To the OP:
That P.Audio 2" driver that you linked above in your original post will cover 400 Hz-17 kHz (with some Ti diaphragm ringing at 12 kHz). I'd actually recommend that one. The predecessor driver to it ("series I") is the Klipsch K-69-A. That's the driver that I'm currently using in my "new center" multiple-entry horn without a tweeter. I plan on upgrading later.
Chris
That P.Audio 2" driver that you linked above in your original post will cover 400 Hz-17 kHz (with some Ti diaphragm ringing at 12 kHz). I'd actually recommend that one. The predecessor driver to it ("series I") is the Klipsch K-69-A. That's the driver that I'm currently using in my "new center" multiple-entry horn without a tweeter. I plan on upgrading later.
Chris
Last edited:
Long post - I've been on this quest myself 
I'm assuming this is for your man cave double garage?
HF: 18sound NSD1095N 1” compression drivers
MIDS: ???
Low: Etone 1525 15” woofers
On paper, your Audax PR170M0 would be perfect as mids - did you already try them and find them lacking, and how?
I'll address your questions below, but my short answer: buy or build a big horn with a 2" throat.
Try your various cone midrange drivers on it.
Only buy a compression driver if these don't work out.
I get about the same results 700-7kHz with a $50 cone driver as I do with a JBL 2445J. The cone driver gives 5dB less sensitivity, but either solution requires stacks of padding to match non-horn woofers, so it doesn't really matter.
+++++++++++++++++
"Or a 2" exit driver from P.audio"
Probably a good deal at $200. In 'straya, you'd be lucky to get a 2" driver for less.
I like this vendor, Cannonsound. My main hifi is built with stuff from them.
Have you seen this thread? Dude uses that driver in a 2-way, from 500Hz.
https://community.klipsch.com/index.php?/topic/117131-p-audio-ph-4525-horn-and-bm-d750-aka-k69/
Note: Cannonsound sell similar JBL clone horns for just over $100 each. Tres convenient. However, I don't like the original design - too short for a 700Hz cutoff, and the diffraction slots are something I'd rather avoid.
"Second hand JBL or similar would be good but they are rare and expensive in my part of the world."
Used units seem to pop up reasonably often. I got some JBL 2445Js for $200 each in December - from the same ebay bloke who is selling the EV driver you noted. Kind of an impulse purchase / curiosity. I also got some EV 18" drivers for less Not sure if they are actually better than new P Audio stuff, or if they just have USA / historic cachet.
With my JBLs I am getting better results, particularly 1-2kHz, than the klipsch dude was measuring with P Audio drivers. However, there are so many variables that I'm not sure what differences (if any) can be attributed to the driver.
My horn:
The attached graph shows:
Middle (dark grey) 2445J (with replacement diaphragm) on cheap horn.
Top (dark red) 2445J (with replacement diaphragm) on modified horn. I added 3cm to the throat (I stacked and glued rings of 5mm HDF) to smooth the transition between driver and horn flares**. This modest effort is enough to make a clear improvement to the low end.
For both graphs: Crossover 500Hz, 10dB attenuation below 6kHz, no other eq.
Bottom (blue) is a P Audio 4" cone driver on the same horn (this is the stock horn, without the throat addition, so the best comparison is with the middle, grey, graph). Crossover 500Hz, 10dB attenuation below 3kHz, no other eq.
In all cases the horn was on the floor, pointing up, and I was holding the mic 1m over it. Horn is in this link:
11" x 17" ABS 2" Bolt on Long Throw Horn 90° x 40° for Many 2" Exit Driver | eBay
...and for anyone seeing this thread after the link dies:
Horn Dimension Front: 17.5"(444.5mm)x 11"(279.4mm)
Front to Rear : 12"(304.79mm)
*One of mine measures better than stock, with a no-name aftermarket diaphragm. This was a very welcome surprise. Ribbed diaphragms (new from JBL, or aftermarket) have more extended HF than the originals, it seems.
