BMS 4592 vs Radian Beryllium

wow this is really eye opening.

when i look at his charts really closely i see he measured the exact same IMD for both single and dual diaphragm but he read them differently because in one case he was measuring in a quiet room and in another case in a loud room

they are the exact same % IMD in both cases but he deemed one driver "distortion free" ( loud room, distortion below noise floor ) and another "unacceptable" ( same exact % IMD )

i will never take this guy seriously again. he means well but he's just too sloppy.
 
  • Like
Reactions: Gill.T
i mean if you look at this chart

View attachment 1431057

it is pretty clear the driver should be used from 500 hz to 3 khz

then he just pushes 20 khz into it to measure IMD at 4 khz and declares the driver is bad
You may have a point here.

Nonlinearities may creep in when IMD is measured outside of the specified range.
The increasing HD above 4 kHz is an indicator.
 
  • Like
Reactions: Dissident Sound
wow this is really eye opening.

when i look at his charts really closely i see he measured the exact same IMD for both single and dual diaphragm but he read them differently because in one case he was measuring in a quiet room and in another case in a loud room

they are the exact same % IMD in both cases but he deemed one driver "distortion free" ( loud room, distortion below noise floor ) and another "unacceptable" ( same exact % IMD )

i will never take this guy seriously again. he means well but he's just too sloppy.
It's quite remarkable that he deemed THD to be extremely low while IMD was 'unacceptable'.
And the comparison is moot given the environmental noise.
 
  • Like
Reactions: Dissident Sound
i honestly do not understand it but because i don't and to be safe i would stay with a single diaphragm.

it should also be cheaper, easier to replace the diaphragm and easier to find parts ( since the diaphragm is shared with coaxials, which is not the case for dual diaphragm driver that has dedicated diaphragms ).
Rather than looking at Crowe's indoor measurement discrepancies, we can look at the BMS spec sheet.

The dual-diaphragm 4599 and single diaphragm 4591 are measured on a 40° x 20° 2242 horn, at 1W/1m.
Such a narrow dispersion horn increases sensitivity considerably over the more typical 90°x45° used for home use. Still, the 4599 only achieves only ~115dB from 1-3kHz, the 4591 is more sensitive over a wider bandwidth, though the 4599 has ~+3dB at 200Hz, below the 2242 horn's Fc of ~275Hz.
BMS 4599:4591.png

The 4599 appears to have slightly less second order, but more third harmonic distortion, not a good trade for "hi fi" use.
Overall it is less sensitive than the 4591 (-6dB above 4kHz), and rated for only twice the peak power, so wouldn't achieve more output in anything but the 200 Hz range (with a horn with an Fc that could support it) if the BMS charts are correct.

Anyway, the dual diaphragm approach in mid/hi drivers seldom lives up to the marketing hype, the JBL D2 used in the M2 reaches around 1% distortion at 5kHz at only 96dB @1meter:

JBL D2:M2.png


Given that it's distortion rose by ~10dB relative to it's 86dB level, I'd expect it to reach around 10% at 126dB.
Pretty similar distortion to the old JBL 2445 on a big "old school" 2360 horn:

Screen Shot 2025-03-05 at 4.52.17 PM.png

https://www.audioheritage.org/vbull...5-D2430K-compression-driver-information/page3
The JBL 2450 and 2435 were improvements over the 2445, here we can see them compared to a D2 on a JBL PT-F95HF horn:

Screen Shot 2025-03-05 at 5.06.32 PM.png
Screen Shot 2025-03-05 at 5.19.43 PM.png


While lots of press has been given to the wonders of the D2, I doubt you will find JBL producing any actual direct comparisons of their old drivers to them.

Art
 
wouldn't achieve more output in anything but the 200 Hz range

but that's precisely what you want in a driver like this - you want more output at lower frequencies, because basically all compression drivers can go down to 2 khz so it is only what the mid can do BELOW that point that matters.

this is why BMS switched from CCAW to Copper wire for the 4599. the coaxials use CCAW because it's lighter allowing them to reach 7 khz XO to the supertweeter but nobody is crossing 4599 to a supertweeter so it didn't need to go high and could use heavier copper VC that produces more force and that combined with the doubling of area is what allows it to play lower ( plus i am sure there are other tweaks as well ).

but yes finding a horn that can actually realize that potential could prove challenging.

the driver is useful because even the loudest paper midrange can't keep up with SPL of a compression driver like D2. as i previously mentioned the paper mid in JBL VTX F35 is HORN LOADED to be able to keep up - and i don't mean in a waveguide -it's an actual horn with phase plug and 4 X 1 compression.

my current concept for this driver ( 4599HE ) is to cross it over to a curved array of 8" midbass drivers at 800 hz on the low end at to a D2 at 2.5 khz on the high end.

i am thinking six of these:

https://www.beyma.com/speakers/Fich...ers-data-sheet-low-mid-frequency-8MC500Nd.pdf

from 200 hz to 800 hz. then onto 15" woofers below 200 hz.
 
