Mark,
The B&C 4NDF34 sensitivity used to be appropriately rated at 88 dB sensitivity, then it was inflated to 92dB
.
I still prefer to see both the impedance and raw response of the device driven with a voltage appropriate to it’s nominal impedance that equals 1 watt, rather than an “averaged” post-processing single number using what amounts to an arbitrary voltage (2.83 volts = 1 watt into an 8-ohm load) for drivers of different nominal impedance. I've re-posted your sensitivity chart with the "1 watt" equivalent sensitivity, though the actual impedance is a question.
You have chosen appropriate bandwidths for your drivers, if that had not been done, an averaged post processing sensitivity could give an inflated idea of the usable potential.
How does the quad B&C 4NDF34-16 and BMS 4594HE impedance on the SYN9 compare to the 4594HE (on an unnamed 90 x60 horn) posted below?
Art
The B&C 4NDF34 sensitivity used to be appropriately rated at 88 dB sensitivity, then it was inflated to 92dB
.I still prefer to see both the impedance and raw response of the device driven with a voltage appropriate to it’s nominal impedance that equals 1 watt, rather than an “averaged” post-processing single number using what amounts to an arbitrary voltage (2.83 volts = 1 watt into an 8-ohm load) for drivers of different nominal impedance. I've re-posted your sensitivity chart with the "1 watt" equivalent sensitivity, though the actual impedance is a question.
You have chosen appropriate bandwidths for your drivers, if that had not been done, an averaged post processing sensitivity could give an inflated idea of the usable potential.
How does the quad B&C 4NDF34-16 and BMS 4594HE impedance on the SYN9 compare to the 4594HE (on an unnamed 90 x60 horn) posted below?
Art
Attachments
Mark,
The B&C 4NDF34 sensitivity used to be appropriately rated at 88 dB sensitivity, then it was inflated to 92dB
I still prefer to see both the impedance and raw response of the device driven with a voltage appropriate to it’s nominal impedance that equals 1 watt, rather than an “averaged” post-processing single number using what amounts to an arbitrary voltage (2.83 volts = 1 watt into an 8-ohm load) for drivers of different nominal impedance. I've re-posted your sensitivity chart with the "1 watt" equivalent sensitivity, though the actual impedance is a question.
You have chosen appropriate bandwidths for your drivers, if that had not been done, an averaged post processing sensitivity could give an inflated idea of the usable potential.
How does the quad B&C 4NDF34-16 and BMS 4594HE impedance on the SYN9 compare to the 4594HE (on an unnamed 90 x60 horn) posted below?
Art
Hi ya Art,
Well, i don't much know about OEM sensitivity ratings...cause i generally think they are full of crap...but i do know what i measure.
I think trying to gauge sensitivity off 1 watt nominal impedance, and then extrapolating that onto the impedance curve ,is the utter root of the error in sensitivity specs.
I personally wish wattage specs would utterly disappear....
And I mean, who cares about eyeballed sensitivity off a response curve, often out of bandwidth.....
Processed and flattened.
(I do like to see the processing curve to know how much boost might being applied to hokey up the spec.)
Straight to the point, who cares about anything other than linear usage, flattened response, within the intended bandwidth?
I don't get it ....Why do we make this so hard? .....
When linear response sensitivity through the intended bandwidth is so easy to measure?
Hi Mark and Art
Maybe we are using sensitivity-ratings to figure out different things, maybe?
For me it´s not about, how loud can it get. I was asking for average sensitivity to figure out how the combination of four drivers in parallel works with a given amplifier, for instance in respect to hiss or other issues at low volume. I guess that there is a fixed ratio when the drivers are coupled directly to the amp, and post-processing will not lower an eventual hiss/noise-issue, rather do the opposite. I will likely not surpass 1 Watt very often in my living-room!
Actually I had just read on a germane forum about noise-issues with highly efficient CD´s and amplifiers, making me nervous!
It´s really just speculation and uncertainty from not having build and tried things my self! When I have my first MEH done I will know a lot more and have my own experience to draw on. But thanks again for answering and sharing.
I will re-attach the data-shet I got from B&C for the 4NDF34-16, then we have it on the same page. It says 91 dB/W/m at 4 Volt.
Steffen
Maybe we are using sensitivity-ratings to figure out different things, maybe?
