For the fans of the BMS coaxials drivers, let us not forget that they are not a compression driver in the style of most other compression drivers discussed here. In fact they are two ring radiators in a single housing. The do not have the classic dome diaphragm and phase plug of typical CDs.
The phase plug is a bullet shape and extends into the driver throat, as is typical of ring radiators. Also note that the midrange sound comes off the back of the ring, not the front.
The phase plug is a bullet shape and extends into the driver throat, as is typical of ring radiators. Also note that the midrange sound comes off the back of the ring, not the front.
My experiences using the BMS 4592ND on a K-402 (the K-402-MEH prototype to be exact) indicated that the HF diaphragm channel needs to be delayed one full wavelength at the crossover frequency:
https://community.klipsch.com/index...ax-xo-sugestion/&tab=comments#comment-2416945
I would also add that this is a candidate application of using PEQs only to cross the two diaphragms in the 4592ND, as a first order set of crossover filters inserts 90 degrees of phase lag on the lower frequency diaphragm/channel. Without that all-pass phase delay, I can't really tell the difference between the TAD TD-4002s on K-402s (each side of the center MEH) and the K-402/MEH with BMS 4592ND.
Chris
https://community.klipsch.com/index...ax-xo-sugestion/&tab=comments#comment-2416945
I would also add that this is a candidate application of using PEQs only to cross the two diaphragms in the 4592ND, as a first order set of crossover filters inserts 90 degrees of phase lag on the lower frequency diaphragm/channel. Without that all-pass phase delay, I can't really tell the difference between the TAD TD-4002s on K-402s (each side of the center MEH) and the K-402/MEH with BMS 4592ND.
Chris
Last edited:
For pro use no but for audiofools like us an active DSP to fine tune the crossover seems necessary ..
True, and that insight took me some years, but I made the 4592nd sing in the end (imho) 😊For pro use no but for audiofools like us an active DSP to fine tune the crossover seems necessary ..
FWIW, Bert Doppenberg from BDDesign in Holland modifies these drivers, but his modification is a business secret. It has something to do with the cavities inside the drivers or with the removal of rough edges. I don't know anymore, but it was something physical. I heard his system once and it sounded excellent, but I could not test side by side with an untreated driver and to me, I do not have complaints with mine. Maybe he has the better ears however.
That would be about 0.16 millisecond, right?the HF diaphragm channel needs to be delayed one full wavelength at the crossover frequency:
It depends on what frequency you choose to cross. I have used 0.145 ms (as seen in the linked thread, above) and used only PEQs to cross the midrange and treble diaphragms at ~6.9 kHz, as shown in the XConsole screenshots. I've also experimented with inverted phase of the treble diaphragm channel, and using half the delay (0.072 ms). There are tradeoffs to each approach, each having an effect on the resulting acoustic amplitude and phase response. Since I'm not using my setup for PA-service SPLs, I have a bit more flexibility in the crossover design to minimize phase growth.
The goal was to match timbre with the TAD 4002s/K-402s, and that was easily achieved.
Chris
The goal was to match timbre with the TAD 4002s/K-402s, and that was easily achieved.
Chris
Last edited:
When you have such quality compression driver it finally begs for a good enough sounding multi channel dac. Because all the delays needed : the compression motor itself, the horn with the woofer, plus EQ.
I'd like to listen to those BMS one day, they have many positive testimonials reviews indeed ! 🙂
I'd like to listen to those BMS one day, they have many positive testimonials reviews indeed ! 🙂
Not the improved GSU from GT sound. The GSU are 6000$+ each driver if memory serves me right.
4 slit phase plug, 2,0 T, good inductance control with copper shorting ring, copper plated motor, really thin and lightweight alu diaphragm. very interesting large format drivers indeed.
Make it $8,000.
G.T.-Sound has put extensive R&D efforts into the magnet structure and phase plug design. They found that an optimized 4-slit plug results in less distortion compared to the 5-slit plug used in the Sony SUP-T11 and large TADs.
