Hi just a guy,
For almost all subwoofers in 'free field' count that when they are stacked, the response within the passband improves. That also counts for Tapped Horns!
I just showed a pic of a simulation of a tapped horn single vs stacked. It doesn't look like an improvement to me. I can show you a dozen other pics of other models of tapped horns that will look quite similar and a dozen examples that are much worse when stacked.
I agree that stacking changes the response. As far as "improves the response", that depends entirely on whether the change is viewed as an improvement or not. Clearly the flh sim I showed is an improvement, while I would not consider the tapped horn sim I showed to be an improvement at all. (But as I mentioned, this particular tapped horn does not measure as the simulation indicates - not sure why not - so this may not be a bad thing for this particular tapped horn.)
I've simulated some of Danley's tapped horns and seen measurements. Clearly those were designed with stacking in mind (at least as a possibility, not as a necessity). Those tapped horns do improve when stacked. This is not true of all tapped horns though, and not necessarily true of any particular example of any alignment.
If you want to argue again that Hornresp is not accurate for simulating multiple cabs, I'm not interested in going down that road again.
Hi just a guy,
Arguing is not necessary considering the words of the creator/writer of HornResp:
Last edited:
Without the context of the discussion that quote came from, it's basically meaningless. What did he want to improve on the existing model? If the existing model is so terrible why doesn't he remove it from Hornresp?
I'm aware that it's said that stacked tapped horns do not play the 2 or 3 hz lower than a single cab than the sims would indicate. Since I don't have any multiple tapped horns to stack and measure, I can't comment on that specific phenomenon, and that 2 or 3 hz of extended response is really not a big deal and not what I was commenting on here.
But as far as getting a clear picture of the response curve shape, the multiple speakers tool is way better than we need it to be. Think about what this feature does - it multiplies the drivers, chamber volumes and cross sectional areas by the amount of multiple speakers. That's all it does (as far as I know). If that doesn't work then it shouldn't be able to simulate a single cab accurately either. The method that the multiple speakers tool uses to simulate multiple speakers is exactly how we would simulate a stack if this tool were not available.
I'm willing to concede that Hornresp isn't perfect but I don't think you give it nearly the credit it deserves - it's way better than we need it to be.
I'm aware that it's said that stacked tapped horns do not play the 2 or 3 hz lower than a single cab than the sims would indicate. Since I don't have any multiple tapped horns to stack and measure, I can't comment on that specific phenomenon, and that 2 or 3 hz of extended response is really not a big deal and not what I was commenting on here.
But as far as getting a clear picture of the response curve shape, the multiple speakers tool is way better than we need it to be. Think about what this feature does - it multiplies the drivers, chamber volumes and cross sectional areas by the amount of multiple speakers. That's all it does (as far as I know). If that doesn't work then it shouldn't be able to simulate a single cab accurately either. The method that the multiple speakers tool uses to simulate multiple speakers is exactly how we would simulate a stack if this tool were not available.
I'm willing to concede that Hornresp isn't perfect but I don't think you give it nearly the credit it deserves - it's way better than we need it to be.
Last edited:
Just to be very clear about what I'm getting at here, stacking multiple speakers changes the response curve shape (compared to a single cab). I don't think anyone is arguing that.
It stands to reason then, that if the response of a single cab is already flat, changing the response curve shape will make it less than flat. That also means that if it is not already flat, changing the response curve shape could potentially make it more flat.
But it's just not true that stacking ANY speaker will improve the response. What you end up with depends entirely on what you start with. Even if you don't trust Hornresp I don't think you can argue with that.
It stands to reason then, that if the response of a single cab is already flat, changing the response curve shape will make it less than flat. That also means that if it is not already flat, changing the response curve shape could potentially make it more flat.
But it's just not true that stacking ANY speaker will improve the response. What you end up with depends entirely on what you start with. Even if you don't trust Hornresp I don't think you can argue with that.
Hi just a guy,
Because of the difficulty in determining the theoretical directivity of a multiple speaker array (detailed information on the relative positions of the radiators is required) the Hornresp multiple speakers model calculates overall power response, not pressure response. This is why predicted results can be different to actual measured pressure response.
