W5-704 + ND20 can work together well?

[critofur]

Tang Band W5-704:
http://www.partsexpress.com/pe/pshowdetl.cfm?&Partnumber=275-035

and

Parts Express "Dayton" ND20F: (the new model)
http://www.partsexpress.com/pe/pshowdetl.cfm?&Partnumber=264-850

Anybody made crossovers for this pair?

How would you cross them and why would you choose to do it that way?

 

[dlneubec]

I have not used these two together, but do have a project that is underway using the W5-704 and the Dayton ND28F, the big brother to the ND20. Here is a photo of the project and a link to a thread on it over at HTGuide. The W5-704 can be crossed pretty high, so I could see the ND20 working. In my case, the ND28F is rear mounted in a 1/2" wavegide like Zaph suggests for the Vifa D26ND55.

The Duo's

 

[critofur]

Ah, yes, I had been looking at your designs on HTguide.

The ND20 is sooo cheap, and looking at the FR chart on Zaph's site the latest version's response looks so nice:

An externally hosted image should be here but it was not working when we last tested it.

The ND28 response looks fairly bumpy, no?:
http://partsexpress.com/pdf/ND28F-6_specsheet.pdf

I was thinking that the W5-704 is a little too large to be crossing it at ~3Khz or higher (poor [narrow] dispersion) and also the top end maybe a little rough? (though not nearly as nasty as metal cones...)

With the ND28 it prob. can cross just low enough to pair well with the W5-704 without having to go sharper than 2nd order slopes?

It's looking more and more to me like I can't get a very good design w/1st order slopes without using some kind of 2-3" upper midrange in between the tweeter and midrange.

I've got a bunch (about 15 or 20 or so) of Audax TM20J8 3/4" Neo soft dome tweeters which have a fairly similar response to the Dayton ND20s, that's one other reason I had an interest in them [ND20], I figured a design with one could easily be tweaked to work with the other.

[EDIT] Oh, and I wanted to participate in the discussion over on your HTGuide "Duo" thread, but, one of the more veteran members and a mod over there were real jerks, ended up getting me banned from posting even though we had moved our discussion to PMs!

[EDIT#2] I had been thinking of doing something like that w/the tweeter, but with a 3/4 roundover - I'd like to see how the two compare (3/4 vs 1/2). But what I'd really like to see is, if you use a 1" roundover except on a 3/4" thick board so that you'd be using the "top" 3/4 of the roundover resulting in a slightly shallower [slope] and broader waveguide.

How about a design like this for your tweeter housing above the woofer?:

 

[critofur]

Here's what I mean by using the "top" 3/4" of the 1" roundover.

I guess I would do it by temp. attaching a 1/4" "backer" board to the 3/4" material that we'd use for the front board/corian/etc.

I was thinking of this idea because most of us can't afford CNC machines and working with pre-made plastic wave guides is a bit of a pain, and the idea of using the wood baffle as the wave guide has been intriguing for a while...

I wonder if this method could also be used with the Dayton RS52? Has anyone taken the face plate off and know if it could be rear mounted on a baffle this way?

 

[critofur]

Hmm, how about this router bit for making wave guides?:

I don't know enough about wave guide angles, maybe the slanted portion is too steep and the beginning of the throat will be too much like a pipe?

[EDIT](I would make a template backer board and use the bit on 3/4", or, possibly, 1" thick (cut in two 1/2" steps, using two templates) baffles.

Maybe I should have made a new thread about cutting wave guides on baffles, instead of tacking on this one?

 

[augerpro]

I've just started using the W5-704 and I'm happy with it so far-considering the low price. It can do 2nd order, but just barely and it should really be crossed fairly low to keep the grunge around 4k Hz from causing problems. I'm crossing LR2 @2.5k Hz to a Vifa DX25 and almost nailed the crossover right off the bat. The preliminary version sounds very good, just had to drop the tweeter level a bit. The woofer itself does an amazing thing you don't find at this price often- it doesn't do anything annoying. Not the bottom line in clarity of course, but it doesn't do anything to really screw up the sound either, just matches the DX25 quite seamlessly.

