"our bass player might not be so happy with this"
Measure the spectrum on a single note, I would wager that the fundamental is about only half the total voltage (about 6dB down).
Measure the spectrum on a single note, I would wager that the fundamental is about only half the total voltage (about 6dB down).
Hi Iand
I went back to look at the TH-115 measurement with the 4 ohm driver we switched to recently.
The text portion of the data sheet is correct, that was measured at 28.3Vrms at 10 meters, the same as the 8 Ohm version was.
This makes the graph about 3 dB higher than 1 Watt nominal as you mentioned.
I will mention that to the shop, the graph text should reflect that and should have been revised when the change was made.
Given the "in box" impedance curve, one box is not quite high enough to call it an 8 ohm system, 6 is atypical and potentially confusing so it's a 4 ohm box.
“Given that the TH-115 is measured using bandlimited pink noise then the 2000W stated power handling figure is perfectly believable and accurate, just so long as it's understood that it won't handle this power as a single note anywhere in the passband without hitting the driver Xmax limit at some points -- our bass player might not be so happy with this, but it's still probably better than most (all?) other boxes on the market.”
Most important is that we rate our products in a way that is consistent with out competition and if anything to error on the conservative side.
Nobody I know rates subwoofers based on a 0dB sine wave and “program” is normally defined as +3dB over the continuous (sine equivalent) rating anyway.
If one did sound cancellation, then a sine wave makes much more sense than the AES convention in pro sound of using b.l. noise to emulate music.
Also, how much load there is depends partly on how many boxes there are in that one location, all that effects the results too.
I have not extensively tested one box alone to see where the sine wave limits are at each frequency other than driving it with a slow swept sine with the TEF.
We are pretty much stuck with what the driver does in each case, like all speakers and it’s how they work side by side with others that matters most.
That being said and I tire of this point but what you measure (in a proper measurement) trumps the computer model, you have to alter / expand the model to fit what you measured (strong hint).
The gain in the system comes from reducing cone motion, an acoustic load, in the case of a DTS-20, the Tapped horn portion raises the sensitivity about 10dB over the raw driver, reduces the excursion for a given SPL by about 3.
Do you know what the compression ratio does here?
So far as horn response, I have not used that program yet, I probably should but I use Akabak mostly, with “compensations” to the model that makes what you measure pretty close to what it predicts.
Hopefully the explanations have not sounded like these are magic, they are not, they work by the physics involved. How well they work relative to what else is available is what I ‘d like you to hear and compare more so than predictions.
Comparisons with other real speakers are our most powerful sales tool.
Best,
Tom
I went back to look at the TH-115 measurement with the 4 ohm driver we switched to recently.
The text portion of the data sheet is correct, that was measured at 28.3Vrms at 10 meters, the same as the 8 Ohm version was.
This makes the graph about 3 dB higher than 1 Watt nominal as you mentioned.
I will mention that to the shop, the graph text should reflect that and should have been revised when the change was made.
Given the "in box" impedance curve, one box is not quite high enough to call it an 8 ohm system, 6 is atypical and potentially confusing so it's a 4 ohm box.
“Given that the TH-115 is measured using bandlimited pink noise then the 2000W stated power handling figure is perfectly believable and accurate, just so long as it's understood that it won't handle this power as a single note anywhere in the passband without hitting the driver Xmax limit at some points -- our bass player might not be so happy with this, but it's still probably better than most (all?) other boxes on the market.”
Most important is that we rate our products in a way that is consistent with out competition and if anything to error on the conservative side.
Nobody I know rates subwoofers based on a 0dB sine wave and “program” is normally defined as +3dB over the continuous (sine equivalent) rating anyway.
If one did sound cancellation, then a sine wave makes much more sense than the AES convention in pro sound of using b.l. noise to emulate music.
Also, how much load there is depends partly on how many boxes there are in that one location, all that effects the results too.
I have not extensively tested one box alone to see where the sine wave limits are at each frequency other than driving it with a slow swept sine with the TEF.
We are pretty much stuck with what the driver does in each case, like all speakers and it’s how they work side by side with others that matters most.
That being said and I tire of this point but what you measure (in a proper measurement) trumps the computer model, you have to alter / expand the model to fit what you measured (strong hint).
The gain in the system comes from reducing cone motion, an acoustic load, in the case of a DTS-20, the Tapped horn portion raises the sensitivity about 10dB over the raw driver, reduces the excursion for a given SPL by about 3.
Do you know what the compression ratio does here?
So far as horn response, I have not used that program yet, I probably should but I use Akabak mostly, with “compensations” to the model that makes what you measure pretty close to what it predicts.