For me it´s not about, how loud can it get. I was asking for average sensitivity to figure out how the combination of four drivers in parallel works with a given amplifier, for instance in respect to hiss or other issues at low volume. I guess that there is a fixed ratio when the drivers are coupled directly to the amp, and post-processing will not lower an eventual hiss/noise-issue, rather do the opposite. I will likely not surpass 1 Watt very often in my living-room!
Actually I had just read on a germane forum about noise-issues with highly efficient CD´s and amplifiers, making me nervous!
It´s really just speculation and uncertainty from not having build and tried things my self! When I have my first MEH done I will know a lot more and have my own experience to draw on. But thanks again for answering and sharing.I will re-attach the data-shet I got from B&C for the 4NDF34-16, then we have it on the same page. It says 91 dB/W/m at 4 Volt.
Steffen
Last edited:
Linear response sensitivity is easy to measure at 2.83 volts, the problem is it may not be linear at a voltage well within rated power specifications in the bandwidth a design has called for.
We want to see the processed curve, but most packaged systems won’t show it, often for good reason :^).
Without seeing individual raw frequency and impedance response curves, really hard to determine if the processing chosen would be appropriate for it’s intended usage
.Linear response sensitivity is easy to measure at 2.83 volts, the problem is it may not be linear at a voltage well within rated power specifications in the bandwidth a design has called for.
We want to see the processed curve, but most packaged systems won’t show it, often for good reason :^).
Without seeing individual raw frequency and impedance response curves, really hard to determine if the processing chosen would be appropriate for it’s intended usage
Yep to all that.
We need both raw and processed curves, as well as measured sensitivity with whatever processed curve is given by the driver manufacturer (which of course includes all EQs and any high pass &/or low pass in place.)
Personally, i think the 2.83V processed curve i get after tuning, should stay linear all the way to rated wattage........IF....
I've done my homework correctly and I'm only trying to use the driver within the bandwidth that it can do such.
Tis the point of multi-way, imo,... to be able to use drivers in the meat of their operating range, and put together a fully linear, high-dynamic & SPL capable system.
Mark,
The “homework” you have done demonstrates that very expensive co-axial mid/high compression drivers rated for hundreds of “watts” as low as 300 Hz can be beaten in terms of linear output, better sound, and more SPL by relatively inexpensive cone drivers loaded on the same horn in the 300-1kHz range.
Art
The “homework” you have done demonstrates that very expensive co-axial mid/high compression drivers rated for hundreds of “watts” as low as 300 Hz can be beaten in terms of linear output, better sound, and more SPL by relatively inexpensive cone drivers loaded on the same horn in the 300-1kHz range.
Art
update....
Just finished a pair for stereo....first stereo i've heard since syn9 project started back in July ..yikes...
Ended up, after tons more listening and measuring the syn9 proto, and the versions 2 and 3,
with a 90x60 pattern, 36" x 22". And no secondary flares.
First task was to play with temporary secondary flares, felt around mouth edges, felt in horn various places, etc ....all in attempt to get polars as nice as possible. Well, all that stuff helps...helps the measurements at least, but i'll be damned if i ever really heard any improvement.
So to heck with secondary flares !!!!! Was time to simply compare the versions "as is".
The 60x60 version 2 was the easiest to eliminate. Sound simply seems to come from too deep in the horn. I think Ro808 has described it as a "cupped" sound. My guess is that it would work better in stereo than mono, putting more apparent sound between two speakers.
Harder choice was between 75x60 version 3, and the rag-tag thrown-together 90x60 thread-starting proto.
On paper, for SQ, this should have easily been won by version 3, which has 12pr320's and 4mdf34's, vs the proto's lowly 10fe200's and 4fe35's.
Not so easy.
Believe it or not, and despite nearly identical mag & phase, and very similar polars within +/-30 degrees, the proto out-punched the much stronger version 2 in the low 100Hz area.
Both boxes had their low ports close to the same distance from the throat, about 11 inches.
Version 2 used the traditional placement with 2 ports per driver in the horn corners.
The proto used my favored method of a single large port as close to cone center as possible
So I decided to go back to the single, centered port, strategy.
And keep the box 90x60 which allows a smaller overall size and weight than the 75x60 pattern.
Anyway, finished pair are using a bms 4594he, four 4mdf34's, and two 10pr300's (from 12 inches to help reduce size and weight)
Also found the mids, the 4mdf34's sound better to me unenclosed, open baffle so to speak. Since they do not share a cavity with the 10" lows, it works fine other than leaving them exposed.
