The JX6 4-driver line array has been in our office for a few months now. This spiderless driver's strengths are tremendous coherence and seamlessness. Within it's range it is lightening quick and never fatiguing.
Each driver has about 24sq/cm of surface area. x4 drivers would give about 100sq/cm or roughly equivalent to one 5 1/2" full range driver. Dynamics must be limited then to the amount of air that limited surface can move. By comparison the 10" Ciare we have here is a noticeably more dynamic-sounding (despite it limited Xmax) but it has ~320sq/cm. So it should be expected to move a lot more air with more than 3x the surface area, within its range. But in every other respect the Jordan beats it by a country mile. Our listening area is large, >1000sq/ft so ability to move air is an issue. Interesting that the 4 driver array has the same subjective efficiency as the Ciare - at the listening position.
What if we were to double or triple the number of drivers in our Jordan line array? 8 drivers would give ~200sq/cm of cone, roughly equivalent to an 8" driver (Lowther, Supravox territory) . x12 would be almost 300sq/cm, a little more than a 9" driver.
These little wonders are expensive so for this project I've acquired 16 more drivers, 2 x 8driver arrays. If it seems advantageous, I can always add the existing 4 drivers to make it a 12 driver array.
Noted are some comments about full-range drivers being problematic in line arrays. I haven't found that to be the case so far and frequency distribution in our room measures roughly similar to to single full range driver. No combing is evident. The JX6 do mount fairly close together. I tried the suggested EQ from the Jordan site but then removed it; it was for the JX53 which is no longer made and quite a different driver. The JX6 sounds more lively without EQ.
We will see what the 8 driver array brings.
The Jordan website suggests using the linked design for the box. It's not the typical configuration for a line array but those that have heard it at Ted Jordan's home report that it sounds great. It would certainly suit the small listening environments we have here in a typical Hong Kong flat.
So wondering what you think? Has anyone built the cabinet as shown? Or any comments on it? Caveats etc welcome.
Whichever cabinet I go with will be built in stackable 4-driver enclosures.
Brian
Each driver has about 24sq/cm of surface area. x4 drivers would give about 100sq/cm or roughly equivalent to one 5 1/2" full range driver. Dynamics must be limited then to the amount of air that limited surface can move. By comparison the 10" Ciare we have here is a noticeably more dynamic-sounding (despite it limited Xmax) but it has ~320sq/cm. So it should be expected to move a lot more air with more than 3x the surface area, within its range. But in every other respect the Jordan beats it by a country mile. Our listening area is large, >1000sq/ft so ability to move air is an issue. Interesting that the 4 driver array has the same subjective efficiency as the Ciare - at the listening position.
What if we were to double or triple the number of drivers in our Jordan line array? 8 drivers would give ~200sq/cm of cone, roughly equivalent to an 8" driver (Lowther, Supravox territory) . x12 would be almost 300sq/cm, a little more than a 9" driver.
These little wonders are expensive so for this project I've acquired 16 more drivers, 2 x 8driver arrays. If it seems advantageous, I can always add the existing 4 drivers to make it a 12 driver array.
Noted are some comments about full-range drivers being problematic in line arrays. I haven't found that to be the case so far and frequency distribution in our room measures roughly similar to to single full range driver. No combing is evident. The JX6 do mount fairly close together. I tried the suggested EQ from the Jordan site but then removed it; it was for the JX53 which is no longer made and quite a different driver. The JX6 sounds more lively without EQ.
We will see what the 8 driver array brings.
The Jordan website suggests using the linked design for the box. It's not the typical configuration for a line array but those that have heard it at Ted Jordan's home report that it sounds great. It would certainly suit the small listening environments we have here in a typical Hong Kong flat.
So wondering what you think? Has anyone built the cabinet as shown? Or any comments on it? Caveats etc welcome.
Whichever cabinet I go with will be built in stackable 4-driver enclosures.
Brian
An externally hosted image should be here but it was not working when we last tested it.
