Simon's new model.

I'm interested in what you all think about Simon's model of the WSX. Read carefully, especially bits about horn mouth area and manifold chambers (didn't get those first skim)

This from Simon:

Ihad to do some checking on the wsx sims due to some comments on
another post. I found that the plots were made using a version earlier
than 17.xx, and did not include the low frequency enhancements in later
versions, so were very low. Also, I thought that the model used for the
plots didn't look right. After checking back with the plans, I found
that this wascorrect, and that there were some errors in the model.
The first section of horn directly in front of the driver should,I
believe, be considered as throat chamber and not horn. The horn proper
starts at the end of the lower panel at the point wherepath starts to
bend round. I think this first section iscalled a manifold?. The
figure that should be entered for the volume of throat chamber is 13000
for the manifold and then whatever is actually in front of the
diaphragm. Normally about 4 to 5 litres for most drivers.So 18000
should be entered for VTC. The average cross section of the throat
chamber should be entered as 544, as this is the smallest and hence
most important restriction, andforms the greater part of the path from
the letterbox entry along to the start of the flare out to the throat.
Next is the throat itself, at no time does the horn path or manifold
narrow down to 404!. The Throat is 688 at the start, and S1 should be
enetered as this. Next is the actual length of the horn and thisis
never 200 (or was it 210), even including the manifold!. The model for
the horn should be something like as follows:
<div>VRC: 88</div>
<div>LRC: 20</div>
<div>ATC: 544</div>
<div>VTC: 18000</div>
<div>S1: 688 S2: 1238 exp/con 62. </div>
<div>S3: 1238 exp/con 42.5</div>
<div>S4: 3078 exp/con 39</div>
<div>S5: 4300 exp/con 19</div>
<div></div>
<div>Closest approximation for this as a single section horn would be :</div>
<div>S1: 688 S2: 4300 Hyp: 1630 T: 0.7/0.8or possibly 0.9. </div>
<div>Does anyone bother with the grilles any more?, as the original
design might well have had allowances made for the restrictions
imposed, and would also suggest that the mouth area should be
considered as a much lower figure than the total front area.This model
does not include the top panel of the flare, as this is so close to
vertical that the difference in tuning to the horn would be minimal.
Why exp or conical?,.The sides are actually flat and not curved, as a
true exponential flare should be. The difference is not that great,
it's up to you. A caution message will pop up saying atc < sd. Well
it is, isn't it?. The difference in plots between entering the driver
sd and the manifold area is nearly all at the top end , which is
probably why these cabinets are supposed to go higher than most, check
it for yourselves and compare them.</div>
<div>After looking at the plans and drawing them out on cad, I noticed
something that I will probably be very unpopular for mentioning, but
here goes anyway.</div>
<div>There is,for all intents and purposes, very little difference
between the wsx and the 186/1850 horns. The manifold is the feature
which puts them apart. This does make an improvement in the upper
frequencies from about 130 hz upwards.It opensup to 3db at 250hz !.
All other things considered, it's the same breed of animal, just
different coloured spots. When considering the 186, it has been said
that this horn can be used up to 200hz, but that is probably the pd186
driver. As far as which one is louder, or goes lower, then the guy with
themore powerfulamp and the most efficient drivers that can handle
the power is going to win!. Anyone want to discuss any of the above,
or feels that I've just had a senior moment, then maybe we should move
this over to the advanced discussion forum, and let anyone who wants to
chew on the bone have their tuppence worth, we might get somebody who
knows what they are doing give us a few pointers. happy Building!.

</div>Comments, please? I'll add mine after a few meaningful posts.
Edited by: Deadbeat

I would like to add that a correction was made later to the throat chamber area, and should read ATC: 435, not 544. I'd like to quote Tonys' defence and say I didn't have the right glasses on when reading my list of figures.


The reference to 404 for the throat chamber is made because the area at the rear of the throat chamber is of course 404. However, this does not impede the the path of the horn or the 'manifold'. It forms the end wall, and imposes no restrictions of the path. I think the path should be considered as rotating through 90 degrees from the front edge of the 'letterbox' inthe baffle that leads into the manifold, and the smallest restriction is 435.

