B&W FST mid

[5th element]

Has anyone else noticed Zaph's test of this driver?







It's extremely impressive performance if I do say so myself, the high sensitivity is also a nice bonus. I do wonder why there aren't more mid-range drivers with similar parameters as this, the only one I can really think of is the ridiculously priced Accuton. According to this website, the current FST from the 800 series only costs a little over $140 each, that's if they will sell them to you. Not a bad price all things considered.

[stuofsci02]

You can buy the FST driver if you like from B&W as a spare part.. You would have to pick which B&W speaker you want it from, and that price is about right.

Cheers

[Hdale85]

Looks pretty nice...and I agree the price isn't bad at all! Their profit margin must be gigantic.

[5th element]

Well after Zaph's measurements I just had to get me a pair. I've been waiting for a mid like this to show up for a few years now 5-6" diameter, high sensitivity and low distortion between around 400-3000khz. The 6.5" accuton that the two J's (Jon and Jed ) praise highly has always been a bit too over budget, so when this B&W unit was measured I had to enquire about getting hold of a pair. I was going to wait until the other parts of my system were complete (mainly finishing the building of the DSP), but in light of the sudden price hike of neodymium I didn't want to wait a couple of months to find the driver had taken a hefty price hike. Regarding the price I paid just under £380 for the pair (including VAT and shipping), still almost half the price of one C176-96 And who knows how much I'd have to pay for the accuton in the UK.

The build quality of the driver is first rate, with a small neo magnet and thin framed cast basket, a very open back to the cone, venting under the spider and as I can't see any coil above the front plate it would appear to have an under-hung motor. The phase plug is black and considering the pictures of the 800D series the phase plug has always looked to be made out of plastic (necessary because the grille clipped onto the tip of the phase plug), the standard one however is black painted solid metal, so the phase plug will provide additional heat sinking for high drive levels. The M6 screw for mounting from the backside of the magnet is a nice touch, helping to keep the front of the cabinet looking neat and I would imagine it also gives some additional strength to the cabinet.

FR measurements are similar to Zaph's, the distortion however is, as Zaph's measurements showed, excellent to say the least. My measurements were performed at around 50cm from the cone with the behringer ECM8000 and at a drive level of 2.3volts RMS. The driver is nominally 4 ohms, so this is, in reality about 1.3 watts. The driver is very sensitive though so we're looking at about ~95-6dB.



Just look at that :E You don't see performance like that every day, it looks like it would be a great match for the air circ or similar illuminater tweeters. B&W certainly knows what they are doing with driver design and my respect for them has somewhat increased. Both the motor and cone construction are top rate. Kevlar is a material they've been using for many years and it appears that their experience has paid off, the usual quite severe resonances that can be associated with the material seem to have been dealt with quite well. Although their design philosophy regarding crossover points and such seems to differ from my own, apart from the off-axis performance trade offs, the driver can most certainly support the higher xover point that B&W like to use of around 4khz. I am going to be pairing it with a wave-guide loaded scan speak D2608/913000 and depending on how things go, probably an xover point around ~2.3khz, the goal being to match the directivity of the two together. The wave-guide I am using is the monacor WG300, this has exactly the same external diameter as the B&W mid so a directivity match should work fairly well.

If the rise in neo prices really have halted accutons production of the C173-96, this could be a viable alternative. Quite how B&W would feel about supplying lots of FST midranges to the DIY community I don't know, one would think they'd be thrilled as it's great publicity and at the end of the day profit is still profit.

B&W also produces 5 and 4" versions of the FST, this does have me wondering how they perform too. Considering the low end performance is excursion limit by the FST surround, I'd imagine that the 4 and 5" versions would have to be crossed over a little bit higher on the low end.

[stuofsci02]

I have a set of 804s with the bigger FST and also a set of CM7 with the smaller FST. Both are amazing performers. I did notice a much better midrange when going to a B&W speaker with the FST. Other things improved as well on the speaker, so I cannot attribute it all to the FST, but I know that I will try to stay with speakers that use it.

[5th element]

I can't comment yet with how it sounds because I haven't got the DSP configured, besides it's only really good for a 2 way at the moment, I need to build another DAC channel. Run full range though the FST mid sounded quite decent I must say, gonna be interesting to see what it's like with a tweeter added to the mix.

The voice coil is made of Kapton too. This is the same as what Audax used in it's higher sensitivity aerogel prestige drivers.

[5th element]

Now that I've had a bit more time I thought I add to the thread a little. First some high quality speaker porn. :T

The driver from the front.



The driver from the rear.



More from the rear.



More from the rear.



As can be seen the driver has a very open basket and small neo motor giving a lot of breathing room behind the cone, it might interest the OB crowd.

A close up of the voice coil former with the phase-plug removed, the former isn't made of metal so is probably Kapton, for some reason it still has a slit cut in it though.



