Hi just a guy,
While the throat chamber can't contribute to the horn flare directly, because it is assumed in the simulation model to be cylindrical, it can be used to increase the effective horn length, as shown in the attached 'extreme' test examples (chosen to clearly illustrate the effect of the length of the offset driver horn throat chamber). It is also possible to use Ap1 and Lpt in an OD system to further increase the effective horn length.
The unmasked power response results for the 100cm offset driver horn (with 60cm throat chamber) are virtually identical to those for the 160cm front loaded horn, as compared in Attachment 3.
Attachment 4 is a representation of the complete acoustical path from the driver diaphragm to the offset driver horn - the red line is the throat chamber length Ltc, the green line is the port tube length Lpt, the pink line is horn segment length L12 and the blue line is horn segment length L23.
Kind regards,
David
Attachments
Well I have to agree that in this extreme example the result is the same because the Nd and Od examples are the same. That is NOT the case in the TH118.
By using L12 of .01 cm in this example there is effectively no segment so your Nd and Od examples really are exactly the same (except for a .01 cm long segment that effectively isn't really there since it's so small). In this case the pink line in your drawing would be so small you wouldn't even be able to see it, and your Od example is basically just an Nd horn, that's why there's no difference in the example. You've created an Nd sim from an Od layout by using a throat chamber but they are both exactly the same.
Once you start getting into a 23 cm L12 which is about as short as you can possibly have with an 18 inch driver in this layout things are going to change. At that point your drawing in attachment 4 would be a bit more valid, you would have two stubs at the end of the horn, each with a significant length. And since the chamber is in parallel with the horn the pink stub doesn't add length to the horn, it's just acting as a 1/4 wave trap, that's what happens with any closed end offshoots of the main line. In the TH118 the chamber (if you attempted to use one) would be short and stubby, only a few cm, and that short stubby chamber would reflect the actual line length. The actual L12 part would be much longer, 23 cm at least, with a triangular shape, and it would NOT add any length to the horn, it would function as a 1/4 wave trap. Vaguely similar results maybe, but not the same and not accurate. You can't simulate a long triangular shaped L12 by using a short, stubby tube shaped chamber. It just isn't the same thing.
By using L12 of .01 cm in this example there is effectively no segment so your Nd and Od examples really are exactly the same (except for a .01 cm long segment that effectively isn't really there since it's so small). In this case the pink line in your drawing would be so small you wouldn't even be able to see it, and your Od example is basically just an Nd horn, that's why there's no difference in the example. You've created an Nd sim from an Od layout by using a throat chamber but they are both exactly the same.
Once you start getting into a 23 cm L12 which is about as short as you can possibly have with an 18 inch driver in this layout things are going to change. At that point your drawing in attachment 4 would be a bit more valid, you would have two stubs at the end of the horn, each with a significant length. And since the chamber is in parallel with the horn the pink stub doesn't add length to the horn, it's just acting as a 1/4 wave trap, that's what happens with any closed end offshoots of the main line. In the TH118 the chamber (if you attempted to use one) would be short and stubby, only a few cm, and that short stubby chamber would reflect the actual line length. The actual L12 part would be much longer, 23 cm at least, with a triangular shape, and it would NOT add any length to the horn, it would function as a 1/4 wave trap. Vaguely similar results maybe, but not the same and not accurate. You can't simulate a long triangular shaped L12 by using a short, stubby tube shaped chamber. It just isn't the same thing.
Last edited:
It should be accurate enough as long as the wavelengths in question are large compared to the length of the section in question, i.e. the same type of assumption that is made almost everywhere else when looking at the impact of a horn's geometry on its frequency response.
Using a 1/4 wavelength criteria for evaluation purposes, for a 23 cm section, we're talking about frequencies over 360 Hz. Even if we're to double that path length, we're still looking at frequencies over 180 Hz, which is at or above the passband of almost every TH bass-horn I know about.
