For very modest output of around 85dB, it has been pointed out to me that some hifi tweeters can be crossed as low as 1khz or so. And this is true; they'll do so even on the manufacturer's FR charts, at 1watt. Though the lower distortion ones (in my area of the world) are quite costly (e.g. scanspeak D3004, though the SB26 looks reasonable for less) and of course they are being used outside the expected range.
So it has me wondering about compression drivers with small 1" throats, yet a much bigger diaphragm. Even some modest ones would do it 'very' easily, with a tiny fraction of a watt and without any need whatsoever for horn loading at a mere 85dB or so. I'm also happy to use DSP/EQ to flatten and smooth their frequency response. However this is for a compact speaker, there is no way that a big <1khz horn or waveguide could be used; I'd be looking at something much smaller - essentially round-overs, just big enough to avoid bad edge diffraction. I accept this doesn't control directivity, but listening near-field I'm mostly interested in the on-axis results (within reason).
Can anyone help me understand how a compression driver used like this would compare with a hifi tweeter?
Just to anticipate: Yes I realise there are AMTs and full-range/mid-tweeters etc. etc. that would cover the <1khz and higher range. But I want to ask the question if, at a low SPL, compression drivers are also a viable and possibly even better option?
Thanks,
Kev
So it has me wondering about compression drivers with small 1" throats, yet a much bigger diaphragm. Even some modest ones would do it 'very' easily, with a tiny fraction of a watt and without any need whatsoever for horn loading at a mere 85dB or so. I'm also happy to use DSP/EQ to flatten and smooth their frequency response. However this is for a compact speaker, there is no way that a big <1khz horn or waveguide could be used; I'd be looking at something much smaller - essentially round-overs, just big enough to avoid bad edge diffraction. I accept this doesn't control directivity, but listening near-field I'm mostly interested in the on-axis results (within reason).
Can anyone help me understand how a compression driver used like this would compare with a hifi tweeter?
Just to anticipate: Yes I realise there are AMTs and full-range/mid-tweeters etc. etc. that would cover the <1khz and higher range. But I want to ask the question if, at a low SPL, compression drivers are also a viable and possibly even better option?
Thanks,
Kev
Last edited:
Ignoring the horn you have about 0.5mm excursion before the dome hits the phase plug. This gives you about 117dB/1m from a 4" dome @ 1kHz. From a typical 1.7" dome for a higher end 1" throat compression driver your getting about 102dB/1m.
From my understanding of radiation impedance in the case of a very short and small mouthed horn the result will aproximate that of a piston.
From my understanding of radiation impedance in the case of a very short and small mouthed horn the result will aproximate that of a piston.
Thank you! I'm not very familiar with horns, or their theory, so that is very helpful. It makes sense that the near absence of an effective guide might result in a simple pistonic effect.
The maximum SPL you mentioned is approximately what I'd expected, which is reassuring. More than I need of course, but hopefully that might mean even lower distortion combined with a healthy amount of headroom. Compared to a hifi tweeter anyway, which would require much more excursion and be working close to its limits in the same situation.
In fact a compression driver might even allow a lower or shallower crossover, should I wish. Some look intended to go lower than most hifi tweeters, admittedly with a horn but also up to deafening levels. So in a low-level application I might even be able to use them somewhat like a woofer-assisted wide-range; at least, with a lower crossover the implications for wider driver spacing would offer greater flexibility of baffle/short-waveguide designs.
Thanks again,
Kev
The maximum SPL you mentioned is approximately what I'd expected, which is reassuring. More than I need of course, but hopefully that might mean even lower distortion combined with a healthy amount of headroom. Compared to a hifi tweeter anyway, which would require much more excursion and be working close to its limits in the same situation.
In fact a compression driver might even allow a lower or shallower crossover, should I wish. Some look intended to go lower than most hifi tweeters, admittedly with a horn but also up to deafening levels. So in a low-level application I might even be able to use them somewhat like a woofer-assisted wide-range; at least, with a lower crossover the implications for wider driver spacing would offer greater flexibility of baffle/short-waveguide designs.
Thanks again,
Kev
Hi, why not use a waveguide? You could use minimal one as well, like a donut, or very shallow and small in a box, but it would be better than no waveguide. The driver itself, without one, is waveguide as well, just not optimized!🙂 and the more you optimize, the better the sound.
I guess the OP actually meant soemething like a minimalist waveguide when he was talking about the use of a "roundover". This would mean mounting the driver from the backside of the baffle and using the hole in the baffle with roundover as a simple waveguide. I am convinced I have even seen this being done before (Westlake ??).
