I have revived this thread because it contains the details of my project. I have completed construction (it took almost 2 full months more than I had anticipated) and have been listening for a couple of weeks now to make sure it is fine the way it is, and complete, at least for now. It sounds very nice, but I made a couple of changes during construction.
GM, are you still out there? To complete my research I would very much like to see my current alignment charted with MJK's spreadsheets, if it is not too much to ask. It would be a very interesting comparison to my audible impressions and would allow me to complete my report.
I am using the Dayton Quatro 15, same sub you modelled for me in this very thread. If you would like me to post the specs again, just ask, it's not a problem.
The change that I made was a taper. It is a 2:1 ratio (closed end 2x larger than the open end). The closed end is 170 inches and the open end is 85 inches. My ears tell me it is tuned at 16 hz, which is exactly what MJK's alignment tables told me it would be before I built it. The line is still 171 inches long and I am using no stuffing. Driver offset is 25%. I am using the transflex/Danley loading style but would very much like to see how this would model in a traditional loading style that the spreadsheets allow so I can compare.
Thank you in advance for your time, whether you have the time to do it or not. Just let me know if you don't and I'll try to figure something else out so I can complete my report.
GM, are you still out there? To complete my research I would very much like to see my current alignment charted with MJK's spreadsheets, if it is not too much to ask. It would be a very interesting comparison to my audible impressions and would allow me to complete my report.
I am using the Dayton Quatro 15, same sub you modelled for me in this very thread. If you would like me to post the specs again, just ask, it's not a problem.
The change that I made was a taper. It is a 2:1 ratio (closed end 2x larger than the open end). The closed end is 170 inches and the open end is 85 inches. My ears tell me it is tuned at 16 hz, which is exactly what MJK's alignment tables told me it would be before I built it. The line is still 171 inches long and I am using no stuffing. Driver offset is 25%. I am using the transflex/Danley loading style but would very much like to see how this would model in a traditional loading style that the spreadsheets allow so I can compare.
Thank you in advance for your time, whether you have the time to do it or not. Just let me know if you don't and I'll try to figure something else out so I can complete my report.
I don't have much time now and due to internet connectivity issues of late I can't check in often, but I will give you a couple of impressions quickly, and more later. I haven't been around too much lately and don't know if GM is checking in either. If anyone else wants to do this spreadsheet model, I would appreciate it.
First off, I have no experience with subs tuned this low, so this is all new for me. Also I have no SPL meter.
Sensitivity is decent, I'm guessing 90 db/1 watt or maybe a bit less. The bottom end is strong, but my cheap amp and source show power rolls off below 25 hz. The LED's on the amp indicate 25 hz is getting ten times more power than 15 hz. This is something I will have to fix but it still sounds good for movies if I cross over low enough.
Bandwidth sucks. The useable range is about 16 - 40 hz before the peaks and valleys start. Looking at the TOP FR charts, my peaks seem to be bigger and harder to tame. This could easily be fixed with stuffing, but since I don't crave the extra bandwidth I would rather cut the output at 40 hz than muddy up the low end with stuffing. Actually, I would much rather cut as low as possible, using mains down to 25 or 30 hz and the sub below that. I would really like a steeper cut than my active XO can provide because it can't COMPLETELY kill the peaks, but it's not a big issue right now.
I have tried crossover points as high as 60 hz, and while this is ok (but certainly not perfect) for music, it sucks the life out of the LFE's in movies. Since this was made to be an HT sub, I have been crossing over as low as possible.
I am using my integrated computer sound card as a source and one channel of a cheap 20 year old Technics amp (110 watt peak), so at this point I have no idea what it will be capable of.
So, to recap, I like it alot, but bandwidth is limited until I break down and add some stuffing. I will know more about what it is supposed to be like when I get the "as built" spreadsheet model and have a chance to compare. What about you, Qi, do you have the MJK spreadsheets?
If I am using the MJK alignment tables right, I figure my CSA is about 2x the recommended area, which should be giving me a hefty boost around tuning to counteract (at least a bit) the amp and source LF rolloff, hearing rolloff, etc.
First off, I have no experience with subs tuned this low, so this is all new for me. Also I have no SPL meter.
Sensitivity is decent, I'm guessing 90 db/1 watt or maybe a bit less. The bottom end is strong, but my cheap amp and source show power rolls off below 25 hz. The LED's on the amp indicate 25 hz is getting ten times more power than 15 hz. This is something I will have to fix but it still sounds good for movies if I cross over low enough.
Bandwidth sucks. The useable range is about 16 - 40 hz before the peaks and valleys start. Looking at the TOP FR charts, my peaks seem to be bigger and harder to tame. This could easily be fixed with stuffing, but since I don't crave the extra bandwidth I would rather cut the output at 40 hz than muddy up the low end with stuffing. Actually, I would much rather cut as low as possible, using mains down to 25 or 30 hz and the sub below that. I would really like a steeper cut than my active XO can provide because it can't COMPLETELY kill the peaks, but it's not a big issue right now.
I have tried crossover points as high as 60 hz, and while this is ok (but certainly not perfect) for music, it sucks the life out of the LFE's in movies. Since this was made to be an HT sub, I have been crossing over as low as possible.