Adding length to the horn does affect Fb, not just frequencies over 360 hz. Even if you only add 1 inch it shifts Fb. Not a lot but this is undeniable fact. Your path length as shown was several inches too short and that does affect tuning. Deny it if you like but it's fact.
We are only looking for a few hz, and there's at least 4 sections of the horn that are quite different than your simple sim. A couple hz here and a couple hz there and you are at your goal.
You already said mass loading at the stub area decreases tuning. You already said mass loading in the mouth area decreases tuning IIRC (I can't remember if you specifically agreed with this but it would be pretty uneducated to disagree). You continue to deny that adding path length in the beginning of the horn decreases tuning, but it does, adding path length anywhere in the horn ALWAYS lowers tuning. So all those little things add up and you end up with lower tuning than your wildly inaccurate sim suggests.
I agree that the 1/4 wave trap will affect frequencies that are quite high, but adding path length ALWAYS affect tuning. It's assumptions like this that are the problem here. Assuming you know what's going to happen isn't getting you anywhere closer to the answer because your assumptions are dead wrong. This is frustrating for me because even a rank amateur to this site should know that adding path length ALWAYS affects tuning.
Do you mind if I ask why you are so unwilling to sim this properly? Is it a time issue, or too much work for you, or are you afraid you will be proven wrong? I realize at this point you are so invested in defending your wildly inaccurate sim it would sting a bit, but why this all out war on accuracy? It doesn't make any sense to me at all. It would take about 8 hours to sim this properly (less if you are good at math) so why not just do it?
We are only looking for a few hz, and there's at least 4 sections of the horn that are quite different than your simple sim. A couple hz here and a couple hz there and you are at your goal.
You already said mass loading at the stub area decreases tuning. You already said mass loading in the mouth area decreases tuning IIRC (I can't remember if you specifically agreed with this but it would be pretty uneducated to disagree). You continue to deny that adding path length in the beginning of the horn decreases tuning, but it does, adding path length anywhere in the horn ALWAYS lowers tuning. So all those little things add up and you end up with lower tuning than your wildly inaccurate sim suggests.
I agree that the 1/4 wave trap will affect frequencies that are quite high, but adding path length ALWAYS affect tuning. It's assumptions like this that are the problem here. Assuming you know what's going to happen isn't getting you anywhere closer to the answer because your assumptions are dead wrong. This is frustrating for me because even a rank amateur to this site should know that adding path length ALWAYS affects tuning.
Do you mind if I ask why you are so unwilling to sim this properly? Is it a time issue, or too much work for you, or are you afraid you will be proven wrong? I realize at this point you are so invested in defending your wildly inaccurate sim it would sting a bit, but why this all out war on accuracy? It doesn't make any sense to me at all. It would take about 8 hours to sim this properly (less if you are good at math) so why not just do it?
In anticipation of the question "Why don't I sim it myself if I'm so sure?", I'll tell you. I don't care even the least bit about the details of this particular cab, I'm arguing the concepts at play. Adding path length decreases tuning, mass loading decreases tuning, these are concepts that even beginners should be well aware of and you seem to be discounting them or outright dismissing them. My busy season is starting and I don't have time to devote several hours to a sim of a cab that I don't really care about. On the other hand you seem deeply invested in why this cab measures the way it does and I've told you why.
It's entirely possible that the measurement is wrong or at least inaccurate but as I've said many times now if the sim is accurate and the measurement is accurate they will agree to a reasonable degree like they always do. There is so much marketing and half truths on DSL product pages that I wouldn't be surprised if the measurement was wrong. For example, referencing sensitivity to 2.83 V (or 28.3 V at 10 m) on ALL products regardless of the product's impedance is cheating and so close to an outright lie that it's shameful. And referencing sensitivity at the peaks is just as bad, why is that number even there? The peaks are a problem that needs to be eq'ed out, not a selling feature. This is misinformation solely invented to prey on the uneducated. And sadly DSL is more honest than most in the industry. So I wouldn't be surprised if the measurement is wrong but as I've said if the sim is accurate and the measurement is accurate they will match.