I wiil post more a bit later, got to stoke the boiler.

I've also got to check my horse's shoes for stones before I go to work tommorow. In my buggy-cart with whip.

Happy birthday, simon. Sorted out your community buys on the other horn yet?

Just reignitin this thread:

Can you explain to us what a manifold is and how you realised it to be there?

Your model makes a few assumptions that I might dispute, but the real test would be to line up the 186/50 and WSX and do a proper measurement with the same woofers. As mike has said on the light side (well, the old light side), there be a difference of a horn going up to 200hz on paper, and it actually doing that without sounding like a bunch of geese.

www.speakerplans.com/forum/forum_posts.asp?TID=561

WSX and 186/1850 have always been different beasts to me anyway.
According to Martin,
WSX: 7ft (2.1m) Hyperbolic bass horn
35-150hz +-3dB
-10dB at 28hz
Originally powered by the RCF L18S800 (or custom, I can't remember, someone correct me)
I'd argue that the almost vertical bit at the end increases the virtual mouth size. Horns in a stack have increased mouth size without additional barn doors or expansion, though this aids it. I'll try modelling the horn again including the manifold, without the manifold, with the big mouth, without the big mouth, and with and without each other, whatever fits an extended horn path. (Take note that I haven't measured up either of the horns, this is just speculation) I've seen what my RTA says on a quick sweep and the cutoff is lower than in your model. modelling all of this tommorow morning, all with the V18-1000, which is undisputedly a good driver for bass horns.

According to Rog,
1850 horn - 50-200hz +-2dB ,sharper lower rolloff, Horn length approx 1.6m (from looking at it lol)
This horn is optimised for large stacks and does
186 horn - 51-215hz +-3dB slower higher rolloff
They seem to be pretty different to me, though on the 186 as you said the higher extension may be due to the driver. The 1850 is used as a drop in replacement for the V18, and is on par or better than the RCF in the WSX, so I'm having some doubts, really. Rog also says the 1850 would have the same sound and plot really in the 186, albeit at much higher levels.

I'll sort it out tommorrow morn, when my brain isn't fried on coffee at 4AM. WSX vs 1850 vs 186, in singles and in groups of 4. No promises, but data up by the end of the day after tommorrow, hopefully. Please point out any errors, discontinuity, or things that are just plain wrong (I can't think straight, so why am I posting this anyway [img]smileys/smiley36.gif[/img][img]smileys/smiley12.gif[/img])...
[img]smileys/smiley12.gif[/img]

From the man himself:


Hi Peter,</font>

Well, I don’t know about the respected bit, but I’ll try and explain where I’m coming from with my horn designs.</font>

If
you look at the 1850 horn on the speakerplans site you will see that
the response does drop off like a rock below its cutoff point. The horn
length is 1.59 meters or 62.59 inches. This equates to a full
wavelength frequency of 216.77 Hz. So divide by 4 to get the quarter
wavelength figure of 54.19 Hz. Again if you look at the plot you will
see it fall sharply below 54 Hz, I also quote the cutoff frequency and
state that the design is good down to 54 Hz. I also assume that at
least 2 horns will be used in a stack and preferably 4. So I can’t see
the problem or see why you could have a problem with this. If I quote
that 2 horns have a cutoff of 30 Hz then I would be lying, but I don’t
do this. </font>

Concerning the HD 15 horn, I do quote that the design
will operate down to 52 Hz which for one cabinet would be impossible,
but it does state that it must be used in a stack of at least 6
cabinets. When used as such the combined horn mouth and extra horn
length that can be achieved by combining horns will give a new cutoff
of 52 Hz. Also note that I still say you should use a reflex or
bandpass type cabinet below these horns to get a fully extended bass
response below 50 Hz.</font>