Regards
Charles
Regards
Charles
Maybe but I'd be more concerned that a too small waveguide could be worse than a simple radius. I'd have to know which to begin with.
Oh yes of course it would need to be sized to suit application, what ever it is. Naked driver above cabinet, or peek hole on a wall, or something in between.
well a radius is a waveguide it will still have an impact at greater than a few kHz. Here is a waveguide I designed for >3kHz, its very shallow: https://www.diyaudio.com/community/...ture-tweeter-array.388439/page-3#post-7300287
I recently used JLPCB for some small 3D prints at very low cost with exccelent quality so perhaps using a 3D printing service you could make a more sophisticated small waveguide?
I recently used JLPCB for some small 3D prints at very low cost with exccelent quality so perhaps using a 3D printing service you could make a more sophisticated small waveguide?
Most decent quality 1” compression drivers can easily play to 1khz in a home hifi environment for those that actually value their hearing. Modern horn geometry has come a long way in a just a few years but available baffle space is still a prerequisite sadly…..no sidestepping physics here. Put a well designed system with a proper dome side by side with a compression diver system in a modest size room and run it through the gamut of polite singer/songwriter to heavy metal at sane levels and most listeners will prefer the dome. I believe it’s an as yet unidentified and unintended consequence of the power compression that occurs within that 1” throat……there’s harmonic content clarity in complex music that measurements won’t reveal.
what material you choose?
Have you tried to print large horn with this service? I want to print a large horn for DCX464 but not sure if they can bear the weight of it...
Well, there would be something that might incidentally cause some guidance, but I wouldn't dignify it with the term waveguide. Certainly nothing like the size that (e.g) Geddes uses for his compression drivers. More just a rounded doughnut type of thing, or a contoured baffle, to manage edge diffraction. I actually don't think a proper waveguide would be a major advantage for my usage, neither increased SPL nor narrowed directivity are needed. But there isn't the space for a decent one, in any case.
Mayhem13, yes I've heard similar. Although also the opposite too (tweeters sounding strained, unless crossed high). Either way though, I'm personally happy enough with tweeters for normal hifi use. The extra complexity of compression drivers is only of interest here for their larger diaphragms and lower frequency potential. By comparison, tweeters being pushed too far could possibly be worse.
Thanks, Kev
Mayhem13, yes I've heard similar. Although also the opposite too (tweeters sounding strained, unless crossed high). Either way though, I'm personally happy enough with tweeters for normal hifi use. The extra complexity of compression drivers is only of interest here for their larger diaphragms and lower frequency potential. By comparison, tweeters being pushed too far could possibly be worse.
Thanks, Kev
I've decided to try it. So I've got a pair of Faital Pro HF108 compression drivers on the way. This isn't a risky purchase; even if I don't like them for the purposes of this thread, I know from several reports that they'd work very well for the bigger waveguided room-sized speaker project that I'm intending in the future.
Anyway, back to the present. They seem well regarded, will go low enough (at my modest SPL) and also have a useful 31-degrees throat angle. That should give a head-start in transitioning smoothly between their exit and the baffle. At the moment I'm envisaging a kind of jukebox shape for the cabinet, with a rounded doughnut affair in the curved end for the compression driver, transitioning to a woofer in the straight bit.
Anyway, back to the present. They seem well regarded, will go low enough (at my modest SPL) and also have a useful 31-degrees throat angle. That should give a head-start in transitioning smoothly between their exit and the baffle. At the moment I'm envisaging a kind of jukebox shape for the cabinet, with a rounded doughnut affair in the curved end for the compression driver, transitioning to a woofer in the straight bit.
Thinking things through a bit more, turning a round doughnut into more of a juke-box shape will allow a nice little 2-way. Something like the following:
Just showing the outline of the front of the baffle (in green). I've not yet thought about the inside, or shape of the back of the cabinet.
I found a mention by Linkwitz that round-overs start to become effective when their radius is about 1/8th of the wavelength. These are about 1.5" or 38mm radius, which should mean they start being effective by about 1100hz. Dr Geddes feels that the region about 1khz or lower is better for compromises than above it, so I'm happy enough with 1.1khz. They make an overall cabinet width of about 215mm or 8.5" which seems about right to me, for a desktop speaker.
Driver spacing here is about 125mm or 5" centre-to-centre. That would be within a quarter of a wavelength if crossed around 700hz. Likely I'll cross a bit higher but it shows that things are at least reasonable. Here I've used an example of a 5FE120 for the woofer; it is about the right size and shape, has a narrow frame/bezel and will do okay in a small sealed enclosure. I could go for a smaller woofer, as this speaker is intended for use with subwoofers, but actually smaller doesn't gain very much and I'm interested in it having sufficient dynamics and impact.
