These differences could be explained by differences in the baffle design and mounting of the driver, as that's the frequency range critically affected by whether the driver is flush mounted, recessed, raised, how thick the panel is (cavity resonance behind the faceplate due to an exccessively thick panel) how close it is to the edge of the baffle etc. (Or indeed whether it has any significant baffle)
That's why I try to go by large infinte baffle mesurements when comparing drivers, but of course I use the response in the box/baffle when it comes to equalizing the driver.
It will be interesting to see how mine measure when I get them - I don't have an infinite baffle at the moment but might be able to make one out of some 6mm 2.4m x 1.2m plywood I have with a little bit of bracing around the edge.
I can take a measurement in the same box my Coral drivers are mounted in now which would give a direct comparison to those drivers, but wouldn't give a measurement that could directly compare with anyone elses..
Unless you put it on a really wide baffle, add some BSC or put it in a horn, the lower midrange is going to sound lean. The Coral's are the same without any EQ. No surprise there.
Not a very fair test to be honest, on an open baffle with (presumably) no EQ the low Q 1772 is going to sound very thin on bass. It was never designed for that use. Most people are going to pick the one that gives more bass in that situation.
A better test might have been to electronically roll them both off below about 250Hz so that neither driver had any bass, then compare the quality of the midrange and treble.
I bet.
Yes (!!) a large baffle is the ticket
It may seem counter-intuitive, but I can tell you from a recent experience that a small/narrow baffle which seems to be the popular choice (for decor) these days, results in the well known phenomenon as baffle step loss.
I have taken my {beastmode} large horn system down for a while, mostly to re-work the mounting baffles, and am listening to some Altec ER-10's for mid range.
The "box" they are mounted in is only 11 inches wide. That just ain't gonna get it. 11/12= .916 ~~~ 1130/.916= 1232Hz, and it's definitely audible.
The mathematics of baffle step loss dictates the baffle should be as wide as one can tolerate, unless of course, the choice is to use a compensation circuit.
It may seem counter-intuitive, but I can tell you from a recent experience that a small/narrow baffle which seems to be the popular choice (for decor) these days, results in the well known phenomenon as baffle step loss.
I have taken my {beastmode} large horn system down for a while, mostly to re-work the mounting baffles, and am listening to some Altec ER-10's for mid range.
The "box" they are mounted in is only 11 inches wide. That just ain't gonna get it. 11/12= .916 ~~~ 1130/.916= 1232Hz, and it's definitely audible.
The mathematics of baffle step loss dictates the baffle should be as wide as one can tolerate, unless of course, the choice is to use a compensation circuit.
Not sure if there is a hint of sarcasm in there or not
but baffle step loss is absolutely real and needs to be accounted for. It's just a change in directivity with frequency.I'm absolutely on the side of wide baffles, always have been. However unless they're REALLY wide or the speaker is right in the room corner (or it's a horn etc) then you still need baffle step correction, the only difference is at what frequency you have to apply it.
For my next build I'm planning for the cabinet to be 50cm wide and aprox 90cm tall (with a tweeter pod on top) so by todays standards that's a fairly wide baffle. 115/0.5 = 230Hz approx for the baffle step frequency - actually below the midrange and very close to my intended 250Hz crossover frequency to the woofer. That way no explicit baffle step correction contour needs to be added in the individual pass bands - it can be achieved just by adjusting the attenuation of the midrange/tweeter relative to the woofer.
By the way your calculation above is way out - a first approximation of the 3dB rollover point on a rectangular baffle is 115/width in metres. So in your 11 inch example the rollover point would be 412 Hz not 1232.
And yes, very audible if not equalised.
Last edited:
So the 1772's arrived today, and I thought I would give my first impressions.
This is not a review as such as I don't have a proper cabinet or even a flat baffle to test them on at the moment, so just some observations about their construction and build quality, along with a bit of quick listening with them on their backs on the floor pointing up nestled within a "cushion baffle".