I am using my integrated computer sound card as a source and one channel of a cheap 20 year old Technics amp (110 watt peak), so at this point I have no idea what it will be capable of.
So, to recap, I like it alot, but bandwidth is limited until I break down and add some stuffing. I will know more about what it is supposed to be like when I get the "as built" spreadsheet model and have a chance to compare. What about you, Qi, do you have the MJK spreadsheets?
If I am using the MJK alignment tables right, I figure my CSA is about 2x the recommended area, which should be giving me a hefty boost around tuning to counteract (at least a bit) the amp and source LF rolloff, hearing rolloff, etc.
just a guy said:
Greets!
Currently not nearly as much as I was, 'life' is increasingly consuming my quality time. I still have the last WS I did (http://www.diyaudio.com/forums/attachment.php?s=&postid=926782&stamp=1148791792), so changing it per the latest specs:
GM
Attachments
Thank you so much, that last chart explains a lot. It shows exactly what I am hearing, right down to the EXACT location of the peaks and nulls. It shows very clearly that just a bit of stuffing would even out the response to almost perfectly flat up to around 100 hz. It shows why my crossover point at 40 hz clears up the first problem peak, and why the peak just after 100 hz is so hard to tame, no matter where the crossover is set.
There is a surprise for me too. I was completely expecting to see a nice bump around the tuning frequency, which is clearly not there in the graph. The only way I can reconcile the graph and what I hear is this explanation. I set up the sub position and listening position to boost 16 hz as much as possible. There are other spots in the room that are much louder higher in the audio spectrum but my whole setup is based on getting the most out of the bottom end.
So now the question is this. Is the couple of extra hz extention in the bottom end worth the couple of db loss? (By changing the design to tapered.) I really don't know. But it did seem to tame the upper harmonics nicely (on paper at least).
And the other question would be - Did the transflex/Danley loading method help? Without an SPL meter or previous tl experience I can't answer that one either. Sorry.
But I can say this much for sure. I discovered (to my dismay) that I need a rumble filter set at maybe 15 hz, or definitely no lower than 10. I bottomed out my excursion a couple of times so far during the really low frequency effects during movies. This prompted me to test how much power it could take at 16 hz with a sine wave. The good news is that the house sounds like it is going to fall apart by the time I reach my excursion limits. The bad part is that it seems that I am running out of excursion with under 100 watts. That's another thing I wasn't expecting, but the silver lining on that is that I don't think I will ever be watching movies at that volume.
I have a couple of tricks up my sleeve yet (more like fumbling around in the dark) but no immediate plans, as I am relatively happy with it as is. I have considered adding some weight to the cone to get fs more in line with the pipe tuning. I have thought about mass loading the open end and tuning with ports to get down to 10 hz and using the sub from only 10 - 30 or so. I have also wondered what would happen if I closed the open end and opened up the closed end to reverse the taper, making it almost exactly like Danley's PB12 (higher tuning to match driver fs but a whole lot louder). Or possibly any of the above, but with a smaller driver with lower q and more excursion, to really bring the bottom end up a few notches.
But like I said, I am relatively happy with it and won't be making any changes without a lot of thought and possibly getting the MJK spreadsheets for myself. The only reason I haven't done so already is that I am very unsure that I will actually be able to use them. I tried to figure out Hornresp once and got so confused I gave up on it.
Thanks again to everyone who has helped me so far, and to those who have commented, questioned and otherwise furthered my project.
There is a surprise for me too. I was completely expecting to see a nice bump around the tuning frequency, which is clearly not there in the graph. The only way I can reconcile the graph and what I hear is this explanation. I set up the sub position and listening position to boost 16 hz as much as possible. There are other spots in the room that are much louder higher in the audio spectrum but my whole setup is based on getting the most out of the bottom end.
So now the question is this. Is the couple of extra hz extention in the bottom end worth the couple of db loss? (By changing the design to tapered.) I really don't know. But it did seem to tame the upper harmonics nicely (on paper at least).
And the other question would be - Did the transflex/Danley loading method help? Without an SPL meter or previous tl experience I can't answer that one either. Sorry.
But I can say this much for sure. I discovered (to my dismay) that I need a rumble filter set at maybe 15 hz, or definitely no lower than 10. I bottomed out my excursion a couple of times so far during the really low frequency effects during movies. This prompted me to test how much power it could take at 16 hz with a sine wave. The good news is that the house sounds like it is going to fall apart by the time I reach my excursion limits. The bad part is that it seems that I am running out of excursion with under 100 watts. That's another thing I wasn't expecting, but the silver lining on that is that I don't think I will ever be watching movies at that volume.
I have a couple of tricks up my sleeve yet (more like fumbling around in the dark) but no immediate plans, as I am relatively happy with it as is. I have considered adding some weight to the cone to get fs more in line with the pipe tuning. I have thought about mass loading the open end and tuning with ports to get down to 10 hz and using the sub from only 10 - 30 or so. I have also wondered what would happen if I closed the open end and opened up the closed end to reverse the taper, making it almost exactly like Danley's PB12 (higher tuning to match driver fs but a whole lot louder). Or possibly any of the above, but with a smaller driver with lower q and more excursion, to really bring the bottom end up a few notches.
