New Jordan JX92S mini-monitor design

[AntM]

Hi,
I thought I would share with you a new loudspeaker design I am building, based on the Jordan JX92S. The design evolved from a desire to build a speaker with a high Wife Acceptance Factor .

I built my first pair of diy loudpeakers around 20 years ago as a school woodwork project, based around a Maplin fibre-glass drive unit and an Audax polycarbonate tweeter. Over the years, I upgraded the drive units to Focal, then Morel, then eventually replaced both driver units with the Seas L18, inspired by Zaph's design Zaph|Audio

I was never completely happy with the sound, which was harsh, particularly with modern over-compressed music, possibly due to inadequately attenuated breakup nodes of the L18 driver and 27TBFCG tweeter (no disrepect to the Zaph design which I cannot claim to have followed too closely)

Anyway, when my wife offered to trade the complete removal of all children's toys from the lounge against the removal of the old boxes, a deal was struck...

The biggest constraints of the new design are size and placement. The boxes must fit on or above an existing fireplace, which is pretty shallow (around 19cm). The listening position is relatively far away from the speakers (relative to the distance between the speakers) which means that for stable stereo imaging I expect to have to toe-in the speakers more than usual. The speakers will also be quite a bit higher than the listening position. All in all, a tricky set of constraints.

[AntM]

Based on the constraints and the desire to keep the boxes as small as possible the internal volume is likely to be between 3 and 5 litres.


After a long scout around I've opted for the Jordan JX92S.
Why? Several reasons:
- I've always been intrigued by the possibility of using a single driver, eliminating the crossover. Ted Jordan waxes lyrical about the advantages of this E.J.Jordan Designs, loudspeaker consultants, designers and manufacturers.
- This driver is designed to be toed in - it has a rising high frequency response complemented by a falling off axis response to give a roughly constant power response. Using the driver slighly off axis should yield a flat response at the listening position. This meets with the requirements of toeing in the design to get a stable stereo image.
- The driver has a better high frequency response, a lower fs and better power handling than most of the fostex full-range drivers I looked at
- There are oodles of diy and commercial designs using this driver. In places it is raved about and can be seen in pretty expensive systems.
- The driver has good distortion (measured here Zaph|Audio)
- It has a relatively flat response. I should be able to get down to 100Hz with a sealed box design. With such a small enclosure there is little advantage is going to reflex (according to the frequency response modelled from the Thiele-Small parameters) and colouration of lower mid may be an issue with a reflex design.
- Although it has an metal cone, it is designed to flex, thus reducing break-up nodes. This satisfies my concern about having to design matched notch filters and living with the associated phase issues
- Jordan has some very interesting example designs on the web site. He suggests mounting the loudspeakers close to the back wall (citing reduced time difference betwwen the primary and reflected signals). As an engineer, I'm surprised that this is an advantage (it just means that the resulting comb filtering will probably affect the midband), but my enthusiasm as a diyer sometimes overrides reason.
- I like the simplicity of the example designs e.g. his 18cmx18cmx18cm monitor. Again, this seems like a crazy design (standing waves emphasised by the cube, diffraction problems a-plenty) but again I see no reason as to why Ted Jordan would promote a bad design if he is trying to sell his loudspeakers
- I'm hoping this will put an end to the constant tinkering I did with my old system (there are just too many degrees of freedom with an 8-element crossover. With this design I only have two components (baffle-step compensator) to deal with. Not a real reason, but there you go.

more to follow

[AntM]

On to the box design. The first choice to make is that of cabinet material. My initial thoughts were to use mdf for the back and sides, with a real oak front-baffle. I like the idea of being able to round-over the edges to spread the diffraction frequencies, which would require more tricky veneering if MDF was used. Eventially I was put off over concerns for diffferent rates of expansion and contraction between the MDF and wood (with humidity and temperature( causing splitting of the wood.

So, bring old MDF it is! Fortunately B&Q do a pretty good wood cutting service which makes the cabinet design that much easier.
After umming and arring over the MDF thickness, I've chosen 18mm. This is a very small cabinet, so hopefully the panel vibrations should be minimal.
For construction, I'm using glue-blocks on all the internal joints, with screws screwed through the blocks to ensure good stability and pressure during glueing. This has the advantage of not needing clamps and also offering more stability at the edges of the cabinet. Also, a few experiments with screwing directly into the MDF had poor success, with some of the MDF panels splitting if I screwed in too hard.

Pictures to follow, once I can get my laptop to read the memory card...

[Solve]

And the price is....mucho.

[richie00boy]

Look forward to seeing the results.

If you have not yet started building, try a #8x40mm chipboard screw with a 3mm pilot hole (4mm clearance hole in the through panel). I never have problems with MDF splitting this way.