Hopefully the explanations have not sounded like these are magic, they are not, they work by the physics involved. How well they work relative to what else is available is what I ‘d like you to hear and compare more so than predictions.
Comparisons with other real speakers are our most powerful sales tool.
Best,
Tom
Tom Danley said:
I think we're in complete agreement here Tom 🙂
Cheers
Ian
P.S. Yes I do know what effect changing the compression ratio has on cone travel and how much the tapped horn decreases it compared to many other enclosures -- however I'll bet good money that the Hornresp predictions aren't that far out, which still doesn't stop the TH-115 from being probably the best box around for its intended application...
FlipC said:
Yes. I would love to, but its layout is what I consider to be a true TH that's probably much closer to DSL's offerings than the simple end loaded, folded back on itself expanding TLs that me and others have posted for non-commercial use, so I'm not willing to maybe inadvertently give away anymore of DSL's design details for others to (illegally IMO) profit from than I might already have by example and/or explanation.
Like as always seems to be the case, the few ruin it for the many.
GM
GM said:
I'd guess that most people reading and contributing to this thread are interested in building their own THs for their own use, and wouldn't even consider buying DSL -- partly because of the cost (expensive, but not for the performance they deliver), but also because they *want* to build their own.
Anyone who then starts selling them for profit is definitely crossing the line both legally and morally, especially given the amount of information Tom has released about how the TH works on this forum.
But maybe from Tom's point of view enthusiastic home-builders could be a good thing, especially in Europe where most people have never heard of TH or DSL -- "Brilliant technology but I built this for myself, why don't you buy one from DSL?"
Ian
Tom Danley said:
Hi Tom,
I guess, the "compensations" are the losses mentioned in the first quote. The main source of losses i can see are the horn path foldings and maybe vibrating enclosure walls. One could use the AcouResistance, AcouMass and AcouCompliance elements to simulate this. As i doubt i have the skill to derive the information from a "build, measure and modify speakers" iteration process, maybe you could give some hints as how to do this right?
I can see quite well, that you probably dont want to tell all the secrets and thats really understandable. Its just that i like the idea of optimising a design.
MaVo said:
After too many cabinets to count...... I'ts a pain and the most obsessive project I have ever dealt with. Then again I am in the position of having no chance in hell of buying any for a year or two so that makes it worse. Lot's of wood scraps!
There are lots of tricky things that don't show up in sims, and bends have a funny way of driving you nuts just going one way.
In a tapped horn they go BOTH ways.
iand is quite right.
I do sports retail for a living. Make about 28K a year. So I am not classified as a DSL customer just by that. Sorry. Unless you want to sell me a SPUD on the cheap Mr Danley? But then I would have to have the surround system hardware to go with it that I don't have either. ): I am going to be working on that this year.
Also
The UK guys definitely know about TH's. The TH seems to fit in with their thinking on sound quality to boot. I assume its more about ROI.
Least until DSL opens a fab plant in the UK (Poland is cheap)
GM
I don't think companies in the speaker business would troll forums looking for something to clone. They would just buy it, dismantle, measure, change something cosmetically and call it new.
I do sports retail for a living. Make about 28K a year. So I am not classified as a DSL customer just by that. Sorry. Unless you want to sell me a SPUD on the cheap Mr Danley? But then I would have to have the surround system hardware to go with it that I don't have either. ): I am going to be working on that this year.
Also
The UK guys definitely know about TH's. The TH seems to fit in with their thinking on sound quality to boot. I assume its more about ROI.
Least until DSL opens a fab plant in the UK (Poland is cheap)
GM
I don't think companies in the speaker business would troll forums looking for something to clone. They would just buy it, dismantle, measure, change something cosmetically and call it new.
I cant figure this one out...
Lets say you build a cabinet, test it, then remeasure it physically, retransfer to hornresp, tweak it to match your test results.
You then tweak a copy of said cabinet by simply extending L12 to enhance the upper end at 100hz, manually put s2 right back where it was and calculate the result. More 100hz.. O.K.
We're talking a an extra 30 cm here.
You now go back to the physical box, extend L12 reseal the whole mess and retest.....
Now you wind up with a db or two loss at 40hz and absolutely no extra gain at 100!!!!!!!! What the heck? I've noticed this on a couple of boxes so this time I was carefull to be precise. The only difference between the two boxes is the L12 extension. S2 through the mouth all stayed exactly the same physical dimensions and flare rate. Any ideas?
Lets say you build a cabinet, test it, then remeasure it physically, retransfer to hornresp, tweak it to match your test results.
You then tweak a copy of said cabinet by simply extending L12 to enhance the upper end at 100hz, manually put s2 right back where it was and calculate the result. More 100hz.. O.K.