But like I said, I am relatively happy with it and won't be making any changes without a lot of thought and possibly getting the MJK spreadsheets for myself. The only reason I haven't done so already is that I am very unsure that I will actually be able to use them. I tried to figure out Hornresp once and got so confused I gave up on it.
Thanks again to everyone who has helped me so far, and to those who have commented, questioned and otherwise furthered my project.
EDIT TO THE ABOVE POST (it won't let me edit this one for some reason).
"It shows very clearly that just a bit of stuffing would even out the response to almost perfectly flat up to around 100 hz." - WITH ROOM GAIN AND CORNER LOADING TAKEN INTO CONSIDERATION.
There's not much truth to that statement the way it was UNLESS room gain and corner loading are added to the sub's response.
"It shows very clearly that just a bit of stuffing would even out the response to almost perfectly flat up to around 100 hz." - WITH ROOM GAIN AND CORNER LOADING TAKEN INTO CONSIDERATION.
There's not much truth to that statement the way it was UNLESS room gain and corner loading are added to the sub's response.
I've been wondering for days now if my excursion problem is to be expected or if something has gone wrong somewhere. After all, I am blasting a 16 hz sinewave with only one driver with 10 mil xmax (and exceeding xmax with under 100 watts).
So can someone answer this one for me? Have I messed something up or is low power handling to be expected at these frequencies?
So can someone answer this one for me? Have I messed something up or is low power handling to be expected at these frequencies?
OTH
On the other hand you are pushing 110dB at 16Hz with a single 15" driver! Possibly your movie was giving you significant sub 16Hz energy. The modeling that I have done indicates that driver excursion increases dramatically below the line's -3dB point. A possible approach would be to lengthen the line, increase the taper (or a little of both) to trade off some efficiency in the upper part of the passband for more damping at the extreme low end allowing you to use more power below 16Hz. Yust a thought. Or Just build another one.
mike
On the other hand you are pushing 110dB at 16Hz with a single 15" driver! Possibly your movie was giving you significant sub 16Hz energy. The modeling that I have done indicates that driver excursion increases dramatically below the line's -3dB point. A possible approach would be to lengthen the line, increase the taper (or a little of both) to trade off some efficiency in the upper part of the passband for more damping at the extreme low end allowing you to use more power below 16Hz. Yust a thought. Or Just build another one.
mike
16 hz is inside my passband of usable frequencies and as such, I kind of expected it to handle at least a bit closer to the driver's rated power handling. It is plenty loud enough in the medium/small room that it is in - but - the TOP can handle gobs more power and the tuning frequency is not that much higher than mine. I just wonder if this is normal operating procedure for a tl of this size and tuning. Maybe it is pointless to expect any more with only 10 mil xmax.
My options do not include changing the pipe length or taper, those are fixed. I can, however, mass load to get lower. But that will afford me even less sensitivity and even larger xmax problems down lower I think. The other option is to reverse the taper by opening the closed end and closing the open end. I believe this option would increase sensitivity to about 95 db and reduce excursion, as the tuning would change to around 22 or 23 hz. Unfortunately, this would end it's short career as an HT sub and turn it into a music sub that is much louder than I need it to be.
So I guess what I am asking is a two part question.
1. If you look at the last graph GM posted above and keep in mind that it is excursion limited at under 100 watts at 16 hz, is it decent, mediocre, or just plain bad performance for a tl? Comparing to the TOP mine doesn't seem too special, but then the TOP is not your typical tl either. I've seen lots of tl FR charts, but have no idea what to expect as far as excursion goes.
2. Any suggestions on where I should or could go from here? Or maybe leave it alone? Mike, those are good suggestions but neither is possible. The line length and geometry can't be changed. The driver offset is fixed at 25% but could be traditionally loaded outside the box on either side, depending if I keep the decreasing taper or reverse it. Mass loading is a complete unknown for me, other than I know it will lower tuning. I tried searching on the subject and gets loads of info but none of it has anything to do with tl's.
Just looking for ideas or opinions, thanks guys.
My options do not include changing the pipe length or taper, those are fixed. I can, however, mass load to get lower. But that will afford me even less sensitivity and even larger xmax problems down lower I think. The other option is to reverse the taper by opening the closed end and closing the open end. I believe this option would increase sensitivity to about 95 db and reduce excursion, as the tuning would change to around 22 or 23 hz. Unfortunately, this would end it's short career as an HT sub and turn it into a music sub that is much louder than I need it to be.
So I guess what I am asking is a two part question.
1. If you look at the last graph GM posted above and keep in mind that it is excursion limited at under 100 watts at 16 hz, is it decent, mediocre, or just plain bad performance for a tl? Comparing to the TOP mine doesn't seem too special, but then the TOP is not your typical tl either. I've seen lots of tl FR charts, but have no idea what to expect as far as excursion goes.