The 186 horn is meant to be used singly and
with higher Qts drivers than a horn normally requires. This and an over
sized rear chamber can give you a shallower rolloff below cutoff. This
is depicted in the response, but you will also see that the cutoff has
been raised in frequency from what the horn length dictates it should
be. This is again due to the rear chamber being too large, a smaller
rear chamber with a lower Qts driver will have a lower cutoff (or a
cutoff that behaves predictably and is true to the 4/WL rule) and an
over sized rear chamber will tend to shift the cutoff up in frequency
and give a shallower rolloff, especially with drivers with a higher
than recommended Qts and low BL. While not ideal I do prefer listening
to horns with drivers with a higher than required Qts, good for the
lounge but not good if you need a quick transient response and loads of
SPL like at a concert. </font>

So I hope that has explained the figures
behind the designs. I always assume that my designs will be used in
multiples. This is large scale PA for big venues and no one is going to
use just one small 15” horn in there PA system, well I hope not. So my
mouth areas are calculated assuming that a minimum of 2 1850 horns will
be used together and 6 of the HD 15 horns will be used in one stack. </font>

After
liasing with David McBean, his wonderful Horn Response program (ver
5.60 and above) now lets you calculate the response for multiple horns.
While not usable for everyone it’s really helpful in the PA world where
multiples are the norm. </font>

each.</font>

Best wishes and happy resonating,</font>Rog Mogale.

</font>

Simon, where are you?

I'm only modelling if you reply....
Edited by: Deadbeat

Hi there, it seems that this is now down to a two way discussion, and nobody else wants to add their own tuppence worth. I had hoped that a few more people with more advanced knowledge than myself might givesome help. I know I'm fairly ignorant as far as theory goes, and a little basic knowledge will always lead the uninitiated into making hugely exagerated errors when dealing with the more advanced subjects. But I'm gradually stumbling along and there does seem to be some light at the end of the tunnel.
I have a solid background in engineering, and when considering using a new piece of equipment, a good look at the spec. sheet is the first step on from quantifying what is actually needed to do the job. When I read that something has a certain parameter, I get extremely frustrated when the equipment does not seem to match the given figures. Itappears to the beginner, that for most specifications with regard to horns, there is a certain laissez-faire regarding physical rules, and one has to take a trip into virtual reality to finally get to the point where the design figures have been originated from. I guess that at some points along the theory trail, the rules are open to a bit of interpretation, and that it depends on how the designer has interpretedand applied those rules. For a beginner revisiting an original design, and trying to make sense of the final figures, it's a bit like doing the times crossword for the first time. If you aren't familiar with the setters personal take on things, a lot of the clues make no sense at all. It is only after familiarisation with how they work that you finally get the true meaning. I see! said the blind man. Or at least I hope I do!. I suppose the clues were there all the time, but I was too busy looking inside the box (literally) to see the bigger picture (again a bit of irony!). The trip into virtual reality was also literally true, as well as apocalyptic.
So I guess it's apologies all round, and can someoneplease follow me through my reasoning so thatI may rest a bit easier. At least I could then offer up a better opinion as regards to some other topics.
So, I think what I am proposing is a general set of rules for everyone to apply to their own sims, so that when posting some comparisons, everyone is playing at the same level. Does this make sense?. Does everyone agree that there is a need for a set of ground rules?.
I will put forward my own version and ask for comments, suggestions, and ammendments where it is deemed necessary.

