The KVAR looks a bit like an E-JMLC with the lips but it measures nothing like it. The 280 that you linked to is much more like a JMLC in that it is quite high directivity on axis but with a much steeper slope so the difference between on axis and off axis sound will be much more pronounced with that one, the Rad Imp fell off at 1K with the 280 so it should load a little lower than the ST260 with more constant directivity.
-well, if not going with a compression driver then my suggestion is easy 🙂 :
https://essspeakers.store/collectio...air-of-great-ess-heil-amt-mounting-kit-bundle
That one each time I listened to it in a 2 ways...it never worked fine, whatever says the datasheet, it should not be cutted off that low... 1800 hz mini beginns to be its confort zone...
around 1.2 kHz elliptical
https://techtalk.parts-express.com/forum/tech-talk-forum/43372-cauer-elliptic-filter-based-crossover
That should stop the resonance and that "brittle" sound when it's pushed to low and loud.
Getting the transition (off-axis) from 12" (if not a 3-way) to the Great Heil with and elliptical is critical, it might be lower than 1.2 kHz - but because the high-pass is that "steep" it shouldn't be a problem. It will also be critical to get the loading correct on the 12" (..can't be to over-damped nor under-damped).
https://techtalk.parts-express.com/forum/tech-talk-forum/43372-cauer-elliptic-filter-based-crossover
That should stop the resonance and that "brittle" sound when it's pushed to low and loud.
Getting the transition (off-axis) from 12" (if not a 3-way) to the Great Heil with and elliptical is critical, it might be lower than 1.2 kHz - but because the high-pass is that "steep" it shouldn't be a problem. It will also be critical to get the loading correct on the 12" (..can't be to over-damped nor under-damped).
Although the BMS 4550 is a great sounding driver, it's annular diaphragm does not have the displacement capability to keep up with a 1" exit 2" diaphragm like the Eminence PSD 2002 when crossed low.
https://www.diyaudio.com/community/threads/high-frequency-compression-driver-evaluation.212240/
In the 2 meter dual sine wave “lo” test (about one watt at 523 & 932 Hz, two watts total), second harmonic distortion (2ndHD) of 530 Hz was >32% for the BMS 4550 at 85.2 dB.
The PSD2002 was a surprise, with 6 dB more power, it hit 95.7 dB on the same test at 8% 2nd HD.
In the dual sine wave “mid” test using 1046 & 1865 Hz, there was little difference in clean output between the two drivers.
For home use with a passive crossover in the 800-1000Hz range I'd prefer the extended smooth response of the BMS 4550 over the PSD2002, but there was no question that the PSD2002 could hang together to a lower frequency.
Art
Art, may I ask your extensive experience, please ? Knowing the BMS 4550, do you believe it can be as clean at ears for my modest domestic use than for illustration, a classic 6" or 8" PA + tweeter?
I know the main interest is the controlled directivity of good horns, but is the THD will make a worse difference at average, 75/80 dB spl over a pro Mid made to be clean ? I believe a 2 ways passive has a theorical advantage, but how is the reality for hifi use in the sofa at 3 meters ?
I know the main interest is the controlled directivity of good horns, but is the THD will make a worse difference at average, 75/80 dB spl over a pro Mid made to be clean ? I believe a 2 ways passive has a theorical advantage, but how is the reality for hifi use in the sofa at 3 meters ?
There are some very positive comments about the BMS drivers in this thread, the 4538 is like the 4550 without the xmax.
https://www.diyaudio.com/community/threads/your-favorite-1-cd-for-1-5khz-and-up.337608/post-5784369
The THD of compression drivers is predominantly 2nd order the higher orders are generally very low and you will see that in the test you linked before. So a high THD number with a compression driver does not mean there is a lot of audible distortion.
https://www.diyaudio.com/community/threads/your-favorite-1-cd-for-1-5khz-and-up.337608/post-5784369
The THD of compression drivers is predominantly 2nd order the higher orders are generally very low and you will see that in the test you linked before. So a high THD number with a compression driver does not mean there is a lot of audible distortion.