Obviously in this condition they won't produce any bass and will only go down to around 300Hz since it's basically a very lossy open baffle. However, I'm only listening to the midrange and treble to evaluate how neutral they are and whether they have any obnoxious (and potentially incurable) upper midrange or treble resonances that are common in many full range drivers, and I have listened to several other full range drivers in a similar fashion in the past as a rough and ready test so sort of a fair comparison.
In short, in the couple of hours testing I did I'm extremely impressed by the midrange and treble quality of these drivers. I think it would be fair to say in terms of midrange and treble smoothness they are probably the best 8" full range drivers I've ever heard, and I don't say that lightly. Perhaps even the best full range driver I've heard, full stop. The frequency response measurements while good, just don't do them justice.
Yes I still love my vintage Coral drivers, and for their age (designed in the 70's using paper and pencil without modern CAD and measurement equipment) they are still wonderful but these are better. No question about it, at least for midrange and treble.
And that's a good thing, because I've been looking for many years for a modern driver which follows the same basic design principles as the classic full range Coral's but which are affordable, in current production, and don't suffer from excruciating whizzer cone resonances common to many other other full range drivers.
And I've tested some duds in the past. Particularly disapointing was a pair of Fostex FE207 which I bought in the late 2000's before the W8-1772 existed, which was well regarded at the time but which suffers from a severe and inexcusable resonance around 2.5Khz which totally ruins the driver, IMHO. A resonance that can't be notched out or ignored. (It's not a symetrical peak, it's a razor sharp high Q peak and notch side by side due to some sort of phase cancellation effect between the cones I think)
Within minutes of listening to the FE207's all those years ago my heart sank and I knew I had wasted my money. I soldiered on with them for a few weeks but eventually gave up on them and went back to the Coral's. Also disapointing with the FE207's was the whizzer cone was ineffective at improving upper midrange dispersion compared to the Coral's.
That bad experience is partly why I've held off for so long in taking a chance on these. But I'm glad I finally took the plunge.
I tried a variety of music listening for any trace of upper midrange or treble harshness and couldn't hear any at all. Clean, clear, smooth. Not even any sibilence. No random "cone breakup hash/noise" audible either.
Granted, it's not completely neutral, at least on my cushion baffle....
There is some emphasis in the presence region (later revealed in measurements) and the on axis treble is a bit hot, as widely documented, but not unreasonably so. Both easy to tweak with a little bit of EQ.I was particularly impressed by how far off axis the treble response goes. It's remarkably wide dispersion for an 8" whizzer cone driver with usable treble right out past 20 degrees, and there is a decent amount of treble in the room reverberant field.
By comparison the FE207E has very poor off axis treble more akin to an 8" midbass driver, the Coral Flat 8 have considerably better off axis treble than the FE207E but these are on a whole other level when it comes to off axis treble dispersion. Quite impressive.
I then did some very rough measurements with the microphone pointing down at the speaker and noted a few things.
There is indeed no "surround" dip at 1Khz at all, agreeing with other measurements posted in this thread. Every other 8" full range driver I've measured has had a distinctive surround dip near 1Khz - some worse than others but always there.
This is the first one I've measured with no dip at this frequency. That is quite some engineering as it would take precise mechanical impedance matching between the cone material and surround with just the right damping in the surround to achieve this. Well done whoever designed this driver, surround dip is a hard problem to solve!
There is a peak of about 3dB at 2.2Khz with approximate 1/4 octave width which explains the slightly forward presence. I tried notching this back a bit with active EQ and sure enough it made the driver more neutral in tonal balance and a little bit more laid back. To be honest it sounded good with or without the notch though - just a bit more forward or laid back respectively. In neither configuration (notch or not) did it sound harsh or dull.
On axis a peak at about 3.5Khz is evident and this is shown in some other people's measurements, however this peak is not there off axis, and also not there when I change the baffling around the driver. Likewise there is a significant dip/hole around 4.3Khz on axis which disappears off axis or if I change the baffling. However the 2.2Khz peak is always there roughly the same at different angles and with different baffling.