Basics.
To bring everything down to the same level, and to ignore theidiosynchratic nature of environment, all plots should be made in half space.Anything else is an artificially enhanced environment, and would be open to too many different interpretations. I'm forgetting free space,so if you are doing some high-flying, then apply as appropriate.
Use 1w/1 meter for sensitivity.This will give everybody an idea of how efficient it is.
Next would be a plot at the highest diphragm displacement that thexmax of the driverwill allow, or ifxmax is not exceeded, a plot at the RMS rating of the driver.
Horn parameters.
First, the compression chamber. Hornresp has a flatrepresentation of the diaphragm.
For most drivers this already poses a question. There is a volume of air in front of the diaphragm that is not accounted for. Also, there will normally be a baffle that has a cutout for the cone, if rear mounted, that also adds volume. If one is presented with a problem like the WSX and the rather complex compression chamber, there needs to be a general 'rule of thumb' that can be applied so that this may be represented as accurately as possible.
Sohow does Hornresp deal with themodelwhen you enter a smaller chamber area than the SD of the driver?.Would it be more accurate to assume that the chamber is all at the same area as the SD of the driver, sum up theenclosed volume of the whole chamber including the volume in front of the diaphragm and baffle, and then use the smallest area of the exit into the horn as the throat area?. In the case of the WSX, this might mis-represent the actual distance of the diaphragm from the throat of the horn, but should we concentrate on the compression/volume side of things, or do we ignore the reference to area of the chamber on axis to the horn?. Comments please. Bear inmindthathave been discussing low frequency bass, and should we apply some different rules for higher frequencies?. If so, whatdo you propose?.

A theoretical horn can be modelled easily enough. Producing a modelof an existing (folded) horn without the original design parameters is a completely different kettle of worms. Most of the time, if a manufacturers plot is available, use it. How do you know it is accurate? and what happens if you want to find out how it behaves with a different driver?. Well these are the questions that have prompted this section in the first place. If we can all agree to a set of rules for this, then we can all share and add to a database of designs that we can use for comparisons tonew designs, without any controversy. If anyone has a measured plot usingreasonably accurate measuring equipment that disagrees with published figures, then please get them posted. But also make sure that the measuring environment is correctly noted!. It can give others an idea of performance under certain conditions.

Plotting the horn segments.
Most horns are made up of flat sided plywood, and do not curve. The horn is built up of flat sided sections and so resemble morea series of conical segments, thanslightly curved exponential segments. What is a more accurate representation?.Is it acceptable to plot a series of conical segments as a single hypex or exponential horn, and if so, to what degree of accuracy do we have to work to?.Comments please gentlemen.

Throat to end of first section, and subsequent sections. S1-S2 etc.
I would normally work along the central axis of the horn, and if going round a fold , look for the narrowest radial restriction. After finding all the narrow points, I would then compare each in progression from throat to mouth. If a narrower restriction is foundfurther down the horn path than a preceeding one, therestriction further down will takepriority in deciding the end of the first section. Each subsequent section then has to be derived from the next narrowest restriction further down that takes priority over preceeding restrictions.
A lot of horns will provide more restrictions than we have available sections to enter.
This would suggest that an Hyperbolic curve is more appropriate to simulate this horn.
The actual length of the path of the horn around the folds. ?. If we were to assume a rotational transfer of the volume of air going around the fold, thenwe would have to assume that some of the air must travel a lot further and faster than the rest, all the way around the outside of fold, while some remains on the inside and slows down. Just measuring the outside length of a hornaround the folds etc. is not the correct method.A lot of people have a misconception about this, and I suppose at first glance this may be a natural assumption to make. The volumetric transfer is more linear,the airacts more like a liquid, albeita liquid that is compressable, hence propagation delay equal to the path length.
However, if youimagine a ball inside the horn at the start of the fold. The balls' diametercan expand to suit the radialwidth of the hornat any givenangle.If we take away any outside influences like gravity, friction etc, and then imagine that the ball can movealong without rotating,then have theball in contact with the inner, or the outer surface and slip it along all the way round. Every part of that ball will travel exactly the same distance.The centre of the ball willdescribe the path length of the horn. Simple. But, in the real world, surface friction, eddies,turbulance etc. will detract from the overall efficiency, and most horns will have voids where the outside surface is not a perfectly expandingcurve that matches the inner surface. It seems that for all intents and purposesthese are ignored, and provided the volumes of thevoids are not excessive, will not have much effect on the horn?.This would then require that theenclosed volume of the fold has a radial cross-sectional area, at any given angle, equal to (or slightly greater than) the projected linear expansion rate of thehorn flare, and does not pose any restrictions to this. Does anyone care to commentso far?. Would anyone like to comment on void volumes that would appear to be excessive in some horns, and would imply that some form of correction should be made?: and if so, how is the correction applied ?.Do these voids add some delay on the path, and should a correction to the length of the horn path be made in proportion to the volume ofthe void?.What is everyone's opinion?.