🙂
... and Veeneth recent ST260 input is promising & encouraging !
Could it be made passive with the less flat BMS 4550 according to you ? I am not sure to understand if such fine EQ can be made passive after the computer modelling made in Vituix active filter. Is there an active to passive filter translator in Vituix that model the filter with the parts values needed or has one to re do it with the passive Vituix design tool since knowing the cut-offs and EQ positions ?
... and Veeneth recent ST260 input is promising & encouraging !
Could it be made passive with the less flat BMS 4550 according to you ? I am not sure to understand if such fine EQ can be made passive after the computer modelling made in Vituix active filter. Is there an active to passive filter translator in Vituix that model the filter with the parts values needed or has one to re do it with the passive Vituix design tool since knowing the cut-offs and EQ positions ?
There isn't anything less flat about the BMS driver it depends heavily on the waveguide.
Here is a measurement (taken from stereonet) of one on the nice QSC waveguide you can't get anymore. If you can't make a passive crossover for a response like that you need to avoid passive crossovers 🙂
For comparison this is Vineeth's with all the EQ taken away, different scales but hopefully you can optically adjust for that
Here is a measurement (taken from stereonet) of one on the nice QSC waveguide you can't get anymore. If you can't make a passive crossover for a response like that you need to avoid passive crossovers 🙂
For comparison this is Vineeth's with all the EQ taken away, different scales but hopefully you can optically adjust for that
No, in a way that doesn't make much sense... however there is an optimiser which allows you to specify a target and it will try to get there.
The problem with an optimiser is you need to know the type of circuit to use or it will not fit. Therefore an optimiser works best when you already don't need one.
For those interested, here is an easy tutorial on how a response shape is created from the circuit type used - https://www.diyaudio.com/community/...overs-without-measurement.189847/post-6910468
There is this that you can try to at least get you in the ballpark but be cautious that it might give you some things that aren't ideal from a practical perspective even though the response looks good.
https://www.diyaudio.com/community/threads/automatic-crossover-designing-with-xmachina.316283/
https://www.diyaudio.com/community/threads/automatic-crossover-designing-with-xmachina.316283/
Thanks AllenB and fluid for the inputs. Seems I have a lot of homeworks going for the winter!
Just as an example with Vineeth's data it doesn't take that much to get a good response to match an LR4 at 1200Hz. There is no impedance data so it is not reliable but as a basic example I think it illustrates the point.
Just in case the impedance data is helpful, Here is a comparison between the impedance of the compression driver as it is and with the ST260KVAR horn attached to it 🙂
Blue is impedance of the CD
Green is impedance after attaching it to horn (in its current form bare printed in PLA form. Some postprocessing work on the horn itself is pending like sealing minute gaps after glueing etc)
Blue is impedance of the CD
Green is impedance after attaching it to horn (in its current form bare printed in PLA form. Some postprocessing work on the horn itself is pending like sealing minute gaps after glueing etc)
Thanks for chimming in Vineeth. I have to follow the ATH4 thread as well.
At 75/80 dB SPL, crossed at around 1000 Hz into a 90 degree nominal horn, the BMS 4550's would be clean enough for most anyone I've ever met. Even at 3 meters, I'd expect fewwould notice any distortion at levels approaching 95 dBA (slow).
At levels approaching 95 dBA (slow) room flutter echos and ear overload, not distortion, would have me putting on ear protection.
The ATH4 thread has many horn models simulated, pay attention to the excursion vs frequency, make sure it does not exceed 0.4 mm at the SPL and frequency you want to achieve, but keep in mind that the BMS 4550 Sd may be more or less than the drivers simulated.