The removed phase-plug.



Behind the vented spider, a close-up of the voice coil former showing zero visible coil, in what must be an under-hung motor.



The driver with the required M6 headless bolt screwed into the back plate and then secured with an M6 nut around the rear of the enclosure.



And finally the loudspeaker that the driver is used in.



This is currently a 3 way using an RS225 on the bottom with a Scan Speak built Peerless HDS tweeter, in a Monacor WG-300 wave-guide.

Now for some measurements



The red trace represents the FST midrange, as can be seen it far more sensitive then the RS225 and even manages to outplay the tweeter in the wave-guide. Of course the FST is really a 4 ohm unit so knock of 3dB, but regardless of that the thing goes loud with very little power. At a normal listening levels it was using about 0.05 watts peak.

It's also interesting to see the increased resolution on the frequency response compared to Zaph's which squishes everything pretty flat. The cone is obviously going through a region of controlled breakup past about 4khz and before this at just over 3k there's a small highish Q resonance. In a normal driver this would create an issue, but in the FST it's not so bad, this is mainly because it's HD plot is nominally so clean. If you look at the HD plot, the cone resonance adds a 10 fold increase in distortion across all the harmonics except the second. This behaviour is nothing unusual, the SEAS excel magnesium cones display similar behaviour, except with the W15 its 3rd order is already above 0.1% before the peak hits, with the cone resonance amplification raising it up above 1%, this naturally sets the upper xover frequency of the driver. In the FST though it's 3rd order product is already so amazingly low that by the time the resonance hits the peak only manages to hit about 0.06%.

If you look at Stereophile's recent review/measurements of the current 800D2 flagship you can clearly see the FST drive units signature plastered across the response of the loudspeaker.

First there's a high Q bump in the low end roll off with the driver crossed steeply to the bass units at around 400hz. Then there's the peak @ ~3khz which they do nothing to suppress before crossing over, what appears to be, very gently to the tweeter, creating a hump and ragged top end almost mirroring the drivers raw response. All I can say is that it's a damned good thing that the FST is a clean as it is, it is quite rare in that it can actually support a design like this. The FST unit is clearly B&Ws main selling point and perhaps this gives their speakers what could be called the 'B&W' sound.

In my own design I follow B&Ws trend of using a steep xover @ 400hz as the FST really cannot go down much lower then this. However I've chosen to crossover @ 2.5khz with a 2nd order slope to the wave-guide.

This obviously requires quite a lot of work and is something the DSP makes light work of. To make things easy I popped an LT on the FST to equalise the high Q low end response and then went for a 4th order target between the RS225 and FST. The 225, being in a small sealed cabinet also has an LT on it to EQ the bass.

Due to my cabinet shape the FST has a diffraction bump @ around 1500hz, this required a notch filter when arriving at a 2nd order slope, as does the peak at ~3khz but apart from that it's easy enough to work with.

The wave-guide loaded tweeter obviously needs a lot of work to hammer it flat (3 notches), the end system uses digital delay to bring all the drivers into phase with one another at the listening position.

I tried a 4th order xover @ 1800hz, a 4th order @ 2500hz and a 2nd order @2500hz before settling in on a preference for the latter. The DSP allows me to change between one crossover and the next at the push of a button on the remote control so it makes it rather easy to compare one crossover with another.



The crossover is by no means finalised, it's just a quick mock up with less then ideal measurements and when I finish the DSP completely I'll go at it again, at least with taking better off axis measurements to confirm the design goal is working properly. I might not exactly change the xover much though as I don't think I'll really need to. It was a bit of a gamble going for the FST in the first place, but I am pleased to say that it was well worth it, the system sounds amazing and so far doesn't do anything wrong. Once I integrate the subs it should only get better.

You often wonder sometimes if a commercial manufactures products are nothing more then a product of marketing hyperbole, however I am pleased to say that so far the FST midrange certainly manages to live up to the hype and it makes me wonder what the smaller versions of it might be like.

It appears that the htguide forum might not rescale the images, so the LspCAD exports are rather large, I guess no one can complain about the graphs showing too little!

[Bear]

Good work! Cool looking cabinet, too. :T

Looking at the price list, it looks like the 800D2 mid has a bit of Accuton envy -- almost $400/ea. if I'm reading the price list right.

[5th element]

Good work! Cool looking cabinet, too. :T

Thanks The cabinet came about because I wanted something reasonably attractive but also functional. I like to make changes to the system so having a separate cabinet for each driver mean that I only really have to rebuild one cabinet should I want to change a driver.

Looking at the price list, it looks like the 800D2 mid has a bit of Accuton envy -- almost $400/ea. if I'm reading the price list right.

Holy Cow! I knew there was a good reason for buying them when I did. It appears that the neodymium prices have finally hit B&W. The price on the current D2 range was $140 a couple of days ago and the price on the LF00264 that I've got (from the 803D) was $170.