Your sim is not even remotely close to accurate.
It's entirely possible that the measurement is wrong or at least inaccurate but as I've said many times now if the sim is accurate and the measurement is accurate they will agree to a reasonable degree like they always do. There is so much marketing and half truths on DSL product pages that I wouldn't be surprised if the measurement was wrong. For example, referencing sensitivity to 2.83 V (or 28.3 V at 10 m) on ALL products regardless of the product's impedance is cheating and so close to an outright lie that it's shameful. And referencing sensitivity at the peaks is just as bad, why is that number even there? The peaks are a problem that needs to be eq'ed out, not a selling feature. This is misinformation solely invented to prey on the uneducated. And sadly DSL is more honest than most in the industry. So I wouldn't be surprised if the measurement is wrong but as I've said if the sim is accurate and the measurement is accurate they will match.
Your sim is not even remotely close to accurate.
No-one's denying that. However there is a MASSIVE difference between shifting Fb 1% to 3% downwards, and shifting it 20% downwards, particularly when we're talking about frequencies 34 Hz and below, which is what is required here. You continue to suggest that somehow my sim is "wildly inaccurate", however you also continue to ignore that all of your suggestions as to what that "wildly inaccurate" means in actual numbers is nowhere near what is required to get a match with Danley's measured results, and that's based on my actual experience testing these things (including one I performed and provided the results for yesterday concerning the impact of using a driver with a larger volume @S4), not just fiddling with sims.
I thought the answer to that would have been blatently obvious based on what I've told you so far. If I see something different between the attempted sim and the actual build as picture that could possibly account for that 20% difference between sim'd and measured Fb, then I might try to sim it. As of this moment, I don't see anything that would get us anywhere near to that goal.
So, why don't YOU go and do the Akabak sim and prove me wrong?
Wrong again. I'm suggesting, based on my experience with performing similar changes to THs, that the amount of changes involved is simply too low to account for the 20% difference between sim'd and published results.
I think it would be hard to get an impedance measurement wrong. Maybe when they graphed it they shifted the X axis up by 10 Hz?
Sensitivity is always quoted @ 2.83V/1M. If DSL is referring to the SENSITIVITY of their systems, than that is in fact the correct way to do it. And measuring horns at 28.3V @10M is a well-established means of getting around the issues that can happen when you try to measure horns @1M (i.e. when the horn's mouth is large compared to the distance at which the measurement was performed). It also provides a better idea of what a PA horn can be capable of when in actual use.
Which is actually why I've been looking for independent measurements of their TH118, which seem to be somewhat elusive.
Oh, and FWIW, if the impedance measurement is in fact correct, my hypothesis at the moment is that there's a lot more to that "adapter plate" than is portrayed in that cross-section you provided. Maybe the path at that point is split by a central panel (which won't be visible in that cross-section view) and S1 is moved to the front rather than the rear. We are looking for an extra 60 cm or so, which interestingly enough is quite close to the length of that adapter plate.
I'm not ignoring anything. Your sim is several inches too short in the throat area and there's 3 separate areas where mass loading is present in the cab but not in your sim (the stub section, the driver volume and the grill and grill rails). That stuff all adds up. Add to all that the fact that you've made no attempt to get the actual flare right (it could be considerably different that the quick estimate you've made because you never actually measured the dimensions on the pic I provided) and that's at least 5 factors that your sim doesn't account for. That's the definition of wildly inaccurate in my opinion.
I guarantee you'd get a lot closer to the goal if you sim'ed it properly.
So, why don't YOU go and do the Akabak sim and prove me wrong?
I told you why. This reminds me a lot of my nephew's potty training. I told him that if he did a little bit more work he could save himself a lot of misery sitting in his mess. This isn't something I could do for him and it's not something I'm prepared to do for you. I'm not going to pour several hours of work into proving that common sense and physics are correct and wildly inaccurate guesses don't usually prove very fruitful. He was a stubborn little guy and he spent a lot more time than he had to sitting in his own mess. This is pretty much exactly what is going on in this situation too. Go ahead, be stubborn, it doesn't hurt me but it doesn't help you.