For a high output (semi-pro) 3-way, these drivers should be more than sufficient.

If you want to build a slightly smaller/more modest 3-way, this 6" midrange looks interesting, although the response plot is misleading.

View attachment 1430868View attachment 1430869

"lightweight paper pulp, specifically developed for this application. This unique cone provides the ideal weight to strength ratio."
(It's got my favorite/preferred cone & dustcap structure)

Resonant frequency: 150 Hz
Re: 6.9 ohm
Qes: 0.38
Qms: 4.58
Qts: 0.35
Vas: 2.9 liters
Sd: 143.1 cm2
Mms: 11.1 grams
Bl: 13.7
Le: 0.16 mH
Sensitivity: 96.79 dB

Considering the hypothetical 3-way, a small tractrix horn would suffice for the the top end.
Crossed from about 1800 Hz, depending on the midrange used and preferences.

1741453849305.png
1741454095332.png



Response with the BMS 4540ND (unfortunatedly only on-axis and heavily smoothed - still looking good though):

1741454372057.png
 
Considering the hypothetical 3-way, a small tractrix horn would suffice for the the top end.
Crossed from about 1800 Hz, depending on the midrange used and preferences.

i am currently thinking:

BMS 4599 HE on BMS 2236 horn ( 800 hz to 2.5 khz )

1741455930380.png


JBL D2 on Beyma TD385 horn ( above 2.5 khz )

1741455884111.png


the BMS horn is more Midrange optimized with a more exponential shape ( albeit realized with two steps ) while the Beyma is more HF optimized with a more constant directivity shape.

Constant Directivity Horn seems to be the preferable solution sound quality wise if you have enough output capability to reach your crossover point, which in this case using a strong ( albeit VHF optimized with its very narrow ring diaphragms ) driver to reach down to 2.5 khz shouldn't be a problem.

i really wish there was a better 2" midrange horn than the BMS but the BMS seems the only one widely available at a good price that i know of at least.

will the D2 fit on that Beyma horn ? does it use the same throat size / bolt pattern ?

800 hz XO for BMS chosen based on rule of thumb i read somewhere that a good XO point is 2X the "cutoff" frequency of horn, and the cutoff frequency is listed as 400 hz for this horn. i wish i had better data to go on but 800 hz should work.

wavelength at 2.5 khz is 5.4 inches and we can visually compare that to the 4" bolt circle pattern on the BMS horn and see that at the point where the horn has its "step" it is about 1 wavelength tall and wide ... which means it will probably beam pretty hard above that frequency but we won't use it above that frequency anyway ...
 
Last edited:
hello @phase_accurate do you have any links to D2 specs that show the horn and bolt dimensions ? when i tried to google it seemed like it is 1.4" ... i am sure you are right but wonder where that information is coming from.

also, do you have a link for that "PT95 waveguide" ?

EDIT: did some more googling 1.5" horns seem to be rare to non existent.

i did find this one:

https://reconingspeakers.com/product/jbl-m2-horn-lens-5025594/

but it's kind of pricey and not clear whether it has all the parts needed to mount the driver. seems like you would need this metal washer:

1741534415950.png


of course i suppose you can make the washer yourself ...

you could probably also ignore 0.1" mismatch in horn size ? would at least the bolt pattern match ? i suppose i could drill the extra holes if the flange is large enough ?
 
Last edited:
also have Radian be and Radian standart drivers. i do not have enough experience them on yet to say much and I use Radians half an octave narrower than I used bms. Radian sounds relaxed and detailed.
Relaxed sounds very inviting. Have you since used the RadianBe (745Be? 950Be?) in a two way system, perhaps with midwoofers like this? https://josephcrowe.com/blogs/news/altec-416-8b-in-100l-sealed

If the 745Be with a 15" woofer, did you end up crossing at around 700Hz?

But as there's likely no diaphragm that's more revealing than Be, was your less than pristine recordings (e.g. too much applied dynamic range compression and/or analog tape saturation) of your favorite music still enjoyable or become fatiguing?