The price on the FST midrange drivers used in the CM series and in the CI wall installation series has remained the same, so they must use a standard ferrite magnet. I do wonder how they perform, could be very similar.

[Bear]

Holy Cow! I knew there was a good reason for buying them when I did. It appears that the neodymium prices have finally hit B&W. The price on the current D2 range was $140 a couple of days ago and the price on the LF00264 that I've got (from the 803D) was $170.

The price on the FST midrange drivers used in the CM series and in the CI wall installation series has remained the same, so they must use a standard ferrite magnet. I do wonder how they perform, could be very similar.

Yeah, at $350/ea., these aren't really a bargain anymore. That is solidly into Accuton C78/79 territory, and well above the competing offerings from SEAS, ScanSpeak, and even Zaph, himself.

That being said, the woofer from the 805D2 is still showing at just over $100/ea., and I would be curious why it is so much cheaper than the woofer from the 805S and the rest of B&W's FST drivers (ferrite magnet?).

[5th element]

I did compare the ferite to neo magnets in the D series (I thought all the woofers in the D2 series used Neo magnets and the ones in the standard D used ferrite) and they didn't appear to differ too much in price. The only thing I could imagine is that the driver pricing is done on batches of loudspeakers made. I wouldn't mind betting that the motor in the FST 800 series is pretty much identical so perhaps they are all made at the same time, thus the price is reflected the last batch made or something. That and with the 800 series driver being common to all the loudspeakers they should use way more of them.

The driver I am using costs $311 now and I am comparing it against the 6.5" Accuton mid, the C173-6-096E, which madisound lists as being $615. Note that the Accuton still hasn't changed in price to reflect the new neo prices. Even ignoring that you can almost buy two LF00264s for the cost of one C173, so I would still consider it a bargain. I'd also rate the LF00264 a notch above the C173 too, that is if you can crossover at ~400hz. The LF00264 is certainly the more capable driver on the upper end, easing and giving more options for integrating it with whatever tweeter you choose.

In my opinion the parameters of the C173 and the FST are somewhat unique and it is the design approach of producing a dedicated midrange that combines high sensitivity and a highly linear motor, that perhaps makes them sound somewhat special. Neither SEAS nor Scan-speak offer a driver that in anyway compares to this, which is a shame, they always seem to go for more low end extension rather then optimising the driver for 3-400hz and up.

RE the 805D2/S etc, it could very well be that B&W price the replacement parts competitively and they change the price per batch made. Certainly the price of $140 for the D2 midrange before the price hike was ridiculously good value when compared to anything else. It could also be that the D2 series being current means that more drivers are made so perhaps this pushes the price down a bit, the S drivers should after-all be old stock.

If I ever buy another FST mid I'd probably go for the 5 or 4" version used in the CM series, mainly just to satisfy my curiosity at how the more 'budget' driver performs.

[Jonasz]

Very interesting speaker you have there! :E

Could you tell me the outer- and cutoutdiameter? Need to check if they fit in a pair off baffles I have.

[5th element]

Very interesting speaker you have there! :E

Thanks, I've only had the cabinets done for a little over a week now, I'll get around to painting them soonish I think. I'm agonising over whether they should be black or a very light grey/off-white. I guess I could try both.

Could you tell me the outer- and cutoutdiameter? Need to check if they fit in a pair off baffles I have.

Yes they are 168mm. The flange is quite deep too at 1cm. The cut-out diameter...bahh, I don't know, I always wing-it on the inner cut-out, I'd say 15mm off at either side so something around ~140mm. With the driver mounting from the rear though, making sure you've left enough room for screw holes/tnuts isn't important, it really lets you keep the rear of the driver open to the inside of the box, which is after-all what B&W want with it firing into their moulded cabinet.

[Jonasz]

Looks like it's going to be ok then, the baffle is made for a Seas W18. I wonder if the SS8565 in the baffle will be able to reach up to 400hz? Maybe the FST can do 300-350hz with a steep crossover?

[5th element]

I think you'd be fine with a 300hz 4th order.

[5th element]

Here's the system distortion played at what is my normal listening volume (it's not particularly loud). I had to splice in a near field measurement for the RS225 below 200hz as room artefacts were screwing the thing up.

[5th element]

And here's the same thing but at an elevated listening level. This is actually as loud as it will go as limited by the RS225 xmax with the LT. This is one reason I'm going to limit the thing to 60hz once the subs are up and running. There's only so much a single 8" driver can do. Not to mention there is significant chuffing occurring through the air gap provided by the phase plug when done like this. Of course if I wanted the RS225 to handle the bass on its own then it wouldn't be in a ~10 litre sealed box!