 

[critofur]

Also remarkable, is that most expensive woofers do do annoying things, in one area or another. What interested me in the W5-704 in the first place was the combination of low price + lack of giant spikes of cone resonances on the top end.

Now, if we find some modifications we can make to the top end of the W5 smoother such as: removing the dust cap and putting a phase plug, "glue mods", embossing the cone, etc... maybe we can make the crossover even nicer?

Hmm, I wish I knew how MarkMCK did his cone breakup analysis, so I could have some idea how to figure out exactly where to, say, cut slits in the cone then use thin strips of that thin aluminum heating duct tape to seal them...

You said you had to drop the tweeter level a bit with the Vifa X25, do you think if you used a pair of W5's their level would match the tweeter without having to do any R padding?

I wanted to reply in your thread also, over on HTguide - I guess I'll have to make a new account over there since I don't have any un-ban hax.

 

[Moondog55]

Hi Critofur, I know very little about wave guides, only what I have gleaned from the forums.
I know that in his "Tweeter mish mash" JK has a small article on the Vifa neo tweeter and he use's ( from memory ) a 1/2 inch roundover.
Maybe this deserves a thread all of its own??

My thinking was that a bigger driver might need a different profile, flared starting steep and then getting shallower but all I can find in router catalogues are Roman ogees and roundover bits.
My thinking and it is probably a little naive, is that a biggish round over would allow a smallish woofer ( 6 to 10 inches ) to couple just a little better to the air and be a little more efficient.

Is the reason it's not done the paucity of suitable router bits??
As stated in the other thread the bits would have to be great big ones and then you have all the associated costs of big high horsepower routers.
Good question , needs serious discussion,
Regards Ted

 

[critofur]

Here's the bit from the picture in my earlier post, I meant to add the link, but my time to edit expired...

http://www.amazon.com/Freud-85-001-...f=sr_1_20?ie=UTF8&s=hi&qid=1208813120&sr=1-20

 

[Moondog55]

That looks like it should work, I have seen nothing like that here in Australia, would need a big router to drive it though, so smaalllll bites of the cherry, I used a 1/2 inch round over on my 10 inch woofer but project not finished yet so no SQ test done

 

[augerpro]

Also remarkable, is that most expensive woofers do do annoying things, in one area or another. What interested me in the W5-704 in the first place was the combination of low price + lack of giant spikes of cone resonances on the top end.

Now, if we find some modifications we can make to the top end of the W5 smoother such as: removing the dust cap and putting a phase plug, "glue mods", embossing the cone, etc... maybe we can make the crossover even nicer?

Hmm, I wish I knew how MarkMCK did his cone breakup analysis, so I could have some idea how to figure out exactly where to, say, cut slits in the cone then use thin strips of that thin aluminum heating duct tape to seal them...

That would certainly be interesting. DLR should grab one and start modding.

Originally posted by critofur You said you had to drop the tweeter level a bit with the Vifa X25, do you think if you used a pair of W5's their level would match the tweeter without having to do any R padding?

In my thread at htguide I posted the response for a quick MTM "sketch" I did. No padding was needed, so they would be quite sensitive. I think final voicing may require just a bit though. We'll see when I finish the cabinets.

 

[critofur]

Originally posted by augerpro "...Not the bottom line in clarity of course, "

Not trying to be negative or mean at all by this, but: are you certain that's not just your bias to the cheap price? Any scientific explanation as to why it would be audibly less "clear" than a $150 midbass driver? Aside a little roughness above 3Khz, I'm guessing distortion in the pass band may not be audibly significant.

Bias has remarkably powerful effect on the "sound" of speakers. It's mind boggling how much some people will pay for speaker cables, for instance, when, the net result of the effect that the wire has on the result is probably 1/10th of 1%

Let me finish by saying: I'd guess you're probably correct, there may be an audible improvement in the sound quality when using Scan Speak Revelator drivers, though I've not made the comparison myself. It'd just be interesting to see, if, in a blind listeng test, speakers built with the W5 as a midrange, and another woofer to support the low end (as the revelator has greater x-max and power handling) could sound as good or better.

