View Full Version : JBL xover & horn compensation primer

02-12-2005, 11:49 PM
Hi I read the recent thread Quick and Dirty 4430 'shttp://audioheritage.org/vbulletin/showthread.php?t=3838'
thread with great interest. However as a JBL novice it has raised some questions that may be found in a JBL tech note etc on the evolution of their xovers etc. Can anyone help me locate this info?

Anyway my specific questions are as follows

1. There appear to be three series of Pro passive xovers
(a) 3105, 3106, 3110, 3115 & 3120

(b) the "A" models 3110A, 3115A, 3120A and

(c) 3150, 3160, 3180, 3152, 3182 etc

My understanding is that (a) are for use with pre CD horns and that (b) have a HF boost option which I guess is for some CD horns (eg 2360, 2380)

I have no real understanding of what (c) can be used for other than that they appear to have higher power ratings and more tap options.

I would appreciate advice on the suitability of all three types for home use. It can be taken as granted that I understand the need for differing freq xover points and therefore different xovers depending on which driver and horn combination is used. I have some CD horns I am currently using (2380A) but also some earlier 2" (eg 2390 and 2397 Smith horns) and 1" (2392) that I want to play with.

2. Would also appreciate some advice on the evolution of the JBL active xovers.

(a) What is the essential difference between the 5234 and 5235 xovers?
Can they both use the same card plug ins? and

(b) Are the later M552/ M553 significantly better? I am aware these have options for CD horns (2360 & 2380) but I noted from the earlier thread these options are not considered ideal for 2344.

Would also appreciate any insights into earlier or later JBL active xovers other than the digital ones which are really beyond, and in a different ball park to, the above.

3. Running the risk of asking too much in one post, I would really appreciate advice on the efficiency of the earlier pre CD 2" (eg 2390, 2397 ) and 1" (2392 etc) mid difraction horns say for simplicity assuming they were all driven with a 2440, or a 2420 (in the case of the 1" ) horns. Believe I read in an earlier post that they were all about 108 dB/1W/1m. Also do any of these require any equalisation?



02-13-2005, 06:56 AM
I can help you with 2

The diferences between the 5234/5234A and 5235 are

They can all use the same cards

The 5234 is the one with the least amount of flexabillity Screw down Strip terminations

The 5234A adds the DIP Switch selectable LF boost/cut and has 1/4 Jacks plus Screw Down Strip Terminations

The 5235 adds bypass caps has XLR terminations and is the latest and greatest. You have to watch the polarity on the XLR's not the norm used today I believe it has Pin 1 Hot where Pro Gear now is Pin 2

All three manual are up on the JBL Pro site so have a read to see in detail what's up with them.

The M552/553 are very good, quiet and transparenet has only XLR terminations. I have never directly compared the M552/553 to the 5235. They do not have the DIP selectable bass curves in the earlier units. They do not use cards and are variable frequency but fixed 24dB L/R slopes. The earlier crossovers gave you a choice of slopes 12 and 18. The CD curves are set up for a movie house ie PRO applications to meet the curve they use as standard in them. Above 10K things are going south. I use the M553 with 2344 but just use the passive compensation lifted from the 4435 schematic.

Hope this help's


02-13-2005, 11:04 AM
Here's my simplistic understanding of it. Compression drivers have an inherent rolloff in response of 6 dB per octave. You can see it on the published curves for the various drivers.

The earlier exponential horns compensate for this by progressively narrowing the beamwidth as the frequency increases. Thus, at higher frequencies, the sound energy is concentrated on axis, which is fine if you're sitting where the horn is aimed; you can hear it all from the "sweet spot."

Constant directivity (CD) horns, on the other hand, are designed to distribute the sound energy at all frequencies over a wider, controlled beamwidth, which is specified for each horn type. Thus, it is necessary to provide progressively more energy as the frequency increases to compensate for the driver's rolloff, as that energy is (presumably) distributed evenly over the wider sound field.

There are hybrid combinations of both approaches, as well. See, for example, the 2370A. Designed to require minimal compensation, it allows the the vertical beamwidth to narrow (within limits) with increasing frequency to provide a uniform horizontal field.