At it's loudest point the FST is taking around 8 volts peak to peak , which is 2 watts into a nominal 4 ohm load. This is about as loud as I'd ever want to play it. I don't have a huge room to fill either though which will affect how loud you'd nee do crank it.



As you can see distortion is still very low so no worries if you want to crank it even further.

[jkrutke]

Nice looking design, and I concur, nice midrange driver. I also have to say that I've seen some inner workings of some of B&W's high end tweeters and was equally impressed. Their diamond domes really are diamond, or at least a platinum and carbon sandwich. A while back I sent a fragment to a testing lab and got hardness and composition testing. I've seen some claimed exotic material domes end up just being spray painted plastic.

New prices on Ne: not so nice, heheh.

[5th element]

Nice looking design, and I concur, nice midrange driver.

Thanks

I also have to say that I've seen some inner workings of some of B&W's high end tweeters and was equally impressed. Their diamond domes really are diamond, or at least a platinum and carbon sandwich. A while back I sent a fragment to a testing lab and got hardness and composition testing.

I've always been somewhat impressed by B&W, they've always appeared to have sound technological reasoning for designing their products in they way they do. They also manage to do this at a price point that is sometimes significantly below a lot of the competition. Of course when you reach a certain price it all tends to become rather meaningless anyway.

I've seen some claimed exotic material domes end up just being spray painted plastic.

Isn't that illegal? I have seen some of the older Audax tweeters where they specifically state that the dome is titanium deposited on plastic. Although I do remember the whole beryllium fiasco a couple of years ago :roll:

New prices on Ne: not so nice, heheh.

It seems I decided to buy just at the right time. The price hike on the LF00264 seems in line with an eminence driver mentioned here that went from $160 approx to $315 CDN. I will be interesting to see how the neo prices pan out and if they stay expensive I can imagine that product lines will have to be overhauled. Can't really be having a woofer costing $300 that was initially released to compete with the $150 price bracket.

[jkrutke]

Isn't that illegal? I have seen some of the older Audax tweeters where they specifically state that the dome is titanium deposited on plastic. Although I do remember the whole beryllium fiasco a couple of years ago :roll:

That's funny, I had Audax in mind when I mentioned that. "Vapor deposited" is just another name for spay paint. And as much as I dislike Steve Mowry for his clear bias on the issue, A few vendors were indeed misleading about their beryllium dome materials.

[Hdale85]

For what those speakers cost, I'd sure hope it's real diamond.

[5th element]

For what those speakers cost, I'd sure hope it's real diamond.

Well the individual tweeter prices are around $1200 as a replacement part. This may seem expensive but considering how much the accuton dia tweeter costs $2900, it again seems like a bit of a bargain

If I had money to burn I'd probably buy a B&W diamond tweeter, just to see if there's anything to it. It might work well in a wave-guide too. If B&W changed the marlan midrange pod to include a 6" moulded wave-guide + diamond tweeter + crossover to match the directivity of the FST to the wave-guide, they'd be onto something imo.

[5th element]

With regards to the high pass on the FST I ran some extra distortion sweeps with different xovers just to see what it was capable of.

All sweeps were first done at a drive level which is about 8vp-p delivered to the driver, which would be about 2 watts and probably give around 95dB. I then increased the drive level by 5 dB, then by 10dB just to see if anything peculiar jumped out. These were done with the RS225 disconnected so all you're seeing is the raw bottom end roll off of the FST + its crossover.

The graphs are rather large, but they need to be so that you can read the legends, you might want to zoom out so you can directly compare one graph vs another though.

First up are the 4th order xovers. The graphs on the left show the current xover I'm using with a target of around 400hz 4th order. The graphs on the right drop the frequency down to 300hz.



Generally speaking this is a very impressive set of numbers. The level of the harmonics generated remains flat through all of the FSTs passband. The level of the harmonics remains flat as the SPL decreases at the low end roll off, obviously as a %, distortion would increase, but here is where the bass driver should be filling in and keeping the SPL constant so THD should also remain constant unless the bass driver craps out. To put these dB numbers into perspective here's a % graph for the 400hz 4th order @ what I assuming is around 105dB.



THD is going to be pretty much less then 0.5% for most of the range that the FST and tweeter are playing in, not bad considering the drive level. I also think it's clear that the wave-guide is also showing how necessary it perhaps is in this application. The tweeter and the FST seem particularly well matched and lets not forget I'm using a 2nd order xover at 2500hz. This places a large demand on the tweeter and even though the wave-guide reduces the demand on the tweeter considerably you can still see a gentle rise in the 3rd order product as the tweeter is brought in. This isn't really apparent at the 8vp-p drive level, but as the drive level increases it becomes quite noticeable. Quite clearly the HDS tweeter would probably be unhappy @ these drive levels if naked. This is also quite possibly why B&W use quite a high xover point with their designs. They don't use steep filters either in the 800 series, so with a naked tweeter you'd be forced into using a higher xover point if you wanted the speaker to remain clean at higher SPLs. I know I said that B&W might be onto something if they chose to mould a wave-guide out of marlan above the FST and then pop their diamond tweeter in it, but it seems like the wave-guide could be necessary if they wanted to lower the xover point by much.