 

[augerpro]

Oh I understand what you're saying. Subconscious bias is tough to avoid. I make every effort to recognize it and avoid it. One thing that helps is making comparisons between speakers that are level matched with an A/B switch. The differences are harder for the mind to color when you can switch back and forth instantly.

Now I've only had two sessions with these speaker so I don't want to make any concrete judgments, but during A/B testing versus teh Onix Rocket RS250 (Vifa XT25, HiVi M6N) there were a couple times where I thought the lower midrange wasn't quite as crisp and clear as the Rocket. Now I haven't went through a lot of music yet, and sometimes a speaker just sounds good (or not)with certain music. Sometimes the "good" speaker is suddenly not so good, and the "mediocre" speaker suddenly sounds the better of the two. And you never know when a little frequency response difference just happens to be pleasing (or not) to YOUR ear on the particular music you listen too. Not saying the driver is better in any objective sense, but it suits YOU.

I agree the difference between a good $25 woofer and a $150 one is not always obvious. Often the difference is better dynamic performance- more xmax, better optimized suspension and motor, even and low Le(x) so amplitude modulation of the passive crossover is minimized, better cooling of the coil, etc.

 

[critofur]

I got one of those "bowl" bits, I noticed the Bosch 85662M for less than 1/2 price ($29 w/free shipping) @ amazon.com:

http://www.amazon.com/85662M-3-Inch-Diameter-Carbide-Undermount/dp/B000GYE1FA

So, it's basically like a 1/2" round over bit, except, the bottom doesn't go to the full 90 degrees vertical; and, it extends down a little further in a straight line at a 19 degree off vertical axis angle.

I'm working on building a router table, I guess I could use this w/my handheld "hulk" of a router: Hitachi M12V (I'm so lucky I got this back when it was $150, now they're selling for $600?!? !!)

So... I tried to use Illustrator to make a nice little picture, I guess the first time you use that program in more than 10 years is not going to be quite as fast as a pencil and graph paper That project will have to wait for another day...

 

[14G-Dutch^]

Lol. Crit.

Still trying to figure a route to go with these? I haven't been in the forums for a while. 'You find anything interesting?

I have yet to get started also.

Jon

 

[dlneubec]

In my comparisons with the Duo-T (W5-704D/ND28F) and Duo-S (Scan15W/VifaD26NC55), the Duo-T sounds a bit warmer and more laid back, but quite good. The Duo-S is clearly stronger in the bass and is cleaner/more accurate and clearer sounding in the midrange and treble. I sounds more accurate and refined. Whether that bass depth and clarity is worth the $400+ difference in driver cost is a personal decision. One of the reasons for doing this design was to see how much difference drivers can make in the same box. That said, I'm can't switch between one pair and the other immediatly, since they share the same box. My comparisons have been a single Duo-S versus a single Duo-T.

This design, as most of the other omni's I've done is extremely sensitive to the tweeter baffle position, both height and front to back position. It effects both the woofers FR and the tweeter FR. The higher you crossover, the more sensitive it becomes. I can't predict what any change to the tweeter baffle would do to the design.

As far as the waveguides go, I would imagine a deeper waveguide for the tweeter will work, but you won't know the results until you make it and measure it. I don' think that the RS52 would benefit much, if at all, from a waveguide that shallow in depth and that small in width, simply due to the wavelengths is used to cover. I've done some measurements of the RS52 in a 8" waveguide, in another omni project and the change was certainly measurable, but that was probably about 3" deep and 8" wide.

I don't know whether the tapered throat you are proposing with the 19º angled bit with the 1/2" roundover will work well or not.