I was initially trying to see if I could change the model to better reflect results obtained by Art (weltersys), but realised that this was the wrong approach to take, as there was little point in trying to directly compare "apples" (predicted power) against "oranges" (measured pressure).
Because I am comfortable with the existing multiple speakers model as a predictor of overall radiated power.
Kind regards,
David
Because of the difficulty in determining the theoretical directivity of a multiple speaker array (detailed information on the relative positions of the radiators is required) the Hornresp multiple speakers model calculates overall power response, not pressure response. This is why predicted results can be different to actual measured pressure response.
I was initially trying to see if I could change the model to better reflect results obtained by Art (weltersys), but realised that this was the wrong approach to take, as there was little point in trying to directly compare "apples" (predicted power) against "oranges" (measured pressure).
Because I am comfortable with the existing multiple speakers model as a predictor of overall radiated power.
Kind regards,
David
I think what you are saying here is that there is no problem whatsoever with the response curve shape prediction but there may (or may not) be a problem with directivity (depending on overall dimensions, which could theoretically change the measured response).
If that's what you are saying, it makes no difference whatsoever if you simulate a stack of 8 speakers or one large speaker with 8 drivers, 8x the chamber size and 8x the cross sectional area of the horn flare, as long as the mouth shape is the same in both instances (and as close to circular as possible). And that stands to reason, since there's really no physical difference (except the stack will have dividing walls, which doesn't really change anything, since we all know a single large speaker can have dividing walls called bracing with no detrimental effect).
Djim is arguing that this is not the case and the Hornresp multiple speakers feature is inaccurate.
Just to be sure, I simulated the same tapped horn model I did a few posts ago as a stack of 8 with the multiple speakers feature, captured the results, and then redid the sim as a single speaker with 8 drivers, 8x the chamber volume and 8x the horn flare csa. Then I compared the results and they overlay perfectly.
This means the multiple speakers feature is no less accurate than the regular simulation mode. As I said previously.
Does anyone have a link to this discussion? I don't know what was measured and I don't know if the speakers were even simulated so I'm not sure what you would want to change.
In case it wasn't clear that was rhetorical. I'm very comfortable with the existing multiple speakers feature as well. As I said, it's way better than we need it to be.
Assuming the mouth shape of the physical speaker is as close to circular as possible and the dimensions aren't too large compared to the wavelengths in the passband, there's nothing wrong with the multiple speakers feature - and even if there was a problem, this same problem exists in the regular simulation mode as well since they give the same results.
Djim, I didn't want to get into this but since it's started and we have Mr. McBean's attention, perhaps you can tell me what exactly you think is wrong with the multiple speakers feature. I'd like to clear this up now instead of having you state that it's not accurate every time I show a simulation of multiple speakers.
Last edited:
Discussion link here:
http://www.diyaudio.com/forums/subwoofers/204472-multiple-cabinet-combined-response.html
Hornresp assumes the mouth shape of the physical speaker is circular, the mouth exit is the size of the horn, and that the entire construct is buried in a hole with the exit flush with the ground (for half space modeling), assumptions hard to match in most real world situations.
As David McBean is aware, those details seldom match the physical array of boxes, which is why he says:
"Because of the difficulty in determining the theoretical directivity of a multiple speaker array (detailed information on the relative positions of the radiators is required) the Hornresp multiple speakers model calculates overall power response, not pressure response. This is why predicted results can be different to actual measured pressure response."
DJM and I have no quarrels with David (we both find Hornresp useful in many respects) but do have a problem with your presenting models as a reality, since the simulation is often is at variance from the actual measured response of the constructs.
Art
Thanks, I'll read that when I get some spare time. I'm building sub boxes today.
This is NOT specific to the multiple speakers feature. The regular simulation mode makes these assumptions as well and I don't see anyone complaining about regular Hornresp sims. (Actually, you did complain about Hornresp's regular mode accuracy, but you admitted that you never attempted to simulate your horns accurately in the first place so it's no wonder that the sims didn't match the measurements.)