Next up is the FST run with a 2nd order filter, again at 400 and 300hz and at the same drive levels.



As you can see this is not recommended. :rofl: Poor FST.

You could perhaps get away with a 4th order @ 250hz, but I really wouldn't push it any lower then that, you can already see in the 4th order plot @ 300hz that the distortion on the lower end is rising ever so slightly. This wont really be an issue, but it's clear that the driver is working close to its optimum lower end limit.

[tktran]

Hi,

Can you tell us a bit more about the suspension? Is it foam, cloth or some other kind of material?

IMHO it's a stellar midrange, and everything I could ask for - higher sensitivity, low distortion, large Sd, it's got a very open chassis which would be great for airflow.

With only 1mm of x-max, the lower end is limited, but it doesn't really matter
as many woofers can cross comfortable at 300Hz.

The upper end breakup is manageable. I don't know why B&W like to cross at 4KHz, but my suspicion is that they prefer to use a single cap on their tweeters, so they need to cross higher.

Have you thought about a 3-way with bigger/higher sensitivity woofers?

regards,
Thanh.

[5th element]

Hi,

Can you tell us a bit more about the suspension? Is it foam, cloth or some other kind of material?

The suspension is made of foam. Really all it is is a ring of foam that is glued to the basket and then glued to the cone. The ring however isn't exactly small, I mean what you see from the front implies a very thin ring of something that could be quite delicate and prone to damage, this isn't how it's constructed.



As you can see here there is quite a thick ring of foam that is glued to the back and sides of basket. The cone then 'sits' or rests on top of the ring with the outer rim of the cone being glued to the foam.

The upper end breakup is manageable. I don't know why B&W like to cross at 4KHz, but my suspicion is that they prefer to use a single cap on their tweeters, so they need to cross higher.

Well the FST itself is still a capable performer all the way up to 4khz. I find it odd that they like to cross over so high to the tweeter myself, but at least they aren't using something like a SEAS W18 all the way up to 4khz just for the sake of using a single cap on the tweeter.

You are quite right though, B&W do like using just a single cap on the tweeter. The service manuals for lots of B&Ws previous ranged are available on the B&W group website for North America, they contain all the crossover schematics. B&W seem to prefer the less is more approach.

There is also a white paper on the 800 flagship that contains a lot of information and from what appears to be quite a technical point of view. I'd imagine they go into discussions about the crossover in that.

Have you thought about a 3-way with bigger/higher sensitivity woofers?

regards,
Thanh.

I have indeed. As the system will end up using multiple subs the lower end extension of the main loudspeakers isn't vital. I'd prefer to keep the system smaller rather then bigger, but a higher efficiency 8" woofer from somewhere like 18sound does seem interesting. The trouble here is that there are very few measurements of pro drivers and of the measurements that I've seen they've been a bit hit and miss. Still if someone were to measure something and it fit the bill, I'd probably give it a try.

[Face]

Which side of the pond are you on?

I may have a pair of AE TD10M's for sale soon.

[5th element]

Thanks for the offer, but the wrong side of the pond^^ I'm not even sure if I'll change the bass to anything yet. The AE stuff sure is quite tempting, I need to get the rest of the system completed before I make any decisions about whether I'd like to change anything though.

[Jonasz]

Finally received my pair. They will go into these cabinets and be partnered by an Eton ER4 and a couple of Eton 7". (unless I decide to go dipole which is very tempting...)

This project will have to wait though until my current build is done. Sure going to follow this thread closely!

[5th element]

Nice to see that someone else is going to start experimenting with these at some point. Due to the tests that I did regarding low end distortion/low end xover point, I did lower the crossover point from 400 to 300hz. May as well have the high sensitivity driver covering as much of the low end as it's capable of.

Large scale orchestral stuff sounds amazing with dynamics exploding from the speakers. Of course with my small room I don't need the dynamic peaks to be any greater then around 95dB.

The FST fits quite nicely into that cabinet Jonasz, could be better, but it could also be a whole lot worse. What xover do you have planned, active, passive? Also have you measured the distortion of the ER4 to get an idea of what its low end capabilities are like? Eton's own measurements don't go down below 2k.

[Face]

Is that a circus clown's watch you're wearing?

[Jonasz]

Nice to see that someone else is going to start experimenting with these at some point. Due to the tests that I did regarding low end distortion/low end xover point, I did lower the crossover point from 400 to 300hz. May as well have the high sensitivity driver covering as much of the low end as it's capable of.

