Jim,
I appreciate the work you have done and shared.
Like you, I like to make sawdust, then see if results match predictions.
Sorry you seem defensive about a detail that could easily be confirmed with a simple a/b test.
Sonic memory is rather fleeting, I prefer to see actual side by side tests done rather than believe simulations. The simulations only show a 3 dB difference however, while guns and knives are more like 10 dB different :^).
Directionality does increase on axis sensitivity. However, for a horn to start to impart directionality, the mouth size needs to be about 1/4 wavelength wide. That size would happen in multiples, but a single horn the size of the SS15 won’t have much directionality below 140 Hz.
Directionality follows the inverse square law, it makes no difference the distance the measurement is taken at.
It is easy to see this effect when looking at polar charts of any horn, once the mouth size is less than a wavelength, the pattern starts widening.
If you have measured greater directionality of a single SS15 at low frequencies, please direct me to where I could see them.
Art
art:
you missed olivers point with his horn resp.... (it was what I was getting at with my 'show me the hornresp comment from earlier.')
A 220liter reflex with a 3015lf doesn't tune well to do an A/B vs the SS15. You either tune it lower than SS15, or you make the box smaller.
an A/B won't work.
On the directivity, what happens (in general) when you put 2 sources close to each other?
you missed olivers point with his horn resp.... (it was what I was getting at with my 'show me the hornresp comment from earlier.')
A 220liter reflex with a 3015lf doesn't tune well to do an A/B vs the SS15. You either tune it lower than SS15, or you make the box smaller.
an A/B won't work.
On the directivity, what happens (in general) when you put 2 sources close to each other?
Last edited:
Frankly, I really don't care exactly how the low end of the 3015LF in a box compares to the SS15, even a sealed box test would show the sensitivity of the driver between around 120 -200 Hz (it should be about 99 dB, as the specs indicate), an A/B would determine whether your dB meter is correct, or reads a couple dB high as it appears.
Directivity depends on the size of the sources, but two stacked sources will reduce the vertical coverage while maintaining the horizontal coverage.
Since the ground plane reflects the vertical height, a vertical sub stack is acoustically twice as high as its physical height, imparting greater directivity than one would think based on physical height alone.
This point is often missed by proponents of “center cluster” subs, by the time they are put in a long line across the front of a stage, the horizontal coverage is reduced, so side coverage is lacking, while the subs are still omni in the vertical dimension, putting out nearly as much energy behind as in front.
Last edited:
But how does Behringer measure it?
I'm sure I am not the first to think of this, but I don't think anyone has mentioned it:
if Behringer tests in an anechoic chamber, and you test on the ground... the 2 dB difference to the specs could be accounted for by this?
Regards, Ben
I'm sure I am not the first to think of this, but I don't think anyone has mentioned it:
if Behringer tests in an anechoic chamber, and you test on the ground... the 2 dB difference to the specs could be accounted for by this?
Regards, Ben
The amount of sound sources would be the amount of drivers contained in the tapped horn, though at low frequencies closely spaced drivers are virtually the same as a single larger driver.
The output of a tapped horn is a single acoustical source comprising the summed response of the speaker cone front, rear, and internally reflected radiation.
Testing in an anechoic chamber would imply a free space measurement, which would be up to six dB less than a half space (ground plane) measurement at low frequencies.
For frequencies in wavelength less than the center of cone to baffle edge, the baffle creates a half space, so measuring high frequencies in half or quarter space shows less increase than would be predicted by pure physics, which use a single point source radiating 360 degrees as the model, which never happens in the real world.
Having no response curve to look at, my speculation would be Behringer may have merged a ground plane measurement with a free space measurement, this makes the upper portion of the graph look smoother, and shows more LF, commensurate with the speaker placed next to a wall as is common in many small room situations.
If the speaker actually had a flat response to 60 Hz or so 89 dB in free space, it would measure around 95 dB in half space. A ported speaker the size of a B2031P would not be that sensitive that low, hence my previous assessment.
Manufacturers use various amounts of smoothing (or averaging), more smoothing can reduce +/- 6 dB peaks to +/- 3 dB (the usual range) or less.
Since a dB meter responds to the loudest frequency peak, it will always read higher than the average level of a speaker, unless the speaker happens to be perfectly flat response.