The major challenge in trying to make a successful two-way speaker system like 4430 with compression drivers serving the mid and high frequencies is that their response falls off even more rapidly above approximately 15 kHz, i.e., in the VHF/UHF region. Thus, even more HF boost than the standard CD 6 dB/octave must be provided in an attempt to replicate the performance of a system having a dedicated UHF driver, as in a three-way system. This is accomplished in the high-pass section of the 4430 passive crossover via selectively attenuating the mid versus HF and VHF frequencies in the requisite controlled fashion. Even so compensated, some find the biradial two-ways like 4430 lacking the UHF "sizzle" of true three-way systems.

At a certain point, of course, at the performance limit of available technology, it all runs out of gas. While some horns may be found to have better high frequency response than others, the primary limitation is in the compression drivers themselves. The 2344A is not significantly different from other horns in this respect, but we're always looking for an alternative that might provide more UHF "edge" at these performance limits.

If you're NOT trying to push the VHF envelope, i.e., using CD horns as midrange only, then the compensation built into the the 31xxA, 5234/5235 or M55x crossovers is likely adequate for the task. You have other judgements to make there as to whether you want to use active or passive compensation, or a combination of both. 31xxA allows "flip o' the switches" selection of compensation within its design limits. It could benefit from installation of bypass capacitors, tho. 5234/5235 provides facility for precise tailoring of compensation for specific dirver/horn combinations, AND an independent LF "bump" filter (@ 20, 30, or 40 Hz) on 5235, but swapping out cards (assuming you have or build different ones) is inconvenient. M55x has lovely, clean 24 dB/octave filters with "turn o' the knob" frequency and gain settings, plus mute buttons on all channels, and is readily available for cheap on eBay (~$100 for M552). An RTA will tell the objective numeric tale, but your own ears must be the final arbiter of what works for you in your listening environment.

With respect to the larger question of what is appropriate for home use, there are those who assert that compression drivers are not appropriate at all. Others will argue that their unique performance characteristics are particularly well suited to home theater use. If you use them, then, there is the secondary decision of whether you want the precise imaging conferred by a narrow "sweet spot" or the more diffuse soundstage provided by constant directivity horns, and if so, how wide. Indeed, in some of JBL's high-end Synthesis systems, horns are provided for home theater use, but cone midrange drivers are switched in for music playback.

Me, I like the 4430 for everything. Everybody on the couch gets the same fine JBL experience. :applaud:

02-13-2005, 05:22 PM
Rob & Zilch

Thanks for your thoughtful replies

I am currently using a M553 with my 2235x2, 2450/ 2380 & 2404 main setup and it seems to work fine. The only thing I find a little annoying is the 5K limit on the HF freq cut off though I guess I could mod it to make it higher if for example I wanted to use 2405. I am impressed with the way JBL have made detailed info available on their web page on these xovers. Wish all audio companies were as helpful.

Rob, I guess from what you have advised, that the 5235s are worth acquiring because of the added flexibility and fiddleability for tweakers. Thanks again for the useful summary of the active xovers.

Zilch, I understand what you mean about the sweet spot vs broader coherence that CD offers. There was an interesting discussion on this in the context of waveguides v tractrix horns on the Audio Assylum High Efficiency Speaker forum a few months ago which is worth a read. It was headed "Revisiting Horn Soundstaging & Imaging" and was initiated by Mark Seaton if anyone wants to check it out.

I was very interested in your comments on home theatre v stereo. My limited experience with compression horns is that I find them great for stereo. The main difference I have found compared with friends with typical modern domestic Hi Fi speakers is the extended freq range and effortless dynamics that a 2235x2, 2450/ 2380 & 2404 system offers. Some of them clearly prefer a more restricted range.

The reason I raised the xover issues is that I want a rear set of speakers to complement my main speakers when used for HT. I picked up a pair of 4331As recently. I was planning on replacing the 2420 / 2392s with 2441s/ 2397s and using 2405s or 2404s with a passive xover if possible. Am trying to avoid active xovers for the rears mainly to avoid the added complexity.

For my main system, like Doogster who posted on the topic in December, I am planning on trying out some 500Hz conical horns/ waveguides when construction time permits.

Zilch, on the "sweetspot" thang, I suppose the most important thing is to be aware of it and as you so well put it to decide what is best for you in your own room. Personally I am inclined toward the alternative as I like to share my music with others and find it a little distracting to have to sit in a precise spot. An extreme example of this I found with a pair of QUAD ESL57s which to my ears have a very precise sweet spot.