I think it would be hard to get an impedance measurement wrong. Maybe when they graphed it they shifted the X axis up by 10 Hz?
Yes, it might be hard to screw up an impedance measurement. But for all we know this is a measurement of a prototype that is not the same as the production cab. And there could be any number of other issues with the measurement too.
BUT if the sim is accurate and accurately reflects the build the sim will match the measurement to a reasonable degree. This is universal truth.
But even if we assume the measurement is garbage and throw it out, at least an accurate sim would provide an accurate representation of real world response. Your wildly inaccurate sim can't do that.
If your cab impedance is 1 ohm you think it's fair and accurate to state sensitivity at 2.83 V? Lets say you have two cabs each with multiple drivers. One cab has 1 ohm impedance and the other has 16 ohms impedance but other than the driver wiring the cabs are identical. Those cabs have the same sensitivity with 2.83 V input?
I'm not arguing against measuring at 10 meters, it's a good idea. But cabs should be measured with one watt input referenced to their impedance, not 2.83 V regardless of cab impedance. This is cheating, especially when you consider that ALL the cabs are well under 8 ohms, even the ones that are 8 ohm "nominal". All the B&C 8 ohm drivers are closer to 5 ohms and NONE of the DSL cabs should be measured at 2.83 V because from the ones I've looked at none of them are remotely close to 8 ohms, even the 8 ohm "nominal" cabs are a lot closer to 4 ohms than 8. Seriously I think you are looking for stuff to disagree about now. This is common sense.
And this is the problem. Instead of doing on accurate sim you are looking for things that are not there. There's no room for central panels and other conspiracy theories inside that throat section. Danley doesn't like to use compression ratios higher than 2:1 in general, in this cab it's a lot higher and what you are proposing would produce an incredibly high compression ratio and a massively uneven load on the cone. Maybe there's leprechauns inside the throat that spin around really fast and produce a vortex which simulates a longer path length. Have you considered that possibility too? Isn't it easier just to do an accurate sim and see where that puts the tuning than to make up things that are not there?
Last edited:
Yes, because that is how sensitivity is DEFINED, JAG. It should not be stated any other way. Have a look at this: Loudspeaker Sensitivity & Impedance Explained | Audioholics
Sensitivity has nothing to do with the impedance. You are confusing sensitivity with efficiency.
I actually agree with DSL's approach to using sensitivity instead of efficiency. Almost all good modern amps are almost like voltage sources (within their minimum output impedance ratings), and using the sensitivity ratings will give much better idea of how much dB of difference you can expect when you hook up different cabs, when powering them with an amp operating within its operating range.
I'm going to go ahead and admit I haven't studied the formal definitions on these things but even in the article you linked to is this gem.
Whenever sensitivity is quoted, the nominal impedance must also be stated. This will prevent the manufacturer from cheating in cases where the lower impedance speaker is able to draw more current making the speaker appear more efficient. Of course the efficiency has NOT increased, but the sensitivity has, which is why it is important to include nominal impedance into the spec.
He specifically states that sensitivity increases with lower impedances, which does contradict the article itself, but there you have it.
Then a quick google search gave these three top hits.
"Sensitivity is most easily defined as the speakers’ ability to effectively convert power into sound. The traditional way of measuring a speakers’ sensitivity is using the standard of 1 watt/1 meter." - Guide to Speaker Specifications - PSB Speakers
This is my understanding of how it should be done. Is that the formal definition? Not sure.
"Speaker sensitivity is measured in decibels per 1 watt per 1 meter, but is usually referred to as just decibels [source: JBL]." - Speaker Sensitivity Explained - HowStuffWorks
Again, same thing, and apparently the source on this is JBL.
