Ardent Speaker Camp

[JonMarsh]

The Illuminator 6620 is now $1 more than the similar Air Circ 6600, which is good news. Thought I'd let you know. Also, pretty much the entire Illuminator line is going through price changes-- for the better.

well, that's all good news! And a lot friendlier price than I paid!

I'll have to keep a closer eye on things- only been checking every couple of weeks or so. I placed an order with Mad this morning, but the only item from SS was the D2608/9130, which still isn't listed in the SS Discovery page. How weird, or tardy.

Now if they could cut the price of the illuminator midwoofers a bit, (say, in half?), that would be fairly interesting. Great looking frame and impedance curve, but the value proposition still seems weak to me.

BTW Jed, have you ever looked at the SB Acoustic's 10's, that is, measured one?

[Jed]

BTW Jed, have you ever looked at the SB Acoustic's 10's, that is, measured one?

I have the 8" here now (has a shorting ring I'm told). When I get around to it I'll measure and post the info. I haven't tried the 10" yet, but I probably should to see how my results compare to some of the results I've seen.

Good find with the SS HDS tweeter... I didn't see that one listed until you just posted that it's at madisound already.

I've been eyeing the small format SS Illuminators as well to go with some Accuton C44s I picked up. The good news is those small format tweeters are much more affordable as well.

[Paul Ebert]

Given that the Le of the 8" is lower than that of the Rs225, it would appear to have some shorting ring(s). The Le of the 10" is quite a bit higher (1.6 vs 0.6), so I would guess that it does not.

The 8" does look quite nice, not to mention a fairly good deal. I look forward to your results.

[JonMarsh]

Google is my friend...

Voice Coil SB Acoustics tests...
This is a report on the test results of the new SB Acoustics 10" woofer in Voice Coil magazine. In my eyes, I'm a little disapointed, particularly with the Klippel results. Bl(x) looks pointed and asymmetrical, and Le(x) varies from 1.75 mH in the inwards position to 0.75 mH in the outwards position. Given that Klippel info, as expected the harmonic distortion sweeps don't look that great either. THD exceeds 2% at a few frequencies above 100Hz and probably averages 1.5% between 100 and 400 Hz. This motor is nothing better than what you see in cheap Asian crap. It makes a Seas ER18RNX look like gold, and even a Dayton RS270 out performs it in just about all ways. Results like this are not what I expect from the former Scan Speak guys.

Sigh....

Not exactly what I'm looking for....

[Jed]

Sigh....

Not exactly what I'm looking for....

You can see why I wasn't exactly jumping to buy a pair. I may still try a pair for myself though. Although John Krutke mentioned in a blog entry awhile back that he might try a pair, and I was holding out for him to drop the dough on a pair instead of me.

[JonMarsh]

Yeah, I MUCH prefer to learn something bad about a driver on someone else's money. I've already done too much of that this year on my own nickels, so I've turned rather cautious- but not too cautious to order some ZA14W08's. And there's an SS 8" model I have my eye's on for a new application. I'll be curious to see what you think of the SB23NRXS45, though glancing through it's parameters, it may not work in the size box I'd prefer to have. If the motor is as good as the impedance curve hints at, it could be a good value proposition.

[Jed]

I wouldn't trust SS's published T/S parameters.

Edit: see explanation below is post 463

[TacoD]

I wouldn't trust SS's published T/S parameters.

Scan Speak?? In my view the Danish are rather consistent with the parameters, also every woofer I've got from Scanspeak (no matter what age) is doing exactly what is specified with small deviations due to temperature/ equipment.

[Jed]

Scan Speak?? In my view the Danish are rather consistent with the parameters, also every woofer I've got from Scanspeak (no matter what age) is doing exactly what is specified with small deviations due to temperature/ equipment.

FR and Impedance, I agree. Other parameters like Vas and QES are often different.

Just take a look at zaph's T/S parameters that were tested on various drivers and you'll see what I'm talking about. My measured parameters closely match zaph's but not SS's published numbers. The result is often a driver that needs a larger ported box because of the higher QES.

[JonMarsh]

FR and Impedance, I agree. Other parameters like Vas and QES are often different.

Just take a look at zaph's T/S parameters that were tested on various drivers and you'll see what I'm talking about. My measured parameters closely match zaph's but not SS's published numbers. The result is often a driver that needs a larger ported box because of the higher QES.