If you are placing against a wall you can more than likely forget about baffle step.

[Jim Griffin]

AntM,

If you wish to cut to the chase, I started several threads on DIYAudio that cover my various JX92S designs over the past 8 years. These include a full range mini-monitor in a 7 liters enclosure, the mini-monitor with the addition of a ribbon tweeter, and a MLTL with the ribbon tweeter. You can search for threads and posts by me or start at these points:

Jim Griffin's JX92S designs

http://www.patcave.com/diy2002/s11b6945.jpg

Jordan with a Ribbon MLTL

JX92SG2siDesignPak - eSnips, share anything

Many multiple page threads and such to address whatever you need to know.

Jim Griffin

[AntM]

Solve,
Oh, yes the price is high, around £200+vat for both drivers, but I'm hoping you get what you pay for. If I get a cheap driver, I'll probably regret it in a few months and always wonder what I could have got for another £100. Not a huge amount, when spread over the lifetime of the speaker.

Jim,
Thanks for the links. I had noticed your designs before - I like the look of this one... SingleDriverSpeakers Forum - Re: Ping: Jim Griffin - DIY Atlanta speaker - Jim Griffin, November 27, 2004 at 21:14:31

The cabinets are around half completed but there are still a couple of design options I am considering:

- whether to add a 6mm birch-plywood layer on top of the MDF front-panel. If the recess depth of the driver is 5mm, using the MDF-only approach reduces the depth to only 13mm where the driver is mounted. Adding the layer should improve the stiffness of the baffle, but will it make a difference? The panel is 27cm-by-20cm. The cost? around £20 for an 8ftx4ft sheet, plus a slight complication in the design in having to glue on the sheet after I have screwed on the front panel. This will mean I will have to route the baffle after mounting it on the cabinet, which means I run the risk of damaging the cabinet if something goes wrong.

- whether to provide the option of porting. According to the box simulations, adding a port will offer no LF benefit, and have the disadvantage of consuming valuable internal volume. Any comments of whether there are any "second-order" benefits of porting the box.

[planet10]

Originally Posted by AntM

- whether to add a 6mm birch-plywood layer on top of the MDF front-panel. If the recess depth of the driver is 5mm, using the MDF-only approach reduces the depth to only 13mm where the driver is mounted. Adding the layer should improve the stiffness of the baffle, but will it make a difference?

If you are going to get ply to laminate to the baffle, why not just build the whole thing out of birch ply... you'll end up with a better cabinet. Given the size of the cabinet you propose you could probably get enuff pieces out of your 4x8 to do a triple laminate of 6mm.

An example of a sealed 5.5 litre cabinet is here ... we use 15mm, and with the strategically placed bracing have no issues with cabinet resonance. It does require using good multi-ply thou... MDF is just not stiff enuff or well enuff damped. With the driver brace, the 15mm baffle (even with a rebate) works fine.

With you (laudable) philosophy wrt driver price it seems strange that you'd cheap out and use MDF for the box.

dave

[Colin]

Sometimes MDF is all you can easily get in the UK. IOne supplier locally (part of nationwide builder's chain) has stopped because of concerns about tracking it to sustainable sources.

The sheet of ply I have I'm saving until THE project comes along!

btw, may be a concern if you have too deep a front panel. I would have thought the 13mm perfectly ok. Go for the largest volume you can for the cabinet. 5 - 8 litres is ideal for a sealed box for the JX92S. 5 litres may be too small for porting. Do you plan to use with a sub at all? A small one can be hidden quite well and a sealed cabinet may make a better match.

As the JX92 is highly directional, rounding the corners may be less important. I tried putting felt round mine and it didn't seem to make any difference.

On a side issue, whereas the old Beeb designs used thin wall ply damped to drop resonance below the mid-range, I think Harbeth are now using thin wall MDF. Perhaps MDF is actually better suited to this one particular technique as it is inherently less stiff? Anyone know any more on this?

[acarinanders]

Hi Colin. I have some experience with building speakers of BB and a combination of mdf+soft fiber board. (it is normally used for notice boards)
My A126 horns are made of BB. They sound great, but I have had some problems with a resonance which I think are related to the material.

30 years ago I built a mini horn for a Philips SQ5061 completely built of soft fiber board and then covered with a layer of mdf. No resonances at all and no midrange came out of the mouth.

I have also built a variovent based Dynaudio two way system with the same technique. Absolutely stone-dead cabinet.
There are others that are using soft fiber board for their horn, like HM.
The soft fiber damp frequencies below 200hz.

Wooden Hard Boards, Medium Density Fiberboard, Moisture Resistant Fiberboard, Fire Retardant Fiberboard, Exterior Medium Density Fiberboard

/Anders