Here's couple of picts, and the sections' response curves. xovers are 96dB/oct LR, ...120Hz, 300Hz, 700Hz, 6300Hz. (sub not shown)
Using 4 sections, and only in the meat of their bandwidth, required the least EQ i've experienced...could almost simply use steep linear phase xovers without any EQs.
I'm super happy with the clarity, imaging, spl, and dynamics.....
...and importantly for my indoor & outdoor listening.... size and weight are easily manageable....came in at only 49 lbs.
Just finished a pair for stereo....first stereo i've heard since syn9 project started back in July ..yikes...
Ended up, after tons more listening and measuring the syn9 proto, and the versions 2 and 3,
with a 90x60 pattern, 36" x 22". And no secondary flares.
First task was to play with temporary secondary flares, felt around mouth edges, felt in horn various places, etc ....all in attempt to get polars as nice as possible. Well, all that stuff helps...helps the measurements at least, but i'll be damned if i ever really heard any improvement.
So to heck with secondary flares !!!!! Was time to simply compare the versions "as is".
The 60x60 version 2 was the easiest to eliminate. Sound simply seems to come from too deep in the horn. I think Ro808 has described it as a "cupped" sound. My guess is that it would work better in stereo than mono, putting more apparent sound between two speakers.
Harder choice was between 75x60 version 3, and the rag-tag thrown-together 90x60 thread-starting proto.
On paper, for SQ, this should have easily been won by version 3, which has 12pr320's and 4mdf34's, vs the proto's lowly 10fe200's and 4fe35's.
Not so easy.
Believe it or not, and despite nearly identical mag & phase, and very similar polars within +/-30 degrees, the proto out-punched the much stronger version 2 in the low 100Hz area.
Both boxes had their low ports close to the same distance from the throat, about 11 inches.
Version 2 used the traditional placement with 2 ports per driver in the horn corners.
The proto used my favored method of a single large port as close to cone center as possible
So I decided to go back to the single, centered port, strategy.
And keep the box 90x60 which allows a smaller overall size and weight than the 75x60 pattern.
Anyway, finished pair are using a bms 4594he, four 4mdf34's, and two 10pr300's (from 12 inches to help reduce size and weight)
Also found the mids, the 4mdf34's sound better to me unenclosed, open baffle so to speak. Since they do not share a cavity with the 10" lows, it works fine other than leaving them exposed.
Here's couple of picts, and the sections' response curves. xovers are 96dB/oct LR, ...120Hz, 300Hz, 700Hz, 6300Hz. (sub not shown)
Using 4 sections, and only in the meat of their bandwidth, required the least EQ i've experienced...could almost simply use steep linear phase xovers without any EQs.
I'm super happy with the clarity, imaging, spl, and dynamics.....
...and importantly for my indoor & outdoor listening.... size and weight are easily manageable....came in at only 49 lbs.
Hi Mark
Congratulations to Stereo. Nice to hear something new.
Interesting that you have chosen against secondary flares and felt. I guess you have tried the towel-trick?
What made you choose the 10PR300 over the 10PR320? Is that part of your choice to raise the crossover-frequency from 100 Hz to 125 Hz? And why did you choose to raise the crossover-frequency?
What is the max spl with 125 Hz and 10PR300?
Steffen
Congratulations to Stereo. Nice to hear something new.
Interesting that you have chosen against secondary flares and felt. I guess you have tried the towel-trick?
What made you choose the 10PR300 over the 10PR320? Is that part of your choice to raise the crossover-frequency from 100 Hz to 125 Hz? And why did you choose to raise the crossover-frequency?
What is the max spl with 125 Hz and 10PR300?
Steffen
Hi Steffen ! Thx
Yep, i tried the towel trick, the felt trapeziums trick, every dang 'trick' i've seen on the net..
Yeah, they all help some ....well like said, they helped the measurements at least.... Lol
(I'm beginning to think CEA2034 has got us all drunk on working for perfect polars ....haha)
I would rather have the 10PR320's. I got the 300's back for syn7 from Part's Express, when the 320's weren't in stock. Wound have needed to buy 320's from Europe.
The 300's play fine down to 100Hz. In fact, i redid that xover point right after posting this morning. Funny, & cool, you bring it up !
Using the piston excursion calculator, and 5mm xmax, i get about 120 dB @ 100Hz, and 125 dB @ 125Hz.
I don't know for certain if there's is any acoustic gain from the 8:1, sd to port ration, or any horn gain.
But i think so, ..... guessing around 2-3dB extra above the calculator.