Large scale orchestral stuff sounds amazing with dynamics exploding from the speakers. Of course with my small room I don't need the dynamic peaks to be any greater then around 95dB.

The FST fits quite nicely into that cabinet Jonasz, could be better, but it could also be a whole lot worse. What xover do you have planned, active, passive? Also have you measured the distortion of the ER4 to get an idea of what its low end capabilities are like? Eton's own measurements don't go down below 2k.

It's encouraging to read about your experiencies with the driver, makes me wanna start tinkering with it right away! :B

The only measurement I've seen of the ER4 is this one and I think a 4th order in the 2.5-3kHz region could work well? I've read about designs using them all the way down below 1500Hz so...



Now reading about the Accuton price increase, these at $170 must be considered a steal! :B :P

[Jonasz]

Is that a circus clown's watch you're wearing?

Not a clowns watch but a Casio G-Shock G-9100R-4 Men In Rescue Orange Gulfman! :P

[5th element]

The ER4 looks like it'd manage a 4th order at 2khz, I don't know how loud it'd go like that though. Where you cross over the FST really depends on what you're happy with in the off axis. Crossing at 2.5-3khz will create an off axis dip with a normal tweeter at the xover frequency. In my case the wave-guide matches to the directivity of the FST reasonably well giving a uniform off axis response.

[Jonasz]

Thanks for the offer, but the wrong side of the pond^^ I'm not even sure if I'll change the bass to anything yet. The AE stuff sure is quite tempting, I need to get the rest of the system completed before I make any decisions about whether I'd like to change anything though.

What is the status of your build today?

Embarrassing but I still havent started a design with my FST's yet...

[5th element]

New cabinets :B Rough, still need painting, a crossover redesign and the longer M6 threaded bolts to arrive, but nevertheless, even with the old xover they are singing sweetly.

[JonMarsh]

I like! Very hard core! :T I mean that in a good way... you seem to be trying to out do me even from a few years ago with long periods and no updates. What's the status of the range below the mid?


Now, you know, as far as their crossover philosophy, we can take the high road and say that they figure they don't want more than one cap in the tweeter signal path, OR, one might point out that it's just less expensive.

I took a set of warranty DM6 drivers in the mid 70's and built a larger set of speakers with the auto former and much different crossovers, that made the original B&W DM6 sound chesty, constricted, and closed in... but I didn't spare expense in the crossover, compared with the factory design. Remember, they're in it to make money...

[5th element]

When ones health isn't to be relied upon, updates can be few and far between sometimes. It is nice to be able to do stuff again.

The mids are a floor mounted RS225 that was originally under the old midrange enclosure, I then migrated it to the floor, under the stand, similar in concept to that Ferrari red job from the other thread and now it sits just to the side of the stand.

At some point I am going to rebuild the bass section to replace the stand - hence the requirement for something different and hopefully more sensitive. So far I am seriously thinking about buying a U22 just to see how it actually measures. A pair sim quite nicely in a ported 40-50 litre box, which would fit well as a stand replacement too. Not a lot of 8 inchers sit well in a ported box so small, but the pair do end up at 96+dB, which is in line with the rest of the system.

I think B&W have a philosophy as well as a budget in mind. It also lends itself well to the marketing department, where they can put uses X boutique range of xover component in an advert. KEF seem to go the opposite route and use generic parts, but have very complex xovers in their flagship products. I definitely think that at the price point the KEF and B&Ws are set at, B&W would be far better off using generic parts and going more complex to sort out the 'issues'. There is something to be gained from engineering some issues into a flagship range, then in 5 years time engineer them out as a reason for people to upgrade. It puzzles me somewhat though because B&W obviously put a lot of engineering know-how into their drivers and cabinet design, yet screw up the xover in a number of different ways.

When I hear single cap and coil systems I think of $50 JPW mini monitors, not flagship products.

[Jonasz]

This is where my system was at a couple month ago (please ignore the santa :B). I have a big system with removable m/t baffles and right now there's a W22 with Waveguided RS28A in it's place. I have invested in a DSP since messing with the B&W and I now feel the urge to pop them back in again! :yeah:

[5th element]

Hmm, interesting. I'd expect the waveguided W22 to sound better than the FST and the RAAL. AFAIK the RAAL needs a 2.5k-3k xover to sound at its best and the B&W is just like any other 7" in that it starts to lose directivity quite quickly and if crossed to a standard format tweeter would need to be xovered below 2k for the best results. Of course this doesn't mean it's going to sound bad, just not optimum and I simply love what waveguides do.

The W22 is very capable in the bass, the B&W is not, so you'd have to be careful there, especially OB'd. I'd be tempted to make a new baffle (if possible) that uses the W22 from say 100-400Hz, where the B&W takes over up until ~2-2.5k and then crosses over to the RS28 in a waveguide.