That's been my experience in the past, too, with what they now call their standard series drivers. And there is at times reports of variation in measured data for T/S parameters depending if it was measured at mW power levels or a watt or two.

[mkc]

Hi All,

From what I know SS data should be very consistent. As far as I remember, they use a current source instead of voltage source when doing T/S measurements. To my experience VAS is difficult to measure as measurement method might influence results.

Also, as Jon mention, signal level also influence.

I would like to state that I have tried to persuade a certain member of the danish forum: http://forum.speakerbuilder.dk to come and join this forum. He is a former SS senior engineer and is now working for SB Acoustic. He goes by the name usxx in the danish forum. He has deep knowledge about these matters. Try to make a post in the danish forum and see if you can smoke him out .

Best regards,
mkc

[DancesWithBeers]

Hi All,

From what I know SS data should be very consistent. As far as I remember, they use a current source instead of voltage source when doing T/S measurements. To my experience VAS is difficult to measure as measurement method might influence results.

Also, as Jon mention, signal level also influence.

I would like to state that I have tried to persuade a certain member of the danish forum: http://forum.speakerbuilder.dk to come and join this forum. He is a former SS senior engineer and is now working for SB Acoustic. He goes by the name usxx in the danish forum. He has deep knowledge about these matters. Try to make a post in the danish forum and see if you can smoke him out .

Best regards,
mkc

The deal is that when Jed or Zaph measure their drivers, model them, and then put them in enclosures, their tested results in the enclosures line up with their own measurements, but not the published measurements. If it were the case of SS being correct about their published specs, then actual results in enclosures would bear that out, but they do not.

-DWB

[Jed]

The deal is that when Jed or Zaph measure their drivers, model them, and then put them in enclosures, their tested results in the enclosures line up with their own measurements, but not the published measurements. If it were the case of SS being correct about their published specs, then actual results in enclosures would bear that out, but they do not.

-DWB

And sometimes it works with both set's of parameters. If that makes any sense.

For example... say a driver has a set of parameters from the manufacturer that model well in say 10L. Then, I measure the TS parameters and it suggests the optimum would be 20L, but modeled in 10L with the same tuning frequency or a bit lower everything is okay minus a slight bump in the response.

I like Scan Drivers, but when an Fs is off by some 10hz it's pretty obvious (after break-in). I agree Vas is difficult to get right.

A good practice is to model the specs given by the manufacturer and taken by yourself. Then compare the results. Find a box volume that will work well for both. Listen, and measure the box tuning with an impedance sweep. Then, measure the port output and woofer output to see if it matches the model. Usually it turns out just fine.

My comment that I wouldn't trust the mfg parameters really was to caution that the parameters really should be measured independently after breakin and evaluated later, rather than relying solely one set of data to see if it will work optimally in a target box size. Just thought I'd clarity that point.

Jed

[mkc]

Hi,

There was a discussion in the danish forum. http://forum.speakerbuilder.dk/forum...lukket-kabinet It's in danish, but I can do some translation if needed. Here a member also have different measurements than SS specifies. Especially Fs is higher.

The member do some experiments with various signal levels. According to USXX, using constant current instead of constant voltage gives a slightly higher Fs. Also he says that increasing level will lower Fs, but at a certain point it will start rising again, due to the fact that speaker-elements are non-linear.

Beside this the member(kappen) also discovers that clamping the unit firmly while doing measurements also have great influence on the results.

Please, understand that I do not have deep experience with this as I also measure using constant voltage. I just followed the thread and picked up some good information. However, I do intend to do some experiments sometime.

Sorry Jon for going off-topic on your thread. I look forward to follow the build and see some measurements.

Best regards,
Mogens

[JonMarsh]

This is a "good" type of off topic discussion as it is useful knowledge and discussion for everyone- this kind is appreciated, Mogens. :W

[Jed]

One last reference to illustrate my point:

By Vance Dickason in VC magazine.

"I commenced my anaylsis using the LinearX LMS analyzer and VIBox to produce voltage and admittance (current) curvers with the driver clamped to a rigid test fixture in free air at .3V, 1V, 3V, and 10V. As has become the protocol for Test Bench testing, I no longer use a single added mass measurement and instead used actual measured mass, but with the manufacturer's physically measured Mmd data......

Because most T/S data provided by OEM manufacturers is being produced using either a standard method or the LEAP 4 TSL model, I additionally produced using a LEAP 4 TSL model using the 1V free air curves."