I think we're all aware of that, and again, this is NOT specific to the multiple speakers feature, this will affect ALL horns when the mouth does not conform to the specifications you mentioned above.
As I mentioned the last time we discussed this, if you take the time to simulate what you build ACCURATELY and make the mouth shape as close to circular as possible, the sims will be very very close to the measurements. Much closer than we have any right to expect. This has been proven over and over again by people that take the time to simulate accurately, as I mentioned, soho54 did this all the time and proved it to be true. Unfortunately he's MIA so he can't add to this discussion.
As far as I can tell, the only issue being argued here is directivity of a large mouth (or combined mouth) but this is very old news by now. It's been repeated over and over that as long as the mouth is not ridiculously large in comparison to the wavelengths in the passband and care is taken to make the mouth (or combined mouth) as circular as possible, this is NOT an issue. Even if you COMPLETELY mess up the mouth shape the worst that can happen is the on axis measurement won't match the sim, but the power response is still accurate.
I appreciate you taking the time to discuss this again (even though I don't agree with you). I do NOT appreciate Djim dropping in every time I show a sim of multiple speakers to say it's not accurate and then leaving no further comment on why he thinks it's not accurate. That doesn't help anyone, and as far as I can tell from his comments, he doesn't even understand that the issue here is one of directivity and it's really not much of an issue at all.
There are few who build round exit subwoofers for PA use, they just don't pack in the truck well
.Although the sims would be very accurate if anybody could build a multiple speaker box that had a single round exit buried in the ground, that does not happen in reality.
The directivity factors which are not simulated in Hornresp have very large real world effects, as you will find when (if) you measure your subs in various arrays.
Good luck with your build, I should get back to mine, only 72 cones and 3 more cabinets to go...
So I was interested enough to stop what I'm doing to take a quick look at the thread you linked to.
I hope this is not news to anyone, especially the parts in bold (emphasis provided by me).
With comparisions of stacking 8 cabs horizontally vs 8 cabs vertically, and every combination in between those extremes, the difference within the range of interest (below 100 hz) is only about 3 db.
Wow, you guys are really making a big deal about nothing.
I'm going to study these graphs in much more detail tonight but I don't see anything noteworthy at first glance (I don't have enough time to study which graph belongs to which array right now), but it shouldn't surprise anyone that different stack shapes produce different results on axis. The variations of about 3 db max below 100 aren't anything to get excited about, and power response for ALL these arrays should be the same.
I hope this is not news to anyone, especially the parts in bold (emphasis provided by me).
With comparisions of stacking 8 cabs horizontally vs 8 cabs vertically, and every combination in between those extremes, the difference within the range of interest (below 100 hz) is only about 3 db.
Wow, you guys are really making a big deal about nothing.
I'm going to study these graphs in much more detail tonight but I don't see anything noteworthy at first glance (I don't have enough time to study which graph belongs to which array right now), but it shouldn't surprise anyone that different stack shapes produce different results on axis. The variations of about 3 db max below 100 aren't anything to get excited about, and power response for ALL these arrays should be the same.
Last edited:
There are few who build round exit subwoofers for PA use, they just don't pack in the truck well
Although the sims would be very accurate if anybody could build a multiple speaker box that had a single round exit buried in the ground, that does not happen in reality.
Getting close enough with a square mouth (and also factoring in boundary reflections as we discussed the last time we went over this) is more than close enough.
Clearly at 1000 hz it makes a huge - tremendous - difference. At 100 hz the variation in your measurements looks like about 3 db max.
Parking your truck behind the array would probably cause more difference in the measurement than these array shapes.
Just a little sub box for my car. Not even a horn. Yours sounds a lot more ambitious, so post pics.
Last edited:
Originally posted by weltersys
The directivity factors which are not simulated in Hornresp have very large real world effects, as you will find when (if) you measure your subs in various arrays.
Scale the frontal dimensions up as in a "normal" horn size of around 45" x 22.5", which become large relative to a 100 Hz wavelength, extend response to 35-40 Hz, and the differences in upper to lower response between array configurations become far more extreme than shown in my "little" test.