2. Any suggestions on where I should or could go from here? Or maybe leave it alone? Mike, those are good suggestions but neither is possible. The line length and geometry can't be changed. The driver offset is fixed at 25% but could be traditionally loaded outside the box on either side, depending if I keep the decreasing taper or reverse it. Mass loading is a complete unknown for me, other than I know it will lower tuning. I tried searching on the subject and gets loads of info but none of it has anything to do with tl's.
Just looking for ideas or opinions, thanks guys.
I would think that going to an expanding line would make the "problem" worse. Raising the tuning frequency would increase the frequency at which the driver excursion skyrockets would it not? Above the tuning frequency efficiency would increase but by raising fc the excursion at 16Hz would (I think) actually increase for a given power level.
But I guess the long and short of it should be not how much power you can absorb at the lowest frequencies of interest but rather what SPL you can sustain at those frequencies. If you are cracking the plaster at 50W do you really need to handle 100? If you can't lower the tuning or add another sub you may in fact need to go ahead and high pass it at or around your fc. Also, how confident are you of the actual power output that you are using (are the output meters acurate at 16Hz)?
For me 110dB at 16Hz would be a fantastic music sub and a huge improvement over what I am currently using.
I am in the process of building a TL sub within the constraints of a 15" organ speaker that I have on hand and the scrap wood available. The current design sims a max power at 16Hz of 15W (yes 15W) and 95dB assuming an xmax of 6mm (complete guess). Of course xmax and xmech are not the same thing and will affect the total output.
However because my fc is about 25Hz that same 15W should give about 105dB at fc rising to over 110dB at 60Hz (not accounting for room gain and crossover). Now if I limit frequencies below 22Hz I can put in up to 85W before reaching 6mm excursion and produce SPLs of about 115dB.
If I lowered the tuning (and I will probably build a bigger line later when I can get the materials) I could get higher output at lower frequencies according to the sims but I would sacrifice ultimate SPL at higher frequencies and would require higher power input to acheive the same output levels.
For example I just ran the numbers with the same taper but lengthening the line enough to get fc of about 16Hz (adjusting So for best response). Since my driver Fs is only 45Hz the reponse is not nearly as flat (looks kind of like EBS) but I can take up to 50W at 16Hz now and more importantly produce 108dB at 16Hz. However I can no longer reach those 115dB levels overall because I now get 6mm displacement at 28Hz and 108dB also. The rising response gets me to 119dB just before the bottom falls out but in practice the crossover and/or stuffing will lower that down to the 110dB range. So I have given up 6dB or so overall maximum output to increase the maximum output of that 32' pedal stop by about 12dB.
Judging from GMs response plot it looks like you are tuned almost right at 16Hz. I note that the excursion plots tend to have a deep null just north of the tuning frequency. I bet if you fed it 18Hz instead of 16Hz you would notice a rather large increase in power handling. When you are operating right near fc it seems that it takes only a small movement in frequency to make a large difference in output capability and power handling.
I think that you have a very nice sub there and just need to roll off the subsonics a bit to protect your driver. Just for giggles I sim'ed a dayton quatro15 in a 6.7m line with taper factor of 2 which tunes to about 12Hz. This gives xmax limited maximum output at 10Hz (and 16Hz) of 108dB at 150W. That should be pretty decent as a HT sub I think.
mike
But I guess the long and short of it should be not how much power you can absorb at the lowest frequencies of interest but rather what SPL you can sustain at those frequencies. If you are cracking the plaster at 50W do you really need to handle 100? If you can't lower the tuning or add another sub you may in fact need to go ahead and high pass it at or around your fc. Also, how confident are you of the actual power output that you are using (are the output meters acurate at 16Hz)?
For me 110dB at 16Hz would be a fantastic music sub and a huge improvement over what I am currently using.
I am in the process of building a TL sub within the constraints of a 15" organ speaker that I have on hand and the scrap wood available. The current design sims a max power at 16Hz of 15W (yes 15W) and 95dB assuming an xmax of 6mm (complete guess). Of course xmax and xmech are not the same thing and will affect the total output.
However because my fc is about 25Hz that same 15W should give about 105dB at fc rising to over 110dB at 60Hz (not accounting for room gain and crossover). Now if I limit frequencies below 22Hz I can put in up to 85W before reaching 6mm excursion and produce SPLs of about 115dB.
If I lowered the tuning (and I will probably build a bigger line later when I can get the materials) I could get higher output at lower frequencies according to the sims but I would sacrifice ultimate SPL at higher frequencies and would require higher power input to acheive the same output levels.
For example I just ran the numbers with the same taper but lengthening the line enough to get fc of about 16Hz (adjusting So for best response). Since my driver Fs is only 45Hz the reponse is not nearly as flat (looks kind of like EBS) but I can take up to 50W at 16Hz now and more importantly produce 108dB at 16Hz. However I can no longer reach those 115dB levels overall because I now get 6mm displacement at 28Hz and 108dB also. The rising response gets me to 119dB just before the bottom falls out but in practice the crossover and/or stuffing will lower that down to the 110dB range. So I have given up 6dB or so overall maximum output to increase the maximum output of that 32' pedal stop by about 12dB.
