For the 4mdn34's, i recessed 3 7/8" diameter, 6mm deep.
Both to allow for excursion, and to reduce the thickness of the ports. (Horn is 12mm BB)
The ports are 15/16" diameter, with centers at 3 1/2" from throat.
They lie tangent to the horn's top and bottom flares as seen from front view posted earlier.
One quick indoor measurement showed no change to their response when enclosed.
Which surprised me, as I thought something would change....but no....
A similar surprise occurred testing different 12"s.....kappalite 3012LF, faital 12pr320, and rcf mb12n351.....I could not tell their response traces apart.
The horn seems to dominate things somehow....
Making sawdust is a truth serum, huh?
So is listening...still have no clue why adding the small mids brings so much extra SQ to the party.
Can't see a bit of difference in mag, phase, decay, THD, etc..... between taking the CD straight to 12"s, versus adding the small mids as a bridge.
About to get serious with modulation testing like Art Welter has done.....would love to find some objective answers.
Both to allow for excursion, and to reduce the thickness of the ports. (Horn is 12mm BB)
The ports are 15/16" diameter, with centers at 3 1/2" from throat.
They lie tangent to the horn's top and bottom flares as seen from front view posted earlier.
One quick indoor measurement showed no change to their response when enclosed.
Which surprised me, as I thought something would change....but no....
A similar surprise occurred testing different 12"s.....kappalite 3012LF, faital 12pr320, and rcf mb12n351.....I could not tell their response traces apart.
The horn seems to dominate things somehow....
Making sawdust is a truth serum, huh?
So is listening...still have no clue why adding the small mids brings so much extra SQ to the party.
Can't see a bit of difference in mag, phase, decay, THD, etc..... between taking the CD straight to 12"s, versus adding the small mids as a bridge.
About to get serious with modulation testing like Art Welter has done.....would love to find some objective answers.
Owe a couple of corrections:
In #117 i said displacement increases 16x (for same SPL) per octave decrease.
Meant to say 4x.
(16x error came from thinking in two-octave terms, trying to use each section in 5-way for as close to two-octaves as reasonable.)
And i keep getting the name of the B&C 4" mids wrong.... it's 4NDF34.
Ok, back to current discussion...
Steffen, did you mean (1) volume between the cone and the port?
Or (2) the volume of the rear chamber behind the cone?
Because if (1), that should create a higher frequency low-pass filter.
Since the driver is already strong all the way to 1kHz as is, I haven't bothered with reducing space there. So it was the same for both unenclosed and enclosed.
If (2),
I just ran Hornresp for a sealed 4ndf34....
Shows it only makes some difference below 200Hz, and then very little. So sim ties to experience nicely, i think. Box size doesn't matter here, yea!
In #117 i said displacement increases 16x (for same SPL) per octave decrease.
Meant to say 4x.
(16x error came from thinking in two-octave terms, trying to use each section in 5-way for as close to two-octaves as reasonable.)
And i keep getting the name of the B&C 4" mids wrong.... it's 4NDF34.
Ok, back to current discussion...
Steffen, did you mean (1) volume between the cone and the port?
Or (2) the volume of the rear chamber behind the cone?
Because if (1), that should create a higher frequency low-pass filter.
Since the driver is already strong all the way to 1kHz as is, I haven't bothered with reducing space there. So it was the same for both unenclosed and enclosed.
If (2),
I just ran Hornresp for a sealed 4ndf34....
Shows it only makes some difference below 200Hz, and then very little. So sim ties to experience nicely, i think. Box size doesn't matter here, yea!
Hi Mark
At (2) my thought was/is to create a natural roll-off, to hand over to the woofers. I am still hoping to stay in the IIR-domain and maybe use Chris´s approach? I dunno yet?
I am still trying to wrap my head around, what happens with filters and PEQ´s in regard to phase in SYN´s! What does the bandpass created by the volume under the cone and the port do to phase, I guess it is just behaving like a filter!? What happens to the phase at the notch-frequency? Bla bla bla...... My current knowledge about filters is mainly in respect to Duelund Syncron Filters, and how a loudspeaker with those filters is created. With a Duelund-crossover you create target-curves for each driver that you have to match acoustically, i.e. by combining the acoustical filtering with electrical filtering to generate the desired frequency-response for each driver. If you do it right, the the woofer, midrange and treble combine coherently, because they share the same phase-curve/rotation. Maybe SYN´s are more forgiving in that respect, because the drivers are placed axially and not stacked vertically!?