From that my initial conclusion (until I can do properly controlled measurements) is the peak at 2.2Khz is a genuine (but fairly benign and damped) cone resonance which could perhaps do with notching, while the peak at 3.5Khz and dip at 4.3Khz I initially measured is just a diffraction artefact due to the driver not being properly flush mounted on a low diffraction baffle.
An unbaffled driver is a bit of a worst case scenario as the frame of the driver behaves as a circular baffle which is much worse than a rectangular baffle since every ray leaving the middle of the driver reaches the edge at the same time, stacking all the standing waves all around the perimeter at the same frequencies... The worst measurement result was achieved with the driver sitting on it's back with the frame in mid air, and the response with cushions acting as a partial baffle was much flatter. At some point in the future I will get a chance to measure it on a proper baffle.
In a second post I'll talk a little bit about the physical design of the driver and add some closing thoughts, as this post is getting a bit long now.
Last edited:
1772 follow-up
I'm please that you like the 1772's. They are indeed something special, especially at their price compared to the big-boy names.
When you achieve an air-load on the back of the cone, the response aberrations will be changed.
Mine exhibit a very slight dip(recessed sound) at 1.0 and 1.6K
To get the VERY BEST mid-range performance from them, please consider a wave-guide, or, a small Bruce Edgar style tractrix horn.
I'm please that you like the 1772's. They are indeed something special, especially at their price compared to the big-boy names.
When you achieve an air-load on the back of the cone, the response aberrations will be changed.
Mine exhibit a very slight dip(recessed sound) at 1.0 and 1.6K
To get the VERY BEST mid-range performance from them, please consider a wave-guide, or, a small Bruce Edgar style tractrix horn.
Regarding the physical design and build quality, I'm also very pleased overall.
Negative points first - one driver has a tiny bit of damage to the edge of the whizzer cone - barely visible but it is there.
My Coral drivers have had far worse damage to the whizzer cones accumulate over the years and still work OK so it's not something I'm going to worry about, as it's only visible on a close inspection and is a bit of a hassle to try to claim warranty on.
One driver also had a small stray bit of hard black something (plastic ? glue ? not sure) clinging to the back of the surround just inside the edge of the frame which I noticed and plucked out.
No harm done and not a big deal, however slightly disapointing. I always inspect new drivers carefully when I receive them though.
As an aside I received my Aurum Cantus G2's last week and close inspection of them right out of the box showed a foreign object (a speck of plastic or foam) stuck in the gap between the ribbon and magnetic gap on one of them - behind both the front grill and the protective adhesive tape applied over the grill for shipping. The speck was also bigger than the gaps in the grill so must have been there from the factory. I was able to remove the front plate and remove the spec (it fell out with an upside down shake so fortunately wasn't magnetic!) but a little disapointing there too. Inspect your drivers when they arrive!
Now onto the good things, of which there are many. The drivers shipped with a protective plastic cap over the front to provide additional protection beyond the cardboard box, this is welcome as whizzer cone drivers are fairly delicate and easily damaged in my experience, and significant damage to whizzer cones (creasing/crumpling) can cause significant response changes.
The cast frame is much better in design than the pressed steel frame of the Coral's and is also designed with a very simple, easy to flush mount round frame edge.
The frame on the Coral's is stepped and not round at the outside and is designed for behind the panel mounting (very old fashioned) and is nearly impossible to flush mount, so a frame which can achieve a perfect flush mount from the outside with a very simple round rebate is extremely welcome. Furthermore the surround termination is itself perfectly flush mount with the frame, all of which along with the reverse roll surround helps minimise diffraction at the edge of the driver.
The cone design is very interesting and diverges somewhat from the Coral's.
The main cone on the Coral's is straight conical, but with three "stiffening rings" stamped into the outside to improve control of radial and circumferential modes.