The stackable is probably a good move. I suspect the short array works because of the driver size, and that it's intended for the far-field (the opposite of most line-arrays). If you increase the array lenght, you'll also shift the transition between near and far-field further back, and I suspect because of this (i.,e you'll probably begin to end up in the nearfield), you'll start to run into lobing issues in the HF. Tred carefully. Going stackable means you'll have a fall-back. Other option would be to create a curved array. I'm not a fan because of the tiny sweet-spot they provide, compared to a regular arrays huge 'sweet area' as Jim Grifin describes it. But it should help focus the HF a bit.
Best
Scott
Best
Scott
Hi Brian
I'm also running the JXR6 but as singles. To throw a spanner in the works, I believe that the next step up from a 4-unit array is actually a 9-unit set up, then the 16 after that. I have the write up about it somewhere and will dig it out.
The box design on the Jordan site for the array is designed to mount against a wall. Sounds good but the best I heard was a prototype using the JX53 array in a closed box similar in size and shape to the VTL (wide and shallow). Maybe scaling it up would be too visually overpowering for a 9 or 16 array. The Jordan drivers are generally quite sensitive to back reflections so anything you can do to reduce this, the better.
Colin
I'm also running the JXR6 but as singles. To throw a spanner in the works, I believe that the next step up from a 4-unit array is actually a 9-unit set up, then the 16 after that. I have the write up about it somewhere and will dig it out.
The box design on the Jordan site for the array is designed to mount against a wall. Sounds good but the best I heard was a prototype using the JX53 array in a closed box similar in size and shape to the VTL (wide and shallow). Maybe scaling it up would be too visually overpowering for a 9 or 16 array. The Jordan drivers are generally quite sensitive to back reflections so anything you can do to reduce this, the better.
Colin
Thanks Scott and Colin,
Scott:
I will pay attention to the nearfield/farfield issue as we have a big room for testing but if I take them home they most definitely will be listened to closeup about 8ft from the drivers. At the office I can get up to 30ft from the speakers.
Colin:
Interesting, would really appreciate knowing more.
That sounds interesting as well. Was that used with EQ and against a wall? Is there a source for more info? The 4 driver array I have is also in a wide flat sealed acrylic box, firing straight ahead. Not against the wall though.
Internal cabinet reflections? I held the diaphragm of the JX6 and it is is almost imperceptibly light. It makes sense that back reflections could easily excite the back surface and muddle up the sound radiating from the front.
Thanks, gentlemen, for the comments.
Brian
Scott:
I will pay attention to the nearfield/farfield issue as we have a big room for testing but if I take them home they most definitely will be listened to closeup about 8ft from the drivers. At the office I can get up to 30ft from the speakers.
Colin:
Interesting, would really appreciate knowing more.
That sounds interesting as well. Was that used with EQ and against a wall? Is there a source for more info? The 4 driver array I have is also in a wide flat sealed acrylic box, firing straight ahead. Not against the wall though.
Internal cabinet reflections? I held the diaphragm of the JX6 and it is is almost imperceptibly light. It makes sense that back reflections could easily excite the back surface and muddle up the sound radiating from the front.
Thanks, gentlemen, for the comments.
Brian
That's two of us. The polar response is going to look very ragged, with severe lobing in a nearfield array, and pretty dramatic attenuation of the highs. I know the Jordans aren't large and havce narrow surrounds, which will push it a bit higher than most other units, but you can't bypass the laws of physics. As I say, I suspect the only way to mitigate against this will be to construct a curved array rather than a line array, which will help focus the treble energy. It's probably the only thing you can really do, nearfield array-wise, with full rangers. I'm still not convinced, but I think that way you'll at least be giving yourself a better chance of something workable.
I honestly hope I'm wrong. But I'm pretty sure I'm not.
Interesting remarks re the rear of the cone. Probably lining the rear of whatever enclosure with thick foam or rubber would be a good move?
I honestly hope I'm wrong. But I'm pretty sure I'm not.
Interesting remarks re the rear of the cone. Probably lining the rear of whatever enclosure with thick foam or rubber would be a good move?
Hi Brian
I'll dig out details about the 9-unit array later today and try to post this evening.
Ref the VTL box, this was a prototype, plain sealed box with no EQ or anything fancy. It was the extra volume and the wide baffle which made the difference and re the latter, it was out in the room, so using the cabinet you describe against a wall would take care of that.