The project I am working on now are unusual very narrow dispersion mid-high cabinets (160-16Khz), I'll post results on the full range forum when they are finished and tested.
Art
The directivity factors which are not simulated in Hornresp have very large real world effects, as you will find when (if) you measure your subs in various arrays.
The directivity array tests used small monitors, only 11.25 x 22.5 inch frontal dimensions, with differences as you mention of about 3 dB in less than an octave range.
Scale the frontal dimensions up as in a "normal" horn size of around 45" x 22.5", which become large relative to a 100 Hz wavelength, extend response to 35-40 Hz, and the differences in upper to lower response between array configurations become far more extreme than shown in my "little" test.
The project I am working on now are unusual very narrow dispersion mid-high cabinets (160-16Khz), I'll post results on the full range forum when they are finished and tested.
Art
For the full horizontal test, 22.5 x 8 cabs = 180 inches from end to end (15 feet). I think a 15 foot horizontal array is about as bad as it's ever going to get. That's as wide as a horizontal array of 8 Labhorns.
I don't think there's anything wrong with your little test.
For the record, I appreciate this info very much, I think it's a great test, the results are very interesting and informative, well worth studying and a great resource for anyone considering stacking speakers.
BUT I do not think it constitutes a convincing argument that Hornresp's multiple speakers feature is inaccurate. Since the multiple speakers feature provides exactly the same results as the regular mode, Hornresp as a whole is either good enough or it's not. As far as I'm concerned it's way better than good enough. It's true that you can do all kinds of things with different shapes and sizes of horn mouth but that's beyond the scope of Hornresp and really not that important until mouth size gets pretty massive. It's good to be educated about dispersion effects but nowhere near important enough to proclaim Hornresp is inaccurate every time a stack is simulated and posted. A simple link to any number of discussions about dispersion is more than adequate and (IMO) much more tactful than repeatedly claiming Hornresp is flawed.
Post a link please, I don't visit that forum.
Hi just a guy,
What I think of how much credits David deserves for all his efforts and devotion to give everyone the opportunity to model an enclosure with HornResp does not come close to your personal interpretation of my words. Actually I didn’t question HornResp but only your claim, nothing more nothing less.
A power response shows the amount of sound energy, per frequency, coming from a loudspeaker. It doesn’t take account of directivity influences (radiation pattern) that results in sound pressure differences which you can see in a Polar response or a 3D radiation pattern. In other words the power response from the ‘multiple speakers’ function is the theoretical acoustic output power per frequency for a loudspeaker given in decibels (dB’s).
When you setup an array of Tapped Horns in 'free field' (outdoors) conditions, the sound is ‘steered’ into a certain direction. It’s the result when subwoofers are clustered into an array and they 'mutually couple'. This 'steering' also has an effect on the sound pressure around the stack and that can be measured in a pressure response with a microphone (pressure device). I believe that is what Art tried to show you and what sound engineers in general consider as 'real world' results.
But there are more differences between 'multiple speakers' and 'mutual acoustic coupling'. I can’t speak for David but his word choice 'multiple speakers' could be an indication it is not meant to suggest it the same as 'mutually coupled speakers'.
David, please correct me if I am wrong (anywhere) but the current HornResp approach is basically modelling a larger enclosure with a larger driver in a different acoustic environment. This approach means that all drivers, ports, mouths ect fully operate in the same enclosure volume and the same hydrodynamic nearfield since every element is considered 'one'. The response of the 'multiple speakers' function, is therefore the theoretical sum of the radiated power of the multiple drivers coming from 'one' system. HornResp does not use a specific model that actually models the complexity of ‘mutual coupling’ that is different for each type of enclosure but also different for each type of array-setup.