So I'll need to do a lot of testing first (no idea how a compression driver will work low in a doughnut) but that is the rough concept, at the moment.
Just showing the outline of the front of the baffle (in green). I've not yet thought about the inside, or shape of the back of the cabinet.
I found a mention by Linkwitz that round-overs start to become effective when their radius is about 1/8th of the wavelength. These are about 1.5" or 38mm radius, which should mean they start being effective by about 1100hz. Dr Geddes feels that the region about 1khz or lower is better for compromises than above it, so I'm happy enough with 1.1khz. They make an overall cabinet width of about 215mm or 8.5" which seems about right to me, for a desktop speaker.
Driver spacing here is about 125mm or 5" centre-to-centre. That would be within a quarter of a wavelength if crossed around 700hz. Likely I'll cross a bit higher but it shows that things are at least reasonable. Here I've used an example of a 5FE120 for the woofer; it is about the right size and shape, has a narrow frame/bezel and will do okay in a small sealed enclosure. I could go for a smaller woofer, as this speaker is intended for use with subwoofers, but actually smaller doesn't gain very much and I'm interested in it having sufficient dynamics and impact.
So I'll need to do a lot of testing first (no idea how a compression driver will work low in a doughnut) but that is the rough concept, at the moment.
Last edited:
Dougnut works fine for
If you put the woofer behind the waveguide and use just small hole for it's exit (acoustic low pass filter), it would be close to Multiple Entry Horn -concept.
Definitely use roundovers on the high and mid transducers, to reduce effects of edge diffraction on listening window.
Donut, circuilar profile on the waveguide, works fine, although it beams some. You can optimize bit further if you wish, see the ath4 thread.
If you put the woofer behind the waveguide and use just small hole for it's exit (acoustic low pass filter), it would be close to Multiple Entry Horn -concept.
Definitely use roundovers on the high and mid transducers, to reduce effects of edge diffraction on listening window.
It is interesting to hear your thoughts, thanks. I had wondered about making it somewhat like a small MEH (albeit without an actual horn) or maybe like a Cabasse spherical affair, but it seemed to be getting over-exotic. However, maybe not; it would only be an oddly shaped bandpass enclosure, so could still be modelled easily enough at a basic level.
I'd not gone down the waveguide route because a proper effective one would be huge, and I don't particularly need to restrict the directivity in any case. However if my anti-diffraction doughnut would actually cause more beaming then it is something I'll need to look into, thanks again. Looks like 'another' thing I'm going to have to learn about!
Cheers,
Kev
I'd not gone down the waveguide route because a proper effective one would be huge, and I don't particularly need to restrict the directivity in any case. However if my anti-diffraction doughnut would actually cause more beaming then it is something I'll need to look into, thanks again. Looks like 'another' thing I'm going to have to learn about!
Cheers,
Kev
In researching this, I've seen a few projects using compression drivers reversed; just like you describe. It isn't what I want for this particular project, a small throat but bigger diaphragm is an advantage here. But it does show how capable and versatile compression drivers can be. I've only heard them in professional sound systems, and have not until now used them for DIY HiFi, but they clearly have a number of interesting possibilities.
Ha, yes. Well that is closer to what I'd want for my next big room speakers, though with quite different wave-guides. And less hair.
Anyway, I've been testing one of the HF108 drivers 'very' crudely just to see roughly what it can do. I'm quite encouraged; with no waveguide (or even a baffle) it'll go easily loud enough for my near-field situation. Surprisingly, I can also give it down to below 500hz with my 50W amplifier turned pretty high, without any worrying diaphragm sounds.
Of course, the frequency response is rubbish used like this. Though even with crude manual EQ it is surprisingly listenable. But any form of sound quality tests are going to need a reasonable approximation of the baffle that I'd like to use, and lots of measured FR correction.
Anyway, I've been testing one of the HF108 drivers 'very' crudely just to see roughly what it can do. I'm quite encouraged; with no waveguide (or even a baffle) it'll go easily loud enough for my near-field situation. Surprisingly, I can also give it down to below 500hz with my 50W amplifier turned pretty high, without any worrying diaphragm sounds.
Of course, the frequency response is rubbish used like this. Though even with crude manual EQ it is surprisingly listenable. But any form of sound quality tests are going to need a reasonable approximation of the baffle that I'd like to use, and lots of measured FR correction.