Judging from GMs response plot it looks like you are tuned almost right at 16Hz. I note that the excursion plots tend to have a deep null just north of the tuning frequency. I bet if you fed it 18Hz instead of 16Hz you would notice a rather large increase in power handling. When you are operating right near fc it seems that it takes only a small movement in frequency to make a large difference in output capability and power handling.
I think that you have a very nice sub there and just need to roll off the subsonics a bit to protect your driver. Just for giggles I sim'ed a dayton quatro15 in a 6.7m line with taper factor of 2 which tunes to about 12Hz. This gives xmax limited maximum output at 10Hz (and 16Hz) of 108dB at 150W. That should be pretty decent as a HT sub I think.
mike
just a guy said:
Greets!
You're welcome!
To get the most out of the bottom end requires the vent to load a corner's apex, though down really low, most room's construction become increasingly house shaking transparent with decreasing frequency instead of increasing gain, hence the low power handling at Fp and below.
FWIW, if I have a clue, the driver is at the wrong point along the pipe for max benefit from TOP loading and reverse tapering (mass loading) such an acoustically 'skinny' pipe ~defeats most/all of the loading gain due to port compression. It's a given that port area needs to be at least = Sd.
The driver will be too close to the terminus if you swap ends. A lower Q driver will have more peaking at Fp to offset the room's apparent lack of gain, but the octave above it will suffer even more from lack of pipe gain than it already does.
I have issues with Hornresp also, though it's great for folks who know little/nothing about horns since it has a pretty decent designer. MJK's WSs OTOH are as easy as falling off a log and his Classic TL tables are good enough that many folks will be pleased with their performance without further tweaking.
GM
Bassboy,
You can easily change your taper. Just block the terminus and open up the larger side, then flip the polarity on your connection to the amp. Then you have a tapped horn, which I don't think you have with your current decreasing CSA. That may also help your excursion problem to some extent, and at least you'll be alot closer to the Tower Of Power (TOP) design of Danley.
You can easily change your taper. Just block the terminus and open up the larger side, then flip the polarity on your connection to the amp. Then you have a tapped horn, which I don't think you have with your current decreasing CSA. That may also help your excursion problem to some extent, and at least you'll be alot closer to the Tower Of Power (TOP) design of Danley.
Thanks for the replies, guys. I know my posts here are a bit long winded, and I apologize, but I would like to respond to each of you.
MASHAFFER - "are the output meters acurate at 16Hz?" Good eye for detail, man. This requires a tricky answer. So - they are probably close enough to show what an 8 ohm load is doing. But the amp is not 4 ohm stable to match the Quatro. If you send the amp a signal but do NOT select a speaker output, or do not hook up any speakers then the led's still light up indicating power output. This leads me to believe that the led's are not coupled to the power output stage, but rather the input or volume knob stage of it's amplification process. Further speculation leads me to suspect that the lights may be only representing what would happen if an 8 ohm load were connected, if only a 4 ohm load exists, it is possible that the lights are only reflecting half of the actual power being output. But if this is the case, I would expect it to be louder than it is in reference to the mains, which are very close to the same rated sensitivity, and can be checked A/B style at frequencies where their bandwidth overlaps. So the short answer is that I have no idea how much importance to connect to the led's, other than the fact that I can light up 5 led's at 25 hz with no problem, but only 4 at 16 hz before the bad sounds start.
I realize that tuning higher would make the problem at 16 hz worse, that is why I said that it would be a music sub - not HT. It would need a rumble filter at 20 hz if I reverse the taper, I accept that. But in it's operating range, I suspect that it would handle significantly more power with more sensitivity.
I appreciate your info about elongating the pipe to 20 feet, and although it is not possible with my current box, it is probably something I will do on another project with the goal being response to 10 hz. Since no one even mentioned the possibility of mass loading or tuning with ports to get lower, I assume the option is not a good one. Also, although it may be out of context, I have seen GM state that small pipes give small performance, and at 14 feet, this would be a small pipe for frequencies below what I already have.
GM - With too little information and possibly too much imagination I suspected that the "tapped" concept would work at 25% for any line length and geometry. I have come to believe that the exact location is much trickier to pin down with other lengths and geometries, so I agree that my driver offset is probably not suited to my geometry. This is why I assumed my next logical step would be reversing the taper, as it makes for a design VERY similar to the PB12, the TOP's concept box. Proven design = good results, right? But you said "The driver will be too close to the terminus if you swap ends." I don't understand, as reversing the taper would make it almost exactly the same as the PB12, with the same tap of 25/75%, same length and presumably similar performance. Since the driver is inside the box, it is equidistant from both ends, no matter which end is open or closed.
Finally, you said "It's a given that port area needs to be at least = Sd." Are you referring to this application, or all tl's in general? Is it the high or low end of response that suffers if CSA is not at least = SD?
And last but not least, John - I knew your opinion already, and I wasn't ignoring it, but please understand that the extension down to 16 hz will be missed when I go ahead with this. Also, when looking at the TOP's FR graph (or PB12, whichever is actually REALLY represented in that graph) I don't consider anything above 60 hz to be usable output. There are wild fluctuations in SPL, amounting to 6 db swings up and down and I really don't think I'd cross that sub over much higher than I do with mine. So the only real advantage to reversing the taper is extra power handling and sensitivity at the (LARGE) expense of extended bandwidth, and in particular LF bandwidth.
