Printable View
I have managed to come up with an equivalent bandpass circuit for the 4333/S300/L300 which negates the need to use the very expensive and hard to find tapped autotransformer.
If anyone is interested I will post the schematic.
Has anyone thought of using charge coupled technology in their 4333/S300/L300's?
If anyone is interested I will post that schematic as well.
Hi Giskard
I'm always interested in seeing these circuits, especially transformed/equivalent types .
regards <. Earl K
Here is the schematic.
It's voltage drive is a dead ringer for the original voltage drive as measured on December 16th, 1975.
Most of the components are available from Parts Express. The LF inductor is the 14 gauge perfect layer wound air core. For those who like the signature JBL bass hump at ~ 30 Hz you may need to tune the enclosure up a couple of Hz. All the MF inductors are the 18 gauge perfect layer wound air cores.
4333/S300/L300 with charge coupled technology
Beauty!
I'm the happy owner of four S3100, but nevertheless on the look-out for improvements, so I would like so see the charge coupled cirquitry for this model.
Thanks Rob !
<. Earl K
Alrighty then!
Here's the stock schematic.
Let us know what you think if you decide to give it a try. I wouldn't mess up your stock networks, I'd build from scratch.
Use all polypropylene if you can instead of any NPE's.
For those who care, I believe the S3100 MK II is already charge coupled...
Thanks, Giskard, for the charge coupled version of the NS3100.
One of my S3100 has a problem with static, so I may kill two birds with one stone by building the charge coupled version of the cross-over cirquitry: Getting rid of the static (which is probably due to a defective capacitor) and getting better sound.
Whoa, man, Giskard... is that big HF notch filter/impedence "compensation" network connected to the input terminals of the cabinet there? What exactly does that do for the effective impedence of the whole cabinet?
Regards,
Gordon.
I've never seen nor run an impedance curve on an S3100 so I don't know.
I tried to find the -6db point of an air core inductance(for example, 1.5mh) which its -6db point isn't exact the .75mh and then tried to make a line-out from the -6db point. It seems it works like a taped autotransformer. But I found its frequency response droped several dbs as attached picture. And the small load resistance the more it drops. And I also found the higher inductance was and the less it droped.
Who knows the frequency response curve and could show it here? then we can try to make taped autoformer ourselves.
The attached is freq response with 8ohm load.
Please give some good advice for the drop.
Thanks.
sorry for this large picture. It's the first time to load a picture.
Quote:
Originally posted by Giskard
I've never seen nor run an impedance curve on an S3100 so I don't know.
JBL has published an impedance curve for the S2600 in the owner's manual for the S2600/S3100. The former has a 12" woofer and 92 dB/W sensitivity, whereas the latter has a 15" woofer and 95 dB/W sensitivity, but they are otherwise identical. I can't attach the entire owner's manual as it exceeds the 160 KB limit on this forum.
First, the scale must be wrong -- the impedance can't be between 2 and 3 ohms over most of the frequency range.
Second -- and this is not apparent from the attached file --, the cross-over cirquitry in the S3100 mk II is not only charge-coupled, it also has a resistor in parallel with the woofer with a higher resistance than the resistor in parallel with the woofer in the S3100. What is the consequence of this other than a lower sensitivity in the latter case?
"it also has a resistor in parallel with the woofer with a higher resistance than the resistor in parallel with the woofer in the S3100. What is the consequence of this other than a lower sensitivity in the latter case?"
It helps to reduce the motional impedance of the transducer making the load a bit easier on a driving amplifier.
This is a very important thread for those interested in this group of speakers.
Is it possible to change the title?
Written as it is the search engine does not pick it up for the individual model numbers, and nowhere in the thread are any of those three model numbers mentioned individually. I think hyphens in place of the slashes would do it, or a space on each side of each slash. As it is now it is buried.
Thanks
David
Yea, I agree with Dave. Also, it could be entered into the reference section .
And , the "S300" in the title ought to be fixed to read S3100 .
Giskard, did you ever get a DCR value for the 1.0 mh coil in the highpass circuit for the S3100 ?
Maybe Jan D ( or Bernard ) could measure it for us .
- It should be quite high or else that series LC(r) across the driver , will create a huge notch filter.
regards <. Earl K
I've asked Don to fix the title using spaces between the slashes.
The S300 is the LCS version of the L300.
I want to say it's on the order of 6 to 8 ohms.Quote:
Originally posted by Earl K
Giskard, did you ever get a DCR value for the 1.0 mh coil in the highpass circuit for the S3100 ?
I believe it is supposed to "fix" it so one doesn't blow up an amp. :DQuote:
Originally posted by GordonW
Whoa, man, Giskard... is that big HF notch filter/impedence "compensation" network connected to the input terminals of the cabinet there? What exactly does that do for the effective impedence of the whole cabinet?
