| moray james |
| the Arnold Klayman patent? US # 3 047 090. The second drawing in this patent is especially interesting as Klayman uses a simple but very clever idea to build two lines into one box and manages to load not only the fundamental but also the first harmonic of his 1/4 wave line. Regards Moray James. |
|
|
| cowanaudio |
G'day Paul
What a great idea! It's fun playing with tapped horns, and not too hard coming up with a good result. I'd join you in a heartbeat if I didn't have to drive 16 hours.
This will certainly save having to purchase the drivers yourself. All the drivers I tested were my own. I assume the baffle the driver mounts on is removable, so you can modify it or replace it to adapt different drivers? The bottom layer of the baffle could have a clearance hole for a 15" driver and remain fixed and the driver baffle could then be screwed to that.
What is the length of the line? Also it might pay to mount the drivers the other way around for more space behind the magnet.
The 830515 was very poor in my 30Hz tapped horn. The 830500 was a good performer.
Measurements are good if you can do them.
Cheers
William Cowan |
|
|
| moray james |
| I see you peaking out from behind those woofers! I would be interested in your comments on the Klayman dual duct design in the above patent. Since you have been sucessful in obtaining two full octaves from a single tapped horn do you think that the dual version would or could stretch another half octave? Regards Moray James. |
|
|
| rick57 |
Hi Paul
The PHL I measured IIRC as 390 mm centre – centre, can check on the weekend.
William’s suggestions were in one of the threads here, I search I think on ‘tapped’ and ‘Vas’.
Hi William,
Peerless– on the 5th graph at www.diy.cowanaudio.com/th.html you got more deep bass with the 830515 than the 830500 - -did you prefer the ‘top end’ of the 830500?
Is the left hand vertical line on those graphs, 10 or 20 Hz?
Thanks |
|
|
| cowanaudio |
G'day Rick
I didn't notice this until now, but those graphs are labeled incorrectly. The red trace is the 830500. Many people must have thought I was crazy preferring the brown trace! The red 830500 trace is very easy to eq flat, with only one channel of PEQ. I will change the page ASAP. That plot is 20-200Hz.
Moray, I'm not sure how the Clayman design would work, I havn't played with one.
Cheers
William Cowan. |
|
|
| rick57 |
I tried to look up Klayman’s 1962 patent
http://patimg2.uspto.gov/.piw?Docid...View+first+page
but it doesn’t allow access or display for me.
Hi Moray
Would the extra half octave be at the top or boo tom end? Can you post the drawing?
Hi William
At one stage you suggested Qt of 0.2 - 0.4, FS near LF cutoff, and
VAS low. We mightn’t have time to properly test all - any further thoughts on optimum TS specs?
The 830500 is Qt of 0.17, FS of 18, and VAS of 139.
Any thoughts on how Qt of 0.35, FS of 23, and VAS of 144 the much cheaper but well regarded CSX might be?
Cheers |
|
|
| cowanaudio |
| The plot is 10 to 100Hz. |
|
|
| moray james |
| plus another interesting US patent. |
|
|
| paulspencer |
To keep the other thread focused on the day I'm organising, I'm starting this thread to talk about the technical aspects of this tapped horn project. The thread to discuss the day is here:
http://www.diyaudio.com/forums/show...&threadid=97620
Here is a simulation:
Yes, it looks bad for the tapped horn, but hornresp is not simulating the full story. A tapped horn such as this is a 1/4 wave horn at 18 Hz which is what hornresp shows, but at the same time it is a 1/2 wave horn at 36 Hz. This is what hornresp doesn't show. With the correct driver, the result is that the response is flat and the dips are filled in.
The tapped horn has only one driver while the other horn has two. Note that the voltage input is the same for both at 30V.
How can it be both a 1/4 and 1/2 wave horn? The front of the driver loading into the throat gives the 1/4 wave output, while the rear of the driver located in the mouth actually sends its output towards the throat and then back again.