 

[dlneubec]

I forgot to mention this thought. One of the biggest benefits of a waveguide is that it boosts the low frequency output of a driver and when equalized in a crossover, the driver has stronger and lower distortion in it's low end performance. Well, if you make the waveguide/tweeter baffle too small, as in the semi-hemispherical one you showed above, I would think you would lose that benefit almost entirely. The diameter is so small that the drivers output on the low end would drop considerably rather than increase, since the longer wavelengths would wrap around the baffle. Off axis output should increase, since the small baffle would cause the tweeter to continue to play in full space and would transition to half-space at a much higher frequency than a larger baffle.

 

[critofur]

Hmm ~ I'd like to see what the effect of a small waveguide, like that on the semi-spherical housing, would be without the benefit of 2pi radiation.

I guess changing the size of my semi-sphere + waveguide enclosure design will necessarily shift the crossover frequency, and also the design would contribute to a steeper slope.

Perhaps housing the tweeter in a ~ 5" - 8" sphere could have multiple benefits and may work well in your Duo. You experimented some with the Duevel style waveguide/deflector thingies, have you done any experiments with spheres over the midrange like they do with their "Planets" speakers?

I had experimented briefly with some smaller elliptical/spherical deflectors some years ago trying to get an omni radiation pattern out of a regular dome tweeter without much success.

Since I went ahead and ordered that router bit, I will be trying at least some sorts of waveguide experiments within a week or two (I hope ).

Lol. Crit.

Still trying to figure a route to go with these? I haven't been in the forums for a while. 'You find anything interesting?

I have yet to get started also.

Jon

Interesting, yes, but... I don't want to compromise on the sound quality of the finished speaker because I'm using cheaper drivers. I want to use the Tang Band W5s but end up with a finished design that sounds at least as good as one using hundres of $ more of Scan Speak drivers.

I think that within a certain passband, the TBs should not be inferior to any other existing [conventional front firing cone] driver.

 

[Jay_WJ]

I've just started using the W5-704 and I'm happy with it so far-considering the low price. It can do 2nd order, but just barely and it should really be crossed fairly low to keep the grunge around 4k Hz from causing problems. I'm crossing LR2 @2.5k Hz to a Vifa DX25 and almost nailed the crossover right off the bat.

Hi Brandon,

How did you deal with the AC offset issue for your LR2 design with this driver combo? A ladder delay network?

I recently tried LR2 on my Usher 8945P/Peerless 810921 2-way on a flat baffle. What I found from this experiement is that without a delay network, phase mismatch is too large to obtain a proper listening axis. Another finding is that low Qts tweeters, such as the 810921 and the Seas 27 series, CAN be used in an LR2 design with a 2.25 kHz or higher xover point. I hear no distortions even above my normal listening level.

 

[augerpro]

As far as AC, well, I'd say I kept it in perspective considering overall performance and cost. The reverse null is broad and symmetric, but only about 12-15dB deep. And lobing is decent, 5-15 degrees downward from 1k-3k and teh lobes are large so deviation doesn't change amplitude much. Good 'nuff for the cost. Anymore absolutely perfect phase agreement isn't too important to me. If it's reasonably good on AND off axis I don't sweat it much. I think a smooth power response is more noticeable then going from a 18dB reverse null to a 50dB null- especially since you can generally only get that deep null at one position. Hell, Geddes uses 3rd order (phase quadrature egads!). I can't quite bring myself to do that yet...I would imagine teh bigger Usher woofer (deeper cone/motor)would be more of a problem. And were I to get "perfectionist" like I did with my Softies I would probably do something to account for the AC. But this speaker is $160/pair (drivers+XO) adn sounds very nice. They disappear as well as anythign I've built I think, with large 3D image. Sound is very coherent.

Thanks for the info on the other tweeters. I do need to use the Seas 27TBFC sometime. Probably the DXT version. I think for the design goals I'm happy with the choices. The DX25 is cheaper ($27 versus $42 for the Seas), has good distortion performance, and sounds good despite the top octave. I actually think the dip in the top is probably a good thing, mostly just sibilance and hiss up there anyway. So the performance/price just worked.