**all the big old JBL drivers have a ~180Hz flare rate built into them. Most people hear these on small horns like the (common and widely copied) JBL 2380, which are only 23cm long. The driver throat therefore has a sudden transition to a much more rapid flare in the horn's throat. There's also a diffraction slot.
I wonder if that's why it is often authoritatively stated that these drivers only sound OK if crossed over 1,000Hz
"Previously, virtually every compression driver made had a 180hz flare rate whose origin dates back to the original AT&T Labs designs from the 1930's. This low rate was necessary to accommodate the low cross-over points used in early two-way loudspeakers."
435BE
Also see JBL Technical Note, Volume 1, Number 21. It says that every 2" JBL made last century "was based on the exit geometry of the original Western Electric 594 driver" ...which in turn was used in horns literally 10 times as long as the JBL 2380.
I'm assuming this is for your man cave double garage?
HF: 18sound NSD1095N 1” compression drivers
MIDS: ???
Low: Etone 1525 15” woofers
On paper, your Audax PR170M0 would be perfect as mids - did you already try them and find them lacking, and how?
I'll address your questions below, but my short answer: buy or build a big horn with a 2" throat.
Try your various cone midrange drivers on it.
Only buy a compression driver if these don't work out.
I get about the same results 700-7kHz with a $50 cone driver as I do with a JBL 2445J. The cone driver gives 5dB less sensitivity, but either solution requires stacks of padding to match non-horn woofers, so it doesn't really matter.
+++++++++++++++++
"Or a 2" exit driver from P.audio"
Probably a good deal at $200. In 'straya, you'd be lucky to get a 2" driver for less.
I like this vendor, Cannonsound. My main hifi is built with stuff from them.
Have you seen this thread? Dude uses that driver in a 2-way, from 500Hz.
https://community.klipsch.com/index.php?/topic/117131-p-audio-ph-4525-horn-and-bm-d750-aka-k69/
Note: Cannonsound sell similar JBL clone horns for just over $100 each. Tres convenient. However, I don't like the original design - too short for a 700Hz cutoff, and the diffraction slots are something I'd rather avoid.
"Second hand JBL or similar would be good but they are rare and expensive in my part of the world."
Used units seem to pop up reasonably often. I got some JBL 2445Js for $200 each in December - from the same ebay bloke who is selling the EV driver you noted. Kind of an impulse purchase / curiosity. I also got some EV 18" drivers for less
Not sure if they are actually better than new P Audio stuff, or if they just have USA / historic cachet.With my JBLs I am getting better results, particularly 1-2kHz, than the klipsch dude was measuring with P Audio drivers. However, there are so many variables that I'm not sure what differences (if any) can be attributed to the driver.
My horn:
is 33cm rather than 23cm long
has a customised smooth throat transition
is a simple biradial - no diffraction slot
Also:my drivers have aftermarket diaphragms*
my measuring system is sloppy as
The attached graph shows:
Middle (dark grey) 2445J (with replacement diaphragm) on cheap horn.
Top (dark red) 2445J (with replacement diaphragm) on modified horn. I added 3cm to the throat (I stacked and glued rings of 5mm HDF) to smooth the transition between driver and horn flares**. This modest effort is enough to make a clear improvement to the low end.
For both graphs: Crossover 500Hz, 10dB attenuation below 6kHz, no other eq.
Bottom (blue) is a P Audio 4" cone driver on the same horn (this is the stock horn, without the throat addition, so the best comparison is with the middle, grey, graph). Crossover 500Hz, 10dB attenuation below 3kHz, no other eq.
In all cases the horn was on the floor, pointing up, and I was holding the mic 1m over it. Horn is in this link:
11" x 17" ABS 2" Bolt on Long Throw Horn 90° x 40° for Many 2" Exit Driver | eBay
...and for anyone seeing this thread after the link dies:
Horn Dimension Front: 17.5"(444.5mm)x 11"(279.4mm)
Front to Rear : 12"(304.79mm)
*One of mine measures better than stock, with a no-name aftermarket diaphragm. This was a very welcome surprise. Ribbed diaphragms (new from JBL, or aftermarket) have more extended HF than the originals, it seems.