Cool!Quote:
Originally posted by Giskard
I've asked Don to fix the title using spaces between the slashes.
Thanks,
David
Did you try 6 to 8 ohms in your modelling software?Quote:
Originally posted by Earl K
It should be quite high or else that series LC(r) across the driver , will create a huge notch filter.
Hi Giskard,
No, I haven't tried those dcr values.
I don''t have any modelling software for designing "passives" . I'm not aware of any (spice) packages that run on Macs . Perhaps I just haven't looked hard enough.
When one of G.T.s advanced designs tweaks my interest, I mockup the new circuit portion, measure it, audition it and dismantle the "project" .
As to the real world implementation of series LCR traps, I have yet to pursue acquiring a practical/working knowledge . I'm still studying the effects of "broadband" damping from resistors and/or shunted inductors. So, LCR traps are a type of finessed tweak that I have yet to explore .
<. Earl K
I've been pricing out a charge-coupled crossover for the 4333A and so have been looking closely at the component values for the first time. The charge-coupled pair for the series/blocking capacitor in the MF section is not symmetrical. Is this intentional? The original value is 16.5 uF, so a pair of 33's would do it, but one is labelled 39.
(If someone can get in touch with Giskard on this, I'd appreciate it.)
David
Hi Dave
Sorry, can't help you there but ,,,,Quote:
Originally Posted by speakerdave
- Symmetrical pairing of the two caps is not important. What is important, is the final net capacitance of the biased pair . In this case the series pair ( 33 uf & 39 uf ) equate to 17.875 uf which fullfill the 18 uf spec.. This value ( 18 uf ) is what Giskard worked out for his "Equivalent Bandpass" ( or ; lacking the autoformer ) .Quote:
Originally Posted by speakerdave
- By the looks of those values speced ; they were included because of their availability ( instead of two 36 uf caps ) .
- You're correct ; to equate back to the original value of 16.5 uf as found in your original N3133 network ( with autoformer ) / a pair of biased 33uf caps would be used .Quote:
Originally Posted by speakerdave
:) Earl K
OK, I was comparing it to the wrong circuit. Thanks, Earl.
David
I built the tweeter/mid charge-coupled version for a 3-way project, it is working out very well.
The capacitors are Axon, from http://www.zalytron.com. For the 36uf I used a 12uf and a 24uf in parallel.
Cool. Thanks, John.
David
O.K., so who's got the schematic for S3100 MkII?
[This seems to be a dual-topic thread.]
Gonna build a mini-Everest here.... :p
Click here to see the "network" page of the S3100mkII tech sheet, in jpg form.
Here's the full 2 page pdf.
:)
Thank you Earl!
[Jeez, NOTCH city, that one.... :p ]
- Yeh,,,, IMO; it's quite representative of JBLs' evolved network-design philosophies, as found in their SOTA systems ( the compression driver types ) .Quote:
Originally Posted by Zilch
- It's good reading to compare the original S3100 network with that of this S3100mkII .
- One can learn to design with series LCRs, if one studies the design lessons laid-out in the Band-Pass Filter thread in Project May . The last time I looked, all the pertinant information was still buried within those threads . Fine work was done by Giskard and the Widget on that bandpass section ! I think it would be best to imbed those fomulas into an Excel spreadsheet to keep the "brainfreeze" to a minimum . ;)
Capacitors are charge coupled and bypassed as well.
David
The Project May Bandpass filter thread seems to be locked out for casual readers - thats a shame. I suspect I might learn quite a bit in there ... sigh
Quote:
Originally Posted by Earl K
Hi Heather,
PM Don McR. and ask him to grant you access to that forum .
:)
Hi Giskard.
Some times ago, i decided to make a charged filter for my speakers, i have ordred the components, so i gues next week they will be running with this
I meet some people who said it was a really nice thing to do. Sounds great, but coud not tell me the difference.
It seems you know a lot about lot, so what do you mean about this ?
My config is 077, LE104, 2123H and 2243H with at NAD S300 for the top and a home build brigded 300W amp for the 2243, Active x-over.
The 4344 MK2 is factory build with charged caps, and if it sound so great, why hasīnt more speakers been built with this ? Cost ?
Regards, Anders
In a phone conversation about a year ago Greg wondered the exact same thing. He really didn't know why biasing capacitors hadn't caught on. A few weeks ago he finally got around to biasing the network in his 880 Array center channel and said the result was a marked improvement. I'm currently building five biased networks for Tannoy DC's. They are replacing JBL 4406's in a mini 5.1 HT setup.Quote:
Originally Posted by ABJ
As per a PM today here is the schematic for bi-amp only L300's, S300's and 4333's.
I personally would bias (charge couple) them.
HiQuote:
Originally Posted by 4313B
I was just wondering what the x-over point are? and are all drivers meant to be 16ohm or 8 ohm?
Several people seem happy with this combo, so I think I will try to build it.
