JXR6 design analysis for near field setup

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
OK, my local distributor (Decibel HiFi) just down the road from me has the JXR6 in stock now, and a significant discount to the JX53 (about USD $220 for a pair) which is low enough for me to experiment with a pair.

I'd like to start a new thread that discusses the spreaker design using these drivers. There is some information in the other threads and the EJ Jordan site but not enough for me to feel comfortable with spending time/money on developing a quality speaker with this driver. I have an existing setup using Tang Band 871 speakers (6 litres, drivers mounted high on a thin front baffle) and subwoofer which I'll refer to in this thread as a reference. I'd appreciate input from those more qualified in speaker design & more familiar with the JXR6 to help me build a speaker as optimised as possible. Hopefully this thread will be useful for others also looking to develop a single driver monitor with the JXR6.

Firstly to describe how the new speaker will be used, to give you some context for the design.

Setup - Near field monitors where I'm sitting about 2 -3 feet from the drivers. The speakers will be separated by 2.5 feet and a LCD monitor inbetween (flat surface) slightly recessed from the front of the monitors. The current setup with the Tang Band has excellent imaging, like listening to headphones - imaging is better than any far field setup I have heard. A wall is about 1/2 foot behind the drivers, and the back of the speaker is about 1 inch away from the wall.

Source - High quality computer audio. Please no comments about computers being no good as a source, this was debated in the other JXR6 thread, assume that the source quality justifies investment in the JXR6. I listen to a variety of music, mostly rock/pop, alternative and chillout. Pink Floyd, Peter Gabriel, Tori Amos, Dire Straights, Dave Matthews Band etc.

Amp - Tripath TA2022 DIY AMP3 (similar to Sonic "t-amp" but better). 20W @ 0.02THD Class D. A well executed Class D amp will express lots of detail in the source (eg. able to articulate each instrument in a complex passage), has superb soundstage and precise bass. This particular amp has a very smooth & clear top end in comparison to other good Class D amps. The JXR6 will be an interesting match for this amp, as the JXR6 detail and low distortion should shine in this setup with decent quality recordings although I suspect it will also show up poor recordings as poor.

Subwoofer - I'll be crossing these drivers at 100Hz and I have a single Peerless 8" XLS downfiring woofer in a sealed 20l box. It is driven by a linkwitz transform to achieve 30Hz and has a second order active low pass filter at 100Hz designed to match the active high pass filter and a system Q of 0.5 (Critically damped for better matching to the monitors). This subwoofer has good transient response and a warm, firm bass. It is located close to the monitors and currently integrates well with the TB 871's I'm using, adding a lot more body & authority to music.


Questions about the driver to help clarify the design.

There is some discrepancy with the data posted in this thread
http://www.diyhifisupply.com/docs/JX6HD Prov Para 06.pdf and what is on the EJ JOrdan site. Does anyone know what T&S data is correct? Is Brian Cherry's data a measurement from a production unit? That is also a pretty impressive waterfall chart!

Frequency response curve.
The graph at the EJ Jordan site is pretty impressive, this is one of the flattest curves I've seen for a full range driver, especially considering its a 2" driver. However its not perfect and has 3 areas for concern:
1) Forget anything below 100Hz. OK for me as I'll be rolling off the speaker with a second order active HP filter at 100Hz. I do this on my TB monitors and it does make a difference to its power handling ability and improves higher frequency distortion as the cone is not modulated by strong low frequency energy while trying to reproduce high frequencies. As the JXR6 is a light metal cone I'd assume this HP filtering is even more important than a "sluggish" paper cone like my current 871's.
2) Frequency response above 12K starts rising by a few db. I'd say this is a good thing as it will add more 'air' to music. Most music has little energy past 12K as well as my hearing starts to deteriorate around 15/16K a small amount of boost wont matter here. So nothing needs to be done.
3) THere is a 3db boost from about 2.5Khz to about 5K. Now this is an important area of the frequency spectrum especially for voice. Looking at Brian's plots, this hump is less pronounced, especially if slightly off axis.


4 ohms. Should be no problems as my Tripath TA2021 amp only delivers 10w @ 0.01 THD 8 ohms and 20W @ 0.02 THD with 4 ohms. The extra 4 ohms headroom is nice although not really needed, and the increase in THD wont matter as distortion is already very low and the driver distortion is much higher, even a high quality one like the JXR6.