Attached is a directivity plot of the FST and you can see that it's around 6dB down @ about 45 degrees from around 2k-2.5k. This is where it needs crossing to a waveguide of a similar directivity.

It wouldn't get much better than the W22-FST-WGtweeter crossoved over well with a suitable bass section. The RAAL confuses me to a point because of its large size and requirement of a high xover point, you're going to end up with lobing issues whatever you cross it to. Still some designers don't seem to want to optimise the off axis as well as they could. It's all about choices though.

I've now redone the xover and listened more to the sphere+cone (essentially a copy of the B&W nautilus cabs from the whitepaper) vs the other MDF cabinet and there's no comparison. The old MDF cabs had a colouration present in the lower range of the FST and I was hoping that the spheres would cure it - they did. What isn't apparent is how they are constructed. It's essentially two plastic spheres (one bigger one smaller) inside one another, with the ends cut off and the space between filled with marble dust impregnated in polyester resin. The walls are about 1 inch thick and they weigh a ton.

The FST is very nice, but it is a little harder to work with than your run of the mill mid as it needs to be handled correctly on the bottom and top end to make sure that it isn't pushed too hard down low, that and the rising respose + resonance throughout the presence region. You really do need to get it flat otherwise it's going to sound way too forwards.

[Jonasz]

I'm with you that W22 and wg seems better in theory, but after trying the B&W/Raalcombination again there's no question about it, the transparancy, attack, softness, detail, well most everything about the B&W/Raal is scary realistic sounding. I get a narrower vertical sweetspot but that's about it when it comes to the negative points. After equalizing the B&W I'm using 4th order electric at 300Hz and 2nd order at 3kHz. Seeems to work just fine! ;b>

I think your speakers look fabulous, I wish I had the know-how to build cabinets like that... :P

[5th element]

I'm with you that W22 and wg seems better in theory, but after trying the B&W/Raalcombination again there's no question about it, the transparancy, attack, softness, detail, well most everything about the B&W/Raal is scary realistic sounding. I get a narrower vertical sweetspot but that's about it when it comes to the negative points. After equalizing the B&W I'm using 4th order electric at 300Hz and 2nd order at 3kHz. Seeems to work just fine! ;b>

The B&W does do something pretty special, it's like combining the composure, detail and finesse of a well executed metal cone with the immediacy, speed and presence of a high sensitivity soft cone but without the usual trade offs, it's silky smooth and extremely easy to listen to.

I mean I love the way the W15CY001 sounds and I love the way the 93dB HM100Z0 aerogel mid from Audax sounds. They each have a different type of presentation. The W15 sounding noticeably cleaner than the HM, but the HM having a presentation that is extremely addictive due to its higher sensitivity and low moving mass. The B&W is like combining both.

The RAAL with it's next to zero moving mass is probably a pretty good match even if the power response isn't perfect.

I think your speakers look fabulous, I wish I had the know-how to build cabinets like that... :P

Thanks

It did take quite a bit of brainstorming to come up with a way of executing the design. I do have quite a lot of progress pictures showing how it was done, I've just not uploaded them anywhere yet. Needless to say it's actually pretty straight forward, otherwise it wouldn't have been possible, it just takes a lot of time and effort compared to a standard baffle or box.

[5th element]

Carrying on with this build I have started to make sawdust again.

I've been deliberating for a while on what to replace the RS225 with and have finally come to a decision. One of the absolute requirements was for reasonably high sensitivity. The FST, although a 4 ohm driver, has a minimum 2.83vRMS sensitivity of around 95dB, which only increases towards its frequency extremes, approaching 100dB. Although my system is active, and therefore driver sensitivity mismatches irrelevant, I still wanted to keep the 8 ohm sensitivity to at least the 93dB mark. This might seem somewhat easy, but I also wanted several other things. The first was that the driver needed to have distortion at least comparable to the RS225, no small feat really as it's already superb performer, at least up until the 300Hz that I want to use it for. The second is that I wanted it to have extension down to about 50Hz. Yes I use multiple subs to augment the low end, but I wanted the system to be reasonably self sufficient. The third was that the driver needed to operate well in a ~20 litre ported box, only 20 litres because I need this to be compact and ported because this would be necessary to give me the extension required without needing to resort to a Linkwitz Transform circuit, which I would rather be doing without. The fourth was that, due to size constraints, that I would prefer the driver to be only 8" rather than anything bigger. Not an easy task!

A couple of years ago the drivers on my short list were basically from Acoustic Elegance. The TD10X seemed particularly well suited, but at that point I hadn't completely settled on how much volume the cabinet could actually have. The TD10X sits well in a 30-35 litre ported box so unfortunately wouldn't work fantastically well. Then AE went and released the TD8X, which seemed like it may have fit the bill, but it too didn't come out on top. The TD8X works best in ported boxes around 10 litres, which is unfortunate, it also has an 8 ohm sensitivity of around 92dB, which is a little shy of what I ideally wanted, it also didn't reach quite as low as I had wanted.