Since Jim did not do a frequency response test on the speaker, he only determined that the loudest point in the speaker’s response was 91 dB as read by his SPL meter, which unfortunately tells little about the speaker, or accuracy of the meter.
In the chart below, the same small tapped horn (15” x15” x26.5”) with a 10” speaker are compared, the blue trace is using an Eminence Kappa Pro 10”, a +/- 3 dB response from 53 to 160 Hz.
The orange trace is wimpy magnet 10” speaker from an abandoned Fisher stereo cabinet .
Using a dB meter and pink noise band limited to around 160 Hz, the Fisher speaker would read within 1 dB of the Kappa Pro, even though it is 10 dB down (half as loud) for most of the sub range !
Measuring can be fun, interpreting the results even more so.
Having studied the squiggly lines since around 1977, it becomes easier to interpret results, but without knowing precise details of the test procedure and equipment used, results can be very ambiguous.
Attachments
...or you design it with an expanding vent. This should boost the passband response a bit, if done right.
I've included an example below. This is an 8" design I'm looking at, where an expanding vent is used to counteract that inductance rolloff and flatten the passband. A dB or so is added, and even more can be added if I'm willing to go with a larger box size.
Attachments
Quote:
Originally Posted by jbell
A 220liter reflex with a 3015lf doesn't tune well to do an A/B vs the SS15. You either tune it lower than SS15, or you make the box smaller.
The question of actual sensitivity of a specific tapped horn compared to ported cabinet will remain until an A/B test using the same speaker is performed.
If the actual measured SS15 level from 45-120 using a 3015LF were to exceed the front loaded SPL of the same speaker at 120-200 at the same drive voltage, that would be reasonable proof the SS15 would exceed a normally damped ported alignment.
Art
Originally Posted by jbell
A 220liter reflex with a 3015lf doesn't tune well to do an A/B vs the SS15. You either tune it lower than SS15, or you make the box smaller.
Regardless of the contribution of the vent, any ported or sealed alignment (excluding bandpass ) of a given speaker will be within a dB or so in the region from 120-200 Hz.
The question of actual sensitivity of a specific tapped horn compared to ported cabinet will remain until an A/B test using the same speaker is performed.
If the actual measured SS15 level from 45-120 using a 3015LF were to exceed the front loaded SPL of the same speaker at 120-200 at the same drive voltage, that would be reasonable proof the SS15 would exceed a normally damped ported alignment.
Art
Art:
Since you don't have hornresp up and running on your mac yet.
Here is an A/B with the SS15 vs sealed at 2.83 volt, and also at max input power 9.6mm xmax 40hz on up. (40v sealed, 63v ss15)
(since you said earlier you didn't care ported or sealed....)
Since you don't have hornresp up and running on your mac yet.
Here is an A/B with the SS15 vs sealed at 2.83 volt, and also at max input power 9.6mm xmax 40hz on up. (40v sealed, 63v ss15)
(since you said earlier you didn't care ported or sealed....)
Attachments
Appears your and Oliver’s sims show a dB or so difference in midband sensitivity between ported or sealed, so the difference between the SS15 should be 3 or 4 dB between it and a ported or sealed cabinet in the 45-120 Hz passband. A sealed cabinet of course will run out of Xmax before exceeding Pmax, reducing it’s maximum output.
If you measure that same 3 to 4 dB difference in the real world, you will verify that your cabinet performs as well “as built” as the sims predict.
Or you may find, as Tom Danley has, that they don’t always do what they should.
Tom wrote about the Jericho Horn in #586730 on the LAB, Thu, 07 October 2010 12:44:
PSW Sound Reinforcement Forums: LAB: The Classic Live Audio Board => Jehrico horn
“I hate how weight collects and adds up. Yes these are heavier than before.
We have made several revisions to the cabinet which added weight. While I was pleased with the sound of the generation at the Infocom show, it bothered me deep down inside that the woofer section for that enclosure fell short of the computer model in places.
The “missing output”, as much as 6-7 dB in the upper bass, was recovered by greatly stiffing the enclosure. That involved in part more braces and moving to one inch Baltic. “
Having followed Tom’s work (and using it as a benchmark) since the 1980’s, it is always interesting to read his observations.