**all the big old JBL drivers have a ~180Hz flare rate built into them. Most people hear these on small horns like the (common and widely copied) JBL 2380, which are only 23cm long. The driver throat therefore has a sudden transition to a much more rapid flare in the horn's throat. There's also a diffraction slot.
I wonder if that's why it is often authoritatively stated that these drivers only sound OK if crossed over 1,000Hz
"Previously, virtually every compression driver made had a 180hz flare rate whose origin dates back to the original AT&T Labs designs from the 1930's. This low rate was necessary to accommodate the low cross-over points used in early two-way loudspeakers."
435BE
Also see JBL Technical Note, Volume 1, Number 21. It says that every 2" JBL made last century "was based on the exit geometry of the original Western Electric 594 driver" ...which in turn was used in horns literally 10 times as long as the JBL 2380.
Attachments
Good to have a recommendation on the paudio driver. I'll keep it on the shortlist. What are you planning to upgrade with?
Depends on budget and position of loudspeaker in the array. All of the following are full-range (~400-~20000 Hz) applications with no "midrange":
For the center loudspeaker - I'm considering a BMS 4592ND (bi-amped, not passive) or used TAD 4002 (I've got 4002s on either side of the center on the fronts). Both are about the same price, but the BMS is new while the TAD is used.
I've heard the BMS driver and it needs EQ and crossover compensation to sound its best, IMO.
The TADs just smoke everything of course. The concern is abuse and overheating of the used units. Many of these units are becoming available on the used marketplace since TAD (Pioneer) has phased out the neodymium magnets in favor of the really large and heavy Alnicos in the TD-4001 model. Neo doesn't like get really hot (it anneals easily)--like you see in PA and nightclub environments. Plus neodymium is getting expensive and hard to get.
For surrounds, I'm considering just using the P.Audio BM-D750s (K-69-As) or perhaps the B&C DE750, which is an incremental but audible improvement in sound quality (~$170 US vs. $270 US). That one used to be called the Klipsch K-69 driver. I've heard both, and I prefer the B&C.
The real difference is the last octave (10-20 kHz) in which beryllium diaphragms are unequaled in performance. Ti rings, all of them. So I typically have used a Beyma CP25 on top and tri-amped. However, if you're doing multiple-entry (i.e., "Unity" horn) like I'm going to all around, then having one driver (even a coaxial compression driver) is a big deal. Time aligning coaxial drivers isn't difficult using active crossovers. I presently have actives on all loudspeakers.
Chris
Hi Hollowboy, thanks for all that good info.
Mancave drivers were sold but now I'm back with more cash to throw at diy
Cheers
Starting from scratch? What budget / space / tools?
Good info, thanks
I tried ~10 drivers. The ones that worked best were the current P Audio 4" (SN4-60F) and a no-name 2" salvaged from a bad 5.1 system. Most of what I tried (including the FF105WK, which is very similar in size and specs to the P Audio driver) had a big suckout ~2kHz. I wonder if cone rigidity is the key (the Fostex cone is 2g lighter, presumably less rigid).
These attachments show:
1) the (horizontal) polars for the JBL. Horn and mic were on the floor (hence no highs). It seems to lose pattern control at ~900Hz, which is what "DrWho" estimated on the klipsch thread I linked to earlier (same mouth size on that horn).
So this horn is a little undersized to be perfect for a 700Hz crossover.
If you cover the <900Hz range when looking at the graph, then it looks good, parallel lines - so equalisation to flatten the on-axis response should flatten the response everywhere.