Hoffman's Iron law always comes into play with these situations and from my point of view I would have rather they optimise the TD8X to function in a larger box of around 20 litres and in doing so squeezed a little more extension and sensitivity out of the driver. AE already have the TD8M for small midrange focused applications, so why optimise the TD10X, which is supposed to be optimised for bass applications, to work best in still such a tiny, and therefore performance limiting, box? Oh well.

This moves me on to the driver that I actually decided on which hails from SEAS, courtesy of Troels Gravesen. Jantzen Audio's JA8008
had piqued my interest as soon as Zaph measured it, but until I had ultimately decided that the AE drivers weren't suitable, it had remained on the back burner. It does only come with a very limited amount of Xmax, but simulation with it deemed it acceptable, especially if the driver be high passed to prevent it from destroying itself. Then there's the fact that although the driver only has a small amount of Xmax (3mm one way), if it's anything like SEAS other 8" drivers, then it will have a considerable amount of Xmech. Sure this wont be low distortion performance, but it will prevent the driver from damaging itself should it be accidentally abused or sent an explosion during an HT moment.

One of the interesting aspects of the SEAS drivers is in the way that they design their motors. I am sure we've all seen Zaph's measurements where in SEAS drivers tend to exhibit excellent low end distortion performance, but then have rising third order.

Here for example is the W18EX.



This being different to the typical performance that we see with Scan speaks symmetrical drive motor, where the 3rd order product stays pretty much flat, but doesn't tend to dip down towards the low frequencies.



The above being scans venerable slit paper 18W revelator.

This clearly shows the difference between the two and especially as the SEAS has the phase plug it helps contribute to their drivers typically showing better distortion performance within the lower octaves.

Here is Zaph's measurement of the JA8008 which is certainly no slouch, especially if used only up until around 300Hz.



This is where things become interesting though as Troels has only recently released version two of the JA8008, the JA8008 HMQ. Here we have basically the original 8008, but now with a phase plug, a nice non-conductive voice coil former made of fibreglass, a slightly different surround and a stronger magnet to maintain its high sensitivity at the loss of the radiating area from the dustcap.



Given the performance of the previous driver it stands to reason that the new version would only improve upon, what was already a pretty dang good driver. Version 2 also looks far nicer imo and lets face it, looks ARE important. So I have a pair on order.

First up is how a single driver simulates into half space.



This puts the driver in a slight EBS alignment, which is actually perfect for mating with the blended approached used with the multiple subs. As I will be floor mounting the driver the half space sensitivity is realistic so a pair will quite obviously hit pretty decent SPLs in room and the group delay is decently low enough such that it wont be a problem, especially as we are far less sensitive to these things at low frequencies. The only issue is the cone unloading below the tuning frequency, but a high pass in the DSP would easily sort that out.

And here are two shots of the start of the cabinets.





These are around 60cm tall and are built in a similar fashion to some of the other curved sided enclosures we've seen on the HTguide. Here I am using several pine 'braces' as a substrate to which I am attaching the sides. Here in the UK we have a product called 'bendy MDF', which is 6mm MDF sold precut with slits sliced partially through it to make it easy to bend around a curve. I have used this construction technique before and found that the best way to 'clamp' things is to not use any clamps at all, but to temporarily use screws. Clamping MDF around a curve isn't easy, even with this bendy stuff, so I basically glue a brace up and then screw it into place. After the glue has set the screws can be removed and used again. The first time I used this construction technique was in making a pair of loudspeaker stands, that I ultimately filled with concrete. In this case that isn't going to happen and a single sheet of 6mm bendy MDF does not make for a particularly rigid cabinet. Here In intend to glue a second bendy skin, separated some 12mm away from the first and then fill the gap between the two skins with something suitable, like a fine sand concrete or Plaster of Paris. Hopefully this should make for some pretty solid walls.

I am intending on inserting the oversized port into the base of the cabinets, like with the statements. That is one advantage of using a reasonably high tuning frequency, you're free to use a wide diameter port without it needing to get ridiculously long. In this case I will be using 80mm diameter flared ports.

Over the next few days I will hopefully post progress reports as more and more of this starts coming together!

[Steve Manning]

There never seems to be an easy way to do curved cabinets ...... I wish I could spring for a setup from this place http://www.curvomatic.com/ Looks like it would save a lot of time if you were making a bunch of cabinets.

[5th element]

Wow that looks great, it looks like it'd be a pain to set up initially. You are right though, it would save you a lot of time if you had a few cabinets to make and it could be how companies like KEF and Wharfdale make such cheap curved sided cabinets.

[5th element]

And then there were two + the second side being partially glued onto the first cab.