And one more. "Speaker sensitivity is a measurement of the amount of sound output derived from a speaker with one watt of power input from an amplifier." - What Is Speaker Sensitivity?http://stereos.about.com/od/glossaryoftermss/g/sensitivity.htm
And one more for good measure. A speaker’s sensitivity can be measured in any number of ways. However, the audio industry has established a standard way of doing this so that the sensitivities of various loudspeakers can be easily compared." - GoodSound! "Features" -- Understanding Loudspeaker Sensitivity (2/2008)
And I found a bunch like this. This implies that the definition is flexible but that the industry has come up with a way to show higher sensitivity ratings for lower impedance speaker. Note in the first quote it says "the traditional way is 1w/1m, not 2.83V/1m. This implies (to me at least) that it should be 1w/1m and always has been until the industry decided to change it so they could show beefed up numbers for low impedance speakers.
So please excuse me if I got the formal definition wrong (although I'm not sure that I have, at least half the world agrees with me, except the industry of course, who want to show big numbers and this is an easy way to do that.)
But again, I haven't studied the formal definitions and I really don't care that much. I see it as cheating. But in this case I'll back off since there is sufficient proof that much of the industry cheats in the same way.
ANYWAY efficiency is something else entirely, it's expressed as a percentage, not in db. A common efficiency rating for a common speaker is something like 2%. So I don't think I'm confusing sensitivity and efficiency here. They are completely different things expressed in completely different ways (db vs percentage).
Nobody ever rates anything in efficiency. When is the last time you saw a spec sheet say 2% efficiency? Nobody would ever sell anything if they quoted efficiency instead of sensitivity so efficiency is never even mentioned. Ever. By anyone selling speakers.
But more on point, sensitivity should be expressed at 1w/1m, not 2.83V/1m IMO or you have to do a bunch of mental math to figure out what that number actually means. There is a certain usefulness as you have pointed out here, but it would actually be a lot more useful if amps were rated by their voltage rails (not watts) and speaker sensitivity was expressed as 1w/1m AND max voltage instead of max power.
The industry has done a pretty good job at completely and utterly confusing all specs and making them near useless, which is something that Danley does talk about a lot but at the end of the day I don't like how he does it either.
Last edited:
Brian,
Post #46 shows Langston Holland's independent measurements of his 4 ohm DHL TH118s and 8 ohm (IIRC) TH115 cabinets:
http://www.diyaudio.com/forums/subwoofers/185588-keystone-sub-using-18-15-12-inch-speakers-5.html
Langston is very thorough and accurate in his measurements. They agree closely with DSLs. The Keystone is very similar in phase and frequency response to the DSL TH118.
IIRC, Langston's TH118 probably was using an 18Sound driver. The DSL website has still never updated the TH118 spec sheet from their initial use of 18Sound drivers, even though they switched to BC18SW115-4 shortly after the TH118 introduction, after some production problems with the 18Sound drivers that have subsequently been addressed.
Art
Unfortunately that link given in the post to the measurements in question is not functional. I did search the ProSound site to see if there were any other test results available, but what I found also turned out to be dead ends.
Did he take a look into them to see if they differed substantially in build to the TH115?
That should help a bit, thanks.
Sensitivity for loudspeakers is defined in terms of dB/2.83V/1M. Anything else is incorrect. Unfortunately there is a lot of bullsh*t and misinformation out there posing as truth.
The ONLY reason it may need to be stated is for the purchaser to be able to ensure that the amp he's planning to use is able to use speakers of that particular impedance, and perhaps how much power his amp can provide at that impedance (to compare against the wattage rating for the speakers). The rest is just marketing-speak and nonsense. Because, here's the thing - an amplifier is not going to suddenly shift down to 2V of output at the same gain level just because you replaced the 8 ohm loudspeaker at its terminals with a 4 ohm one. It will continue to provide output at the same voltage (unless it has a stupidly low DF) until such time at it's reached its power output limits, at which time its output voltage will no longer increase. So suggesting that a manufacturer who provides a 4 ohm speaker instead of an 8 ohm one is somehow cheating when they quote the sensitivity of their speakers in dB/2.83V/1M is just nonsense. Sensitivity ratings are independent of impedance.