Since your and Oliver’s sims only predict a 3 or 4 dB difference between a ported cabinet and the SS15, and cabinet build and measurement tools could amount to double that, I continue suggesting an actual comparative measurement to see where things sit in the real world.
Attachments
I agree a face to face showdown would be better. I do have a candidate for a bass reflex:
B&C 18PS100 woofer in 5.2 ft3 tuned to 38hz, 20" sonotube cab, very non resonant, and total weight = 42 lbs. As I approach it's limits, the main speakers are WELL beyond theirs.
I'll have to get around to building a SS15 though.
B&C 18PS100 woofer in 5.2 ft3 tuned to 38hz, 20" sonotube cab, very non resonant, and total weight = 42 lbs. As I approach it's limits, the main speakers are WELL beyond theirs.
I'll have to get around to building a SS15 though.
The 18" will give you almost 3 dB more level than a 15, even with the extra sensitivity of the TH, the B&C 18PS100 woofer in 5.2 ft3 tuned to 38hz should be very close in level, probably would have a bit more at F3.
Now if you built a TH of similar size and cutoff as your sonotube cabinet for the B&C 18PS100, you would have a valid comparison point.
Tapped horns can be built using nested sonotube, though you would need a big diameter for an 18", somewhere around between 25" and 30" ID as a wild guess.
ONLY when the snow is gone and it's a bit warmer, I can do this: (even though I'm sure art will never accept any of my measurements)
Test a ss15, sine waves only 2.83v and 28v at 1 and 10m. Also show a smaart pink noise measurement with a reference spl at 60hz.
Directivity test. Take a 10' string and walk around in a circle and take measurements on 50,70,90hz sine waves.
Make a plate for the front of a ss15, that I can mount a 3015lf, and either 2 or 4 three inch diameter ports. This will use the EXACT same driver, and I will not touch any of the measuring equipment settings between tests.
Here's what you (and only you) must do art:
You must design the ported box and tell me how many ports and at what length they should be cut. (ss15 is about 215-220liters)
I'll use gasket foam and seal the plate air tight to the front of the ss15. This will assure the exact same cabinet volume, same cabinet flex, same driver, same input, same measurements.
THEN, when I post the measurements, you must accept them. period.... if you do not agree to that, then I won't do this test for you.
Until it gets a bit warmer, here's some measurements to chew on....
ProSpeakers Forum - Fitzmaurice Titan 48 - Wayne Parham, October 21, 2007 at 14:31:26
In particular look at what happens at 40hz from 100 watts to 800 watts.... only 3db. Even though this is a 35hz 1/4 wave FLH horn using the same 3015lf driver, the designer will tell you it's operating in 'direct radiator mode' at that frequency. Hmm... direct radiator.... like a sealed or ported cabinet.
If a particular design is only 3db more efficient at 2.83 volts.... guess what, it's like the difference between 28v and 80v in this particular case.... so 3db is nothing to sneeze at.
From the measurements from ported, front loaded, rear loaded, tapped, etc.. that I have built in the past... It should look like the ported cabinet brought a knife to a gun fight.
The reason is that I don't care what a sub can do in the 'midbass or midrange' area. If I need more 160hz, I'll eq it in. A sub cabinet must provide the most output at the hardest to reproduce frequency -- that's 40hz-70hz. That's what I care about. If it can't do those frequencies, then it won't work in the real world. You mention 3db difference ported or sealed to ss15... at 60hz it's 10db.... If the above example shows you anything about what happens real world with real power, and only gaining 3db going from 28v to 80v.... then that 10db difference the sim shows you at 60hz makes it pretty obvious why I like the ss15.
Test a ss15, sine waves only 2.83v and 28v at 1 and 10m. Also show a smaart pink noise measurement with a reference spl at 60hz.
Directivity test. Take a 10' string and walk around in a circle and take measurements on 50,70,90hz sine waves.
Make a plate for the front of a ss15, that I can mount a 3015lf, and either 2 or 4 three inch diameter ports. This will use the EXACT same driver, and I will not touch any of the measuring equipment settings between tests.
Here's what you (and only you) must do art:
You must design the ported box and tell me how many ports and at what length they should be cut. (ss15 is about 215-220liters)
I'll use gasket foam and seal the plate air tight to the front of the ss15. This will assure the exact same cabinet volume, same cabinet flex, same driver, same input, same measurements.