I'm far from understanding it fully, hence this project to try out different drivers and see how what we can measure is different from the simulation. |
|
|
| moray james |
| US # 2 871 972 sorry I can not shrink the picture enough to post here you will have to Goolle the patent and have a look. |
|
|
| rick57 |
Hi William
When philpope in the discussion at http://www.speakerplans.com/forum/f...?TID=8329&PN=10
said
“I think when William said 2:1 compression ratio, he meant the ratio of Sd to the end of the horn, NOT the cross section where the tap is.
I assumed that the lowest bass response would be obtained by tapping the driver into the horn where the cross section is the same as Sd so I make the throat of the second section in the model equal to Sd.”
Was that what you meant?
Cheers |
|
|
| cowanaudio |
G'day Paul
Looking at your sim, I'd guess you'd be about 4dB down at 22Hz. That is a good compromise for reduced size. The tap will smooth out the dip above the low cutoff. How about you sim the different drivers being offered for the comparison?
Rick
The compression ratio is the area ratio between the driver and the throat of the horn. If your driver Sd is 800CM^2 then if the throat is 400CM^2, you have a 2:1 compression ratio.
Cheers
William Cowan |
|
|
| paulspencer |
I'm collecting driver data so that I can simluate them all.
Drivers:
AE speakers AV12 2nd run
Rythmik DS12TC
Peerless CSX 10 & XLS12
PHL 5230
Eminence Kappa pro 12
Eminence Alpha 15
Lambda 15TDM
JBL 2035
For the last 2 drives, I don't have parameters for CMS, RMS and MMS which are necessary for hornresp. A quick search didn't turn up anything. Does anyone have data on them?
So far the AV12 parameters look very good according to the target range. It meets all criteria except being slightly under recommended SD (not a big deal). I have found it also models very well in a bass horn generally. Not surprising since it is similar to the Lab driver which was designed for bass horn use. |
|
|
| paulspencer |
Hornresp simulation
AV12 - red
PHL - green
The other 2 drivers are the Alpha 15 and the Kappa Pro 12. The kappa is very similar to the PHL except that it has a peak at the bottom end due to the much stronger magnet. The alpha has the extreme low peak due to the crazy high Qts and very weak magnet, but the dip is the biggest. The XLS not shown as it is fairly similar to the AV12. The Rythmik is similar to the AV12 except that its bottom peak and dip and less. CSX is similar to the Rythmik but less efficient.
Of them all, I suspect the AV12 will work out the best. My gues so far is that if the magnet is too strong, and hence the Qts is too low, the end result will be drooping bottom end.
Simulation is for 1w of input in a corner. |
|
|
| rick57 |
Paul
What I have:
15TDM JBL 2035
Fs 35 48
Re 6.6 3.9
Qms 4.5 5.0
Qes 0.35 0.36
Sd 855 880
Vas 405 140
Xmax 3.3 7.1
(Le) 0.20 0.25
Cms 0.3 ?
Rms 3.0 ?
Mms 70 85 |
|
|
| cowanaudio |
G'day Paul
Hornresponse will let you back calculate the values you are missing. Enter all the values you have and double click on the empty field where the unkown value should go. It will work it out for you.
It looks like the AV12 will work well in this tapped horn. I hope you have enough power available to explore it's limits. An Aussie Monitor AM1600 could not give an 830500 any grief before going into clipping.
Cheers
William Cowan |
|
|
| AJinFLA |
William,
What sort of GD are we looking at with the 18hz horns?
Did you reverse mount one of the 830500 motors?
Cheers,
AJ |
|
|
| paulspencer |
William,
I seem to find the AV12 to be very good for bass horn use. I have an Europower EP2500 to power it, which has 2 x 650w into 4 ohms. To drive just one in a tapped horn, I have 2.4kw bridged into 4 ohms (yes it's stable down to 2 ohms). Do ya think that's enough?! :D
What REALLY surprised me the other day - I saw it clip when driving them in OPEN BAFFLE! :bigeyes:
It was just once for a brief moment. That's only the 2nd time I recall in a few years that I've had it. Before that amp I had tried out two 2x450w pro amps and both of them clipped before the subs ran out of steam for music. Of course I use a rumble filter ... when I upgrade my mains to keep up, then they will have more of a challenge. |
|
|
| paulspencer |
Ok, managed to sort out the missing parameters (WinIS pro is also useful for that)
It's quite surprising how well the AV12 performs. It is actually more efficient than those larger drivers which would normally have over 10db higher efficiency. Both of the larger pro drivers meet all the recommendations for a driver except they have a large SD and they are very light at around 80g where >200g is recommended for a 12" driver. |
|
|
| cowanaudio |
G'day again Paul
This is one of the better threads I've read for a while!