I didn't do verticals, I expect they'd be as DrWho predicted, with the loss of control happening ~1.7kHz. As I understand it, that's what 'pattern flip' is, and what the big 'lips' on Iwata horns are there to address.
2) how I arrived at the relative efficiency of the drivers:
-I measured their raw responses (at horn mouth, 400Hz highpass)
-I then measured how mach padding was required to play (700-7kHz) with a 12" woofer (SB34NRX75-6), measuring @ 1m
The JBL is shown in grey, I used -17dB, and 2 notches for the 'flat' graph
The 4" is shown in black, I used -12dB and 2 notches for the 'flat' graph
The 2" is shown in green, I used -5dB and 2 notches for the 'flat' graph
The bass was the same throughout - fluctuating dips at 80Hz and 200Hz are due to interactions between me, cats, and a nearby door
3) The JBL running as a 2-way vs the P Audio running as a 2-way. Both are pretty good. There's no eq on the HF, I just removed the 7kHz lowpass, so this is just a smoothed version of all the HF gunk you see on graph 2. I don't think the >10kHz stuff is that bad. Maybe if I was still in my teens and could hear it better I'd object to it.
The JBL has better HF & more potential to be tweaked to excellence as a 2-way (new diaphragms or good luck with 2nd hand, horn mods, careful setup, eq).
The P Audio is cheaper, and would suit a lower range in a 3-way really well (400-4kHz), if you can live with a really big horn, and size is truly no barrier.
For me, size IS a barrier, so I burned the 80cm wide horn I first made for this driver
Attachments
Last edited:
Frangus,
Using a horn like the one described in #5 with your ribbon at the apex, and four TC9 would hit 2.5kHz with injection ports small enough to have virtually no effect on the tweeters response.
You could probably get your Galaxy 5" mids to go that high, but the ports would be larger. Using your ribbons at the throat makes the throat much taller than the usual 1" to 2", which will reduce the (minimal) effect the ports have on HF response.
Since you have the Galaxy drivers, making a cardboard prototype would give you an idea of what having a single point source with (horizontal) pattern control to around 250 Hz does for imaging with about an hour invested in time.
Art
May I have an idiot check please - you mean the combo:
P Audio BM-D750 (2" compression driver)
K-402 (shallow conical horn with curved mouth, 1m wide)
...is what you are using, and is good 400Hz-17kHz?
If I've understood you correctly, that could be a great, simple build, as the BM-D750 is available in Australia for a good price, and the K-402 looks a lot easier to DIY than, say, a Le Cleach Iwata.
Is there any rounding at the K-402 throat, or does it go from the driver's throat angle (10 degrees or whatever) straight out to a 100-degree conical?
If you were gonna build one from scratch, is there anything about it you'd change?
+++
EDIT I think I found the 'secret' to the throat; posted by djk on another forum, quotation follows:
A 2" throat will (normally) have about 62.5° dispersion at 8Khz. A pancake driver (no long snout) will be better than theory predicts if the phase-plug parts come all the way up to the mounting plane of the throat (as they do for the K-69 driver used in the plots). Theory predicts about 31.25° at 16Khz, but the -7dB points on this horn/driver combo are about 60° at 16Khz with the phase-plug geometry. A old-school driver (JBL 2445, etc.) with a long snout (stub) inside the driver will have substantially worse performance above 8Khz or so.
Last edited:
Yes - a P.Audio BM-D750 (Series I or II) with a K-402 horn. It looks like the Series II driver is using ferrofluid, and I haven't heard that one, per se. However, it's a good bet that it was redesigned to correct its most apparent deficiency: 13.7 kHz ringing.
The K-402 throat directly expands from the throat with no transition. The expansion is what most call conical until you reach 12" axially from the throat, then it's tractrix. The angles are 56.6 degrees x 35.3 degrees (half angle to the circular 2" [5 cm] throat entrance). Mouth dimensions are 36 x 21 inches (91.5 x 53 cm) at the intersection with the mouth flange (about 2" [5 cm] all the way around). No other geometries are involved, and the throat is of course the most critical area to get right. I wouldn't change anything.