AAAAARGH I can´t really explain what it is I am trying to figure out! I think I simply need to build something and try things my self, to find out what happens! Asking and
to formulate questions is not that easy, understanding the answer can be equally challenging! I have ordered some books that hopefully can help with my lack of knowledge.
I go to sleep now!
Steffen
At (2) my thought was/is to create a natural roll-off, to hand over to the woofers. I am still hoping to stay in the IIR-domain and maybe use Chris´s approach? I dunno yet?
I am still trying to wrap my head around, what happens with filters and PEQ´s in regard to phase in SYN´s! What does the bandpass created by the volume under the cone and the port do to phase, I guess it is just behaving like a filter!? What happens to the phase at the notch-frequency? Bla bla bla...... My current knowledge about filters is mainly in respect to Duelund Syncron Filters, and how a loudspeaker with those filters is created. With a Duelund-crossover you create target-curves for each driver that you have to match acoustically, i.e. by combining the acoustical filtering with electrical filtering to generate the desired frequency-response for each driver. If you do it right, the the woofer, midrange and treble combine coherently, because they share the same phase-curve/rotation. Maybe SYN´s are more forgiving in that respect, because the drivers are placed axially and not stacked vertically!?
AAAAARGH I can´t really explain what it is I am trying to figure out! I think I simply need to build something and try things my self, to find out what happens! Asking and
to formulate questions is not that easy, understanding the answer can be equally challenging! I have ordered some books that hopefully can help with my lack of knowledge.
I go to sleep now!
Steffen
Last edited:
Hi Mark
I got a data-sheet for the B&C 4NDF34-16 from B&C and today also from TLHP.
That data was not shown on neither B&C´s or TLHP´s homepage.
TLHP says: "The 4NDF34-16 is still in the standard range, we have no information from B&C Speakers, as of today, which precisely indicates that this model is discontinued."
I thought it would be nice to have that information out.
Steffen
I got a data-sheet for the B&C 4NDF34-16 from B&C and today also from TLHP.
That data was not shown on neither B&C´s or TLHP´s homepage.
TLHP says: "The 4NDF34-16 is still in the standard range, we have no information from B&C Speakers, as of today, which precisely indicates that this model is discontinued."
I thought it would be nice to have that information out.
Steffen
Yeah, the whole phase rotation / summation thingy can get confusing.
The good news imo, is once we start trying things out for ourselves and making measurements, it gets clearer quicker.
Single drivers in open air tend to have natural 2nd order roll-offs on both ends of their bandwidth. With these freq response roll-offs comes 2nd order phase rotation.
(Phase is relatively flat within the drivers in-band reponse where magnitude is relatively flat.)
The mids in a MEH have a steeper HF rolloff than 2nd order due to the acoustic lowpass formed by the reduced air volume between the cone and port.
I get about 30dB down, an octave away from the high corner at around 1kHz, with the 4ndf34's.
Which means a boat load of phase rotation above 1kHz...clearly unusable territory.
So it's important to crossover well below that high corner if trying to use low order IIR crossovers.....simply to have room to tie it to the CD in its flatter mag and phase region.
With steep linear phase xovers, the mids can be used all the way up to the high corner if desired.
With regard to what happens to phase at notch frequency, the notch that is made from reflections bounced back from the throat....well, anytime their is rapid mag change, there is rapid phase change.
That's why it's important to xover below notch freq.
How far below again depends on how wide a plateau of relatively flat mag and phase below notch frequency there is to work with.
My take on xovers ....(is first off, folks make them waaaaaaay too difficult, but i'll stay away from there haha)
Ok, my retake on xovers is that they all need to start with a complementary acoustic order in mind. A complementary acoustic order that is achieved after adding electrical filters.
(some folks advocate using non-complementary acoustic orders to steer lobing, but i think that is a lot of overcomplicated horse wallop thinking ..i digress again
)Anyway, like your Dueland example where you start with a target acoustic curve, i always start with a target acoustic curve..... a target acoustic xover.
Whether you massage response towards the target curve using IIR PEQs in-band and out-of-band, with or without "named filters/xovers", or use linear phase xovers, or use any combination of any of those, .....is totally immaterial towards the goal of matching the target curves.