Sensitivity at 89db/m is pretty good, 2 db higher than my 871, so coupled with the higher output of the amp with 4 ohms I wont have to drive amp as hard with the JXR6.

Rated power. 100W "music" power. Huh? Music power means nothing? I think I'd be OK with my 20w amp, with the 100Hz high pass filter, near field listening and reasonably high sensitivity. Xmax at 6mm p-p is pretty amazing for a small driver which should give it some additional resilience (the 871 as a point of comparison is 0.5mm Xmax).

Baffle step compensation & location of the driver. I decided not to do baffle step on the current Tang Band speakers for the following reasons:
- Speakers are against a wall (adding a bass boost)
- The front baffle has rounded edges, with the driver almost on the edge in the horizontal plane, and very close to the top edge.
- high pass at 100Hz and baffle step really only a problem from 400Hz & below
THe same theory applies for the JXR6, in fact looking at Brian's frequency plots there is a pronounced mid bass hump from 150Hz to 700Hz, so not having baffle step compensation will remove this hump (assuming Brian's plots were not measured with a driver in a box?). So sounds like baffle step compensation is not needed. Although I see most Jordan designs with a really wide front baffle, I'll stick with a thin front baffle (as I use now with the 871's) with the driver placed near the top, it looks good, reduces baffle step and has the drivers at ear height.


Toe in & off axis response. The driver will be located at ear height, about 2 feet from my ears. Jordan drivers are suppost to be driven with lots of toe in so that you get the on axis response for the drivers, crossing just in front of you. With my current tang bands I do get excellent imaging with the drivers pointing stright ahead, but I do notice better high frequency response if the drivers are pointing at my ears. For the new JXR6 setup, I'll slope the front baffle so that the drivers cross just behind my head for a couple of reasons:
- Brian Cherry's frequency response shows rising high frequency on axis and falling high frequency @ 30 deg. So slightly off axis should make it about right.
- Initial comments about this driver is that they are bright at the top end
- Listening near field should not make the toe in and axis response as critical as far field.
- beaming the driver near field from 2 feet away directly into my ears will accentuate the top end.
- Looks better.

Compensation circuit. I'd like to get away from any passive components in the chain as possible. As I'll be driving the speakers slightly off axis the mid range hump & rising top end should equalise, avoiding the need for a compensation circuit. However I appreciate that these things are best measured in the real world and not from theory, but no circuit sounds like a good place to start and simple.

Box size. Not sure on this, Win ISD shows optimal about 3l with -3db at 100Hz, but I dont know about this driver without a rear suspension, and how to keep Q down below 0.6.

I'm interested in your feedback, tips & critique of my analysis above.

Anything else I need to think about?

Regards,
Dean
 
Phew, that's a pretty exhaustive analysis of your requirements. And now I see what you mean about high quality computer audio ...

Spec sheets for the JXR6 seem to be variable at the moment - I'd stick with the ones on Ted's site. He developed the drivers and he has his own ways of measuring things (he said once that he derived some of the measurements that others later put their names to). His measurements will be the most rigorous as he's not a fan of rose-tinted specs. It's also possible the manufacture is still settling in.

On that basis, to reach 100Hz at 0.6 Qts, go for the largest recommended enclosure of around 4 litres.

On axis sound - although I haven't experimented with my JXR6s yet, my experience with the JX53s is it has much less variation on and off-axis than the JX92. The 92 is very sharp on axis. The 53 does require slight compensation to balance out the sound, it's part of the 500Hz crossover circuit and may be there to deal with that 2.5-5k rise. Until I have my own units running, I don't know if the JXR6 needs it.

Power handling - the 100 watts is probably in-system use, crossed over to a suitable bass. FWIW, Ted told me he once fed a JX53 with a sine wave at deafening levels and left it like that for an hour or two to see if he could destroy the driver. When he came back, the cone and chassis were too hot to touch but it was still happily churning out the sine wave.

Baffle - the more I play with Ted's drivers, the more I come round to his idea of mounting them on a wide baffle. It does seem to do something magical to the lower mid-range. But if you are using yours close to a wall, it won't be so much of an issue.

Okay, I'm not sure how much this helps you. Not many of us have experience with the new drivers. Others will be more helpful on the technical aspects of boxes and crossovers, my own experience relates to using listening to Jordan drivers for the last 8 years and chatting to him when I've seen him.