It's not apparent in the photos so I wasn't expecting it, but the main cone on the 1772 is actually curvilinear like the whizzer cone, and is curved quite dramatically in fact. This is a good thing for control of bending modes as the stiffness to weight ratio of a curvilinear cone is much better than a plain conical cone which is weak and floppy in one axis and particularly prone to formation of circumferential modes. (which generate "cone breakup hash" as they chaotically travel around the circumference of the surround and interact with radial modes)
Compare this to the main cone on an FE207E which is straight conical and has no stiffening rings either. Curvilinear is best for high frequencies, straight conical with stiffening rings is second best and straight conical with no stiffening rings is worst.
The whizzer cone is also curvilinear and quite similar to the Coral's, however the voice coil diameter is a lot larger (38mm vs 25mm) and the curvature of the main cone quite a lot different, as a result the large gap between the two cones on the Coral at the base isn't there with the two cones somewhat closer together. Too much gap can result in phase cancellation between the two cones, so this reduced gap is probably a good idea.
The whizzer cones have another trick up their sleeve - if you look closely at the back side of them there is a strip about 5mm from the tip of the cone that is darker in colour - this is almost ceratinly additional damping doping added near the free edge of the cone to help control standing waves on the whizzer cone. Good idea! I haven't seen that before. Not on the Coral's, not on the Fostex or others. Once again the FE207E falls particularly short here with a conical whizzer cone which is inferior to a curvilinear type even when the edge is creased.
The voice coil former is very long on the 1772 - much longer than any other similar driver I've seen, with the end visibly protruding in the gap between the cone and phase plug.
No doubt on purpose as the edge of the voice coil will act as a ring radiator at treble frequencies and is probably responsible together with the phase plug for the impressive treble response and treble dispersion. The other end of the voice coil former goes right down into the magnet structure. Voice coil former looks metalic, so might be aluminium ?
By comparison the voice coil former on the Coral drivers is traditional varnished paper, and much shorter with the magnetic gap much closer to the base of the cone, and is also an overhung design vs the underhung design of the 1772.
One area where it isn't quite is good is in suspension travel - although both have an Xmax (voice coil gap, voice coil height difference) of about 3mm, the Coral has considerably more Xsus, more on the order of 6mm. It's quite apparent when gently moving the cones back and forth that the Coral has a lot more Xsus.
However the 1772 is not designed for high excursion use cases, the 1808 has 5mm Xmax and presumably more Xsus to go along with it, so I'd say the 1808 would be a closer comparison with the Coral driver for bass.
I've bought these drivers to only use them down to about 250Hz, so the reduced suspension travel isn't an issue, and I'd rather have the good surround damping and flat response through 1Khz.
Although I couldn't do an accurate comparative measurement without a proper test baffle, pair matching of the two drivers seems to be very good, with just a minor difference in the treble response of the two drivers.
The phase plug is apparently aluminium and I was afraid it might be hollow but a gentle rap on it seems to suggest it is a solid aluminium bullet ? Whatever it is, it looks very nice and seems to do the job.
All in all I can't really find anything to fault in either the design or manufacturing of the driver. This is an extremely well designed and seemingly well manufactured driver, at a reasonable price.
While it sounds very good as is, a small amount of EQ to reduce the peak at 2.2Khz and a gentle rolling off of the treble would both probably contribute to a more neutral ultimately more pleasing response, although this is likely to be very dependent on the cabinet design.
Very happy with these drivers and I can't wait to get started on the cabinets and getting some proper measurements done.
Negative points first - one driver has a tiny bit of damage to the edge of the whizzer cone - barely visible but it is there.
My Coral drivers have had far worse damage to the whizzer cones accumulate over the years and still work OK so it's not something I'm going to worry about, as it's only visible on a close inspection and is a bit of a hassle to try to claim warranty on.
One driver also had a small stray bit of hard black something (plastic ? glue ? not sure) clinging to the back of the surround just inside the edge of the frame which I noticed and plucked out.
No harm done and not a big deal, however slightly disapointing. I always inspect new drivers carefully when I receive them though.