Scan Speak 18WU/8741TOO Woofer

Image not available

For comparison purposes:

Image not available

Edit:

So I was curious to see how this would all model using the 3 sets of parameters in the Scan Speak model:

I first modeled the Manufacturer's specs and Unibox recommended a 30Liter box volume with a 39.5hz tuning frequency.

I then set that box volume as the reference to see how the T/S parameters measured by different methods would compare. I normalized the SPL levels to make it easy to compare the FR graphs.

The results are here:

Image not available

As you can see, the difference between the measured standard T/S measurements (or TSL) is closer to what is really occurring in the bass response as opposed to relying solely on the manufacturer's numbers.

I didn't bother modeling the Seas driver because the specs are so close to the manufacturer's numbers.

This is NOT a slight against Scan Speak. This is just one driver used to illustrate a point about the importance of measurements from multiple sources and to make a final judgment based on that data when venting a driver in a bass reflex enclosure.

Jed

[JonMarsh]

Hoo Rahh!

Scored a couple of Black and Decker portable clamping workstations today, which can also serve as Ardent assembly and finishing stations! Had to drive to Fremont to get these. Should also be useful in another upcoming project.

[chasw98]

And no pictures?

[JonMarsh]

Sorry- the pictures were in progress before I saw your post... just web pictures, no time to unpack, business travel tomorrow, and special project work for the Chancellor this afternoon...

Plastic fantastic, but folds up small, supports plenty of weight, adaptable to many tasks... too complex for the average maroon, so I think they're closing these out. That's why I bought two. :W

You, it's like for me, all the intriguing TV shows get cancelled because they don't generate market share with the masses.

[Hdale85]

Those are nifty looking.

[Evil Twin]

One could even say there's a certain imperial technology characteristic to the design... including a sufficient amount of black...

[bigg]

Sorry- the pictures were in progress before I saw your post... just web pictures, no time to unpack, business travel tomorrow, and special project work for the Chancellor this afternoon...

Plastic fantastic, but folds up small, supports plenty of weight, adaptable to many tasks... too complex for the average maroon, so I think they're closing these out. That's why I bought two. :W

You, it's like for me, all the intriguing TV shows get cancelled because they don't generate market share with the masses.

The average maroon? Oh...must be me. ;x(

[JoshK]

How do they work? I've been eyeing the "Jawhorse". But at $179, it seems a bit steep for what it is.

Did some work on my "bass bins" today. Biscuit jointer is coming in very handy! My first few attempts at box building proved that I am not so great at getting everything to line up mutually square. The biscuits plus edge jointing makes it a lot easier.

My bass bins are a net 4.8cuft, so not at all small (Lambda TD15Xs). Plus I did 1.5" birch lams on all sides except the front where it is 2.25" birch/mdf sandwich. They are going to take a forklift to move. At least the mostly BB box makes it somewhat lighter than all MDF.

[JonMarsh]

The B&D Clamping table comes with clamps that can be used built in; that wasn't the main draw for me- what I particularly liked was it's relatively low height (which many people might not like) so that you can put a tall cabinet on it and work on it at a more comfortable height than a normal table. It's rated for 400 lb- let's see if it handles 125 lb OK.

Yeah, the BB does help quite a bit- imagine what these bass bins weighed, which I constructed just with MDF. Live and learn... that was back in the 90's.

[JoshK]

My bins look like they might be close to the size of that base. Heavy!

What did you use to brace that speaker? I am trying Geddes' suggestion of 2" x 2" oak cross bracing.

[JonP]

Heh... now them's some bins!!

You needed to change out that WAF cover materiel on the furniture dollies to better match the carpet... then it would blend right in with the room. You'd hardly notice it... :W

[JonMarsh]

That room is not mine, but another members, and it has long since been re-done!

The thinnest panels were 1-1/2" thickness, and bracing was done in panels front to back parallel to top and bottom, with circles cut in them. Somewhere I've still got AutoCAD drawings for these. :W

This is what they looked like finished, with lamination work done.



Audax 13" and 15" pro woofers. Eton Midwoofers and Focal TC120 tweeter.

[---k---]

What did you use to brace that speaker? I am trying Geddes' suggestion of 2" x 2" oak cross bracing.

Jon recommends hardwood also (see his AudioXpress article).