I see the glue has finally set on my new boxes so I'll have a pair of JXR6s up and running this weekend and will report on how they sound.
 
Hi guys

I will follow this thread with big intrest because my pair arrived this week. :)

Regards
 

Attachments

  • jxr6hdm.jpg
    jxr6hdm.jpg
    78.7 KB · Views: 1,305
My initial reaction is, if the Tang Bands work, the Jordans will work better.

Here are my votes:

~ 3 liter enclosure, i.e. system Q of 0.7, if you can spare the room.
Wide-and-shallow approach, like shown on Jordan web site.
Consider the front baffles of speakers as extensions ("wings") of your computer monitor.
One comment I've heard about Jordans is that they improve with break-in. You might consider breaking them in at higher-than-normal levels.
 
Thanks for the feedback Colin. Do you think the later JXR6 manufacturing batches may be more optimal as the production process is tweaked based on the testing of the initial batches?

It looks like there is nothing fundamentally wrong with my analysis.

The compensation issue (if or what is required) is best sorted with real world measurements in a box, although the paper analysis looks promising. It is also easy to add later.

I have a large amount of blackwood veneer MDF waiting to be cut :D but as I'm not as familiar with these drivers (or speaker design) I'll hold out until the first DIY reports come in.

Regards,
Dean
 
Hi Dumbass.

Hmm. Good idea on the wide and shallow front baffle, that would look really good on either side of the monitor.

What about baffle step with such a wide baffle? My understanding of speaker design is that the less baffle on the front the better. For example, Dunlavy speakers use felt around the driver edges to absorb edge diffraction.

Regards,
Dean
 
deandob said:
My understanding of speaker design is that the less baffle on the front the better.
Untrue. Read the Jordan FAQ.

A lot of folk are coming around to wide baffle idea.

Larger baffle means baffle-step cutoff freq is lower, i.e. a good thing. Freqs above BSC freq radiate directly into half-space. Freqs below BSC need some sort of boost because they radiate into 360deg, not 180deg.

Jordan himself recommends wall-mounting. What I'm saying is that your video monitor plus two speaker baffles can act as pseudowall.
 
Hi Dumbass (love the name BTW!)

The computer monitor between the speakers and a wide baffle would make it equivalent to having the drivers flush mounted in the wall. Is this a good thing for Jordan speakers (as opposed to near a wall)?

Ref the imaging, Ted's drivers are toed in about 60 degrees and cross in front of you. I've found anything large and solid in the way of my set up in the lounge causes the imaging to deteriorate, so I would think the same would happen in the nearfield. Having the speakers slightly forward of the monitor may work or perpahs a LCD monitor might not create such a strong effect. Otherwise, it's down to finding a way to mask the monitor in some way, soften the edges and make them sound absorbing.
Colin posted this in one of the other threads, I take this to mean that if the monitor was midway between the speaker and the listener in the soundfield it would be a problem? This contradicts what you are saying (and written up on the jordan web site) as the monitor in this situation would be flush or behind the line of the speakers.

But I'm keen to give it a try with a wide baffle - Ted Jordan certainly recommends this and he should know.

Regards,
Dean
 
deanbob

I have to agree with you on your observation of the waterfall
chart produced by Brian DiyHifi.Those graph are truly a thing of
a beauty IF it holds true.
Although i've heard others brag about the high distortion at
200Hz & below (1.5%), this driver is still remarkable.

I hope those with JXR6 able to come up with some of their own measurement especially the waterfall plot in this thread for us..
 
deandob said:
Hi Dumbass (love the name BTW!)
It's an homage to Red Forman.
deandob said:
The computer monitor between the speakers and a wide baffle would make it equivalent to having the drivers flush mounted in the wall. Is this a good thing for Jordan speakers (as opposed to near a wall)?

Colin posted this in one of the other threads, I take this to mean that if the monitor was midway between the speaker and the listener in the soundfield it would be a problem? This contradicts what you are saying (and written up on the jordan web site) as the monitor in this situation would be flush or behind the line of the speakers.

But I'm keen to give it a try with a wide baffle - Ted Jordan certainly recommends this and he should know.

Regards,
Dean
IIRC Colin has his speakers placed directly against the rear wall, so I assume he means something like a couch, armoire, TV, etc, between the speakers.