Regarding the combing issues, I suspect Ted's design takes care of this as the JXR6 was always designed to be used in an array and it's something he's worked on since the mid-70s. The only mild effect I've heard is when you are close to the array (6 feet or so) and you stand so that your ears are well above the array, then the HF tails off.
I wouldn't recommend using a focused array as that may negate the effect Ted was trying to achieve (image stability irrespective of where you stand in relation to the speakers). If a curved array worked, I'm sure he'd be recommending it on his website.
I'll dig out details about the 9-unit array later today and try to post this evening.
Ref the VTL box, this was a prototype, plain sealed box with no EQ or anything fancy. It was the extra volume and the wide baffle which made the difference and re the latter, it was out in the room, so using the cabinet you describe against a wall would take care of that.
Regarding the combing issues, I suspect Ted's design takes care of this as the JXR6 was always designed to be used in an array and it's something he's worked on since the mid-70s. The only mild effect I've heard is when you are close to the array (6 feet or so) and you stand so that your ears are well above the array, then the HF tails off.
I wouldn't recommend using a focused array as that may negate the effect Ted was trying to achieve (image stability irrespective of where you stand in relation to the speakers). If a curved array worked, I'm sure he'd be recommending it on his website.
Agreed to an extent. I dislike curved arrays myself. My concern is that to the best of my knowledge he's not recommending a tall array on his site either. A short array with small full-range units will usually still place the listener in the farfield, so integration should still work, but an extended one shifts this to nearfield, and no matter what you do, the driver centres, which are the bits producing the HF will still be physically too far apart. Checking Jim Griffin's white paper & the other general math, I'd expect attenuation to kick in from ~15Khz, with an increasingly rough response about 1KHz below that.
The image below is a copy of one in Ted's 1963 Loudspeakers book, in which he talks about using multiple, widebandwidth drivers. The aim was to get increased power handling to enable smaller units to be used which would have a better transient response than larger cone drivers. He says:
"the number of units used must have an integral square root. Such numbers are 1, 4, 9, 16, 25, etc. ... the combined impedance is that of a single unit."
The idea of a 9-unit array also cropped up in some of his sales literature in the 90s but there isn't any indication of whether it was necessary to tailor the response in any way. I did hear a 9-unit array when Jordan Watts used his J51 units at a hi-fi show in aorund 1991. I thought it sounded pretty good. Certainly went loud.
"the number of units used must have an integral square root. Such numbers are 1, 4, 9, 16, 25, etc. ... the combined impedance is that of a single unit."
The idea of a 9-unit array also cropped up in some of his sales literature in the 90s but there isn't any indication of whether it was necessary to tailor the response in any way. I did hear a 9-unit array when Jordan Watts used his J51 units at a hi-fi show in aorund 1991. I thought it sounded pretty good. Certainly went loud.
Attachments
Back to cabinets (where I feel more secure), the plan view of Ted's linear array cabinet is something like the drawing below. Its similar to the plan on his website but with a curved side - so no reflecting surface directly behind the drivers. Mounted on the wall, it's very discrete.
Back to curved arrays - if HF combing was a problem, would it be easier to use (say) 12 units, with the central 4 carrying the signal fullrange and the other sets of four rolled off above 500Hz? How do the line arrays using loads of dome tweeters cope?
Incidentally, one of Ted's articles form 1970 or so describes his ultimate system at the time - a 10 foot long, horizontal line of 4 inch drivers.
Back to curved arrays - if HF combing was a problem, would it be easier to use (say) 12 units, with the central 4 carrying the signal fullrange and the other sets of four rolled off above 500Hz? How do the line arrays using loads of dome tweeters cope?
Incidentally, one of Ted's articles form 1970 or so describes his ultimate system at the time - a 10 foot long, horizontal line of 4 inch drivers.
Attachments
Interesting stuff. I hadn't heard of the number of units requiring an integral square root before. Be interesting to study the theory behind it -do you have the articles at all?
Multi way arrays are more common with ribbons than domes, partly because they work out cheaper due to the lengths involved, and also because a lot of dome tweeters have extended surrounds, meaning you can't mount them very close together. You need to keep the driver centres at the most 1/2 a wavelength apart if lobing isn't going to occur, so as frequency increases, the necessary distance between the drivers decreases.