Mutual coupling means the extent of the 'hydrodynamic nearfield'. More important is that mutual coupling is determined by the distance between two or more sound sources. For instance, with the current HornResp method both drivers act like 'one'. It means the distance between the two drivers has become zero and they fully operate in a combined pressure zone (hydrodynamic nearfield). In reality the drivers of two mutually coupled tapped horns have a physical distance of at least 2 times the distance from driver to mouth. This distance is too large for the drivers to operate in a combined pressure zone. For the 'taps' the distance with the current method is zero again. In reality they have distance of at least two times from 'tap' to mouth. Although there is still a significant distance, often it is 'close' enough to speak of a combined pressure zone.
'Mutually coupled' enclosures interact in a complex matter that has an effect on pressure, phase, and radiation and therefore also on its pressure response. In my vision this makes a significant difference between the current method as used by the 'multiple speakers' function and the rules of 'mutual coupling'.
That doesn’t mean there is something wrong with the 'multiple speaker' function as long it is not used for specific analyses of arrays. It is also important to be aware that some types of enclosures show more differences between 'mutual coupling' in the real world and the 'virtual' outcome of the 'multiple speakers' function. In my view the current HornResp method is consistent but the outcome is not. In that sense, comparing different enclosures types, based on the outcomes of the multiple speaker function, becomes even more questionable in my view. Nevertheless, I still think this function can be useful as long you are aware of its limitations.
Like I wrote before, so far all measurements of stacked Tapped Horns show an improved response within the passband. Of course you can debate what the passband of a Tapped Horn is but that is a different subject. It also doesn’t matter if the cab says "made by Danley" or if it is made by a poor guy in Botswana that knows what he is doing. This improvement can be expected since it counts for almost all subwoofers in general (enclosure type independent). In other words, Tapped Horns are no exception for the rules of 'mutual acoustic coupling'.
What I think of how much credits David deserves for all his efforts and devotion to give everyone the opportunity to model an enclosure with HornResp does not come close to your personal interpretation of my words. Actually I didn’t question HornResp but only your claim, nothing more nothing less.
A power response shows the amount of sound energy, per frequency, coming from a loudspeaker. It doesn’t take account of directivity influences (radiation pattern) that results in sound pressure differences which you can see in a Polar response or a 3D radiation pattern. In other words the power response from the ‘multiple speakers’ function is the theoretical acoustic output power per frequency for a loudspeaker given in decibels (dB’s).
When you setup an array of Tapped Horns in 'free field' (outdoors) conditions, the sound is ‘steered’ into a certain direction. It’s the result when subwoofers are clustered into an array and they 'mutually couple'. This 'steering' also has an effect on the sound pressure around the stack and that can be measured in a pressure response with a microphone (pressure device). I believe that is what Art tried to show you and what sound engineers in general consider as 'real world' results.
But there are more differences between 'multiple speakers' and 'mutual acoustic coupling'. I can’t speak for David but his word choice 'multiple speakers' could be an indication it is not meant to suggest it the same as 'mutually coupled speakers'.
David, please correct me if I am wrong (anywhere) but the current HornResp approach is basically modelling a larger enclosure with a larger driver in a different acoustic environment. This approach means that all drivers, ports, mouths ect fully operate in the same enclosure volume and the same hydrodynamic nearfield since every element is considered 'one'. The response of the 'multiple speakers' function, is therefore the theoretical sum of the radiated power of the multiple drivers coming from 'one' system. HornResp does not use a specific model that actually models the complexity of ‘mutual coupling’ that is different for each type of enclosure but also different for each type of array-setup.
Mutual coupling means the extent of the 'hydrodynamic nearfield'. More important is that mutual coupling is determined by the distance between two or more sound sources. For instance, with the current HornResp method both drivers act like 'one'. It means the distance between the two drivers has become zero and they fully operate in a combined pressure zone (hydrodynamic nearfield). In reality the drivers of two mutually coupled tapped horns have a physical distance of at least 2 times the distance from driver to mouth. This distance is too large for the drivers to operate in a combined pressure zone. For the 'taps' the distance with the current method is zero again. In reality they have distance of at least two times from 'tap' to mouth. Although there is still a significant distance, often it is 'close' enough to speak of a combined pressure zone.
'Mutually coupled' enclosures interact in a complex matter that has an effect on pressure, phase, and radiation and therefore also on its pressure response. In my vision this makes a significant difference between the current method as used by the 'multiple speakers' function and the rules of 'mutual coupling'.
