B&W FST mid

[JonMarsh]

Looking very cool, Matt! :T So, you'll only be doing an active crossover with these?

[wkhanna]

Great looking build, Matt.

It looks as though your using pine.
If so, what type?

How did you manage to clamp the braces & keep them parallel?

[5th element]

Looking very cool, Matt! :T So, you'll only be doing an active crossover with these?

I've been only active for years, even my new projects always seem to end up using a DSP somewhere in the chain to handle the xover work. These will be for my main system that uses a DSP + ES9018 8 channel DAC.

[5th element]

Great looking build, Matt.

It looks as though your using pine.
If so, what type?

How did you manage to clamp the braces & keep them parallel?

Yeah pine for the cross braces, I have no idea what type actually, just the cheapo stuff sold at the local DIY store for making shelves out of Making the braces from a template with the router is a pain, as is chopping out the central parts of the braces and pine is a lot easier/quicker to machine than MDF.

To secure the MDF to the braces I used screws. I had tried clamps and/or cloth straps etc in the past but without great success. Screws, as simple as they are, work really well, even if it does mean I have to drill lots of pilot holes and screw in lots of screws. This is the main reason for choosing the pine too, MDF is horrible at accepting screws into its side faces, it splits like mad and even if it doesn't split the screws don't hold particularly well. This is not so with the pine, which is as good a reason as any to use it.

[---k---]

Very nice work.
I'm surprised by the pine. You logic makes great sense for construct-ability.
A nail gun would probably go even faster than the screwing.

[5th element]

A nail gun...now that's an idea... most of the time spent with this type of construction is drilling pilot holes and screwing in and then unscrewing loads of screws.

I've got two sides glued on one cabinet and almost the same completed with the other one. I'll have more pics tomorrow I think when I get that side fully glued on and possibly some spacers for the second skin. I need more bendy MDF for that second skin though.

[kevinm]

Excellent work! I want to build a curved cabinet next, I think. They look so cool.

I saw that PassingInterest uses the Curvomatic system - looks very handy

[Steve Manning]

Excellent work! I want to build a curved cabinet next, I think. They look so cool.

I saw that PassingInterest uses the Curvomatic system - looks very handy

I was just going over his build thread yesterday ...... I want one.

[5th element]

I think we all want one of those!

[Steve Manning]

One of these days I'm going to try another curved cabinet, more along the lines of what your doing rather than the translam I did last time.

[JonMarsh]

Man, if I was into curved cabinets on top of all the other things, I think life would just be too complicated for me... :W

Looks very nice, Matt, very interested to see how this turns out. When I was involved as a B&W dealer in the mid 70's, I had access to their drivers and built a completely DIY system using the parts from their big floor stander of the time, on of the first to use a kevlar midrange.

[5th element]

Ah those were the days! (I wasn't even born yet). I bet by today's standards, from your point of view, you probably think us guys have it easy :roll:

Little update.



Both sides are now glued and set onto one cabinet and the other has its final side setting. The glued cabinet has been rough sanded and now has the first of the second skin spaces gluing in place. As cabinet construction goes nothing has been particularly complicated about these. The bendy MDF comes in 2 foot widths, which is handily the height I want the cabinet at, so that cuts down on some of the hard work.

I am a sucker for curved cabinets, although your faceted designs handily give the impression of narrowing the overall profile of the loudspeaker too.

[Steve Manning]

Both sides are now glued and set onto one cabinet and the other has its final side setting. The glued cabinet has been rough sanded and now has the first of the second skin spaces gluing in place. As cabinet construction goes nothing has been particularly complicated about these. The bendy MDF comes in 2 foot widths, which is handily the height I want the cabinet at, so that cuts down on some of the hard work.

I am a sucker for curved cabinets, although your faceted designs handily give the impression of narrowing the overall profile of the loudspeaker too.

Now that's cheating ..... getting the pre-made stuff, here I thought you were running the mdf through the table saw a few zillion times to get that done. I like your way better!

[5th element]

Okay well my JA 8008 HMQs arrived. Wow. These things are built like a tank, typical SEAS build quality, they are clearly 'excel' based drivers from the copper ring you can see installed beneath the phase plug. They weigh a TON, the magnet on these things is huge compared to what I am used to in 8" drivers. I was surprised at how heavy they actually are. Visually the HMQ version looks very nice with its silver anodised and lacquered phase plug.

[Jonasz]

Yes it's a very appealing driver and one I will probably use one day. Btw using the Jantzen will almost make the FST unnecessary in this speaker...:P

[5th element]

Yes it's a very appealing driver and one I will probably use one day. Btw using the Jantzen will almost make the FST unnecessary in this speaker...:P

I did think that myself, however the JA8008 does have cone edge resonances that I'd rather do without, plus you cannot convert it into a coax easily, at least not without removing the copper sleeve that sits above the pole piece and the screw from behind that holds in the phase plug. The FST is ideal for coax conversion because of the flat surround. Of course the FST is slightly more linear than the JA throughout the midrange, especially the upper midrange. This is after all my cost no object type design.

