2245h w/42.0 grams
* Method= Double Curve - Delta Mass
Domain= FreeAir
Model= TSL
Revc= 5.560 Ohm
Fo= 21.915 Hz
Sd= 129.677m M² (= 1296.77 sq. cm)
Krm= 10.286m Ohm
Erm= 0.739
Kxm= 23.029m H
Exm= 0.704
Vas= 23.089 Ft³
Cms= 273.802u M/N
Mmd= 165.773m Kg
Mms= 192.625 g
BL= 20.929 T·M
Qms= 7.179
Qes= 0.337
Qts= 0.322
No= 1.976 %
SPLo= 94.977 dB

Now, the reconed unit is not broken in yet- which handily explains the 10% difference in Fs, the difference in Qms and the 25% or so difference in Vas . But, most tellingly- LOOK at the Bl product, the sensitivity, and the cone mass! Without the mass increase I added, it would have been IDENTICAL TO WITHIN ONE GRAM of stock. And the sensitivity, by definition, would have been IDENTICAL to stock, since the Bl and cone area are EXACTLY the same (the Bl is actually about 1% HIGHER than stock, actualy, and the cone area is completely insignificanly different in any reasonable measurement system).

Also, I verified some other things, during assembly:

1. The new voice coil is indeed EDGE WOUND, and is in the correct (inverse) polarity like the original (+ on red makes the cone go IN).

2. The dimensions of the aftermarket coil, cone, surround and such are IDENTICAL to an original kit (I directly compared them to a stock assembly). As in, the voice coil overhang is 3/8" in each direction... which would result in a 9.5mm X-max, same as stock.

What this means? The aftermarket kit, in a SUBWOOFER application, should SOUND IDENTICAL to the stock unit, once it's broken in.

And the nice thing? This was a kit that is SPECIFIABLE BY UNIT MODEL NUMBER from one of the most prolific aftermarket reconing parts outfits out there. Nice to see 'em get it right, at least once...

Regards,
Gordon.

[Guido]

Thanks Gordon for this VERY interesting article.

Are these cone kits preassembeld?

It woul be very kind if you could PM me the supplier.

[Ralf]

Hi Gordon,

do you know the price of these kits? And where can I get it?

Ralf

[Earl K]

Hi Gordon

An Interesting Comparison ! Thanks for your very informative post . :yes:

Excuse my questions if they seem dumb but :

(A) BL ; was this actually measured by the LEAP LMS system ? ( or was it a calculation derived from the number of coil turns in the magnetic field, etc ? )

- My reason in asking ; I have some 20 year old 2225 recones sitting in 2220a baskets. Recently, I retired these drivers from any active duty . This summer I measured their Fs . Both were virtually deadon the 2225 Fs, of 40 hz. These 27 year old speakers must be partially demagged by now . I listened to this pair in a MTM setup - I thought they should have a rising mid response, but they didn't . Their midrange response from 300 hz to 800 hz, measured smoother than comparable / actual 2225(s) .

This pair of 2225/2220 woofers operated in @ critically tuned boxes ( Fb of 52 hz ) for 2 decades without ever having the cones over-excursion to my knowledge. The slight, coil to gap misallignment ( coil overhang ) never came into any meaningful play.

- I'd like to get the BL measured on this pair to try to get a fix on what has happened to this pair of speakers . I'm thinking of cutting out the linen surrounds / & installing foam surrounds. This is to create a transducer with a higher compliance . I don't want to add mass rings but I realize I may need to add weight to get to a 25hz Fs. This overall kludge of an effort would be my attempt at creating a TAD 1601a .


(B) Once these hybrid 2245 speakers breakin , and Fs drops some, which direction will the Qms figure go ? Up or down ? And then accordingly, what is your prediction for a final Qts ?
( - as you see, Qms numbers continue to baffle me .) :rolleyes:

regards <. Earl K :)

[GordonW]

Originally posted by Earl K
Hi Gordon

An Interesting Comparison ! Thanks for your very informative post . :yes:

Excuse my questions if they seem dumb but :

(A) BL ; was this actually measured by the LEAP LMS system ? ( or was it a calculation derived from the number of coil turns in the magnetic field, etc ? )

- My reason in asking ; I have some 20 year old 2225 recones sitting in 2220a baskets. Recently, I retired these drivers from any active duty . This summer I measured their Fs . Both were virtually deadon the 2225 Fs, of 40 hz. These 27 year old speakers must be partially demagged by now . I listened to this pair in a MTM setup - I thought they should have a rising mid response, but they didn't . Their midrange response from 300 hz to 800 hz, measured smoother than comparable / actual 2225(s) .