That may be "traditional", but it certainly does not make it any more correct. It is, to put it simply, wrong to state sensitivity in terms dB/1W/1M. That's a term used for loudspeaker efficiency, not sensitivity. I think a lot of the confusion may have arisen from most speakers being rated with a nominal 8 ohm impedance. At that impedance, the sensitivity and efficiency values should be the same, i.e. if an 8 ohm speaker is rated as having a sensitivity of 98dB/2.83V/1M, then its rated efficiency should be 98dB/1W/1M. It's therefore not surprising to see that even to this day there are people mixing up these two terms.
Not true. It can be expressed in terms of both % and dB/1W/1M. In fact, there is a simple and direct mathematical relationship between the two, i.e.:
Efficiency (in dB/1W/1M) = 112 + 10*log(n0), where n0 is the efficiency in percentage form.
Hah, that's likely not to happen, maybe because the marketing people will not have those big wattage numbers to put in their ads
. In any case, it doesn't make much sense, as the max voltage will be limited by what power the amp can actually produce. They only act as near voltage sources until their power output limits are approached. Most amps produce a higher peak output voltage at their output terminals when there's no load connected than when there is.Post #112
Hi Art,
Thanks for confirming that the currrent dirver in the DSL TH-118 is the B&C 18SW115-4. That'll help me w/ the drawing, and then w/ an attempt @ AkAbak (this will take a little, as I'm still not sure about the general geometry). I figure, that Neo Dan's drawing in the Keystone thread will give me the cone dimensions that I need for the TH-118 throat drawing.
Regards,
Hi Art,
Thanks for confirming that the currrent dirver in the DSL TH-118 is the B&C 18SW115-4. That'll help me w/ the drawing, and then w/ an attempt @ AkAbak (this will take a little, as I'm still not sure about the general geometry). I figure, that Neo Dan's drawing in the Keystone thread will give me the cone dimensions that I need for the TH-118 throat drawing.
Regards,
I didn't want to mess up our sensitivity / efficiency discussion with this, so I've reserved it for another message
.First thing, I did get the external dimensions for the TH118 from Danley's site and I've also got the mouth dimensions. Knowing those two, it's not that difficult to work out the rest of the dimensions, particularly if you have access to a spreadsheet that assists with the folding
. I believe that slight differences in the flare rate are not going to add up to appreciable differences in Fb, if the volume remains the same. In fact, the evidence is right there to back this up - my sim comes within 1 Hz of the TH115's measured Fb, and that's supposed to share the same internal layout as the TH118. My sim's predicted Fb is actually lower, so if it in fact introducing an error because of an incorrect flare rate, the error it's introducing is going in the wrong direction - it's decreasing Fb, not increasing it .As to the other stuff, it's clear that the only difference between my position and yours at this point is the AMOUNT of the Fb shift we each believe those additional things would produce, so I'm curious to know on what are you basing your opinion that the effects of those things will impact Fb by as much as 20%? Have you performed your own tests, or are you just basing this on the published impedance response and a cross-sectional drawing that may or may not show the whole picture?
If I see something else in the built TH118 that may contribute to a 20% decrease in Fb, I may redo the sim. At the moment, I'm not seeing that, so I'm really not inclined to do so.
Hi just a guy,
Comparing the drawings from Post #44 and the drawings from Post #33/#60 I noticed that there are quite a few discrepancies. I'm not quite shure which one to use for an AkAbak simulation, do you have a preference, or comment? Or should I combine them with an eye on longest path? For example, the little piece @ the bottom of the V above the throat is missing in 33/60.
Thanks for those posts.
Regards,
Comparing the drawings from Post #44 and the drawings from Post #33/#60 I noticed that there are quite a few discrepancies. I'm not quite shure which one to use for an AkAbak simulation, do you have a preference, or comment? Or should I combine them with an eye on longest path? For example, the little piece @ the bottom of the V above the throat is missing in 33/60.