I may possibly make a 2 way with the JA8008 as it would be interesting to see how it performs as such, of course I'd go with a wave guide like Troels so there wouldn't be too much of a difference between them. The JA has such very low distortion up to around 300Hz it makes it perfect, especially when coupled with its high sensitivity, for three way, floor standing designs as it gives around 50Hz extension in a 25 litre ported cabinet. If you floor mount it you have the possibility of coming up with a true 95dB sensitivity speaker, or go with the usual mounting and have a true 91dB sensitivity with 5dB BSC built in. Either way it's a very nice driver.

The cabinets are coming along nicely, I'll have more pics to post tomorrow.

[5th element]

Slightly delayed on the update of these photos, they were taken yesterday but sometimes stuff gets in the way.

First cabinet with the spacers glued in and the second skin also in place. The bottom of the cabinets has also been installed.



Here's a closer shot of the bottom. Hopefully the seal between the MDF base and the sides will keep in the filler when it's poured in.



Front side shot showing the spacers glued in place. These are simply 12x15mm batons of pine.



Here's a rear side shot of the spacers. Here you can see them having been planed roughly flat along the back.



A more side on shot of the rear showing the planing.



For the filler I've decided on using some small stones and epoxy resin. This is rather expensive, or at least the epoxy is in decent quantity. I did think about using something like cement, but considering the somewhat fragile nature of the MDF skins I wanted something that would actually bind everything together and I didn't think that cement would stick to the MDF particularly well. The epoxy is very low viscosity too so should fill in the groves of the outer skin pretty well.

[---k---]

Wow. That should be very stout when done.
Got a link to share for the epoxy. I'm curious what you're using.
I think you're right to skip concrete. I would fear that the MDF would suck the water out of the concrete and become mush.

While you're going overboard, maybe you would want to consider using wood hardener for the outer skin.

http://www.minwax.com/wood-products/maintenance-repair/minwax-high-performance-wood-hardener

I played with it a few years back. It is an acrylic resin suspended in acetone. I got about 1/8" penetration into the MDF. It made the outer edge rock hard. I'm sure there are benefits, but I've been using mostly pry for my projects.

[5th element]

Hmm a wood hardener, I didn't know those existed and yes that is the idea, they should be quite dead with all the stone and epoxy in there.

The first lot of stones and epoxy have gone in. The first lot is only a really small amount to make sure that the bottom of the enclosure is sealed. The last thing you want is all your expensive epoxy draining out of some small holes.

http://www.easycomposites.co.uk/products/epoxy-resin/EL2-epoxy-laminating-resin.aspx

Is the epoxy I am using, it has a very low viscosity which is great for filling in all the gaps and ensuring that it can drain through all of the stones properly.

[5th element]

So then this happened



I figure that the rest of the construction of these is fairly self explanatory, fill the sides with stones, pour in epoxy, glue on the top, then glue on the front and back.

The port is inserted into the base, hence the chocks of wood elevating them off the ground by a couple of inches. The port itself is 80mm in diameter and flared at both ends.

http://www.falconacoustics.co.uk/cabinet-parts-accessories/bass-reflex-port-tubes-loudspeaker-grille-mounts-dowels-calculation-sheet/bass-reflex-port-tube-reflex-80mm.html

What surprised me about these was the tuning. LspCAD predicted that the port would need to be around 20cm in length, where in reality it only needed to be around 5cm. Now I know that when one builds an MLTL a similar thing happens, but I thought that this effect only occurred when the driver was mounted at one end of the cabinet and the port at the other, clearly not entirely.

Before I mounted the drivers within the cabinets I weighed them and they came in at 15.5kg, so they are pretty solid, with the JA8008 and B&W module on top they aren't light loudspeakers, but they are manageable. This was important, I didnt' want a loudspeaker that required two people to move them and as the B&W sphere part is separate that makes them all the more manageable.

Measurements of the JA8008 HMQ will follow later. The real question now is, how on earth do I finish them?

[---k---]

I've read a couple posts by Jeff Bagby and others that discuss modeling a MLTL for what looks like a typical ported enclosure. It MLTL modeling on my list of to-do's ... some day.

[augerpro]

Do you think the port being that close to the ground is causing a higher loading, and so creating an effectively longer port? MLTL also takes into account port and driver placement versus the line, usually staying away from placement at 1/2 or 1/4 wavelength of the line. Obviously you don't have that, but a port on the end of the line strikes my monkey brain intuition as possibly an issue too?

So have you settled on a tweeter for this? I wonder if an XT19 or SB Acoustics new 19mm would fit.

[5th element]

I actually sized the port with the loudspeaker on its side although I did wonder about how high up I'd need to elevate the speaker to ensure that the port would function correctly.

RE the tweeter it's probably worth linking this thread.

http://www.htguide.com/forum/showthread.php?42114-The-Odd-Couple&highlight=odd+couple

The Dayton ND16 isn't completely set in stone, but mounting the tweeter like this does make finding suitable replacements quite difficult. I may have something in the mail that could do, but I wont post about that until they arrive!