This pair of 2225/2220 woofers operated in @ critically tuned boxes ( Fb of 52 hz ) for 2 decades without ever having the cones over-excursion to my knowledge. The slight, coil to gap misallignment ( coil overhang ) never came into any meaningful play.

- I'd like to get the BL measured on this pair to try to get a fix on what has happened to this pair of speakers . I'm thinking of cutting out the linen surrounds / & installing foam surrounds. This is to create a transducer with a higher compliance . I don't want to add mass rings but I realize I may need to add weight to get to a 25hz Fs. This overall kludge of an effort would be my attempt at creating a TAD 1601a .


(B) Once these hybrid 2245 speakers breakin , and Fs drops some, which direction will the Qms figure go ? Up or down ? And then accordingly, what is your prediction for a final Qts ?
( - as you see, Qms numbers continue to baffle me .) :rolleyes:

regards <. Earl K :)

The LEAP/LMS does Bl calculations based on the measured impedence curve (peaking at resonance), the measured Sd and the measured DCR of the voice coil, using the "differential mass" method in this case (where the assembly is tested with a frequency sweep, with and without a known "test mass" temporarily attached to the cone- we use Silly Putty, as it's easily removable without causing any damage whatsoever, BTW). There's a set of iron-clad equations that can definitively derive the Bl product, given that data.

As for the parameters on the 2245H- the Qms should go down after break-in, since the suspension will become more compliant (and therefore, "cushion" the motion more rather than oppose it like a "spring"). My take, is that Qts will wind up probably within 5% of factory value (within spec, if you read the factory tolerance allowances). Obviously, the Vas will go up- probably to 27-28 cubic feet, I'd predict. In all, though, with the Mms, Sd and Bl being so close to factory (and those parameters CANNOT change with age), it should be the exact right efficiency- which will mean it should respond almost identically to a stock piece, no matter what little variations in Fs, Q and Vas occur.

Ralf/Guido- these kits are, unfortunately, available only to "authorized dealers" of the supplier. Though, it's WAY EASIER to become authorized by THIS supplier than by JBL themselves... both the place I work at and the shop friends of mine run in the next town over, are both authorized dealers AND there's another authorized dealer on the other side of town, while I don't think there's more than one authorized JBL parts dealer in the ENTIRE metro Atlanta area (and frankly, what I've seen of their work doesn't make me want to let them work on ANYTHING I own).

Also, these kits come in completely part form- seperate cones, voice coils, spiders, dustcaps, braided lead-in wire, etc. They have to be jigged for alignment, and assembled by the reconer. A bit more challenging for the hobbyist... there's a lot more opportunity for something to go awry if you're not familiar with what you're doing. But, for us that do it every day, it becomes "old hat" quickly... you get used to it, and it goes quickly and without undue difficulty.

[Alex Lancaster]

If they are the same ones I buy from, I appreciate that they will sell to "reconers only", I have bought about 3K from them, and have been very happy with results; on some kits, the gaskets, for example are not correct, but I substitute them locally, the coils are very well made, but on the "E, K, etc, the formers are not Aluminum, but IMO, they work well, I like the brass strips better than the OEM very thin lead wire, they will sell the different Litz wires by the roll, but I also get them locally.

Gordon, it would be good if You become a distributor, it would be difficult for a forumite to get 2 cone kits directly, but You could do it for a reasonable profit, You know about 20dB more than I do, and You could also help our European friends; It would be too difficult for me from here.

On the cloth surrounds, they connected me to a glue supplier; We have been using with good results, up to now, Vynil "sealant", but have not been able to concoct the OEM sticky stuff, pls let me know what You have on it.

The price diff to OEM is outrageous and JBL had it coming.

Alex.

[Oldmics]

"They have to be jigged for alignment, and assembled by the reconer."

And there is the wildcard!

I was always concerned with total mass of the reciprocating assembly.The distribution of the adhesives in the many required locations could lead to an addition in mass.

The hand dispensed method of adhesives V/S the production line method of measured computer controlled applications.More than likely this difference is accountable within the + or - of tolerances specified in the T/S measurements.

I have also measured some generic kits V/S the originals over the years and found varying preformance differences,with some right generic kits right on the mark.I agree with Giskard that it may be who the generic supplier is sourcing the cones from attributing to these generic kits coming close to original preformance specs.

Kudos to you Gordon for such in depth information.