As an aside I received my Aurum Cantus G2's last week and close inspection of them right out of the box showed a foreign object (a speck of plastic or foam) stuck in the gap between the ribbon and magnetic gap on one of them - behind both the front grill and the protective adhesive tape applied over the grill for shipping. The speck was also bigger than the gaps in the grill so must have been there from the factory. I was able to remove the front plate and remove the spec (it fell out with an upside down shake so fortunately wasn't magnetic!) but a little disapointing there too. Inspect your drivers when they arrive!
Now onto the good things, of which there are many. The drivers shipped with a protective plastic cap over the front to provide additional protection beyond the cardboard box, this is welcome as whizzer cone drivers are fairly delicate and easily damaged in my experience, and significant damage to whizzer cones (creasing/crumpling) can cause significant response changes.
The cast frame is much better in design than the pressed steel frame of the Coral's and is also designed with a very simple, easy to flush mount round frame edge.
The frame on the Coral's is stepped and not round at the outside and is designed for behind the panel mounting (very old fashioned) and is nearly impossible to flush mount, so a frame which can achieve a perfect flush mount from the outside with a very simple round rebate is extremely welcome. Furthermore the surround termination is itself perfectly flush mount with the frame, all of which along with the reverse roll surround helps minimise diffraction at the edge of the driver.
The cone design is very interesting and diverges somewhat from the Coral's.
The main cone on the Coral's is straight conical, but with three "stiffening rings" stamped into the outside to improve control of radial and circumferential modes.
It's not apparent in the photos so I wasn't expecting it, but the main cone on the 1772 is actually curvilinear like the whizzer cone, and is curved quite dramatically in fact. This is a good thing for control of bending modes as the stiffness to weight ratio of a curvilinear cone is much better than a plain conical cone which is weak and floppy in one axis and particularly prone to formation of circumferential modes. (which generate "cone breakup hash" as they chaotically travel around the circumference of the surround and interact with radial modes)
Compare this to the main cone on an FE207E which is straight conical and has no stiffening rings either. Curvilinear is best for high frequencies, straight conical with stiffening rings is second best and straight conical with no stiffening rings is worst.
The whizzer cone is also curvilinear and quite similar to the Coral's, however the voice coil diameter is a lot larger (38mm vs 25mm) and the curvature of the main cone quite a lot different, as a result the large gap between the two cones on the Coral at the base isn't there with the two cones somewhat closer together. Too much gap can result in phase cancellation between the two cones, so this reduced gap is probably a good idea.
The whizzer cones have another trick up their sleeve - if you look closely at the back side of them there is a strip about 5mm from the tip of the cone that is darker in colour - this is almost ceratinly additional damping doping added near the free edge of the cone to help control standing waves on the whizzer cone. Good idea! I haven't seen that before. Not on the Coral's, not on the Fostex or others. Once again the FE207E falls particularly short here with a conical whizzer cone which is inferior to a curvilinear type even when the edge is creased.
The voice coil former is very long on the 1772 - much longer than any other similar driver I've seen, with the end visibly protruding in the gap between the cone and phase plug.
No doubt on purpose as the edge of the voice coil will act as a ring radiator at treble frequencies and is probably responsible together with the phase plug for the impressive treble response and treble dispersion. The other end of the voice coil former goes right down into the magnet structure. Voice coil former looks metalic, so might be aluminium ?
By comparison the voice coil former on the Coral drivers is traditional varnished paper, and much shorter with the magnetic gap much closer to the base of the cone, and is also an overhung design vs the underhung design of the 1772.
One area where it isn't quite is good is in suspension travel - although both have an Xmax (voice coil gap, voice coil height difference) of about 3mm, the Coral has considerably more Xsus, more on the order of 6mm. It's quite apparent when gently moving the cones back and forth that the Coral has a lot more Xsus.
However the 1772 is not designed for high excursion use cases, the 1808 has 5mm Xmax and presumably more Xsus to go along with it, so I'd say the 1808 would be a closer comparison with the Coral driver for bass.
I've bought these drivers to only use them down to about 250Hz, so the reduced suspension travel isn't an issue, and I'd rather have the good surround damping and flat response through 1Khz.