We're talking a an extra 30 cm here.
You now go back to the physical box, extend L12 reseal the whole mess and retest.....
Now you wind up with a db or two loss at 40hz and absolutely no extra gain at 100!!!!!!!! What the heck? I've noticed this on a couple of boxes so this time I was carefull to be precise. The only difference between the two boxes is the L12 extension. S2 through the mouth all stayed exactly the same physical dimensions and flare rate. Any ideas?
Perhaps Flat is not the answer...
Grasping at a straw here...
It occured to me that many seem to try to get flat response out of a tapped. Perhaps for high spl applications that is not the best way to approach the design.
Since many top end cabinets require processing to achieve their advertised response, perhaps it is the sub itself that should provide the initial processing rather than a utopian frequency response plot. More marketing hype......ugggh.
In other words, intentionally designing a "Bad" speaker.
The 15's I'm working with run out of xmax where you need it the most (for live) at 40hz. What if one were to design a cabinet that had a very large hump that corresponded to the point where the driver runs out of xmax. My thinking is you could then cut that frequency band at your eq or dsp to reflatten the box. You would then be limiting the amount of these xmax sensitive frequencies BEFORE they hit the cabinet, and the cabinet in turn boosts them acoustically. You should be able to get much higher levels at your xmax frequencies than if you designed a "perfect" box.
You can laugh now, I promise not to 
Grasping at a straw here...
It occured to me that many seem to try to get flat response out of a tapped. Perhaps for high spl applications that is not the best way to approach the design.
Since many top end cabinets require processing to achieve their advertised response, perhaps it is the sub itself that should provide the initial processing rather than a utopian frequency response plot. More marketing hype......ugggh.
In other words, intentionally designing a "Bad" speaker.
The 15's I'm working with run out of xmax where you need it the most (for live) at 40hz. What if one were to design a cabinet that had a very large hump that corresponded to the point where the driver runs out of xmax. My thinking is you could then cut that frequency band at your eq or dsp to reflatten the box. You would then be limiting the amount of these xmax sensitive frequencies BEFORE they hit the cabinet, and the cabinet in turn boosts them acoustically. You should be able to get much higher levels at your xmax frequencies than if you designed a "perfect" box.
You can laugh now, I promise not to That´s not dumb thinkinh at all!! If you look at what a sub does in real world in a room where reflections and room modes are present, a flat response at low frequenvies is hard to achieve anyway. Just take a look at the pro sound shootouts from the Pi Speaker guys. MOst oft the subs present have strong accentuations in frequency response. TDs Basshorns have it, too - to some extend, it´s just not that obvious in the measurement graphs because of the scale chosen.
On thing to think about that comes up some times when trying to flaten the respone: The series inductor. Some recommend high inductances (5mH)... Remember that in pro sound very high voltages are common and that a inductor gets very big at high values. The ones with a ferrit (or other) core are smaller but reach their saturation points very quick. I don´t know about the prices in the states, but a 5mH inductor which can handle 2kW without being driven into saturation ist veeeery expensive here in germany (actually, I´d have to check if there even is one available for this kind of power with 5mH).
There was a recent test in a german DYI magazine where this matter was investigatet quite thoroughly and many common available inductors weren´t even able to cope with a few hundred watts without reaching more than 10% THD... And those tests werde done with smaller than 5mH types... (Hobby Hifi was the magazine, in case someone is interested)
On thing to think about that comes up some times when trying to flaten the respone: The series inductor. Some recommend high inductances (5mH)... Remember that in pro sound very high voltages are common and that a inductor gets very big at high values. The ones with a ferrit (or other) core are smaller but reach their saturation points very quick. I don´t know about the prices in the states, but a 5mH inductor which can handle 2kW without being driven into saturation ist veeeery expensive here in germany (actually, I´d have to check if there even is one available for this kind of power with 5mH).
There was a recent test in a german DYI magazine where this matter was investigatet quite thoroughly and many common available inductors weren´t even able to cope with a few hundred watts without reaching more than 10% THD... And those tests werde done with smaller than 5mH types... (Hobby Hifi was the magazine, in case someone is interested)
Sabbelbacke said:
Those inductors do exist, and they are big and heavy and expensive -- for example, if you look at lautsprechershop.de you can find the Intertechnik transformer core coils, the FE130 4.7mH has 0.08ohms resistance (1% power loss or 0.1dB) and is rated at 2000W/8ohms -- wound using 3mm wire, size is 130x98x116mm and cost is 116euros :-(
Ian
Exactly 🙂 very expensive.