The EQ for this horn with the P.Audio driver, and the Jubilee bass bin can be found here: http://207.67.68.51/forums/storage/4/1554177/DX-38_settings_for_Jubilee_with_K402_%26_K69.pdf
Chris
The K-402 throat directly expands from the throat with no transition. The expansion is what most call conical until you reach 12" axially from the throat, then it's tractrix. The angles are 56.6 degrees x 35.3 degrees (half angle to the circular 2" [5 cm] throat entrance). Mouth dimensions are 36 x 21 inches (91.5 x 53 cm) at the intersection with the mouth flange (about 2" [5 cm] all the way around). No other geometries are involved, and the throat is of course the most critical area to get right. I wouldn't change anything.
The EQ for this horn with the P.Audio driver, and the Jubilee bass bin can be found here: http://207.67.68.51/forums/storage/4/1554177/DX-38_settings_for_Jubilee_with_K402_%26_K69.pdf
Chris
Last edited:
Thanks for the measurements. Seems there are no issues with a 4" driver on a 2" mouth horn then?
I think I will buy the etones again. I already have the ribbons for the top so really I just need big horns and mid drivers. I have 4 sub's already set up. I don't really want to build a horn simply because I don't have the time with 2 young kids and a demanding job.
I think I will buy the etones again. I already have the ribbons for the top so really I just need big horns and mid drivers. I have 4 sub's already set up. I don't really want to build a horn simply because I don't have the time with 2 young kids and a demanding job.
Art, the only issue I have with the smaller horn is the fact that it doesn't control the pattern low enough to take full advantage of the synergy concept. Iam not limited by space so I might try a simple rectangular conical from foamcore. 40" by 20" mouth should be good for 200hz crossover to the sub's?
90 x 45 degrees?
Hopefully I will get some time one day to learn how to use Hornresp
I had a go at hornresp over my lunch break.
I used the TC9 but couldn't find all the parameters so had to fudge some of them. Also couldn't locate the frequency response window so used max spl to get an idea of the FR.
It seems no matter what I try I cant get a conical horn of any reasonable size to play down to 200hz (400hz is about it). I must be doing something wrong...
I used the TC9 but couldn't find all the parameters so had to fudge some of them. Also couldn't locate the frequency response window so used max spl to get an idea of the FR.
It seems no matter what I try I cant get a conical horn of any reasonable size to play down to 200hz (400hz is about it). I must be doing something wrong...
Yes and no, depends on application. As mids in a ~95dB efficient 3-way, to cover 400-4,000Hz, the 4" driver on 2" throat is a contender (as shown in post 27). For a wider range or more efficiency, you'd have to look elsewhere / spend more.
If you're strapped for time, maybe put the ribbons aside, and look at a relatively simple 2-way for your mains?
For ~800Hz and up, one of these combos:
$400 for a pair of the P Audio BM-D750
$100 for a pair of these horns
https://www.vfmaudio.com.au/19780/Horn-Flare-Large-Format-2-Inch-Throat-4-Bolt-Mount/
...and you've got a baby version of what Chris / Cask05 is using.
OR
Something based on the Faital LTH142, example here:
Cornscala Style D | Critesspeakers.com
To keep to a budget, you could look at clones / other vendors, e.g. ebay seller zxpc has a $25 clone of that horn.
Yes, you are doing something wrong, but without screen shots difficult to say what. There are a lot of inputs to get right, and lots of opportunities to get them wrong. Max SPL is one way of seeing the wrong things, unless inputs are correct, in which case it just shows excursion and power limitations on the same screen, which I find an annoyance.
Size does not translate to LF cut off, narrow conical horns will have a lower cut off than wider horns, though pattern control will be lost at a higher frequency. The narrow pattern makes the "Acoustical Power" chart show a HF response that drops steeply, though the wide lower dispersion and narrow HF dispersion result in a near even on axis response.