I'm just curious if anyone has any sound logic of why it sounds better? The braces are just taking axial tension and compression. The panel deflections are so small, the stress on the braces would be very small. I have wouldn't expect their to be huge demand on them, and as such probably anything could work. The difference between the E value for the very best hardwood and an engineered product is at best 26%. While that is significant, I just keep thinking that the stresses in the braces are going to be so small.

EDIT:
That is what I was going to post. But, I couldn't help my geek self. IF you assume a 2"x2" x 24" Long brace, and assume a 1/64" movement at both ends in the panel, you will produce in the brace a tension stress of 1953 psi. Which is up there pretty high - above the limit for structural design.

Humm... Makes me re-think bracing.

[cjd]

This is why it's fun to have non-audio geeks around

What does that reduce to if you instead use standard holey panel bracing? A lot less I presume? I forget what kind of strength standard PVA glue has.

I also suspect that the change in panel resonance from having a brace in play may also reduce the overall chance that it will flex as much.

[---k---]

I started scribbling on a post-it note again. Those numbers can't be right. It is pretty simple math, Stress = E*Strain. Strain = Delta/Length. E for wood is ~1,500,000 psi. Humm. maybe they are My brain is fried, I'm burned out. Too much work. 1/64 at each end or 1/32 is a lot of movement for 24" length of wood. Maybe that was too much. It's linear. Divide it in 1/2 for 1/64 or so forth.

I once did a couple of Finite Element models of boxes, with just strips of wood glued to the inside face, and some with braces going across. The strips of wood glued to the face did work quite well. The window braces were excellent. Just a 2x2 cross bracing varied depending on how big the box, as one can imagine. The results were good from a comparison point of view, but without knowing real input values for loading, they couldn't be used for anything else. And, one has to remember that increasing the stiffness 100% of a panel that has 0.000" of deflection is a 0% gain.


I should put some models together again. They don't take long. Maybe if someone has a specific request. I'm not really interested in one of those discussions of how much bracing is enough.

[stinems]

Ryan, I haven't done the mechanics math and I don't doubt yours. And CJD, I don't know of any wood glue that would stand up to 1900psi either, so either the wood would fail, or the glued butt joint. But you got me thinking about what we use bracing for, and really, its for resonance control and not structural integrity, and I think CJD was on to something. Think about the box panels that a brace is, well, bracing. The panels have to be literally flexing while they're resonating--in other words, the panels themselves are undergoing oscillatory stresses (tension and compression about the neutral axes of the panel due to moment/shear stresses). This, combined with the brace's deformation, we have a statically indeterminate situation, so we must talk about relative E's (Young's modulus) and relative stresses.

Though I'll readily admit that I'm a bit of a charlatan when it comes to acoustical resonance of structures, my conjecture is that even if the brace isn't absolutely rigid (modulus less than infinity), it's still rigid enough to "convince" the box panels to resonate in such a way that a flexure node would appear at the braced point of the panel. This is because the panel will deflect (and this is the key to my point) in a moment/shear type stress relatively easier than deforming/straining the brace axially. In practice, say a panel has a fundamental resonance at 800hz and we brace in the dead center of the longest dimension of it. I would bet the new fundamental will be more like 1600hz and lower in magnitude. And since the braced point is a node at the fundamental, it's not moving and therefore, in theory, not causing mechanical deformation (and no oscillating tension/compression axial stress) on the brace.

Now that I've thought about it, perhaps thats a good rule for deciding when you have "enough" bracing. When the stress required to deflect an unbraced portion of a panel a given amount is greater (and it becomes greater as the unbraced portion becomes smaller) than the stress required to axially deform the next brace...don't bother with the extra brace because you may not be moving the resonance much higher. Granted, I haven't tested this idea, and I'm no expert. Forgive any imprecise use of terms in the above as well, I hope my point shows through.

Now, as for why oak or any other hardwood is "better" for bracing over other materials in terms of sound quality...I don't know. As long as wood has the same general moisture content, is free of knots/defects and load is applied in the same way with respect to the grain, most woods we would use have similar moduli. Has Geddes ever gathered empirical data that the oak brace works better than say, a pine or MDF brace? If he hasn't, then knowing him, he should be the first to admit its merely an opinion or preference of his.



Sam

[Bear]

Hoo Rahh!

Scored a couple of Black and Decker portable clamping workstations today, which can also serve as Ardent assembly and finishing stations! Had to drive to Fremont to get these. Should also be useful in another upcoming project.

Jon - These are only showing as being at Lowe's on B&D's website. Where'd you snag yours? I've got three weeks off, and getting the garage cleaned out and the shop set-up is towards the top of my to-do list.