The failures that I always found in generic kits were

1-Poor assembly procedures

2-Voice coil failures.

I attribute this voice coil situation to lesser quality of material used in comparision to original coil material.This is only my opinion with NO supporting data.

3-Adhesive failure from thermal (or again poor assembly)

Some of these generic adhesives seem to have a lower flash point than original adhesives.Again only my opinion with NO supporting data.

Oldmics-No supporting data-Just 22 years in the biz.

[Earl K]

Good Points about JBL - I do agree there's too much "JBL bashing" around here this week . :)


Same compliance, cone and voice coil but a different spider. You should be able to get the right spider characteristics from a 136/2231 (they're easier to find than 124/2203's).


FWIW:

Speaking of different spiders. I have one le14a ( original LansaLoy surround ) that won't drop below 25hz ( no matter how much the surround is treated ) . I checked the relative stiffness of the spider. Definately , at least 1 grade stiffer than any other le14as I own. about the same , as a 2225h. This le14a also appears to be a newer (free JBL fix ). Cosmetically, the surround is more of a the creamy-white , compared to the creamy-yellow/orange type.

Point being ; the "flexature" of this spider makes at least a 5 hz difference in obtainable Fs .

Thought someone should know .

<. Earl K

[Mr. Widget]

I for one am very impressed that in these days of profit hungry multi-national corporations, JBL still maintains replacement parts for many of their legacy products.


Widget

[GordonW]

I'm definitely in favor of JBL making parts. It'd just be helpful if it wasn't such a minefield to navigate to actually GET those parts FROM JBL!

Given that situation, necessity being the mother of invention and such, those of us that have to be the "ones that keep stuff going", well, we have to cope however we can to keep OURSELVES with jobs, income, etc. If that means using an aftermarket supplier... well, that's what it's going to be, if it's necessary.

As for a 124 kit- sounds do-able. I think the spider replacement should be a cinch. The spider that's used in the 2245 is virtually identical to the 136A spider... and there's one grade softer than that, if necessary. We could also probably use a 2235 voice coil, or something similar... the parts catalog has a few variations available, all edge-wound, same voice coil former dimensions, with slightly different winding lengths and such. The hang-up may be finding a suitable foam-edge 12" cone with a 4" voice coil opening... I'll try to check on that. If there's an aftermarket 121 or 124 cone listed (can't remember right now), it'll be no problem, otherwise, some research will be required.

In any case, if you wanted to send me a 121 cone assembly to use as a reference standard, that'd be good. I'd be glad to measure, weigh, and otherwise analyze it to determine what parts and/or parts modifications we need to do. G, if you want, you could send that cone assembly to the same address as you sent that 128H-1 cone to a while back...

As for cost- well, it's not just a matter of one being cheaper than another, frequently they're not. It's a matter of BEING ABLE TO ACTUALLY DO THE JOB when it needs to be done. But, in some cases, such as 2235s, LE5s and such, it is possible to budget a good hour of labor per driver (a complete recone shouldn't take more than that, actual active work time) and still come out with a final price using aftermarket parts, that's SIGNIFICANTLY lower than the factory kit, not including labor for install...

Regards,
Gordon.

[GordonW]

Originally posted by Alex Lancaster
If they are the same ones I buy from, I appreciate that they will sell to "reconers only", I have bought about 3K from them, and have been very happy with results; on some kits, the gaskets, for example are not correct, but I substitute them locally, the coils are very well made, but on the "E, K, etc, the formers are not Aluminum, but IMO, they work well, I like the brass strips better than the OEM very thin lead wire, they will sell the different Litz wires by the roll, but I also get them locally.


Sounds like the same outfit... the people we primarily use are from Arizona. Sound like them to you?



Originally posted by Alex Lancaster

Gordon, it would be good if You become a distributor, it would be difficult for a forumite to get 2 cone kits directly, but You could do it for a reasonable profit, You know about 20dB more than I do, and You could also help our European friends; It would be too difficult for me from here.


I've been thinking about that. I really haven't had the time to work out the logistics for something like that, but I may investigate the possiblity. It will take some more LMS/LEAP testing, to verify parts accuracy and consistency before I'm willing to commit to it, but at least this one test was promising in that regard.


Originally posted by Alex Lancaster

On the cloth surrounds, they connected me to a glue supplier; We have been using with good results, up to now, Vynil "sealant", but have not been able to concoct the OEM sticky stuff, pls let me know what You have on it.

Alex.