Yep, i tried the towel trick, the felt trapeziums trick, every dang 'trick' i've seen on the net..
Yeah, they all help some ....well like said, they helped the measurements at least.... Lol
(I'm beginning to think CEA2034 has got us all drunk on working for perfect polars ....haha)
I would rather have the 10PR320's. I got the 300's back for syn7 from Part's Express, when the 320's weren't in stock. Wound have needed to buy 320's from Europe.
The 300's play fine down to 100Hz. In fact, i redid that xover point right after posting this morning. Funny, & cool, you bring it up !
Using the piston excursion calculator, and 5mm xmax, i get about 120 dB @ 100Hz, and 125 dB @ 125Hz.
I don't know for certain if there's is any acoustic gain from the 8:1, sd to port ration, or any horn gain.
But i think so, ..... guessing around 2-3dB extra above the calculator.
Hi again
Where do I find that excursion calculator? That could be a useful tool for me, when "breaking in" drivers. I think Art Welter says that is a very good thing to do, to soften the surrounds etc.
When you did SYN9v3 you said that it was not so important, that the porthole is NOT centered over the driver. Have you changed opinion on that matter?
Where do I find that excursion calculator? That could be a useful tool for me, when "breaking in" drivers. I think Art Welter says that is a very good thing to do, to soften the surrounds etc.
When you did SYN9v3 you said that it was not so important, that the porthole is NOT centered over the driver. Have you changed opinion on that matter?
here ya go http://www.baudline.com/erik/bass/xmaxer.html
Yes, transfer function measurements of syn9v3 did not show any difference ..... but i was looking for how the ports' location effected the polars of the CD, not really looking hard at the response of the low driver.
We know that the low port distance from throat determines low end loading. https://www.danleysoundlabs.com/wp-content/uploads/2012/01/The-Tapped-Horn.pdf
I thought bottom end response would be the same, ports in corners vs ports in center, as long as the distance is the same.
Maybe it is, but it doesn't sound like it ! The single center-port plain sounds punchier, more dynamic, to me.
I think DSL might be getting some punch from the reflex ports they use.....ports i've been trying to avoid if possible.
Yes, transfer function measurements of syn9v3 did not show any difference ..... but i was looking for how the ports' location effected the polars of the CD, not really looking hard at the response of the low driver.
We know that the low port distance from throat determines low end loading. https://www.danleysoundlabs.com/wp-content/uploads/2012/01/The-Tapped-Horn.pdf
I thought bottom end response would be the same, ports in corners vs ports in center, as long as the distance is the same.
Maybe it is, but it doesn't sound like it ! The single center-port plain sounds punchier, more dynamic, to me.
I think DSL might be getting some punch from the reflex ports they use.....ports i've been trying to avoid if possible.
Makes sense due to driver upper harmonics are centered around the VC Vs out at the edge where DSL puts them for max practical compression ratio, the ~ default choice for max performance same as the 1st thing done historically to 'hotrod' a motor's internals [thinner head gaskets and/or 'shaved' heads ].
].
Last edited:
Aah, I see what you're saying. Thx, that does make sense.Makes sense due to driver upper harmonics are centered around the VC Vs out at the edge where DSL puts them for max practical compression ratio, the ~ default choice for max performance same as the 1st thing done historically to 'hotrod' a motor's internals [thinner head gaskets and/or 'shaved' heads
Aarg, I´ve read it several times, and I don´t "see it"!? Could I have a bit more explanation. Thx.
Here's my take. I'll let GM correct what's wrong.....
......I have this flexible piston (cone) making sound. What part of the cone is making what sound?
We've probably seen how cones' outer zones are where higher freq breakups, harmonic distortion, etc occurs.
And if you're like me have said 'so what, i'm not pushing the driver that hard....distortion isn't a problem'.
Well, maybe that's true, until using compression. Maybe the compression, right in the weakest area of the cone from the motor's control viewpoint, makes for a less dynamic response. It makes sense to me that ports in the corners beg for a really stiff *** suspension
......I have this flexible piston (cone) making sound. What part of the cone is making what sound?
We've probably seen how cones' outer zones are where higher freq breakups, harmonic distortion, etc occurs.
And if you're like me have said 'so what, i'm not pushing the driver that hard....distortion isn't a problem'.
Well, maybe that's true, until using compression. Maybe the compression, right in the weakest area of the cone from the motor's control viewpoint, makes for a less dynamic response. It makes sense to me that ports in the corners beg for a really stiff *** suspension
Hi Mark
Thanks for the Link to the excursion calculator.