That doesn’t mean there is something wrong with the 'multiple speaker' function as long it is not used for specific analyses of arrays. It is also important to be aware that some types of enclosures show more differences between 'mutual coupling' in the real world and the 'virtual' outcome of the 'multiple speakers' function. In my view the current HornResp method is consistent but the outcome is not. In that sense, comparing different enclosures types, based on the outcomes of the multiple speaker function, becomes even more questionable in my view. Nevertheless, I still think this function can be useful as long you are aware of its limitations.
Like I wrote before, so far all measurements of stacked Tapped Horns show an improved response within the passband. Of course you can debate what the passband of a Tapped Horn is but that is a different subject. It also doesn’t matter if the cab says "made by Danley" or if it is made by a poor guy in Botswana that knows what he is doing. This improvement can be expected since it counts for almost all subwoofers in general (enclosure type independent). In other words, Tapped Horns are no exception for the rules of 'mutual acoustic coupling'.
Last edited:
This is what I've been saying all along. I'm glad we can finally agree. I certainly was not suggesting that Hornresp should be used for specific analysis of arrays, I was just showing the power response of a stack of tapped horns, because that's what Hornresp shows.
And I don't buy your description of an issue with the "current Hornresp method" vs "real world conditions" wrt mutual coupling. What you are showing in the picture makes no practical difference.
Unfortunately we still can't agree on this. Stacking any type of speaker changes the response curve shape. If a single cab is relatively flat, stacking a bunch of the same speaker will result in a response curve shape that is less than flat.
I've only seen one measurement of stacked tapped horns and it was made and measured by Danley, so it's no surprise that it improved when stacked. I've seen no measurements of "bad" tapped horns, which the simulations suggest would not improve when stacked so I don't have any proof with real measurements to show. But I can show you all kinds of simulations. For example, a flh will generally have a rising response but in a stack of 8, it will have a fairly flat response. But in a stack of 16 or 32 the low end becomes overpowering. I can show simulations of this, and I have no reason to believe there's anything wrong with the simulation (other than the fact that directivity would make the situation even worse due to the extremely large mouth with the larger stacks).
If the response curve shape changes when speakers are added, it can't always be considered an improvement when more speakers are added. At some point things are going to get worse, and sometimes that point is going to be passed fairly quickly (with a small amount of speakers.)
Go back and take a look at the TH115 measurement (1, 2 and 4 cabs) in the other thread. I think it was you that posted that pic the last time we discussed this. Use your imagination to extrapolate what might happen if another 4, 8 or 16 cabs were added to that stack. Do you think it would become a perfectly ruler flat line and stay that way no matter how many cabs are added? I don't think so, adding more cabs than the 4 shown would start to result in LESS flat response with a boomy low end (at around 40 hz) and the top end would drop off (80 - 110 hz). In other words, the trends that you already see happening as 1 speaker is added, and then another pair is added, would continue to happen as more and more cabs are added.
And that's a very good tapped horn. What do you think happens when lesser tapped horns are stacked? As I showed last time, you can end up with large holes in response because they have a bad impedance profile for stacking. Of course I don't have measurements of stacks of bad tapped horns and I'm not going to construct them to prove it to you, but since you agree that there's nothing wrong with the multiple speakers feature (as long as it's not used for analysis of arrays), you can go back and look at the simulation I already showed the last time we discussed this. I showed one tapped horn that developed a very bad hole in the response as more cabs were added, and I showed another example that was much closer to a "Danley" tapped horn that did not have this issue.
Last edited:
Just in case you don't remember where those sims are, here's my sim of a "bad" tapped horn for stacking purposes.
http://www.diyaudio.com/forums/subw...oop-folded-horn-subwoofers-6.html#post3419569
Here's a "good" tapped horn for stacking purposes.
http://www.diyaudio.com/forums/subw...oop-folded-horn-subwoofers-7.html#post3421353
The reason these two tapped horns sim differently when stacked is not due to a problem with Hornresp, it's due to their individual differences in their impedance profiles. In the first example, the first really big, narrow TH peak becomes a hole in response when stacked whereas in the second example it doesn't. (In both examples, that first big, narrow TH specific peak that I'm talking about is just above 100 hz.)