The real decision that matters imho, is choosing appropriate target curves to begin with, that can be realistically reached with the electrical filter toolset available.
Differences in techniques are more about semantics than anything else....all approaches do the same thing if done correctly.
Everyone wants flat mag and as little phase rotation as possible.
Which means keeping order low if using IIR.
Or use linear phase xovers if summation benefits from higher order.
That simple really
Hope all that rambling helped
Been simply enjoying syn9 for a while, and trying to give some time to set in, to hear if anything sounds off and needs attention.....
.... before building another one for stereo, although this time i think i'll build 2 more for a LCR setup.
In the meantime, i've made more outdoor measurements trying to figure out where the extra clarity is coming from.
Dual and multitone are easy enough to make...i just don't understand what to look for yet, or how to compare.
Tried some speech intelligibility measurements, STI and STIPA Speech transmission index - Wikipedia
Got all excited when i measured those great numbers at 3m, so i measured a commercial speaker i use for comparisons, a Meyer UPA-1p, and got the same great numbers.
The Upa's score did degrade faster than the syn9 as i cranked them up, but not enough to make me want to continue with this measurement avenue, for looking for why the syn9 has greater clarity.
Plus, the STI & STIPA measurements require sweeps that are at least 1.6 seconds long and echo slap from any even far off structures becomes unavoidable. And may turn into slaps from neighbors
Anyway, here's a good ole THD measurement taken outdoors at a bit over 3m. SPL is correct, so figure the processed fundamental curve is 100dB @1m.
Seems clean for 100dB for sure, but really not that much cleaner than my other builds...so still don't have any solid clues about clarity gains (other than maybe some really good weed Lol)
.... before building another one for stereo, although this time i think i'll build 2 more for a LCR setup.
In the meantime, i've made more outdoor measurements trying to figure out where the extra clarity is coming from.
Dual and multitone are easy enough to make...i just don't understand what to look for yet, or how to compare.
Tried some speech intelligibility measurements, STI and STIPA Speech transmission index - Wikipedia
Got all excited when i measured those great numbers at 3m, so i measured a commercial speaker i use for comparisons, a Meyer UPA-1p, and got the same great numbers.
The Upa's score did degrade faster than the syn9 as i cranked them up, but not enough to make me want to continue with this measurement avenue, for looking for why the syn9 has greater clarity.
Plus, the STI & STIPA measurements require sweeps that are at least 1.6 seconds long and echo slap from any even far off structures becomes unavoidable. And may turn into slaps from neighbors
Anyway, here's a good ole THD measurement taken outdoors at a bit over 3m. SPL is correct, so figure the processed fundamental curve is 100dB @1m.
Seems clean for 100dB for sure, but really not that much cleaner than my other builds...so still don't have any solid clues about clarity gains (other than maybe some really good weed Lol
)Hi Mark
I can not remember whether or not you have done any polar-measurements on this SYN 9v3.
I would be interested if it has had any positiv effect on the CD´s response, that the holes nearby the throat, i.e. the midrange-ports, dont occupy so much area on the horn-wall, i.e. disturb less? That is off course compared to your SYN 7 and SYN 8 with bigger ports near the throat for the woofers.
Have you had a chance to do Art Welter´s test for intermodulation-distortion?
Steffen
I can not remember whether or not you have done any polar-measurements on this SYN 9v3.
I would be interested if it has had any positiv effect on the CD´s response, that the holes nearby the throat, i.e. the midrange-ports, dont occupy so much area on the horn-wall, i.e. disturb less? That is off course compared to your SYN 7 and SYN 8 with bigger ports near the throat for the woofers.
Have you had a chance to do Art Welter´s test for intermodulation-distortion?
Steffen
Hi Steffen,
Yeah, i've worked on the syn9.3 polars some, but not too much really as it doesn't make sense to put lots of fine tuning effort into polars until secondary flares are added.
Here's two sets made outdoors, one that used an average of measurements for each driver at 0-30 degree increment, as the refence measurements for drivers' individual corrections.
And the second set that used just a 15 deg off-axis measurement as reference for their corrections.
(Followed by comparison of the two tunings at just 15 deg.)
Don't know what was going on at about 3kHz...hadn't seen those oscillations on quickie indoor polars...blame gating haha....
It was pretty noisy outside ....but I don't think that's their cause ...
