That is to say - you are nearly on-axis. Which ignores the reflected soundfield, where the midwoofer is beaming from about 2.5k, until the tweeter kicks in, no?
Isn't that reflected field the whole reason we care about directivity? Otherwise we'd just get it where we want on-axis and not worry about all this.
Sadly I don't think that's going to happen - I have a 30" Benq 4K and a 27" Dell set to portrait and I don't have any additional deskspace to bring things forward. I do have some Seas coax drivers (and some Kef ones for ceiling use) that I might try to make some over-monitor speakers with. But not immediately.
Why? Why do you have a room that looks like that?
I have accumulated drivers from ebay and new, and I do wonder what I'll do with the finished experiments.
I know they'll sell for less than the sum of the parts - and I bought all the bits for a RS225/B80 FAST from a guy on that basis recently - all still boxed and cabinets too. And crossover components.
I know in my mind - if not heart - that if I build the speakers I have drivers for, I'll basically have to scrap them. After all, who else will want boxes with drivers set up for external active crossovers? And what do DIY speakers sell for on eBay?
But really - I hope I never have a room that looks like that. Surely if we went to a hifi dealer and it looked like that we'd be 'WTF?'.
Maybe I'll try giving them to local schools for 'educational purposes'.
If the measurement of the speaker is done at the tweeter axis but the offset is done on the tweeter, is that even correct?
I believe its the Graham Chartwell LS3/5A, Xrk shared a pic with both side by side in the Vanguard speaker thread.
And you would be correct. ASR had measured the speaker here and the members of the forum called it "garbage". Xrk mentioned that the unit probably had off spec drivers and offered another unit to be re-measured but Amir ended up not doing it since its too much work for a not "common product" .
That's a conflating of two different things.
The designer meanders all over the place with his description. Way too much hand-waving regarding phase coherence. 🙂
It's a really simple design that should sound pretty decent.
Dave.
Coming from a recording/mixdown engineer, the biggest hurdle is always the environment…..if I can get a monitor thats reasonably flat and minimal distortion, I can make it work AS LONG as I don’t get any excessive early reflections. It’s these cases where a high level of directivity matters…..you actually want beaming to continue down as low in frequency as possible.
I currently have a modded pair of KEF Q150 on my bridge BUT they’re not behind my monitors….the baffles are fwd about 2”…….they’re good for adjusting compression and de essing in the mix as sibilance and narrow peaks jump put through them…….but if I wanted to sit and ‘listen’ for enjoyment?……..nope
Your use case is one of the few where I actually advocate for wideband/ full range applications……with some restrictions…….and that’s an open baffle/pseudo dipole or hybrid cardioid situation……you want that wide band driver to play from 250hz up to 15khz and beam that content right at you as even as possible.
In a box, a 3” or so wide band driver can do that from 4.5khz on up…..but what happens below that is back to omni pole. A 4” driver begins to act like a passband………the content from 10khz and up suffers…..at this point, a tweeter is neccessary and there goes your directivity
What you need is that little 3” to cancel its side and rear radiation as low as possible. You could do that actively with out of phase drivers on the side in a narrow passband and some clever crossover work……lots of experimenting and headache involved here. OR you could simply allow the rear wave of the driver ’leak’ out of the side of the box. The goal here is to use absorbant material in the box and over the vents that act like a 2nd order low pass filter…..this will take some experimenting as sims will only get you so far…..or not at all. There’s some folding inside the box required but it’s doable…..Dutch and Dutch makes it work…..Amphion makes it work. You’ll need a monopole bass solution from 250hz on down…..that’s the easy part…..back to back side firing 6” woofer in a sealed box behind your screen solves that issue wonderfully……your dimensional spread between speakers given your monitors works out up to 400hz given a 2nd order low pass on the woofers at 250 or so. I’d high pass the full range drivers 1st order as this will aid in extending that directivity……but power handling will suffer…..it’s a nearfield application….it shouldn’t be a design concern
All of the above is accomplished wonderfully by Mr Linkwitz with his LX mini and Pluto…..for a ready made solution, look there.
Last edited:
In fact, I am in a near-field listening situation due to the tiny size of my auditorium, which fortunately sounds without noticeable flaws, with reduced reflected field, somewhat like a studio cabin. The maximum off-axis that I have is 30°.
T
i think a 3” unit in a cardioid setup will have a hard time producing 250Hz on a realistic level.
Maybe an array.
The low end "load" of a passive cardioid is very similar to a dipole.
Otherwise an active cardioid with closed cabinets is also possible.
In that case, 250Hz is possible
I believe we’re talking nearfield desktop use so with some boundary gain and crossing over at 250?…..I think it will make it with a 1st order high pass
let's do that with a 2nd order HP, otherwise it's like turning up the volume accidentally once and it's bye bye 3 inch.
Connecting it with the LP is also no fun with just a 1st order HP filter.
It also kinda defeats the purpose of a cardioid when being so close to the wall with such a small speaker.
Might as well just go for a wall mount speaker instead.
Connecting it with the LP is also no fun with just a 1st order HP filter.
It also kinda defeats the purpose of a cardioid when being so close to the wall with such a small speaker.
Might as well just go for a wall mount speaker instead.
Last edited:
Have the curiosity to test the serial configuration 1st Order : you may experience very good surprises - like I did !
T
I'd like to return to my original question which was not actually 'Is this the best thing since sliced bread' but rather:
Its clearly a tradeoff between:
It might be that the trade works only in near-field (or borderline near field). But I find it interesting.
I don't think anyone suggesting that there are better designs has addressed that aspect - let alone whether any better designs use such affordable drivers and crossover topologies: but the absolute value for DIY was not my interest, rather the reasoning behind crossong quite high for a 5". Its arguably high for a 4" if you really care about radiation pattern - but who uses a 3" midrange?
Its clearly a tradeoff between:
- directivity and power radiation at crossover (and phase coherence I guess)
- single driver linear phase through midband
It might be that the trade works only in near-field (or borderline near field). But I find it interesting.
I don't think anyone suggesting that there are better designs has addressed that aspect - let alone whether any better designs use such affordable drivers and crossover topologies: but the absolute value for DIY was not my interest, rather the reasoning behind crossong quite high for a 5". Its arguably high for a 4" if you really care about radiation pattern - but who uses a 3" midrange?
Typical phase variations have been found not to be audible. An exception would be if the phase variation is due to some other problem that is audible. Another exception is if consequent group delay variations exceed the threshold of audibility. A further issue is that a poorly implemented crossover of higher order can hinder attempts at integration, and this is sometimes mistaken for phase issues.
A lot of Japanese constant wide directivity designs - think Coral, Yamaha, Diatone - from the 70s and 80s used midrange domes of about this size. Almost all "classic" continental European 3ways are built around 2 in or so soft dome midrange. Floyd Toole brings up 10-3-0.75 in 3-way as an example of good driver size proportion resulting almost automatically in smooth wide directivity across full spectrum.
Yes - I used to have some ProAc Studio 3 speakers - they had ATC 3" midranges and sounded very good indeed to my then young ears. Last time I saw a measurement of them, they were surprisingly narrow band and needed a lot of eq in crossover (plus the aggravation of installing from behind, which seems to need a removable baffle). Whether those are still the state of the art is another matter - but current catalogues are not exactly full of 3" drivers, whether dome or cone.
The industry strongly gravitates towards 2-way bookshelf in a 5+1 or 6.5+1 format; bookshelves and nearfield monitors, which are essentially the same thing, dominate larger speakers probably 10 to 1, so towers are built as "extended bookshelves". So there is little market for small midranges, despite their acoustic qualities.