I fully agree (as much as is possible without a meter) that I am not getting the benefits of a tapped horn, as my response sounds (almost) exactly like what I see in GM's last graph. The only thing I can not qualify (without a meter) is the severity of the peaks and nulls. But I would guess that I have an ordinary tl performance with the added feature of a hidden and protected driver, and that's about it.
But John, why would I need to switch polarity? When I reverse the taper, I will not be flipping the driver over, it will stay exactly the way it is. Is that what you are referring to?
I could talk about this indefinitely, but I suppose this is a bit long for a forum post so that's it for now. Thanks guys, for the wisdom, opinions and suggestions.
MASHAFFER - "are the output meters acurate at 16Hz?" Good eye for detail, man. This requires a tricky answer. So - they are probably close enough to show what an 8 ohm load is doing. But the amp is not 4 ohm stable to match the Quatro. If you send the amp a signal but do NOT select a speaker output, or do not hook up any speakers then the led's still light up indicating power output. This leads me to believe that the led's are not coupled to the power output stage, but rather the input or volume knob stage of it's amplification process. Further speculation leads me to suspect that the lights may be only representing what would happen if an 8 ohm load were connected, if only a 4 ohm load exists, it is possible that the lights are only reflecting half of the actual power being output. But if this is the case, I would expect it to be louder than it is in reference to the mains, which are very close to the same rated sensitivity, and can be checked A/B style at frequencies where their bandwidth overlaps. So the short answer is that I have no idea how much importance to connect to the led's, other than the fact that I can light up 5 led's at 25 hz with no problem, but only 4 at 16 hz before the bad sounds start.
I realize that tuning higher would make the problem at 16 hz worse, that is why I said that it would be a music sub - not HT. It would need a rumble filter at 20 hz if I reverse the taper, I accept that. But in it's operating range, I suspect that it would handle significantly more power with more sensitivity.
I appreciate your info about elongating the pipe to 20 feet, and although it is not possible with my current box, it is probably something I will do on another project with the goal being response to 10 hz. Since no one even mentioned the possibility of mass loading or tuning with ports to get lower, I assume the option is not a good one. Also, although it may be out of context, I have seen GM state that small pipes give small performance, and at 14 feet, this would be a small pipe for frequencies below what I already have.
GM - With too little information and possibly too much imagination I suspected that the "tapped" concept would work at 25% for any line length and geometry. I have come to believe that the exact location is much trickier to pin down with other lengths and geometries, so I agree that my driver offset is probably not suited to my geometry. This is why I assumed my next logical step would be reversing the taper, as it makes for a design VERY similar to the PB12, the TOP's concept box. Proven design = good results, right? But you said "The driver will be too close to the terminus if you swap ends." I don't understand, as reversing the taper would make it almost exactly the same as the PB12, with the same tap of 25/75%, same length and presumably similar performance. Since the driver is inside the box, it is equidistant from both ends, no matter which end is open or closed.
Finally, you said "It's a given that port area needs to be at least = Sd." Are you referring to this application, or all tl's in general? Is it the high or low end of response that suffers if CSA is not at least = SD?
And last but not least, John - I knew your opinion already, and I wasn't ignoring it, but please understand that the extension down to 16 hz will be missed when I go ahead with this. Also, when looking at the TOP's FR graph (or PB12, whichever is actually REALLY represented in that graph) I don't consider anything above 60 hz to be usable output. There are wild fluctuations in SPL, amounting to 6 db swings up and down and I really don't think I'd cross that sub over much higher than I do with mine. So the only real advantage to reversing the taper is extra power handling and sensitivity at the (LARGE) expense of extended bandwidth, and in particular LF bandwidth.
I fully agree (as much as is possible without a meter) that I am not getting the benefits of a tapped horn, as my response sounds (almost) exactly like what I see in GM's last graph. The only thing I can not qualify (without a meter) is the severity of the peaks and nulls. But I would guess that I have an ordinary tl performance with the added feature of a hidden and protected driver, and that's about it.
But John, why would I need to switch polarity? When I reverse the taper, I will not be flipping the driver over, it will stay exactly the way it is. Is that what you are referring to?
I could talk about this indefinitely, but I suppose this is a bit long for a forum post so that's it for now. Thanks guys, for the wisdom, opinions and suggestions.
Actually I was assuming that the power meters were actually just reading voltage and assuming and 8ohm load. What I was concerned about was the frequency response of the rectification that might be involved. If the meter display is a DC device than the AC signal must be rectified first. Chances are pretty good that they would not go to the trouble of making that circuitry flat to 5Hz. If that is the case you may actually be putting out a good deal more power at low frequencies than your meter might indicate.
As to the mass loading I don't know enough to say anything definite but I wouldn't be surprised to find an increase in acoutic impedence providing a measure of protection. You could always give it a try by blocking off a part of the open end.
mike
As to the mass loading I don't know enough to say anything definite but I wouldn't be surprised to find an increase in acoutic impedence providing a measure of protection. You could always give it a try by blocking off a part of the open end.
mike
just a guy said:
Greets!