Sadly, most of the IT types didn´t even come close to the stated specs in the real world measurements 🙁 The Monacor Trafocoils also failed most of the times concerning very high power. Mundorf worked quite well, Jansen, too. But one realy hast to look close and do some measurements (not easy) on exactly the coil in focus, since the specs are not reliable after all 🙁 (that´s one main reason why the article-series was written after all..) Trying out three inductors really gets expensive🙂
Sadly, most of the IT types didn´t even come close to the stated specs in the real world measurements 🙁 The Monacor Trafocoils also failed most of the times concerning very high power. Mundorf worked quite well, Jansen, too. But one realy hast to look close and do some measurements (not easy) on exactly the coil in focus, since the specs are not reliable after all 🙁 (that´s one main reason why the article-series was written after all..) Trying out three inductors really gets expensive🙂
Sabbelbacke said:
Do you have a copy of the article, preferably in a form that can be translated into English?
(this is all the same reason high-power passive LF crossovers should be avoided whenever possible)
Ian
Why not build your own inductor? Winding up some wire doesnt sound like a difficult task. Besides, lautsprechershop.de has sometimes rather high prices, compared to other shops.
Indeed. Back in the so-called good ol' days me and others with severely limited hobby budgets made our own inductors by the simple expedient of buying from a local industrial electrical supply house a sufficiently large roll of whatever gauge machine wound magnet winding wire that was required and de-wound them using an inductance meter to find the desired value. The scraps made for great wiring and when long enough, DIY antennas, interconnect and speaker wires.
GM
GM
Hi
Use a inductor calculator like bellow:
[http://www.colomar.com/Shavano/inductor_info.html
A 13 wire should be enough. Remember you will play music.
Even if you make the coil handmade, it will not be perfectly aligned, but with a good multimeter, you are able to measure inductance accurately. So if it's a bit bellow add some turns or if it's above minus some turns you will get it precisely.
See how I made mine for two tree way passive crossover for two Davids: the bigger one's are 6,6mH, and the coil is a hollow iron tube, which give me a gain around 36% less wire compared with air coil. My air other coils, you can see in the picture, are a 40mm plastic nuclei (much easier to fix it to me).
http://img72.imageshack.us/my.php?image=s3010393ot3.jpg
See also a coil winder I developed. I can adjust the coil width, and also make it using a little motor for thinner wires e manually for thicker wire which the power of the motor cannot do the work.
http://img383.imageshack.us/my.php?image=img1820ek7.jpg
Some other coils made with the coil winder:
http://img91.imageshack.us/my.php?image=s3011611cj2.jpg
Regards,
Use a inductor calculator like bellow:
[http://www.colomar.com/Shavano/inductor_info.html
A 13 wire should be enough. Remember you will play music.
Even if you make the coil handmade, it will not be perfectly aligned, but with a good multimeter, you are able to measure inductance accurately. So if it's a bit bellow add some turns or if it's above minus some turns you will get it precisely.
See how I made mine for two tree way passive crossover for two Davids: the bigger one's are 6,6mH, and the coil is a hollow iron tube, which give me a gain around 36% less wire compared with air coil. My air other coils, you can see in the picture, are a 40mm plastic nuclei (much easier to fix it to me).
http://img72.imageshack.us/my.php?image=s3010393ot3.jpg
See also a coil winder I developed. I can adjust the coil width, and also make it using a little motor for thinner wires e manually for thicker wire which the power of the motor cannot do the work.
http://img383.imageshack.us/my.php?image=img1820ek7.jpg
Some other coils made with the coil winder:
http://img91.imageshack.us/my.php?image=s3011611cj2.jpg
Regards,
One needs an M6 laminated core for woofer inductors, and large wire.
Dissipation in the inductor is I^2*R, so they can really heat up with high power use.
I tried ferrite drum-core inductors and they burned up with only a Hafler DH500 for drive (255W/8R). The wire got so hot the insulation burned off, and the sintering came apart on the core.
I switched to laminated bar inductors with the biggest wire I could find, and then things were OK.
Did I mention the mounting screws for the drum-core inductor got so hot that they charred completely through the 3/4" plywood the network was constructed on? Smore used to roll out of the port on the speaker!
Dissipation in the inductor is I^2*R, so they can really heat up with high power use.
I tried ferrite drum-core inductors and they burned up with only a Hafler DH500 for drive (255W/8R). The wire got so hot the insulation burned off, and the sintering came apart on the core.
I switched to laminated bar inductors with the biggest wire I could find, and then things were OK.
Did I mention the mounting screws for the drum-core inductor got so hot that they charred completely through the 3/4" plywood the network was constructed on? Smore used to roll out of the port on the speaker!