As far as smaller holes closer to the throat having maybe less influence on polars that bigger ones (when going straight from CD to larger cones).....i dunno....i just don't think the port location matters to polars nearly as much as we tend to think.
I always measure the horn's polars alone before putting any ports in, to know what the polars "best case" can be, after adding ports.
I've kind of come to the conclusion, ports in the corners are more about symmetry and maintaining 1/4WL spacing between all participating ports, than they are about polars.
Besides, i'm getting less and less convinced polars matter anywhere near as much as seems to be currently in fashion.
I know polars definitely matter if there are large or sharp discontinuities in directivity, but otherwise i'm kinda getting reconvinced good processing matters more.
I've played with Art's method for IMD, and want to do more of it....just haven't got a good feel for it yet...
The STI/STIPA is a standard for speech intelligibility often required in public address systems that may be needed in emergencies. (see the wiki link gave earlier).
It uses modulated tone bursts, in an on & off type fashion, to provide a score.
Since the syn9.3 clarity improvement has been most noticeable with lead and backing vocals, i thought a measure of speech intelligibility might be worth a try.
Plus, the second aspect of improved clarity, is the quietness surrounding the music. It's funny, but v9.3 sounds both quieter with louder transients at the same time, and with an overall sense of making less average SPL.
I keep turning up the volume, and it stays great sounding (as in better and better)..without sounding loud.
But when i put the SPL meter on it, it's like wow this is really measuring louder than i would have expected.
Hope that made sense...
Mark,
Although I still think IM distortion in the 300-400 Hz range of the compression driver midrange compared to the cone midrange is the probable cause of the “extra clarity”, another possible cause could be dynamic limiting of the DCX464 co-ax LF annular diaphragm suspension.
If dynamic LF limiting is occurring, louder LF peaks would sound “thinner” than they should.
Have you compared it’s LF response at varying drive levels to determine it’s linearity?
In other words, does a tone burst at 2.83 volts compared to 28.3 volts drive at 300 Hz result in a 20 dB increase or something less compared to a higher frequency?
To eliminate “power compression”, the test would need to be really short time duration, a few cycles or the minimum it takes to get a reading.
Art
Thx Art,
I haven't tried to use either the DCX454 or the BMS coax below 500Hz on any of the MEHs. Both of those have always crossed previously straight to large cones, at 500Hz or higher. I'd be very surprised if any dynamic limiting was going on.
Adding the four small mids, the 4mdn34's, is really taking a little bandwidth from both the CD and the large cones.
I'm using the small mids from around 300 Hz up to about 750Hz.
So the mids take away the 300Hz to 500Hz range from the large cones, and the 500Hz to 750Hz range from the CD.
(Sorry for making my versions too difficult to keep up with....)
I haven't tried to use either the DCX454 or the BMS coax below 500Hz on any of the MEHs. Both of those have always crossed previously straight to large cones, at 500Hz or higher. I'd be very surprised if any dynamic limiting was going on.
Adding the four small mids, the 4mdn34's, is really taking a little bandwidth from both the CD and the large cones.
I'm using the small mids from around 300 Hz up to about 750Hz.
So the mids take away the 300Hz to 500Hz range from the large cones, and the 500Hz to 750Hz range from the CD.
(Sorry for making my versions too difficult to keep up with....)
When listening at loud levels, it can be informative to turn off the lower bands and notice how “harsh” the HF suddenly sounds even though overall SPL has been significantly reduced.
Although I had confused how low you ever used the co-ax drivers, dynamic limiting in the 500 Hz range could still be responsible for perceived “harshness” compared to the four 4NDF34. The four 4NDF34 would have over 20+dB more linear output potential at 500 Hz than the DCX464 on the same horn.
The DCX464 on your 60 x 60 conical horn would require a huge excursion increase between 750 Hz equalized flat compared to 500 Hz.
The DCX464 specification state “mounted on a 320 Hz exponential horn” it can take 37.5 volts at 300 Hz (“220 watt” power calculated on minimum impedance of 6.4 ohms). If the DCX464 can take that without the diaphragm hitting the phase plug (I don’t know if surviving the 2 hour power test actually means it does not..), seems that it’s excursion must be limited by the suspension at higher drive levels, which would cause dynamic peak limiting.
Or not
Art
Hi Mark and Art.
I am not that sharp on quick loudspeaker-math. So, when Art says that:
"The four 4NDF34 would have over 20+dB more linear output potential at 500 Hz than the DCX464 on the same horn."