Although I couldn't do an accurate comparative measurement without a proper test baffle, pair matching of the two drivers seems to be very good, with just a minor difference in the treble response of the two drivers.
The phase plug is apparently aluminium and I was afraid it might be hollow but a gentle rap on it seems to suggest it is a solid aluminium bullet ? Whatever it is, it looks very nice and seems to do the job.
All in all I can't really find anything to fault in either the design or manufacturing of the driver. This is an extremely well designed and seemingly well manufactured driver, at a reasonable price.
While it sounds very good as is, a small amount of EQ to reduce the peak at 2.2Khz and a gentle rolling off of the treble would both probably contribute to a more neutral ultimately more pleasing response, although this is likely to be very dependent on the cabinet design.
Very happy with these drivers and I can't wait to get started on the cabinets and getting some proper measurements done.
Are yours properly flush mounted ? A dip at 1.6Khz will be diffraction from the edge of the frame. There is a deep dip at 1.6Khz on mine in the measurement I took with the frame hanging in free air, (maximum diffraction) however there is no dip at all at this frequency in the measurements I took with the "cushion" baffle around the edge of the driver as flush as I could get it.
So your 1.6Khz dip is almost certainly a diffraction issue, or perhaps a cavity resonance behind the driver frame if the panel is thick and not tapered on the inside, as the dimensions of a cavity formed by the panel cutout would be about the right diameter to cause this too.
I don't see a dip at 1Khz at all in any of my measurements. Unless your drivers are different to mine that's probably a result of a box internal reflection.
If you're following along with the conversation you'll see I won't be using them in a horn I'm afraid.
Maybe next time.
Last edited:
Glad you like your 1772.
I still have mine, gave the extra pair to a friend.
To my ears, they are too shouty, but with either passive or active notches, they sound wonderful, and true, although they can't escape tha laws of physics and do beam at some point, they do manage to do a good job and it is not as bad as many other 8" drivers out there.
I still have mine, gave the extra pair to a friend.
To my ears, they are too shouty, but with either passive or active notches, they sound wonderful, and true, although they can't escape tha laws of physics and do beam at some point, they do manage to do a good job and it is not as bad as many other 8" drivers out there.
A "shouty" presentation is usually just a several dB excess in response from about 500-1000Hz relative to bass and low midrange.
In other words, what you'd expect from uncorrected baffle step loss in most speaker designs. Even if the half space response of the driver was perfectly flat once you put it on a moderate size finite baffle which rolls off the bass and low midrange it's going to sound shouty, more so on a narrow baffle than a wide one. (Especially if the peak the baffle causes falls between 500-1000Hz...)
This is not a failing of the driver, it's simply the laws of physics, and pretty much all midrange or full range drivers in most cabinet configurations are going to be shouty without passive or active EQ to correct this.
My Coral drivers are exactly the same in this regard - the measured driver response on an infinite baffle is flat and balanced from 100Hz to 1000Hz, but the raw driver in a box with no baffle step correction sounds shouty due to the inevitable loss at low frequencies.
I'm not pointing this remark at you but I've always been puzzled and amused by the ultra millitant end of the full range driver enthusiasts who will not entertain the possibility of applying any sort of EQ to their beloved drivers as if it somehow makes them less pure.
On the contrary in the vast majority of cabinet designs some baffle step correction is mandatory to get tonal balance right and get the best out of the driver. By not doing so you're not actually reaching the full potential of the driver.
The only cabinet design (other than sticking a driver in a wall to make a true "infinite baffle" half space response) where you could argue that baffle step correction is not needed is a well designed back loaded horn where the baffle step correction at low frequencies comes from the gain of the horn.
But even then, the amount of gain achieved is unlikely to be exactly right, and further improvements could be made with some more subtle EQ to get the balance just right.
So I take baffle step correction as a given for getting the best out of a driver. However the less corrections you have to make beyond just BSC the better the driver is.
And for the 1772 it looks like it would probably (again, depending on specific cabinet design etc) benefit from a slight notch at 2.2Khz and a very gradual rolling off from low to high treble to level out the top end a bit.