You can look at the loudspeaker wizard to see the horn shape, and how it affects Fc (horn cut off).
You can approximate the response of a conical horn with two angles (say 60 x 40) by adding the two together (60 + 40 =100) then dividing by 2 (100/2=50 degree). The wall angles will be the approximate -6 dB dispersion points, and the acoustical power response of a 60 x 40 will be equal to a 50 x 50.
The unknown TC9FD-18 parameters can all be derived using Hornresp from the known TS parameters given on the spec sheet.
These are the Hornresp inputs:
Sd=36.3
Cms=6.60E-04
Mmd=2.43
Re=6.3
Bl=3.01
Rms=.71
Le=.05
Nd= number of drivers
The TC9FD-18 Xmax is 2.55, rated power is 30 watts, on a narrow two part conical horn (43 cm + 14.7 cm,) with a 270 square centimeter end to the first section, and a 678 square centimeter mouth (26 x 26 cm, about 10.25" x 10 .25") Hornresp predicts 115 dB at 210 Hz without exceeding either, in full space (not on the floor).
That horn design is a bit too narrow for domestic use, but the output is impressive (126 dBA peaks above 900 Hz), and wide range.
Actual raw measurements of the real horn measured below, the top graphs with no smoothing applied, the lower charts with smoothing applied, using two different throat sizes, 2" and 2.5". Measurements were taken outdoors, at 2 meters high, two meters distant.
Art
Attachments
I agree with Art and XRK971; try using a conventional driver instead of a compression driver.
For instance, I have a shelf full of compression drivers, but I generally use an SB Acoustics SB19 instead. Thought it's one-fifth the price of a good compression driver, the distortion is actually *lower* where it counts - in the octave between 1000hz and 2000hz.
I'm no expert on compression driver design, but I believe the improved distortion performance is because conventional drivers generally have higher xmax than compression drivers.
Obviously, a compression driver will have higher output. But I personally prefer lower distortion in the midrange over the ability to hit 130dB. All engineering requires compromise.
Here's an example of a project that I did recently. This is the distortion and frequency response of a AuraSound whisper on a very simple waveguide that I made for my car. Because it's made for a car, it's far from ideal, and the filter is far from perfect. But despite these shortcomings, I'm still getting almost five octaves of bandwidth out of a $20 driver. Even better, the waveguide raises the efficiency into the 90s, and gives me directivity control too. The blue line is the response of the Aura without a waveguide; the red line is *with* the waveguide. You can see it raises the output by over fifteen decibels in the midrange, which is how a $20 computer speaker can compete with a $100 compression driver.
That measurement was from one of my posts over at Diymobileaudio, take a look at some of my threads for more data
For instance, I have a shelf full of compression drivers, but I generally use an SB Acoustics SB19 instead. Thought it's one-fifth the price of a good compression driver, the distortion is actually *lower* where it counts - in the octave between 1000hz and 2000hz.
I'm no expert on compression driver design, but I believe the improved distortion performance is because conventional drivers generally have higher xmax than compression drivers.
Obviously, a compression driver will have higher output. But I personally prefer lower distortion in the midrange over the ability to hit 130dB. All engineering requires compromise.
Here's an example of a project that I did recently. This is the distortion and frequency response of a AuraSound whisper on a very simple waveguide that I made for my car. Because it's made for a car, it's far from ideal, and the filter is far from perfect. But despite these shortcomings, I'm still getting almost five octaves of bandwidth out of a $20 driver. Even better, the waveguide raises the efficiency into the 90s, and gives me directivity control too. The blue line is the response of the Aura without a waveguide; the red line is *with* the waveguide. You can see it raises the output by over fifteen decibels in the midrange, which is how a $20 computer speaker can compete with a $100 compression driver.
That measurement was from one of my posts over at Diymobileaudio, take a look at some of my threads for more data
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.