Here's what Ted Jordan says about wall-mounting speakers:

[D]espite convention, there is strong argument in favour of placing the loudspeakers as close to the wall as possible. This reduces the time lag between the direct sound and the reflection from the 'virtual images' and substantially improves coherence and spatiality. (The ideal, but usually impractical, solution, is to mount the drivers in the wall where there would be no virtual image).
http://www.ejjordan.co.uk/basics.html
http://www.ejjordan.co.uk/faq.html

A lot of food for thought on those two pages.

It just occurred to me, if you do choose to go wide baffle, you might want to place the drivers off-center (mirrored pair), because otherwise you might be forced to have the drivers too far from each other, given the nearfield setup and monitor in the middle. You also might want to have the front baffle leaning back at the same angle as your video monitor, just for kicks.
 
This might explain why I get excellent imaging and soundstage from my current setup with TangBands, the drivers are flush on the baffle and also flush with the monitor, which acts like the drivers are flush mounted in the wall.

Good point about being off-centre, if you have the drivers in the centre of a too wide baffle for nearfield listening the toe in would need to be more accentuated to still be listening on axis.

Another speaker "wisdom" is to have the drivers at different distances to the edge of the baffle. Is this still relevant with Jordan drivers?

I wonder if anyone has tried the opposite to the Jordan wide baffle, stop any side refraction by placing felt or absorbing material completely around the driver, absorbing everything except the directly radiating outward wave. I'll try this on my Tang Bands to see what difference this makes.

Regards,
Dean
 
deandob said:
Another speaker "wisdom" is to have the drivers at different distances to the edge of the baffle. Is this still relevant with Jordan drivers?
The one reference book I own is Weems/Koonce Great Sound Stereo Speaker Manual. They recommend off-center mounting, all of their projects use it. I'm not convinced of the importance, unless in some extreme case of a circular baffle or what. In a normal rectangular situation, the distance from driver to edge of baffle is not uniform over the 360 degrees, there is no sharp baffle step dropoff, it occurs over a range of frequencies.
deandob said:
I wonder if anyone has tried the opposite to the Jordan wide baffle, stop any side refraction by placing felt or absorbing material completely around the driver, absorbing everything except the directly radiating outward wave. I'll try this on my Tang Bands to see what difference this makes.
Weems/Koonce recommend fiberglass on front baffle. This is the whole idea behind narrow baffles, too. Some speaker manufacturers go to extremes, with no real front baffle at all. But you still can't get around the diameter of the driver, eh.

In theory, there is no reflection of directly-radiated sound off the front baffle. Any sound waves that would be "reflected" by the baffle are propagating parallel to that surface.

One place where I can see this stuff being important would be in using a wide-dispersion tweeter (like dome). I know that there are real issues in, for example, whether you flush mount or surface mount, because of the very short wavelengths. So I can imagine surrounding such a tweeter with some absorptive material, and indeed some manufacturers do this.

With cone driver, however, this fails to be important because of beaming. That is, any very high frequencies propagating to the side are already squashed by destructive interference. AFAIK this is true with a driver even as small as 2". (You can calc the freq at which beaming effects come into play, I think compare a wavelength to width of cone but don't take my word for it.)

I know there are complicating factors because of cone flexing (in fact Jordan drivers are designed to optimize cone flex).

Flush mounting simply looks better, in any event.
 
Dumbass,

For a test, I tried wrapping damp towels around the circumference of the tang band drivers (3"), which would stop any sideways reflection.

For this near field setup I have a tall & skinny front baffle, toed in so that the cones are pointing about 6 inches behind my head, and inbetween the speakers is a flat LCD monitor slightly recessed from the line of the drivers, with about a 2 inch gap between each speaker and the monitor. Normally this setup with a class D amp has a very wide soundstage and sounds similar to wearing headphones due to the closeness of the drivers to my ears.

It did not seem to make much of a difference, if anything the sound became a little more focused but the soundstage did not sound as wide or as encompassing. Maybe the effect of side propogation/reflection is less of an issue when listening near field.

I assume the jordans will react a little differently to the tang bands due to their design.

Regards,
Dean
 
Off-centre mounting on the baffle will also mean (I presume) that the drivers will be off-centre with regard to the internal shape of the box, which will help spread the resonances a little.

In theory, the Jordan drivers should be less prone to edge defraction as they are designed to beam at higher frequencies. However Ted once told me that he put felt around the drivers of a commercial design of his and it cleaned things up down into the mid-range.