As driver corrections go, that is pretty minimal. A lot of conventional single cone drivers used in multi-way systems are far, far worse before EQ is applied by the crossover. As I said in an earlier post - while I did notice it was a bit shouty in my quick and ready test (due to a lack of low frequencies in the test configuration) I didn't hear any nasty high Q resonances, even in the treble to be honest.
Low Q resonances like the one at 2.2Khz are mechanically stable and can be easily corrected with active or passive EQ. High Q, high frequency driver resonances (like those in the FE207E) cannot be corrected electronically, and in that case you either put up with the flaws or the driver is a loss.
All this driver needs is a bit of simple, overall EQ to massage the tonal balance to get a really great result.
Last edited:
I agree.
I often use massive boosting, not only for baffle step, but because we often listen way down in the fletchermunson curve (70-80db).
Passively, i think it is a bit tricky to balance it to get a smily faced response (small or large).
Didn't Bob Brines have almost 10db of baffle step to balance out hf climb and baffle step on the 1772 ?
I think 1 coil/resistor doesn't do too much damage removing the magic (transparency) compared to a 10 band eq with a string of sleazy opamps and electrolytics in the signal path.
I often use massive boosting, not only for baffle step, but because we often listen way down in the fletchermunson curve (70-80db).
Passively, i think it is a bit tricky to balance it to get a smily faced response (small or large).
Didn't Bob Brines have almost 10db of baffle step to balance out hf climb and baffle step on the 1772 ?
I think 1 coil/resistor doesn't do too much damage removing the magic (transparency) compared to a 10 band eq with a string of sleazy opamps and electrolytics in the signal path.
No Baffle configuration
I dipped my toe tentatively after my experience with an Audio Nirvana whizzer a decade ago. I have tried many drivers in my no-baffle rig, including Dayton DS100, Tymphany TC09, Alpair 12p, PRV MR450, Faital 4FE array, TB W8-2145 and now W8-1772.
For sure the 1772 are the most refined with a lush treble I am just not used to.
I chose this over the 1808 because of efficiency. Using DEQX I am rolling bass off at 100hz anyway. Crossing with woofers @ 275. Sucking out midrange a few db between 1-4k to tame them, but I am gradually reducing this as they bed in.
Very pleased with result and thanks to all who recommended this unit.
PS...have Lii Audio Fast 8 coming soon to compare
I dipped my toe tentatively after my experience with an Audio Nirvana whizzer a decade ago. I have tried many drivers in my no-baffle rig, including Dayton DS100, Tymphany TC09, Alpair 12p, PRV MR450, Faital 4FE array, TB W8-2145 and now W8-1772.
For sure the 1772 are the most refined with a lush treble I am just not used to.
I chose this over the 1808 because of efficiency. Using DEQX I am rolling bass off at 100hz anyway. Crossing with woofers @ 275. Sucking out midrange a few db between 1-4k to tame them, but I am gradually reducing this as they bed in.
Very pleased with result and thanks to all who recommended this unit.
PS...have Lii Audio Fast 8 coming soon to compare
Attachments
I dipped my toe tentatively after my experience with an Audio Nirvana whizzer a decade ago. I have tried many drivers in my no-baffle rig, including Dayton DS100, Tymphany TC09, Alpair 12p, PRV MR450, Faital 4FE array, TB W8-2145 and now W8-1772.
For sure the 1772 are the most refined with a lush treble I am just not used to.
I chose this over the 1808 because of efficiency. Using DEQX I am rolling bass off at 100hz anyway. Crossing with woofers @ 275. Sucking out midrange a few db between 1-4k to tame them, but I am gradually reducing this as they bed in.
Very pleased with result and thanks to all who recommended this unit.
PS...have Lii Audio Fast 8 coming soon to compare
Ouch. A bit harsh, but true I'd say.
Also a traditional open baffle approach uses a large baffle to put the transition frequency sufficiently low and add sufficient time delay to the first diffraction events.