Even though beryllium is 'theoretically' by far the best material for domes, the special (Ultra-Super?) Duraluminium in combination with the construction of the GSU-4 drivers can easily compete with the TADs in terms of distortion, but sound more natural, according to those who compared both.
Japan has been a leader in the field of high performance aluminum alloys since the first half of the last century.
There are already variants that approach the properties of Be.
Last edited:
I have a JBL 2216ND and even if in the M2 or the JBL4367 they let it run to 700hz , my actual 6.5 Full range driver do a better Job between 250hz and 800.hz than the 15inch JBL View attachment 1171628
Of course it does. The 2216Nd isn't a true midwoofer (too much Xmax, too much cone weight).
The only way to get an acceptable midrange (>500Hz) from a 15" is to sacrifice low frequency extension > minimize excursion, use a stiff and light diaphragm.
Some argue that modern R&D-technologies enable optimization of the magnet structure and moving parts to overcome many problems and arrive at an acceptable solution. They don't!
You can't ignore physics, that's why I strive for 'matching physical properties'.
How many modern 15" drivers have a flux density > 1.4 T?
Last edited:
that's why my interest for a compression driver to cover the 300 /3k with a lighter diagram than my 6.5inch cone driver like the Celestion Axi2050 or the BMS 4592 .. 🙂
Do not forget the BMS4591:that's why my interest for a compression driver to cover the 300 /3k with a lighter diagram than my 6.5inch cone driver like the Celestion Axi2050 or the BMS 4592 .. 🙂
https://www.bmsspeakers.com/fileadm..._4591_2011-04_midrange_compression_driver.pdf
For the bms 4594He the upper section has to be delayed 0.08ms no matter which of the many horns I've tried it on.Does it mean that the tweeter impulse has to be delayed with an active filtering setup or it is non discernible ?
Xover is always at 6300Hz, as there is not much room to vary it.
For the dcx464, it's the same 0.08ms delay for the upper section.
Horn doesn't matter, nor does varying the xover point anywhere between 3-4.4kHz.
I'm of the conclusion the correct delay is purely a function of distance.
0.07ms would be best for both, but my dsp is 48kHz, so 0.083ms (4 samples) is as close as I can get get.
Differences in pink are clearly heard when off by more than one sample, from a spot on delay.
No clue about music, cause i say to self who cares, just do it right anyway.
Last edited:
If we were actually using it for PA, its designed to run to 300hz....but for domestic use, it was suggested that on a large enough horn it could be used to 100hz. I personally used 115db/1m for a benchmark and on my 150hz F6, horn, 200hz with a 48db LR pulled it off. This 115db/1m was for the sake of robust headroom but yet a standard that I picked up and adopted, along the way, supposedly based around the idea of realistic, real life, dynamic potential... At 95db avg with 110db peaks, there is no struggle. This is for a single horn btw. With a 300hz 2nd order HP 120db/1m was no struggle.If you want a LARGE miodrange horn and Xo at 3-400 hz (Axi 2050 distortion starts rising below 350hz, 300hz probably okay for home use), then the AXI2050 does that better then any other single diaphragm driver. It has been designed for that purpose, to go low in PA applications and be abused, 150/300w power rating. It will give you half to one octave lower extension compared to normal 4" diaphragm drivers.
But it has a higher xmax, heavier and large diaphragm (175mm ish) comparatively, larger wire used, larger clearances for more air evacuation and so on, designed to be robust, and it is considerably cheaper to manufacture. And it does not defy the laws of physics, no matter how good the engineering is. The engineering choices made points towards a large format midrange compression driver, to avoid xos in the midrange band. A driver covering 300 hz will also naturally exhibit more excursion and therbey, higher IMD and distortion in the top end.
According to DATS the Fs is 215hz (fs= free-air resonant as in no horn)
Heres a 119db, in room measurement, at 46" inches, on my horn. The distortion at 200hz is ~17% 2nd order. The unique thing about crossing so low, is that room gain becomes a factor.