Thanks for those posts.
Regards,
Last edited:
Right from the Harman.com audio glossary -
A standardized measure of the sound output of a loudspeaker for a known input signal. Originally, the input power was 1 watt. Nowadays, the input is standardized to 2.83 volts (1 watt into 8 ohms).
And although I don't have the formal definitions I'm going to tend to trust Harman. And the link I provided earlier was sourced from JBL. Like I said, the definition used to be 1w/1m. Apparently now the industry has changed that. If you want to believe it's because manufacturers think it's important to spec sensitivity based on a specific spot on the volume knob that must be the same for all speakers regardless of impedance that's fine. I'm a bit more jaded and I firmly believe they made the change so they could publish higher numbers for low impedance speakers. Everything on a spec sheet is purely driven by marketing, not by a desire to help the consumer by providing useful information.
So NOW the sensitivity spec might be independent of impedance but it wasn't always that way. They just switched from sensitivity based on wattage to sensitivity based on voltage. Just for marketing purposes I believe.
And what difference does it make where the amp volume knob is? The sensitivity spec shouldn't have anything to do with that. The lower impedance speaker will be louder at the same spot on the volume knob, so what? Turn the volume up on the higher impedance speaker and that problem goes away really fast. Different voltage, same spl. Rating sensitivity on a common voltage regardless of impedance is backwards, unintuitive, and kind of dumb, IMO. And it wasn't always like this until the industry decided to change it.
We don't have to agree on this, it really isn't that important. Harman and JBL both say the definition used to be 1w/1m but the industry changed it. I think it was much better the way it was, you don't. It seems nobody is wrong here, just a difference of opinion.
My position is only that measurements will match sims when the sim is an accurate representation of the built cab. My secondary arguments are that extra path length and mass loading both lower tuning. My third point is that your sim doesn't take a lot of this into account. My fourth point is that you really shouldn't go looking for things are not shown in any of the data presented (like V chambers or bifurcated paths in the throat area) to explain why the measurements don't match the sims when a proper sim has never even been done in the first place.
Way back in post 47 I stated the measurement may be wrong. I'd prefer to believe it's not, the measurements on these spec sheets are the one thing that don't seem influenced by the marketing dept.
And way back in post 44 I clearly stated I don't know how accurate the dimensioned drawing from the speakerplans site was either.
And a few times I've stated that I really don't care much about this particular cab or how it measures. The only thing I care about is accuracy in simulation.
So really my whole point this whole time is that at least attempting to sim the thing accurately is a lot better than guessing what effect a bit of path length here and a bit of mass loading there (and there and there and there) will have. All these things add up. Do they add up to a 20% difference? I really don't know and I don't care. All I know for sure is they do add up to some amount of difference and you will never really know how much if you don't even attempt to sim it accurately.
These are just pictures I found on the internet, I have no idea who posted them, what they are based on or how accurate they may be. Your picture shows clear differences (it appears all the internal panels drawn in red except the reflector in the last corner are shifted a bit to the left) but they look fairly close, for example the overall path length appears to be the same but the csa is slightly different through most of the horn. Personally if I were interested enough to sim it at all the first thing I'd do is look for more drawings and see if I could find a definitive dimensioned drawing. If not, I'd try to contact the people that originally posted those drawings and find out where they came from and how accurate they are. The difference between those two drawings may be as simple as a scaling error when the pics were saved or something silly like that. It's hard to imagine they would be so close if one or both were just drawn freehand without knowing the actual dimensions. Failing those two options I'd run the sim both ways you've shown and see what happens. The answer is likely that one of them is correct, or if neither are correct the real dimensions are probably somewhere in between.
The biggest reason I didn't want to do the sim myself is because I didn't want to figure out the cross sectional area as the path moves through the cone. I'm not good at math and I don't envy you that task. It would take me a few hours just to plot out the first couple of feet of this horn. I'm sure there's an easy way to go about it, but I don't know what that easy way is.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.