[5th element]

Well I've done some rough distortion measurements of the driver and it has excellent distortion performance just like the JA8008, but it does have some flies in the ointment. At first I was wondering if I was getting some resonances from the room vibrating, like panelling in the bookcase/furniture etc because that does happen when the room rattles, but no, it's a cone edge surround resonance, just like the Satori, but not so pleasant. With the JA8008-HMQ this is centred around 1kHz and the driver itself shows several impedance bobbles throughout the midrange. Luckily most of these bobbles don't show up in the distortion or in the FR so they can be ignored it's the 1k bump, associated with the cone edge/surround resonance, that causes problems.

Typical of cone edge resonances it shows up as high 2nd order, 3rd is impressively low throughout the drivers bandwidth, especially between 100 and 300Hz and higher orders are well suppressed. Zaph's measurements of the other JA8008 show that it too has a resonance at 1k, but this doesn't show up particularly badly in the 2nd harmonic. I do need to do more measurements of these as the 2nd harmonic issue did appear to be less pronounced at lower volumes, it could perhaps be cabinet related and adding more wadding around the driver and cabinet may help it, the cone of these is very thin, but I am not holding my breath. We shall see.

For my intended use they are exactly what I wanted, sensitive, low distortion from 50-300Hz and work well in a cabinet of this volume.

Here is a pic. Dialling them into the DSP was really easy, at least for preliminary testing. The driver should be entirely pistonic, and therefore flat, throughout the range in which I am using it. A 4th order electrical LWR should dial in a 4th order acoustic target and the EBS alignment should blend well with the multiple subs.

[wkhanna]

V impressive, Matt!

I am sure Evil Twin will be impressed!

[---k---]

I'm impressed. I think ET might disapprove of the yellow cone material. :wink:

[5th element]

V impressive, Matt!

I am sure Evil Twin will be impressed!

Thanks

I'm impressed. I think ET might disapprove of the yellow cone material. :wink:

Haha, I was going to say the same thing, although I think the yellow might just be okay providing the rest of the loudspeaker is very black. I mean one cannot appreciate black being black unless there are not black things around it

[5th element]

New old stock! :B



These look like they will just about fit for the coax with some careful milling. I am very partial to pistonic domes, rather than softdomes, so these make perfect sense, plus the fact they've got awesome distortion performance, an underhung motor and a really healthy amount of xmax. The ND16s are good, but they do show increased distortion when the volume is turned up, this isn't that much of a problem for me as I rarely listen that loud, but I would much rather have the capability to do so should I need it.

I've got some nice distortion measurements for the JA8008-HMQ, but I've been quite busy with other things and have stuff planned for this evening too, so will get to those when I have time.

[5th element]

Okay so the Aura NT1s will mill down to around 29mm in outside diameter just fine and work very well as the coax tweeter for the FST. They are a much better choice than the ND16 and have significantly lower distortion, especially around the xover point, they certainly sound nicer and work better in the top octave too.

Here's the response of the NT1 after the various filters have been applied to get it flat. Wave guide loaded tweeters always require a bit of work and this one was no exception, besides the usual high pass elements it required three notches to get it looking like that. Still it works better than the ND16 and gives a much nicer response in the top octave.



Now onto the JA8008 HMQ. First of all these measurements were taken with the driver in the cabinet posted above and with the driver floor mounted. This creates issues with reflections that make the measured distortion a little more peaky than you'd typically find, but the general trends for the driver are obvious.

First up is the 2.83vrms measurement. This driver is very sensitive, we're talking around 95dB. The mic had to be placed a reasonable distance from the driver to ensure that its distortion wasn't impinging on the measurement and so that the drivers response could sum properly.



Generally this is excellent performance except for the elevated 2nd harmonic around 1kHz.

Next up is the same measurement done 5dB lower in level and with the mic closer. This helps with the reflections and gives some sort of impression of how the drivers performance changes as the volume is reduced. Again good performance all around



And finally 5dB lower again, so -10dB over the 2.83vrms level. This corresponds to around 85dB, which is supposed to be a typical listening volume in room for a pair of loudspeakers. Again the mic could be placed even closer allowing for even fewer reflections.



This is a low distortion driver, but imo is better suited to being a high sensitivity bass driver rather than a mid/bass. Of course further measurements should be done in a more typical environment to see if the 1kHz resonance issue clears up somewhat, but I am not holding my breath, high levels of 2nd harmonic are typical of cone edge surround resonances. Interestingly Zaphs measurement of the previous version didn't show any signs of this and Troels FR comparison of the standard vs HMQ version of the driver, show a similar blip in the FR at 1k. Now the surround was changed in the driver 'upgrade' and the dustcap was removed. The surround used to be wider and more sticky glue used to be placed around the edge of the cone, it is possible that removing it worsened the cone edge resonance. Both of my drivers perform in the same way and SEAS are known for their excellent quality control.