Bill

[JonMarsh]

I had to drive to a Lowe's 20 miles away- that's where I got them! The three nearby were all out.

[Dennis H]

Ryan, your numbers for stretching/compressing the braces are WAY too high. Start with the internal air pressure and go from there.

peak pressure = (14.7 psi at sea level) * (1 way cone displacement) / (box volume)

An uber sub in a small sealed box might produce .3 psi at Xmax and most drivers and boxes would produce orders of magnitude less.

[stinems]

Dennis, I agree with your initial assessment from a pressurization standpoint. But let's take it one step further and realize even a .1psi pressure delta spread over a relatively small 100square inch panel is still 10lbs of distributed load on the box panel.

What about the resonance of a box panel just from being mechanically attached with the moving cone? That is, cone to surround/spider to frame to box panel. When that forcing function contains a large frequency component that corresponds to the natural resonance of the box panel, the panel WILL ring, even if it were effectively an infinite baffle (no bulk pressurization).

Maybe the magnitude of this excitation via cone momentum change is much lower than the effect of pressurization. I don't know, I'm just thinking out loud.

As it stands, I also agree with Dennis--my gut says 1/64th of an inch deflection is too large.

Sam

[cjd]

It gets interesting. I remember when someone thought sealed for a speaker box meant sealed - as in marine grade sealing. His sub was pushed almost completely to xmech limits with no signal applied, after use. Ran the numbers on air volume increase with temperature rise and found that was to be expected! It's amazing that a lossy "sealed" system is actually required.

[Hdale85]

Well generally won't you get enough leakage around the driver and through the cone and what not? When I worked in car audio we used to seal up our speaker boxes with fiberglass and resin. I always though it was sealed up about as far as we could take it. I don't know how much air could still seep in but I would think it was minimal.

[Dennis H]

But let's take it one step further and realize even a .1psi pressure delta spread over a relatively small 100square inch panel is still 10lbs of distributed load on the box panel.

Yup and 10 pounds isn't going to compress the length of a cross brace (oak, MDF, whatever) enough to measure outside a lab setting. Hardwood is nice to work with but it isn't really needed for its strength.

[---k---]

Ryan, your numbers for stretching/compressing the braces are WAY too high. Start with the internal air pressure and go from there.

peak pressure = (14.7 psi at sea level) * (1 way cone displacement) / (box volume)

An uber sub in a small sealed box might produce .3 psi at Xmax and most drivers and boxes would produce orders of magnitude less.

Yeah, I figured that I was stretching it too far. Just had no reference point at hand.

When work slows down come December / January, if someone asks, I'll put together some FEA models.

[savage25xtreme]

if Jon will ever release drawings of his enclosures I'm going to model them and drop them into ANSYS and see if I can find some harmonic resonance frequencies. no promises , but it sounds fun to me.

[Curly Woods]

if Jon will ever release drawings of his enclosures I'm going to model them and drop them into ANSYS and see if I can find some harmonic resonance frequencies. no promises , but it sounds fun to me.

I am sure the drawings are complete, as he has already veneered the damn things :-) Come on John, give us slow pokes a chances to get started on the cabinets :-) Even if its next year before you finalize the XO's, some of us will just be getting the money together to buy those friggin' Accutons :-)

[JoshK]

To me it didn't really matter so much about the oak versus other hardwoods. The pieces cost for a $3/lf for a few feet, so not a large expense. I figure it can't hurt, so no reason not to.

The only thing I would be worried about with less rigid woods is not the strength in the direction of bracing but orthogonal resonating like Galloping Gurdy.

[Face]

You can a dowel in the center of the brace and connect it to the next brace, top, bottom, etc...

[Evil Twin]

The Chancellor has seen your request...

I am sure the drawings are complete, as he has already veneered the damn things :-) Come on John, give us slow pokes a chances to get started on the cabinets :-) Even if its next year before you finalize the XO's, some of us will just be getting the money together to buy those friggin' Accutons :-)

The Chancellor has seen your request and is giving it his due consideration- as I have completed the data entry for the majority of the enclosure system (only the base is still in development), it should be possible to fulfill this request within the week.

There is another project which has been occupying my time and effort, which may also be completed for the RMAF meeting. No more on that until there are fruits of the endeavor to discuss.

[Curly Woods]

I was half kidding John, but this would be great! ;x( ;x( ;x( ;x(