From the discussion we've had, we believe the xlim/xmax to be about 0.8mm and that is what? 0.3mm higher than the average 4" dome compression driver. I never found Xlim btw, I mean distortion gets high or whatever but the clicky/clackety never showed up and I got scared because it was way louder than I'd ever play it anyway
Also from our discussions, we concluded that the Surface area is no more or less than the average 4" dome diaphragm... so there's that.
When I do my best to simulate my situation on Hornresp, I get some good numbers on displacement. Considering the lack of clicky clack, Hornresp supports the idea that I never reached 0.8mm at 120db/1m crossed at 200hz. At 95db/1m the excursion peaks at 0.0414mm, give it 110db for true peak, excursion peaks around 0.194mm.... If you think thats an IMD problem, thats funny. I mean this is still for just one channel. Another channel adds 3db, if you catch my drift.
Rergarding Domestic systems, theres no issues of headroom for large Format compression drivers on large horns, but, for sure, the Axi, is likely one of the deepest playing ones, just by having higher xmax. Sd seems to be equal to the rest of the 4" Dome Diaphragms
It doesn't matter if you attenuate the mid or boost the highs. does it? Or maybe to do a little of both? The end result will be the same. All that really matters is headroom. The Axi has enough headroom to reach 20khz for anything domestic concerning at ~90db and 1m distortion is under 1%. I've got outside measurements showing 10khz, under 1% THD at 97db/1m let alone whatevers happening at 90db. I can't see THD above 10khz unfortunately but I think its safe to say that most people don't listen at 90db even..... I've started using an spl meter to keep myself in check as well.The AXI can be padded down over it's midband to make it "flat" but that is 20db+, or the top end can be eqd up with active eq but you are 'amplifying' the worst part of the driver with +20db.
Oh yeah, don't forget, those specs are concerning a single channel, not two lol.
Last edited:
Thanks for the mesures it show more than any subjective review , many of us have just a few watts tube amp for this task so i don't think the axi2050 diagram could be in danger at 300hz even at 100db . for a subjective point of view i am a fan of Full range speaker and always prefer the phase and time coherence than a flat line to 20K and low distorsions.. for me the 300 / 3K band is more crucial than the rest ( except a great 30 /300Hz ) at 50 i can live with a so-so 10 to 20K 😀 the only compression driver i have tried for low mid was the 0.66inch VITAVOX GP1 it was very impressive even if i used it with a long adapter and a too smaller horn for that application.
What mic did you use @camplo ?
I just don’t know how much credence I give REW distortion measurements, at least not using a Umik1. Comparing home measurements using this setup to Klippel using a B&K mic came somewhat as a shock to me.
Not saying this is wrong but I’d like to see a bit more cross referencing before we take these as gospel.
I just don’t know how much credence I give REW distortion measurements, at least not using a Umik1. Comparing home measurements using this setup to Klippel using a B&K mic came somewhat as a shock to me.
Not saying this is wrong but I’d like to see a bit more cross referencing before we take these as gospel.
Can you share? What did you see that was so shocking?Comparing home measurements using this setup to Klippel using a B&K mic came somewhat as a shock to me.
It is a Umik1 but from what I could find, the measurements from various measurement mics are not, night and day. So if its just in the ball park, then thats fine, to me.
I'd love to get to the bottom of it but the lack of measurements suggest to me that theres not a lot to see. The measurements that I have seen comparing several mics, suggested to me, that for general surveying, the Umik is good enough. For example;
BEHRINGER ECM8000 VS. BEYERDYNAMIC MM1 VS. ISEMCON EMX-7150
Im not loosing sleep if the differences look like this when all things else are equal. I wouldn't be surprised if the difference you saw weren't due to the Klippel system and not so much the B&K mic, meaning, if you aren't comparing the the two mics, on the Klippel, what are the chances that the results would be more similar than not....I believe the chances are high. It also seems to trend that the differences are more on the top end?
I found this pic but it would be nice to see phase/thd
Last edited:
- Home
- Loudspeakers
- Multi-Way
- Best Compression Drivers today 2022?