Excursion increases 4x/octave with decreasing frequency, so you run out of excursion quickly with increasing power down low if there's not much room gain. At/around Fb though, power handling is a function of how well the driver is damped by the cab:
GM
Attachments
MASHAFFER - So you already guessed all that. If the LED's are reflecting output before the output stage based on an 8 ohm load then I am probably already near 200 (probably closer to 150) watts at 16 hz. I just assumed that wasn't likely, but using only one channel of the amp is keeping the heat and power supply issues in check, I suppose. So if the output stage is up to the job, it may very well be putting out twice the LED reading in addition to...
The idea that the LED's themselves might be rolling off prematurely. This is something I never thought of but makes perfect sense too. This may explain my previous comments that it appears that 25 hz is getting 10x more power than 16 hz. This is probably why I thought my "as built" graph would have a large spike at tuning, and it clearly did not. I was wondering how 10x could even be possible, but if the source AND the amp's output AND the LED's are all starting to roll off around 30 hz it may explain the LED's visual drop in power while audible output does not drop much, if at all, on the way down to 16 hz.
GM - I'm not familiar with graphs displaying this information, I've probably seen lots of them but paid little attention because I didn't know what they were trying to tell me. That being said, I assume the blue line is regular excursion and the red is the box's influence on excursion. If this is true, it would seem to me that the box is providing plenty of damping to the driver at 16 hz, but without anything to compare it to I don't know if it is as good as it seems.
Also, I got a hint of this with your second last post, and the idea was cemented in my head with this one. I had always thought in terms of how a driver was loading a room, but it never occurred to me to think about how the room loads the driver. So the answer to all my problems may not be anything to do with the box - it may be my lack of 1 foot thick cement lining in my room. It is fascinating to think that the concrete corner is not only a better amplifier, but also loads the driver to afford higher power handling. I wonder how that bit of interior decorating will go over...
So in retrospect it looks like my amp is lying to me, and it appears that even with my last minute changes in design I have come up with a tl with decent specs, although it is not lucky enough to have gained any of the benefits of a tapped horn. I will be reversing the taper sooner or later, to see if I can gain the tapped advantages, and also so I can stand it up again. It did not sound good with the terminus firing into the ceiling 4 inches overhead so I had to lay it down. If I port through the front instead of the end, I can stand it up and fire into the floor corner instead.
But before I do that I want to watch some more movies. Also I will experiment with mass loading a bit to see how low it goes before radically raising the tuning with the reverse taper. Thanks again.
The idea that the LED's themselves might be rolling off prematurely. This is something I never thought of but makes perfect sense too. This may explain my previous comments that it appears that 25 hz is getting 10x more power than 16 hz. This is probably why I thought my "as built" graph would have a large spike at tuning, and it clearly did not. I was wondering how 10x could even be possible, but if the source AND the amp's output AND the LED's are all starting to roll off around 30 hz it may explain the LED's visual drop in power while audible output does not drop much, if at all, on the way down to 16 hz.
GM - I'm not familiar with graphs displaying this information, I've probably seen lots of them but paid little attention because I didn't know what they were trying to tell me. That being said, I assume the blue line is regular excursion and the red is the box's influence on excursion. If this is true, it would seem to me that the box is providing plenty of damping to the driver at 16 hz, but without anything to compare it to I don't know if it is as good as it seems.
Also, I got a hint of this with your second last post, and the idea was cemented in my head with this one. I had always thought in terms of how a driver was loading a room, but it never occurred to me to think about how the room loads the driver. So the answer to all my problems may not be anything to do with the box - it may be my lack of 1 foot thick cement lining in my room. It is fascinating to think that the concrete corner is not only a better amplifier, but also loads the driver to afford higher power handling. I wonder how that bit of interior decorating will go over...
So in retrospect it looks like my amp is lying to me, and it appears that even with my last minute changes in design I have come up with a tl with decent specs, although it is not lucky enough to have gained any of the benefits of a tapped horn. I will be reversing the taper sooner or later, to see if I can gain the tapped advantages, and also so I can stand it up again. It did not sound good with the terminus firing into the ceiling 4 inches overhead so I had to lay it down. If I port through the front instead of the end, I can stand it up and fire into the floor corner instead.
But before I do that I want to watch some more movies. Also I will experiment with mass loading a bit to see how low it goes before radically raising the tuning with the reverse taper. Thanks again.
just a guy said:
Greets!
To compete with the TOP will either require a special driver or multiple sub drivers to increase efficiency/power handling.
1) Down low it's on a par with an equivalent EBS alignment and the latter is smoother higher up and easier to build, so I'm not impressed.
2) Well, heavily mass loading the driver looks promising in general, but will make it work much harder.
Mass loading the cab means to add a vent, which lowers Fp, or add weight to the driver's moving mass (diaphragm), which lowers its efficiency, but not Fp. Anyway, here it is with an added 200 gr (~68.5% increase), which lowers Fs to ~16.2 Hz:
GM
Attachments
I'M NOT IMPRESSED.