There's an array on the Parts Express site using Tang Band 3in drivers that illustrates this- take a look at the FR graphs over ~6KHz. http://www.partsexpress.com/projectshowcase/Kuze3201/Kuze3201.html
Re using 12 drivers, and then rolling off of the upper and lower ones above ~500Hz, that could be a good move. Power tapering the line should do the trick. Personally, I'd do it like this, borrowed from Jim Griffin's superlative white paper. Imagine the 12 drivers of the line numbered, 1 through 12, 1 being the top driver, and 12 the bottom. Wire drivers 6 & 7 together in series. Then wire 4, 5, 8 & 9 together in series. Finally, wire drivers 1, 2, 3, 10, 11, & 12 together in series. Then connect the three groups in parallel. That'll give a 2, 4, 6 tapered feed. I still don't think it's ideal, but it should be better than a uniform feed, and it's more versatile than a curved array. They're both circumventing the problem really, but it could be worth a try.
Multi way arrays are more common with ribbons than domes, partly because they work out cheaper due to the lengths involved, and also because a lot of dome tweeters have extended surrounds, meaning you can't mount them very close together. You need to keep the driver centres at the most 1/2 a wavelength apart if lobing isn't going to occur, so as frequency increases, the necessary distance between the drivers decreases.
There's an array on the Parts Express site using Tang Band 3in drivers that illustrates this- take a look at the FR graphs over ~6KHz. http://www.partsexpress.com/projectshowcase/Kuze3201/Kuze3201.html
Re using 12 drivers, and then rolling off of the upper and lower ones above ~500Hz, that could be a good move. Power tapering the line should do the trick. Personally, I'd do it like this, borrowed from Jim Griffin's superlative white paper. Imagine the 12 drivers of the line numbered, 1 through 12, 1 being the top driver, and 12 the bottom. Wire drivers 6 & 7 together in series. Then wire 4, 5, 8 & 9 together in series. Finally, wire drivers 1, 2, 3, 10, 11, & 12 together in series. Then connect the three groups in parallel. That'll give a 2, 4, 6 tapered feed. I still don't think it's ideal, but it should be better than a uniform feed, and it's more versatile than a curved array. They're both circumventing the problem really, but it could be worth a try.
Thanks Colin and Scott. Must get a copy of Ted's book. Are the 4, 9, 16 etc arrays desirable then just for the sake of keeping the impedance the same as one driver? As generally tube amps like higher impedances I was thinking of wiring them series/parallel to get 8ohms if there are no other issues involved. I do have a spare pr of JX6 if we go with the 9 driver array.
On the other hand I also read there may be some advantage to wiring them all parallel and using an output transformer with a tap for the <1ohm load. Any comments on that? I can get a transformer wound to do the job.
With all these options it may be best to have each driver with its own set of binding posts so impedance, power tapering etc can be configured externally with buss wires.
The curved-wall cabinet also looks like a good precaution. Will have to look into the feasibility of that one with our cabinet maker.
My own measurement of a single JX6 shows a pronounced rising response which is different from the sweep on Ted's site. However hf measurements are very difficult to perform reliably so I don't have confidence in them...but it would be nice for the purposes of an array.
Brian
On the other hand I also read there may be some advantage to wiring them all parallel and using an output transformer with a tap for the <1ohm load. Any comments on that? I can get a transformer wound to do the job.
With all these options it may be best to have each driver with its own set of binding posts so impedance, power tapering etc can be configured externally with buss wires.
The curved-wall cabinet also looks like a good precaution. Will have to look into the feasibility of that one with our cabinet maker.
My own measurement of a single JX6 shows a pronounced rising response which is different from the sweep on Ted's site. However hf measurements are very difficult to perform reliably so I don't have confidence in them...but it would be nice for the purposes of an array.
Brian
Just checked my copy of Ted's book, re numbers of drivers having intergral square root. This is so the array has the same impedance as a single driver, and the power was shared equally between all drivers. The first point was more important then as amp output transformers needed to be matched to the speaker impedance more closely, but today is less important with SS amps. The second still applies unless some steering effect is required for some reason.