Vern
I have built this crossover and all my drivers are 8 ohm. Do I have to change anything? Odd
Quote:
Originally Posted by 4313B
Crossovers are designed to work into a certain load. What happens with a particular change in any particular situation may be a little unpredictable, since most crossovers are tweeked empirically after the math is all done. The 2405/077 tweeters are all about 10 or 12 ohms, whether they are labeled 16 or 8--it doesn't matter--they are all the same. The crossover is designed for that, so you are ok there. It might be worth your while to get 16 ohm diaphragms for your 2420's. The difference in impedance will likely effect a change in the high pass and low pass crossover frequencies for the horn. Getting new diaphragms of the impedance the crossover is designed for will be a whole hell of a lot easier than redesigning the crossover.
Assuming an 8 ohm nominal value for the 2231A/H or 136A/H at crossover point, could one sub in the 2235H or other woofer (like a 2245?) if the impedence were similar?
I would also assume the Zoebel may have to be adjusted if the woofer coil inductance is different.
Bart
Hi Bart.
2235H? I thought that was the current L300 replacement suggestion.
In general, I would think the answer to your question is: it depends...
A nearly matching impedance (curve) is not necessarily
sufficient (sensitivity, power handling, audio frequency response, ...)
Hi Grumpy,Quote:
Originally Posted by grumpy
Yes, that's what I was getting at. Since the 2231 is less available than the 2235, or for those whose L300/4333x's have had the 2235 already subbed in, since JBL specifies the 2235 as the 2231 replacement, the differences are probably small, or are positive enough to be an acceptable replacement.
If someone is rebuilding the L300/4333 with new crossovers, I'd hope there are some acceptable options for the woofer, within reason. But I also agree that assuming any 15 inch JBL driver as a replacement would likely be acceptable or appropriate would be wishful thinking and likely require some re-engineering of the crossover system.
I raised the 2245 question in case someone is thinking about somehow using that in a different size box but otherwise keeping the mid/high end intact. However, on second thought, probably need to move over to the 4343/4344/4345 type crossovers as a starting or end point.
I can see now why the L400 would have been such a dificult project.
Bart
Yes yes, first post and I'm reviving the dead and buried :D
But why start a new thread when this one is perfect.
I'm renovating a pair of home built speakers (2215h/2420/2405) and will be using this crossover so kindly supplied by 4313B.
But I have one question.
As I don't have the variable L pads, could you guide me as to a suitable fixed resistor solution for both the mid and tweeter pads please ?
Thank you.
Quote:
Originally Posted by 4313B
Welcome to the forum... can't help with the specific question, but you need a different network since you are using the LE15A/2215 woofers instead of the 136A/2231A that this network is designed for.Quote:
Originally Posted by OldManWagon
Widget
Thanks, hi.
I'm using the 16ohm 2215h, with a change to the bass section of this crossover.
(zobel 13.5uF 10 ohm, filter 3.3mH 12uF - Copied from the 3110 crossover).
ok, so next question (related) would be, if the 2420 and 2405 are ~10ohm impedence, with the 20 ohm across them gives ~8ohm 'driver' as seen by the crossover, so therefore, if i know the sensitivity of the drivers, I can calculate the required L-Pad values.
So if;
2215 = 96db/w/m
2420 = 118db/w/m (24db pad ?? whats the real output of this driver when mounted to a horn/lens ??)
2405 = 104db/w/m (8db pad)
Can anyone see anything obviously wrong with those numbers ?
Or have any comments..
At this stage I'm going to build the crossovers, with fixed L-pads of around 10db att. in place of the variable ones, and fine tune levels by ear and frequency sweeps.
[QUOTE=OldManWagon;311738]Thanks, hi.
So if;
2215 = 96db/w/m
2420 = 118db/w/m (24db pad ?? whats the real output of this driver when mounted to a horn/lens ??)
2405 = 104db/w/m (8db pad)
Can anyone see anything obviously wrong with those numbers ?
Or have any comments..
....
my experience with 2420 with 2308 Lens and 2307 the sensitivity is about 106~108dB/1m
.....Quote:
Originally Posted by 4313B
Capacitor-"serial capacitor" to the MID section would be well-come in order to protect
242x compression driver
....
Attachment 51323
Cool, Thanks.Quote:
Originally Posted by ivica
Made up the crossover and had a play, Around 9-12 DB Att. on both mid and tweeter seem roughly the area to aim for.
This may be a different thread question but has a S8/S20 charge coupled network been worked out to replace the LX5/N7000-N8000 combo? :confused:
Has anyone put out a diagram for a charge coupled crossover at 500 and 7000 hz? Or can someone note the mods to the Charge coupled l-300 crossover to change it to 500 and 700hz. I am proficient with the restoration and repair thing, but when it comes to design of the crossovers, I wish to leave it to those with more experience. Any help would be greatly appreciated.
Thanks
MB