Well, we're kind of in the same boat... fortunately, we're also affiliated with Waldom, and still can get the AA-75 PVC edge treatment, which seems to be a pretty decent match for the sticky stuff on 123As, Altecs and such. Sure seems to test the same, when applied to drivers. Dunno about any other sources, though.

In all, I somewhat think of these efforts as a "safety net" since I'm personally worried, that JBL seems to be discontinuing more and more of the kits for vintage drivers. I can certainly understand economic forces for a major manufacturer, how it's impractical to cover niche stuff... but OTOH, we need to have a solution available for those people who want to keep this old stuff going, even after the manufacturer forgets about it...

Regards,
Gordon.

[Alex Lancaster]

Same guys.

[Mr. Widget]

Originally posted by GordonW
...I'm personally worried, that JBL seems to be discontinuing more and more of the kits for vintage drivers.

They do. But if you put in an order they keep track of the orders and when enough are taken they will run the parts. They still run 150-4C kits from time to time. That is a woofer that has been discontinued since the 60's! I doubt in the scope of Harman this is a money maker. I think this is a service to our community and I for one am happily amazed that unlike so many other brand names JBL is still really JBL and while they have grown to a lumbering size there are those there that do still care.

Widget

[GordonW]

It's nice that they do occasionally run off batches of older kits. It's definitely laudable, especially in these "disposable society" days. It's a great solution, for the person who just wants to update something to keep his (otherwise working) old system in top shape, and isn't in a hurry for parts.

But the problem arises, when there's a store customer breathing down your neck, that just wants his broken speaker fixed, and JBL has a 5-6 month lead time to produce more kits. Time=money... or in this case, excess time = the potential to NOT get the job, and LOSE income...

Sometimes, we just have to do the best we can, with what we've got to work with at the time. It often takes some really exhaustive parts chasing; other times, it just takes flat-out making or re-making stuff MYSELF. I've gotten pretty good at these processes over the last 25 years... I know I've certainly gotten enough practice! :banghead:

Regards,
Gordon.

[Ian Mackenzie]

Great account of the technical aspect of recone alternatives but how do you stake your claim on the authenticity of your vintage JBL when you decide to sell it off on Ebay with after market spares??

I'll be reconing my 2245 soon and you can be sure it will be the factory spare part.

Given the average foam surround lasts 10-15 years and not acounting for burn outs and abuse you would think it would be time to move on after the 1st or 2nd recone? I mean life''s about moving on and the same appied to speaker ownership.

I think I would rather the JBL ICON of an LE10A cut in half and mounted in perspex in my audio museum than a "has been" that is easily outperformed by better brands current day design and materials technology.

:smthsail:

Ian

[GordonW]

Frankly, there will be ice cubes in hell before I'm likely to sell any vintage JBL stuff on Ebay. I have no interest in "cold call" selling. If I sell something, it'll be to someone who I can get a feel for, that will honestly understand what was done and why it was done. And that can only be done in a one-on-one conversation.

I make it crystal clear, about anything I've reconed, as to where the parts came from and how the work was done. OTOH, I also let the person see and hear, in person, the end result. I've had many times, where the buyer stated that if I didn't say, that they could have had no way, by either looking or listening, of knowing it was a non-factory recone.

In short, craftsmanship can just as easily come from a hand-assembled kit assembled by someone who knows what they're doing, as it can off of an assembly line of a major manufacturer. And as far as percieved value is concerned- in almost every case, once it's non-original parts, it really doesn't matter, in any sort of "collector" hobby, what source the parts came from- only that they LOOK and ACT as the originals. These recones easily fulfill that requirement. It's either completely original or it's not- once you've gotten there, it's just a matter of how well it emulates the original...

Oh, and BTW: How many of these replacement kits are install tested by JBL, by anything remotely equivalent to the LEAP/LMS system I use? Do they still actually QC poll all these obsolete parts kits from the assembly line? If not, then there's a distinct probability that like the LE14 with the excessively stiff factory spider mentioned above, that the QC of a good custom reconer may EXCEED the accuracy of a factory replacement kit, if there's nobody "watching the store" on the assembly line to make sure things stay in spec. Not to mention the cases where JBL "lumps together" different drivers sharing a common recone kit- when there WERE subtle differences in the different driver models. In some of these cases, I've been able to custom assemble a kit that BETTER matched the factory specs than the specified factory replacement...

Sorry if this seems a bit defensive or abrasive- but it kind of irks me, when I (and many other custom reconers) take the time to do things as good or BETTER than factory, and I get lumped in with fakers and charletans...

Regards,
Gordon.