Note this book does not cover his later work on Al foil cones, but is still worth having, being a good easy to read introductory text(as is G A Briggs book of the same name). Having read these, the more modern books with a more mathematical approach are much easier to understand.
It is surprising how much was known then, (found out by experience), and has since been "forgotten". A case in point is the use of resistance in the speaker leads, (Briggs),to compensate for a rising response in some (full range) speakers such as lowther, a point explored much more extensively by Nelson Pass.
Note this book does not cover his later work on Al foil cones, but is still worth having, being a good easy to read introductory text(as is G A Briggs book of the same name). Having read these, the more modern books with a more mathematical approach are much easier to understand.
It is surprising how much was known then, (found out by experience), and has since been "forgotten". A case in point is the use of resistance in the speaker leads, (Briggs),to compensate for a rising response in some (full range) speakers such as lowther, a point explored much more extensively by Nelson Pass.
Hi Colin,
It just twigged; you said 'horizontal'. Was that a slip, or is there a story here?
I ask because one recent visitor to HK tells of the time ( a number of years ago) he went to Ted's home and listened to the Jordan array which was a line of drivers neither horizontal or vertical, but at a roughly 45 degree angle from vertical. Have you heard of that one? The visitor described the sound as VERY good.
I agree with RJB: Hate searching for things that have already been discovered, but we seem to do a lot of that in this business.
Brian
Hi - No, it was a horizontal array, with the left signal connected one end and the right the other, with some kind of network linking the drivers to give (I imagine) the same sort of delay used on the later Quad ESL with its concentric electrostatic panels.
It's in one of his 1973 articles. I think Ted should publish these on his site as they (and his book) are very hard to get hold of. The book, in particular, would be a good candidate for a repirnt. When I read it, I was knocked out by how much it covered and how little seemed to have been developed since.
It's in one of his 1973 articles. I think Ted should publish these on his site as they (and his book) are very hard to get hold of. The book, in particular, would be a good candidate for a repirnt. When I read it, I was knocked out by how much it covered and how little seemed to have been developed since.
Colin,
Would be a fine thing if Ted's writings could be posted and a worthwhile project. If he wanted to reprint we could find a cheap way to do it here, I think. He's one of the full-range pioneers and still with us actively designing drivers so he might even have stuff to add since '73.
If you come across more info on the horizontal array would love to read it.
This new JX6 has more in common mechanically and acoustically with a ribbon that a cone speaker: featherlite diaphragm, smooth hf to 30k, suspended only by the perimeter of the diaphragm. Really it's like a ribbon that goes 100hz - 30k with 3mm of Xmax!
Where we are so far: 8 or 9 driver array in a curved back box, each driver with its own binding posts.
Cheers
Brian
Would be a fine thing if Ted's writings could be posted and a worthwhile project. If he wanted to reprint we could find a cheap way to do it here, I think. He's one of the full-range pioneers and still with us actively designing drivers so he might even have stuff to add since '73.
If you come across more info on the horizontal array would love to read it.
This new JX6 has more in common mechanically and acoustically with a ribbon that a cone speaker: featherlite diaphragm, smooth hf to 30k, suspended only by the perimeter of the diaphragm. Really it's like a ribbon that goes 100hz - 30k with 3mm of Xmax!
Where we are so far: 8 or 9 driver array in a curved back box, each driver with its own binding posts.
Cheers
Brian
horizontal array
A 10' horizontal array with 4" drivers would be about 30 drivers. Do you know what kind of cabinet that was built into?
If the soundstage was coherent this could be the ultimate 'wall-of-sound'. Actually I've heard some highend mono systems that pulsate with such life and have such a big sound that one could forget there is no stereo image. Often wondered how to get that 'drive' quality with 2 channel sound.
Brian
A 10' horizontal array with 4" drivers would be about 30 drivers. Do you know what kind of cabinet that was built into?
If the soundstage was coherent this could be the ultimate 'wall-of-sound'. Actually I've heard some highend mono systems that pulsate with such life and have such a big sound that one could forget there is no stereo image. Often wondered how to get that 'drive' quality with 2 channel sound.
Brian
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.