THEN, when I post the measurements, you must accept them. period.... if you do not agree to that, then I won't do this test for you.
Until it gets a bit warmer, here's some measurements to chew on....
ProSpeakers Forum - Fitzmaurice Titan 48 - Wayne Parham, October 21, 2007 at 14:31:26
In particular look at what happens at 40hz from 100 watts to 800 watts.... only 3db. Even though this is a 35hz 1/4 wave FLH horn using the same 3015lf driver, the designer will tell you it's operating in 'direct radiator mode' at that frequency. Hmm... direct radiator.... like a sealed or ported cabinet.
If a particular design is only 3db more efficient at 2.83 volts.... guess what, it's like the difference between 28v and 80v in this particular case.... so 3db is nothing to sneeze at.
From the measurements from ported, front loaded, rear loaded, tapped, etc.. that I have built in the past... It should look like the ported cabinet brought a knife to a gun fight.
The reason is that I don't care what a sub can do in the 'midbass or midrange' area. If I need more 160hz, I'll eq it in. A sub cabinet must provide the most output at the hardest to reproduce frequency -- that's 40hz-70hz. That's what I care about. If it can't do those frequencies, then it won't work in the real world. You mention 3db difference ported or sealed to ss15... at 60hz it's 10db.... If the above example shows you anything about what happens real world with real power, and only gaining 3db going from 28v to 80v.... then that 10db difference the sim shows you at 60hz makes it pretty obvious why I like the ss15.
Last edited:
Jim,
I can accept your measurements, no need to go to great lengths to do a front load test, any old box you have sitting around is fine to get the relative difference. Just looking for the same speaker in the same place.
Directivity tests- if you measure in proximity to a building it would be more appropriate to rotate the cabinet, rather than moving the microphone.
I can do the same with my small tapped horns, it will be interesting if they show directivity in singles, as I have not noticed any to speak of.
Your observation about small horn cabinets folding up like a cheap suit is something I have measured extensively.
PSW Sound Reinforcement Forums: LAB Lounge => JBL SRX718 - Basic Distortion Measurements
As you said August 2008:
“If 40hz is your low frequency that you need for your PA, that efficiency at that freq sets the overall response you can achieve, as it's the hardest to reproduce and most power intensive frequency in the PA power band. (which we obviously know) A sub that's 13db more efficient at 100hz vs 40hz, helps little. If I need more 100hz out of a sub, I'll eq it in. However, If I need more 40hz... there's only one way to get that, big power and big excursion, and that limits the overall spl you can get out of your cabinet.”
Several years before that I had abandoned my Chorn design for Lab 12” in ported cabinets, like you I was willing to sacrifice the upper response for a gain in the lower response. The Chorn behaves like BFM’s and Jeff Permian's offerings.
I have found in an outdoor side by side test on adjacent stages that four ported 2 x Lab12” had 2 dB more output at 40 Hz than eight Meyers 650P 2x18 cabinets, using a small fraction of the power, space and weight.
The Lab 2x12” work so well and allow a flat to 32 Hz response, it appears to me even with the SS15 I’d be leaving a behind a little LF.
I am about to pull the trigger on a pair of B&C BC-18SW115 , the speaker used in the DSL TH-118.
In the same size cabinet as a TH-115, and very possibly the same path length (phase plots overlay) it goes way deeper, is more sensitive, and due to the larger, higher Xmax driver, has a lot more output.
I have a little bit more volume to play with than the TH-118, my cabinet will be 22.5 depth x 26.5 width x 45” height.
I’ll be sharing my results and hoping for some design help.
Attachments
Last edited:
ahh, but it doesn't... as olivers posts on ported vs ss15 show. I could just set the port tuning to 20hz... then what?
You tell me the ports, and I'll do the plate on the front of the ss15.
I have Acres of land... with one house and one barn... no need to go very far to be very far away from buildings.
The lab2x12 posted on soundforums.net...
Yes the ss15 is leaving a bit of 40hz behind in single cabinets. I always tell anyone building, run pairs or quads. It's flat to 40 in quads, and pretty respectable in pairs. If there were anything I could change in this design, it'd be to get it flat to 40 in pairs.