Hmm, we are back in speculation-territory! I have this idea about differences in sound in regard to dynamics:
The woofer can "breathe" more freely with a single port centered over the voice coil than with two smaller ports over the surrounds (combined port area the same as the one port-option).
I have this imagination, that the port area over the surrounds actually varies depending on the position af the piston/cone, whether its pulling in (negativ current) or pushing out (positiv current), like the cone is partially closing the port if taken to extremes! The air can not pass as freely over the surround and in/out as in the middle over the voice coil!?
Then there is the aspect of symmetry/asymmetry of airflow in the cavity between horn and cone? Maybe the two options have different fase-responses?
Whether there is an asymmetric loading of the cone with the ports over the surround, like flexing the cone, is hard to know, or if there is any asymmetric impact on the surround!? I can´t really determine. This is really getting speculative! And I have to admit, that I do not have any practical experience with this yet.
I have never really been happy with the "one woofer/two ports over the surround" thing. I find it intuitively preferable with the port over the voice coil if possible. That is also the reason that I have been contemplating on using four 16 ohm woofers with ports in the corners and centered over he voice coil.
Sorry for all these words! I think I am seriously infected with a mad-MEH-bug !
Steffen
Thanks for the Link to the excursion calculator.
Hmm, we are back in speculation-territory
! I have this idea about differences in sound in regard to dynamics:The woofer can "breathe" more freely with a single port centered over the voice coil than with two smaller ports over the surrounds (combined port area the same as the one port-option).
I have this imagination, that the port area over the surrounds actually varies depending on the position af the piston/cone, whether its pulling in (negativ current) or pushing out (positiv current), like the cone is partially closing the port if taken to extremes! The air can not pass as freely over the surround and in/out as in the middle over the voice coil!?
Then there is the aspect of symmetry/asymmetry of airflow in the cavity between horn and cone? Maybe the two options have different fase-responses?
Whether there is an asymmetric loading of the cone with the ports over the surround, like flexing the cone, is hard to know, or if there is any asymmetric impact on the surround!? I can´t really determine. This is really getting speculative! And I have to admit, that I do not have any practical experience with this yet.
I have never really been happy with the "one woofer/two ports over the surround" thing. I find it intuitively preferable with the port over the voice coil if possible. That is also the reason that I have been contemplating on using four 16 ohm woofers with ports in the corners and centered over he voice coil.
Sorry for all these words
! I think I am seriously infected with a mad-MEH-bug !Steffen
Very nice results. I look at your horn on the screen, then glance over at mine. Yours is only 20% wider but almost twice as tall. So I'm familiar with the polar response limitations. Like you I forego a secondary flare, just equalizing in room for my primary listening position, knowing the listening window width could be improved by a secondary flare but not caring. Is your decision based on similar thinking or are you saying that +/- 3db FR ripple is not or is just barely audible?
I'd be curious how the polars differed. Perhaps the sound didn't change because off axis the SPL was so low it didn't contribute meaningfully to the combined SPL at the listening position? So a little more ripple might not be audible? I just don't know how low that off axis energy needs to be to behave this way.
I have this imagination, that the port area over the surrounds actually varies depending on the position af the piston/cone, whether its pulling in (negativ current) or pushing out (positiv current), like the cone is partially closing the port if taken to extremes! The air can not pass as freely over the surround and in/out as in the middle over the voice coil!?
Then there is the aspect of symmetry/asymmetry of airflow in the cavity between horn and cone? Maybe the two options have different fase-responses?
Whether there is an asymmetric loading of the cone with the ports over the surround, like flexing the cone, is hard to know, or if there is any asymmetric impact on the surround!? I can´t really determine. This is really getting speculative! And I have to admit, that I do not have any practical experience with this yet.
I have never really been happy with the "one woofer/two ports over the surround" thing. I find it intuitively preferable with the port over the voice coil if possible. That is also the reason that I have been contemplating on using four 16 ohm woofers with ports in the corners and centered over he voice coil.
Sorry for all these words
Steffen
Yep Steffen, we are in pure speculative territory. Glad it's just not me beating my brains out, with the mad-MEH-bug ! Lol
I took the drivers off ver2 &3 with the low corner ports, and put the builds in storage. I may pull one box back out, and just mount the low drivers.
I'm thinking maybe a wavelet burst test might show a difference on how the captured wavelet looks. I'm thinking to test / compare low freq bursts from the different setups, with just the low drivers being driven...