After deciding the path, as enclosed inside the box, let's consider the virtual reality bit.
Projecting outside the box. For an axially symmetrical hornthere would be no real need for forward projection, would there?. If hornresp is giventwice aslarge amouth area, does it automaticallyproject a curved forward wavefront proportional to the area?. If there is a projection to be added in the front of a flat faced assymetrical horn, like a bass bin, does it assume that the forward projection has already been included in the horn length?. If two assymetrical bass bins are put together, so that the horn is symmetrical about the central axis, doesHornresp then automatically add what some manufacturers would claim as their correct projected horn length?. I have made my own assumptions on these questions and I would be interested to find outwhat other people assume when preparing sims for evaluation.
We need to come to a concensus of opinion, and lay down a few rules for evaluation purposes, so that we can all be confident about other peoples' figures. New guys can get a comprehensive guideon how to prepare their presentations for evaluation on the forum. Those of you who are sitting there and saying this is all elementary and saying what the
is this all about- it should never be posted in the first place: all I have to say to you is, when was the last time you learnt anything new?. If the answer is a long time ago, then you can tell me now when the world is going to end.

Come on guys,this forum needs some good input from all of you to liftit above the current level. There are very few people contributingand willing to discuss fresh ideas. Even if you don't know much on the subject, any opinion is worthwhile.I think it is a comment on the effectiveness of this forum, that there have only been two contributions to the advanced section of the forum. Doesn't anyone want to discuss any ideas (however fanciful) they might have?.

I hope some more will come from this diatribe, even complete repudiation would be an improvement.
I will also put up some new models for a few horns that will supercede previous models posted that might have been (almost certainly, I hear you say)in error.

Apologies, I seem to have gone off on a rant.

Any input will be gratefully recieved.

I did find 2 mistakes a while ago and went to post, and as usual it said..... not posted and waisted the whole thing

when i get another min i'll go over it

talking of WSX, WLX, WMX

the WLX and WMX, i would put money on the spec's being wrong

WLX 106db 1w1m?
WMX 104db 1w1m? this maybe, but not 106 for the WLX

Hi guys, I just noticed that a few questions were raised on the other side regarding this topic and I would like to answer a few now. Please read the section below from RM regarding Horns and the extra horn length when used in multiples.

Concerning the HD 15 horn, I do quote that the design will operate down to 52 Hz which for one cabinet would be impossible, but it does state that it must be used in a stack of at least 6 cabinets. When used as such the combined horn mouth and extra horn length that can be achieved by combining horns will give a new cutoff of 52 Hz. Also note that I still say you should use a reflex or bandpass type cabinet below these horns to get a fully extended bass response below 50 Hz.