That is a lot! Then I am wondering, what output are we talking in SPL/1W/1m for those four 4NDF34 16 ohm speakers in parallel? Do you have any estimate? Is it about as hysterically loud as the CD in its optimum passband?
I have always been reading/understanding it as, that part of the reason for clear sound in high-efficiency-systems is, that you do NOT use the speaker (and amplifier) as a heater! That on the other hand demands, that you have good milliwatts to drive them in your living-room!
I am partly writing/asking this, because I still contemplate using four 12PR320´s (16 ohm, parallel) per side to be able to go real low with the MEH, say as low as 50Hz or something. Mark says you need four times the displacement for each octave you go down. Mark uses two 12PR320´s to go down to 100Hz, so for 50 Hz four seems to make sense!? And four woofers just give me some freedom in in how to place them on the horn-walls (horizontally or vertically), and my system is intended to be fixed in the corners of my room. And yes, I know this is way past what "normal" HiFi-guys find reasonable! I have once heart two big bass-horns in my living-room, and can´t forget that clean effortless bass response!
Still dreaming
I am not that sharp on quick loudspeaker-math. So, when Art says that:
"The four 4NDF34 would have over 20+dB more linear output potential at 500 Hz than the DCX464 on the same horn."
That is a lot! Then I am wondering, what output are we talking in SPL/1W/1m for those four 4NDF34 16 ohm speakers in parallel? Do you have any estimate? Is it about as hysterically loud as the CD in its optimum passband?
I have always been reading/understanding it as, that part of the reason for clear sound in high-efficiency-systems is, that you do NOT use the speaker (and amplifier) as a heater! That on the other hand demands, that you have good milliwatts to drive them in your living-room!
I am partly writing/asking this, because I still contemplate using four 12PR320´s (16 ohm, parallel) per side to be able to go real low with the MEH, say as low as 50Hz or something. Mark says you need four times the displacement for each octave you go down. Mark uses two 12PR320´s to go down to 100Hz, so for 50 Hz four seems to make sense!? And four woofers just give me some freedom in in how to place them on the horn-walls (horizontally or vertically), and my system is intended to be fixed in the corners of my room. And yes, I know this is way past what "normal" HiFi-guys find reasonable! I have once heart two big bass-horns in my living-room, and can´t forget that clean effortless bass response!
Still dreaming
Last edited:
Yes to that, .....for sure.
I keep each driver section on its own fader, often for just that purpose.
Great way to find out what's doing the bitchin!
I also adore having the the ability to fix tracks' tonality more handily, and in a cleaner sounding way, than using conventional EQs.
Plus, the indoor vs outdoor listening adjustments are easily made.
With each driver-section covering 2-3 octaves, and phase remaining flat even with relative drive level changes....well...it works great for me.
(The section faders are all DCA controlled...so no hassle at all adjusting overall volume.)
That's been my thinking too....and still is the main "harshness" suspect.
ArtThe DCX464 on your 60 x 60 conical horn would require a huge excursion increase between 750 Hz equalized flat compared to 500 Hz.
The DCX464 specification state “mounted on a 320 Hz exponential horn” it can take 37.5 volts at 300 Hz (“220 watt” power calculated on minimum impedance of 6.4 ohms). If the DCX464 can take that without the diaphragm hitting the phase plug (I don’t know if surviving the 2 hour power test actually means it does not..), seems that it’s excursion must be limited by the suspension at higher drive levels, which would cause dynamic peak limiting.
Or not
makes sense...
Here you go...just measured sensitivity as i've been a little curious about the 4NDF34's myself.
I measure sensitivity with processing in place for the intended bandwidth.
Pink noise is used, and both SPL and driver voltage are time-average measured over the exact same time interval.
It's the most real world way to assess sensitivity i know of...but always looking for a better way if someone has one to offer.
Anyway, here's a chart using the bms CD, at 1m from horn mouth.
Note it's all at 2.83V,......no matter the nom impedances.
The 4NDF34's looked so dang high, i measured a single one on a small open baffle for a sanity check, and got 91.9dB, right in line with B&C's spec of 92dB.
4 of them should raise best case theory sensitivity by +12dB, to 104dB.
But in the 75x60 horn, it's measuring over 110 dB...so it looks like the horn is loading them for sure...using 'loading' in a true sense of the word Lol