When I mentioned anything between the speakers messing up the image, yes I meant something like a table or surface with hard edges. It sounds like Deanbob's monitor makes a continuous curve from speaker to speaker, which should be ok. Only one way to find out ...

Enough of theory - I've now got my 2.4 litre boxes up and running. The JXR6 is a definite step forward from the JX53 in terms of fullrange use. It has a lot more body in the lower-mid and upper bass. According to my diy/noddycalc spreadsheet, the box should be tuned to a Qts of 0.74 with -3dB at around 128Hz. Sounds about right. There's good output to 100Hz, dropping to 75Hz and not much at all below that. (These are just computer test tones and my ears, so not a scientific measurement.) They sound a little lightweight on first hearing but I'm used to a system which goes to below 40Hz. Once I got used to that, they sound excellent and I'd be happy to use them on their own in smaller rooms. I'm using them near a side wall, six feet or so from the back wall and corners. I haven't experimented with positioning nor have I fiddled with the box stuffing - I just shoved in some BAF wadding, uncompressed, and screwed in the drivers. And I will try the comp components from the JX53 to see what that does to the sound.

They're more sensitive than the JX53, on a par with the JX92, which suggests some intriguing possibilities. The next stage is to wire them to my MLTLs with the 92s rolled off around 100Hz and see what happens.

Ref the 92 sound, the 6s are cleaner in the HF, more natural sounding and civilised. Those looking for such a thing may have their 92 'tweeter' at last (one which runs to 100Hz...).

At the moment, they don't image quite as well as the JX92 or the linear array when you are off-centre (the JX92's rising response is supposed to mimic the linear array effect) BUT I have them in ceramic enclosures with the drivers recessed slightly ref the edges, the intention being to fit felt around them, which will be hidden by a grille. (I don't particularly like seeing drivers and I'm even less keen on damaging these very lightweight foil cones.) I'll fit the felt later today and see what that does to the imaging.

That said - they do image very well when on centre line and I tried moving up between them, similar to how Deanbob will be using them, and, like he said, it was like wearing a pair of high quality headphones. Moving even farther, so the speaker line was just behind the listening position, the image was still locked in front, which felt very strange.

I like these little beasties.
 
Hi Deanbob

Now I've spent 48 hours listening to these, they've either broken in or I've grown used to them but I don't think they need any comp at all.

I've also matched them up to my JX92 MLTLs, letting the JXR6s roll off naturally and rolling off the 92s with a 9mH inductor (at around 112Hz according to the Jordan site). The drivers are about 16 inches apart (I simply put the ceramic boxes on top of the MLTLs). These work rather well - if anything the JXR6 might be slightly more sensitive than the 92 so if you can match the bass with an active crossover, you're onto a winner. The sound is very clear, the imaging comes from the JXR6s and the 92s add a welcome touch of bass (although on a singer/acoustic band piece, my partner didn't spot when I disconnected the bass enclosures).

I suspect mine do need felt around the drivers - both driver and cabinet edge are proud of the baffle (the cabinet forms a frame round the recessed baffle) and simply placing a piece of card anywhere near the driver is noticable. However if you recess the driver flush and round the edges, it may not be a problem, especially on wider baffles (my enclosure is 6 inches across). I'll experiment when I find some felt or similar material thick enough to do the job. Everyone round here seems to stock thin, dress-making felt.

Anyway, overall the new Jordans seem an excellent way to get high quality sound out of tiny enclosures - and if the bass enclosures can be hidden away it opens up all sorts of possibilties.

I'll post pics in a day or so.

Colin
 
Colin, my current 3" setup has the drivers close to the edge, flush with the baffle and have the edges sloped at 45 degrees. I can only notice a small change in imaging / soundstage with sound absorbing material placed around the driver.

For felt or similar sound absorbing material, try an auto sound deadening blanket used to line the insides of car doors.

Regards,
Dean
 
Nah, find a quality car audio suppliers, they are run by nutcases, they will think its perfectly reasonable. someone like James at Audioscape (Ipswich) would possibly be a good start.

I'm following this as I'm doing something similar with Bandor 50s. Do you flush mount the driver then cover the face with felt, or could I sink the driver into the baffle say 5mm and cut the felt (5mm thick) to be flush with the baffle at the edge. would I still get edge reflections from the wave moving through the felt?

Piano tuner/repairer may be worth a try too, they use a lot of different grades of felt for piano repairs.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.