[Earl K]

Gordon, I appreciate your sort of custom work and all the effort put into " getting it done right" . I'd also like to continue to hear about your success stories in these areas - so please don't get too defensive about the direction this thread has gone.

For Instance; I'd love to come up with a White or Black Aquaplas substitute. If I ever do, I'll share it and take the consequences .

NOW:
On a different note and direction, Inquiring minds want to know :

- How did you arrive at your conclusion to add 10 grams (?) before you actualy did the addition of weight ?

- And how did you meter out that specified weight ( was this airbrushed on ) ?

- Was the added weight equally distributed ?

Thanks for sharing <. Earl K :)

[GordonW]

Well, more than the weight addition being NECESSARY for the design, it was more motivated by the fact that a) I wanted the cone to be stiffer, and more resistant to moisture wicking from the air (it is to be driven probably pretty hard, and I didn't want to risk the kind of cone failures I've occasionally seen on some ribbed-cone JBL drivers, so I decided sealing and stiffening was in order) and b) I knew that the box volume, passive radiator tuning and the parametric EQ on the sub amp could all be tweaked to get a proper transfer function, all things considered. Given that this woofer was not to ever see any midrange, a bit heavier cone was not in any way detrimental; in fact, it reduced the amount of EQ necessary to achieve flat response in the bottom octave.

As for the application- this was a liquid, brushed onto and soaked into the backside of the cone. I used a small (1/2" wide) brush, and liberally coated the cone with several coats, until I achieved about 10 grams. This seemed to be a good stopping point, as it was the point where the cone started to seem to stop soaking in the material as well, and it started to "puddle up" just a tiny bit on the backside instead of disappearing immediately.

I'm also working on an Aquaplas substitute. I've been thinking of and/or experimenting with various materials, like talc powder, cement powder, fine sand, latex paint, powdered wallpaper adhesive (believe it or not, it DOES seem to have some rather INTERESTING damping properties when mixed with latex paint and fine sand!), and the like. Haven't really gotten anything that meets all the criteria yet, but some combinations have looked at least promising in initial tests...

Regards,
Gordon.

[Mr. Widget]

"liberally coated the cone with several coats, until I achieved about 10 grams."

Was this 10 g. the weight during application or the weight after all of the solvent evaporated?


Have you taken measurements again after the woofer has broken in?

Widget

[Alex Lancaster]

Seriously, people here have tried silica sand and sugar!???.

[Oldmics]

I would advise against the use of anything in the food products group for speaker dabbling.

My daughters made a ginger bread cookie house a long time ago.I coated the house in many repeated layres of clear laquer and thought it was sealed tight.

Went to set it up last Christmas and the critters had enjoyed the desert, laquer and all.Broke my heart!

Critters will eat anything

Oldmics

[GordonW]

Originally posted by Mr. Widget
"liberally coated the cone with several coats, until I achieved about 10 grams."

Was this 10 g. the weight during application or the weight after all of the solvent evaporated?


Have you taken measurements again after the woofer has broken in?

Widget

The 10 grams (actually, to be accurate, it was closer to 11 grams) was after the solvent evaporated. I applied a couple coats, and stopped when it started to not "take" as well. I had set myself a maximum limit of 20 grams (or more accurately, .75 ounce which is like 21 grams), which I would measure BEFORE the solvent dried... I knew that if this was the case, I could kind of "home in" on the right value. Take it to 20 grams, let it dry, measure, apply a bit to get it back to 20 grams, let it dry, measure, rinse, repeat... :D It just happened that 10-odd grams dry was the amount the cone wanted to soak up...

Haven't gotten to break-in the woofer properly yet. Don't have an amp big enough here at home, to do it (the amps here that are easy to access are a whopping 14 and 27 watts per channel :rolleyes: ). Once the woofer is in the cabinet and the plate amp (which will produce 270-odd watts into an 8 ohm load) is installed, that should be short work. I dunno if I'll get to run the woofer back on LMS afterwards, though... I'll just have to base my judgement on how well it tunes out in the box, with the known passive radiator Sd and Mms...

Regards,
Gordon.

[readswift]

Hello, I also read that aftermarkets are very close.
quoting GordonW:
"such as 2235s, LE5s and such, it is possible to budget a good hour of labor per driver (a complete recone shouldn't take more than that, actual active work time) and still come out with a final price using aftermarket parts, that's SIGNIFICANTLY lower than the factory kit, not including labor for install..."

I would be very interested cause I need 4 of em, can someone point out any info ?