You didn't comment on it last time (other than confusing which dips were which), so now is your chance to explain why a huge peak in response turns into a big dip when multiple speakers are stacked in one tapped horn but not in another. (HINT - it's because not all tapped horns improve when stacked - where you end up depends entirely on where you started from.)
But the general trend of the low end becoming stronger is still clear in both examples, as well as the next link.
And here's the measurement of the TH115 that you posted.
http://www.diyaudio.com/forums/subw...oop-folded-horn-subwoofers-6.html#post3421146
All from the same thread.
http://www.diyaudio.com/forums/subw...oop-folded-horn-subwoofers-6.html#post3419569
Here's a "good" tapped horn for stacking purposes.
http://www.diyaudio.com/forums/subw...oop-folded-horn-subwoofers-7.html#post3421353
The reason these two tapped horns sim differently when stacked is not due to a problem with Hornresp, it's due to their individual differences in their impedance profiles. In the first example, the first really big, narrow TH peak becomes a hole in response when stacked whereas in the second example it doesn't. (In both examples, that first big, narrow TH specific peak that I'm talking about is just above 100 hz.)
You didn't comment on it last time (other than confusing which dips were which), so now is your chance to explain why a huge peak in response turns into a big dip when multiple speakers are stacked in one tapped horn but not in another. (HINT - it's because not all tapped horns improve when stacked - where you end up depends entirely on where you started from.)
But the general trend of the low end becoming stronger is still clear in both examples, as well as the next link.
And here's the measurement of the TH115 that you posted.
http://www.diyaudio.com/forums/subw...oop-folded-horn-subwoofers-6.html#post3421146
All from the same thread.
Last edited:
Hi Djim,
The Hornresp multiple speakers model considers just one of the speakers in the array, but with the input voltage and acoustical loading for that single speaker adjusted as necessary to reflect the multiple speaker configuration specified by the user.
Mutual coupling is taken into account to the extent that the resulting bass enhancement at low frequencies is seen in the Hornresp results.
The paragraph below is an extract from a description of mutual coupling of loudspeakers given on the following website:
proAV / data and information, lists, tables and links
" With two loudspeakers next to each other at a distance of more than the wavelength of the reproduced frequency the average measured volume on-axis and off-axis adds up for +3dB (double power). With two loudspeakers close enough to get mutual coupling the volume adds up for +6dB. "
The results produced by Hornresp are consistent with the above description (6dB increase in radiated power at low frequencies and 3dB increase at high frequencies) - see attached screenprint.
Kind regards,
David
The Hornresp multiple speakers model considers just one of the speakers in the array, but with the input voltage and acoustical loading for that single speaker adjusted as necessary to reflect the multiple speaker configuration specified by the user.
Mutual coupling is taken into account to the extent that the resulting bass enhancement at low frequencies is seen in the Hornresp results.
The paragraph below is an extract from a description of mutual coupling of loudspeakers given on the following website:
proAV / data and information, lists, tables and links
" With two loudspeakers next to each other at a distance of more than the wavelength of the reproduced frequency the average measured volume on-axis and off-axis adds up for +3dB (double power). With two loudspeakers close enough to get mutual coupling the volume adds up for +6dB. "
The results produced by Hornresp are consistent with the above description (6dB increase in radiated power at low frequencies and 3dB increase at high frequencies) - see attached screenprint.
Kind regards,
David
Attachments
originally posted by weltersys
The project I am working on now are unusual very narrow dispersion mid-high cabinets (160-16Khz), I'll post results on the full range forum when they are finished and tested.
The project I am working on now are unusual very narrow dispersion mid-high cabinets (160-16Khz), I'll post results on the full range forum when they are finished and tested.
Should have wrote I'll be posting on diyAudio Multi-Way, a speaker rolling off at 160 Hz needs something below it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.