This is the feeling I have been trying to convey in uneducated terms pretty much since the beginning. But since no one wanted to actually say it out loud until now, I finally broke down in my last post and assumed the performance was decent.
GM, I completely understand what you are saying. I threw my driver into WINISD EBS -3 alignment sim again and (without a meter) can see that it is very close to the performance I am getting in a box the same size (or very close). But like you said, the EBS has much more usable bandwidth, and while it is 6 db quieter with 1w, the extra power handling makes up for that at higher levels of volume/power.
From the last graph, it does not appear that adding weight is something that is going to benefit the low end much, although it is amazing how it smooths everything else out, especially the peaks, but also the gradual boost between 25 and 40 hz. And while "heavily mass loading the driver looks promising in general" it appears that my room is not capable of backing up the sub when tuned that low.
This makes me think the John was right about 4 months ago, still in the concept stage, about the taper being the wrong way. At this point, it seems that reversing the taper will be just about the only way to make this into a sub with decent, if not completely impressive performance, at least in the small room it is in.
GM, without reading through this whole thread again, I believe that you have never endorsed this idea of reversing the taper, in fact, I believe you speculated that it would cause it's own problems. So I ask you this. If someone gave this thing to you, and out of respect for the person that gave it, burning it was not an option, what would you do to make it "impressive"? I'm looking for a single, off the top of the head answer, not a series of charts. I know we are all leading busy lives and cannot afford to be unpaid tutors in our spare time. I just want your one idea that would make this into something "better" than the average ported box.
Harsh reality is the only way I will learn anything, if it really needs to be burned I'm ok with starting over. I think reversing the taper is the answer I have been hesitant to explore but the only way to make it worth it's weight. If this was confirmed by BOTH of the wisest people I know on the subject I would be much less hesitant lose my LF extension and start cutting.
This is the feeling I have been trying to convey in uneducated terms pretty much since the beginning. But since no one wanted to actually say it out loud until now, I finally broke down in my last post and assumed the performance was decent.
GM, I completely understand what you are saying. I threw my driver into WINISD EBS -3 alignment sim again and (without a meter) can see that it is very close to the performance I am getting in a box the same size (or very close). But like you said, the EBS has much more usable bandwidth, and while it is 6 db quieter with 1w, the extra power handling makes up for that at higher levels of volume/power.
From the last graph, it does not appear that adding weight is something that is going to benefit the low end much, although it is amazing how it smooths everything else out, especially the peaks, but also the gradual boost between 25 and 40 hz. And while "heavily mass loading the driver looks promising in general" it appears that my room is not capable of backing up the sub when tuned that low.
This makes me think the John was right about 4 months ago, still in the concept stage, about the taper being the wrong way. At this point, it seems that reversing the taper will be just about the only way to make this into a sub with decent, if not completely impressive performance, at least in the small room it is in.
GM, without reading through this whole thread again, I believe that you have never endorsed this idea of reversing the taper, in fact, I believe you speculated that it would cause it's own problems. So I ask you this. If someone gave this thing to you, and out of respect for the person that gave it, burning it was not an option, what would you do to make it "impressive"? I'm looking for a single, off the top of the head answer, not a series of charts. I know we are all leading busy lives and cannot afford to be unpaid tutors in our spare time. I just want your one idea that would make this into something "better" than the average ported box.
Harsh reality is the only way I will learn anything, if it really needs to be burned I'm ok with starting over. I think reversing the taper is the answer I have been hesitant to explore but the only way to make it worth it's weight. If this was confirmed by BOTH of the wisest people I know on the subject I would be much less hesitant lose my LF extension and start cutting.
Greets!
AFAIK you didn't ask for an opinion till now.
Yeah, designing based on 'x' amount of room gain down low is not a good plan without first measuring the room's gain curve. My room is an acoustic sieve.
Correctomundo! Reverse tapering is for reducing pipe length for a given Fp, while a positive taper (horn) is for acoustic gain.
OK, I'll skip responding to any other previous Qs/clarifications and say that short of having an expensive custom driver made, it's about as good as it gets within all the limitations you set for the desired app. I mean you can't set performance goals, then set limitations that insures you can't meet them.
WRT converting it to a horn, it will increase gain above Fs and lose the extra gain BW below it, somewhat defeating the point of a sub if there's no room gain of note to offset the loss.
GM
AFAIK you didn't ask for an opinion till now.
Yeah, designing based on 'x' amount of room gain down low is not a good plan without first measuring the room's gain curve. My room is an acoustic sieve.
Correctomundo! Reverse tapering is for reducing pipe length for a given Fp, while a positive taper (horn) is for acoustic gain.
OK, I'll skip responding to any other previous Qs/clarifications and say that short of having an expensive custom driver made, it's about as good as it gets within all the limitations you set for the desired app. I mean you can't set performance goals, then set limitations that insures you can't meet them.
WRT converting it to a horn, it will increase gain above Fs and lose the extra gain BW below it, somewhat defeating the point of a sub if there's no room gain of note to offset the loss.
GM
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.