When you mount the driver in free air the drivers own frame is the only baffle, so it is very small, (high frequency transition) has a very small time delay (image smearing) and it's a circular baffle, which is the worst possible shape because all rays travelling from the centre reach the edge of the "baffle" at the same time.
This means the standing waves all stack at the same frequencies which maximises the resulting peaks and dips in the response.
If you simulate an 165mm cone with a 223mm frame diameter (which is about right for a 1772) you get something like this:
A big honking (literally) broadband peak of around 4db at 1.2khz, and a rapid fall off below 500Hz.
It's not quite as simple as that as that though as the 1772 is a dual cone driver which does not have the poor high frequency dispersion of a typical 8" driver, it has a wider dispersion in the upper midrange. This can be approximated by using the diameter of the whizzer cone in the simulation:
Oh dear, that makes things even worse, adding a big dip around 3Khz and a response that is more bumpy above 3Khz.
For what it's worth I took some measurements of mine held completely unbaffled when I first got them (I didn't save them unfortunately as I was just messing around) and I saw just that - a big peak around 1.5Hz, a notch around 3Khz and I actually saw more peaks and dips than the simulation suggests. (Which implies the dispersion is wider)
Not a good sound. I also tested them with pillows forming a crude, lossy baffle for a quick rough and ready listening test:
The difference between the bare driver and the "baffled" response with the pillows was night and day, and the measurements were totally different.
@k9vap - if you're interested, try this experiment.
Make a thin wooden frame like you would for a speaker grill, make it reasonably large, say 50cm x 50cm. String a thick wad of polyester batting as used in speaker enclosures across the frame and cut a hole in the middle for the driver. Clamp the edges around the hole together with a couple of circular strips and attach that to the edge of the driver frame. You'll need to support the position of the outer frame of course as the polyester isn't self supporting.
Essentially you have made a lossy baffle out of polyester batting. I haven't tried this, but it should sound and measure a lot better than no baffle at all, and won't have a "boxy" sound. Diffraction should be very low which should result in a much smoother upper midrange and treble.
Interesting experiment to try!
Keep in mind that because the frequency response will change dramatically, you'll need to revise your EQ.Attachments
Last edited:
Not sure what multi-way has to do with diffraction.Jeez.....for a minute I thought I mistakingly posted on the Multi-way forum
Ok....once this Covid thing is over I will see the doctor and get my ears tested
Each to their own I suppose, however in the nearly 20 years that I've been tinkering with speaker designs (only as a hobby) the single biggest dawning realisation that I've had that has resulted in a big shift in my design approach has been taking diffraction at high frequencies (>1Khz or so) really seriously.
And flush mounting of drivers to avoid any diffraction at the driver/cabinet interface is a big part of that and it really does make a difference to not just how smooth a speaker sounds, but how good it's spacial pattern is.
Despite the knowledge being in the public domain for years, I still see so many DIY designs that don't really care about diffraction at all with drivers screwed onto panels not flush mounted, and old fashioned recessed front panels with large frame around the outside, especially on mock vintage full range cabinet designs. Or other oddball high diffraction designs. Just a diffraction nightmare.
I think taking diffraction seriously is so important that I actually went back to an older pair of speakers that I built more than 10 years ago that I've been using for years and flush mounted the tweeter, (previously screwed lazily onto the front with a small lip) and brought the front panel forward so that it was now flush with the frame instead of recessed. Big improvement, and not hard to do.
Mounting a driver in mid air with no baffle at all so that there is a huge diffraction wavefront coming from the abrupt perimeter of the driver frame (it's almost a ring radiator around the driver producing a destructive interference pattern) is frankly the worst nightmare scenario from a diffraction point of view. You pretty much can't do worse for diffraction than a circular open "baffle" that the driver frame forms by itself.
Sure you could attempt to EQ it and remove the most egregious peaks and it will sound better, but it will never be right on and off axis.
While I think there is some merit to large open baffles, (not a design that interests me though) I just can't get behind a circular open back driver suspended in mid air without any baffle.
If you like the result, that's fine by me though, each to their own.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.