Art
Hi,
Considering the BMS4550 8 ohms or 16 ohms (too much high impedance with 16 ohms for electrical phase matching with the 8 ohms woofer?) model with a 24 dB passive around 900 hz, the ST280E is maybe a sligthy wizer choice than the ST260 ? Not sure I understood all about the beaming and impedance. I looked at Marcel both horns numbers here. Following Art's advices about max 0.4 mm diagphragm displacement max for average home needs, it seems ok (again BMS says 800 hz capabilities but some noticed increasing THD at 800 hz. The Faital 12PR320 can handle a 900 hz as most run it with a horn between 1200 to 1500 hz.
But maybe 800 hz is still better for a smooth overlaping according what you said in a post above ?
Also I made a quotation, but not sure which resin to choose. JLCPCB has several choice and the one named 8000 seems to be a good Q/P trade off looking at strength/thoughness (look at pdf)?
So for now, I am at the phase of choosing the best trade off for such cut-off between the ST260 and the ST280E. I am not confident about my skills for the CE360 petals assembly!
Any further tougths welcome 🙂
Attachments
Last edited:
Not understanding your "electrical phase matching" question, acoustical phase match in the crossover region is the goal.
The smoother impedance, frequency response and low frequency output of the ST280E would simplify a passive crossover design.
The HF sensitivity of the ST280E at 20kHz appears to be around 95dB, about the same as the midband sensitivity of the Faital 12PR320.
The simulations show driver impedance above 16 ohms, probably a nominal 16 ohm driver.
The midband sensitivity of the ST280E could be padded down to 95dB, and a HF pad bypass capacitor added to flatten on axis response. The lower DI ST260 would have a bit more HF drop, the 3dB sensitivity gain of using an 8 ohm driver might be considered.
If Marcel's simulations are not using the BMS 4550 specifications, or the same diaphragm Sd, they may be over or under rating excursion.
A 0.5mm excursion is about when most HF diaphragms will contact (hammer..) the phase plug, the simulation shows 0.5mm excursion for the ST260 at about 700Hz at 110dB. In contrast, the Faital 12PR320 could reach 124 dB (or more) at that frequency within linear expectations.
It's interesting to consider clapping your hands together can reach instantaneous peaks over 120dB SPL at one meter.
Art
Hello Art and thanks,
I meant the BMS4550 has both 8 and 16 ohms version. the 12PR320 will have a low end when loaded near 4 ohms. Maybe better to choose the BMS4550 in 8 ohms not to have too much impedance climbing if the loudspeaker is not bi-amped. Is the acoustical phase doesn't suffer at the crossover frequency when the impedance of the two drivers are too much different, or you doesn't care and just ajuste the phase with the filter parts and horizontal setup of the horn vis à vis of the woofer ?
Thanks also for the ST280E thougth. I just looked the fact it seems to have a better lower constant directivity behavior but missunderstanding what fluid told about more beamier VS the ST260 and how it translates at listening spot.
About Marcel sims, I do not know what 1" driver is used for the exursion sims. I assume it is the same diag. Sd as the BMS4550 as it is a comon size (1.75"). AT 0.1 mm/W at 900 hz on the sim, Is it safe enough. With my average listening level and the limited behavior of reccording, I am not sure to need more what I planned. I read somewhere if my memory is correct than most of the discs do not have more than 6 dB dynamic behavior above the average level of the reccording, but dunno for sure if true. If it goes more than 109 db peaks, I surmise you can only handle it if it is very short transcient not to damagingg the ears and also than THD is limiting the confort ? (The less the percieved THD, the more you pump up the volume ?!).
So if I sumarize as I go a little in alll the direction at the same time : 12PR320/8 ohms I own with the ST280E at 900 hz should be easier to work with passive filtering and has the advantage of lower constant directivity control too. BMS4550 in 8 homs the good choice for a little more spl margin if I understand what you meant (but not critical anyway as it needs padding).
I know Zvu member had a project with the same Faital and the ST260 but do not know how he ended to cross over the two units. Mabe a 1100/1200 hz, I saw it in the ATH4 mabat thread.