I think this will show that a trip into virtual reality isneeded when considering multiple horns. How does this work?. For what it is worthhere's my opinion. I would like to get something back, I feel like I'm talking to an empty room here.
A single assymetrical horn like a wsx is stood upright, with the angled front facing down towards the floor.
Thecompression wave will travel outwards and will start leaving the horn at the bottom edge first. This will immediately start a degrading decompression across the whole wavefront as more and more of the progressing wave is able to wrap around the sides. The wavefront does not propagate parallel to theside edges. This effectively reduces the influence of the upper angle at the top of the horn, and reduces the effective mouth area. I am of the opinion that when more than half of the wave is subjected to the degrading expansion, then you will have an exponentialeffect, I would therefore consider this point as the effective cutoff for the mouth area under these conditions.
Now consider when 4 cabinets are stacked together with two on their sides with the 'bottoms' put together and a further pair stacked on top. The flares are placed to form a symmetrical horn about the central vertical axis from front to back. There is now a large radial horn flare formed. As the sound wave startsto emerge from what is now the centre, the wave from the top pair will reinforce the emerging wave from the bottom pair.Both top and bottom waves are laterally reinforced at the central vertical axis.We will now have coherent coupling. The reinforcementeffect of the combined mouth will not only have aneffective area that is considerably larger than 4 single horns, but will project that reinforcement further out from the front of the horn, so increasing the effective horn length. The floor is now acting fully as part of the bottom pair of horns. The reinforcement area to be considered issubject to degradation from the top leading edge of the top pair of cabinets, but the lower pair will have their centre extending a lot further. I suppose if you cuta sphere in quarters and placed one quarter in front of the horn, you would get some idea of area of maximum reinforcement. This is a very poor analogy, the shape is quite complex, but you can work out what I'm trying to get at. From here you will probably see why 6 would be better than four.
The next thing to consider is the rate at which the outer horn panels flare. If the flare rate is greater, it allows a faster expansion at the edges andso will decrease thecompression across the wavefront faster,
and so decrease the effective projection of the horn length. There is a fineline between effective atmospheric coupling and excessive expansion, where the effectiveness of the horn flare has very little effect. A lot of thismight be far moreapplicable to higherfrequencies, where the length of the compression wave is shorter than the actual horn length, but it still has an effect. I suppose any containment is better than none. However, it is how far theflare can beassumed to have a beneficial effect without wasting space, or whether the box design has used the outer panels as an integral part of the intentional design, or has been left as- is because it forms part of the flare further down.It is deciding how far to stretch the virtual reality bit that has been troubling me. It would also appear that this has been applied to the wsx cabinet. If you take into consideration the full length inside the cabinet, count the final section as horn instead of compression chamber, you still wind up a few hundred mm short of the 2.1 meters. When reading the info, it does state that they are designed to be used in multiples, and using the above virtual horn length, it can be seen that 2.1 meters is possible(with an extension ofover 300mm in the centre).You may decide to consider the length of the path from diaphragm to the end of the horn, however if you are adding a compression chamber figure in hornresp,this length is added into themix, and so horn length becomes exagerated.
Finally, when coming to sim the horn in multiples, it is difficult to decide exactly where to finish extending things before the model is invalid.You will notice that if you plot the wsx in conical sections, if you follow the expansion of themouth section to a logical end the cir becomes greater than one. You can reproduce a horn similar tothe wsx horn, by reducing theT value down to about 0.1 for a single section horn, but this will have the effect of raising the lower cutoff frequency quite dramatically. This would also reinforce my ownreluctance about extending the horn length too far, and also my opinion that the rate of flare of the top panel is too eccessive, and hasless of an effect than could be expected in the overall plan.
I would like to ask all of you for some more input, especially some good response plots for cabinets, so that when comparisons are made, there aremore well defined bench marks for guaging the accuracy of sims produced. It might help prevent people frommaking some bigger mistakes thanpreviously, and also make people give a bit more thought to how they prepare the design of their cabinets after simulation. For my own designs I would prefer leaving asmuch horn inside the box, and having the shallowest flare possible at the mouth. For assymetrical designs that form a radial flare when joined, I would like to think that some of the older style radial horn flares that would open out ata rate of 90 degrees still hadsome validity. What do you all think, have any of you got any preferences?.
If I can get some agreement on a few of these questions, or a few answers thatcorrect some of my mis-guided assumptions, I would then feel more confident in comparing a few of my designs with some other well known designs and get the plans posted.I know that the designs I have been working on are reasonably good, but I would like to quantify how good in comparison to others, before I waste everybodies time.
I have come to the end of my little theorising bit, (thankGod!) I hear everyone say. Pretty pictures will be spewing forth back on the more relevant project threads, and I will post my revised wsx plot very shortly.
Have funpoking holes in this littlelot. I will take complete silence as magnanimous approval of my opinions!. (as if).

"I feel like I'm talking to an empty room here."

you are....fking ghost town mate, you're the only one on here lol