I wonder how much spl can reach the 12PR320 at 100 hz as for 2.83V the spl is nearer to 91/92 dB... rising to 95/97 dB towards the 700 hz frequency you say for the 124 dB peaks. Also Dickason said the Xmax calculated by Faital is not the same as his own method. The 7.5 mm Xmax of the datasheet coulb be nearer to 5 mm Xmax. But I also know in PA the Xmax is often outdriven (Xvar number is quite superior). But suffice to say its limits will be hited way after the BMS4550.
Mbrennwa member in all his measurememt (and Paul Vancluysen Leap sims) from the OSMC project make it flat after filtering around 91 dB average.
@fluid , ultimate thougth between the ST260 and the ST280E and final cut off choice here ?
I meant the BMS4550 has both 8 and 16 ohms version. the 12PR320 will have a low end when loaded near 4 ohms. Maybe better to choose the BMS4550 in 8 ohms not to have too much impedance climbing if the loudspeaker is not bi-amped. Is the acoustical phase doesn't suffer at the crossover frequency when the impedance of the two drivers are too much different, or you doesn't care and just ajuste the phase with the filter parts and horizontal setup of the horn vis à vis of the woofer ?
Thanks also for the ST280E thougth. I just looked the fact it seems to have a better lower constant directivity behavior but missunderstanding what fluid told about more beamier VS the ST260 and how it translates at listening spot.
About Marcel sims, I do not know what 1" driver is used for the exursion sims. I assume it is the same diag. Sd as the BMS4550 as it is a comon size (1.75"). AT 0.1 mm/W at 900 hz on the sim, Is it safe enough. With my average listening level and the limited behavior of reccording, I am not sure to need more what I planned. I read somewhere if my memory is correct than most of the discs do not have more than 6 dB dynamic behavior above the average level of the reccording, but dunno for sure if true. If it goes more than 109 db peaks, I surmise you can only handle it if it is very short transcient not to damagingg the ears and also than THD is limiting the confort ? (The less the percieved THD, the more you pump up the volume ?!).
So if I sumarize as I go a little in alll the direction at the same time : 12PR320/8 ohms I own with the ST280E at 900 hz should be easier to work with passive filtering and has the advantage of lower constant directivity control too. BMS4550 in 8 homs the good choice for a little more spl margin if I understand what you meant (but not critical anyway as it needs padding).
I know Zvu member had a project with the same Faital and the ST260 but do not know how he ended to cross over the two units. Mabe a 1100/1200 hz, I saw it in the ATH4 mabat thread.
I wonder how much spl can reach the 12PR320 at 100 hz as for 2.83V the spl is nearer to 91/92 dB... rising to 95/97 dB towards the 700 hz frequency you say for the 124 dB peaks. Also Dickason said the Xmax calculated by Faital is not the same as his own method. The 7.5 mm Xmax of the datasheet coulb be nearer to 5 mm Xmax. But I also know in PA the Xmax is often outdriven (Xvar number is quite superior). But suffice to say its limits will be hited way after the BMS4550.
Mbrennwa member in all his measurememt (and Paul Vancluysen Leap sims) from the OSMC project make it flat after filtering around 91 dB average.
@fluid , ultimate thougth between the ST260 and the ST280E and final cut off choice here ?
There are three DI curves calculated at different angles. The 20 deg ones are very similar in both, the ST260 has all three DI graphs being very similar the ST280E has the curves diverging at HF. You can see in the normalized graphs that the zero degree and near axis curves are trending up over flat in the ST280. So if you flatten the 280 at 20 deg the 0 deg angle will be bright.
If you set the speakers up with high toe in so the 20 degree axis points at the listening position, (with the ST280 tuned as in the graph) some of the sound sent to other angles will be brighter. I find this to sound a bit odd when the reflections can be brighter than the listening axis, but having said that it is only a dB or two at quite high frequency so it may not be a problem in this situation.
It is just a generic representation of a compression driver, I wouldn't try and draw too many conclusions from that aspect.
If you are set that you want to cross low then the ST280 will help to make that easier. As always it is a matter of balancing the priorities and choosing which you prefer.