You should have enough power, but watch your voicecoil. Unless you drive way below cutoff, excursion should not become an issue.
This reminds me of a funny thing that happened the other day. I installed the XXLS drivers, with the help of some adapter rings, into the holes normally filled with the 15LW1401s below the Unitys in my system. (They were in separate sealed boxes of the same internal volume) I was a bothered by some distortion that was present on some tracks at moderate levels. I pulled the XXLS drivers out of their boxes for a visual. They seemed fine. Powering them up with another power amp in free air could not induce the distortion I had heard while the drivers were installed in their boxes. I re-installed the drivers and fired up the system again. Slowly increasing the level until I could just hear the distortion again. At that point I headed down stairs to have a look at the level meters on the AM1600 that was powering this pair of drivers. It's probably an understatement to say I was a little shocked to discover the amp was being driven into clipping! 800WRMS into each of these poor 12" drivers. I'm so used to very efficient speakers that this outrageous consumption of power took me completely by surprise! Had I ever driven the 15LW1401s to even 100W each, I'd have been blowing my head off. Needless to say I am delighted with the low distortion at high drive levels that these XXLS drivers show. There was not a hint of distress to just short of 1KW per driver, then it was the amp that went ugly.
AJ
Didn't measure GD. Should have but didn't reverse a driver. The sub was still very clean.
Cheers
William Cowan |
|
|
| paulspencer |
William, I'm a twit! I was trying to come up with a way to mount drivers push pull, and didn't think of the simple solution you used there with just one driver reversed!
If this does get beyond prototype, then I'll either build two installed under the floor, or one with push pull drivers. The beauty of this is that I can put them anywhere in the room (almost), hence have the freedom to get the best room interaction. Of course, the beauty of push pull is that it's possible to mount them so that their forces also cancel, which is welcomed in a box of this size - anything to keep vibration down without having to brace, make the box thicker and use more expensive materials. The difficulty is getting that to work with this type of box - a bit complex to pull it off, and quite awkward. I need to find out first if it's worthwhile.
What sorts of dimensions do you have with that XXLS tapped horn?
Have you by any chance tried it with push pull?
Do you know what the efficiency with 1w is? Hornresp seems to show very high efficiency.
I'm also curious to know if you have been able to trick hornresp into modelling something like what you actually measure? |
|
|
| cowanaudio |
G'day again Paul
That horn is XLS, 830500. The XXLS drivers are in sealed boxes in my Unity system at the moment (covering 20-300Hz). I will try them in the 30Hz tapped horn some time in the next few weeks (Pending time) The timber in the big horn has since gone into other projects.
Reversing one of the drivers in the dual driver horn will not give a reduction of vibration. It will only serve to reduce even order distortion products. I guess you'd need to do a one up two back type deal to put the drivers in a clamshell front to back type arrangement.
The different tapped horns I measured were between 95 and 100dB/W/M.
No luck with an accurate model in Horn Response, but the guesstimates give you a pretty good idea.
Outside dimensions of the big tapped horn are 2250 by 450 by 515. All material is 25mm. Taper is shallow, under 1.5:1.
Cheers
William Cowan |
|
|
| paulspencer |
When I mentioned the more complicated push pull I meant a clamshell labhorn type mounting, except with one driver inverted, hence as you know, reduced distortion and vibration. As you can imagine, much more complex to build. If I go beyond prototype then I'll see if I can come up with something along those lines.
Did you form any conclusions when it comes to throat to mouth ratios and compression ratios?
Did you discover any particular and significant impact of changing throat to mouth ratio?
Did you find that 25mm ply is adequate without bracing? |
|
|
| cowanaudio |
G'day again
One path up, two back or the other way around will give you an easy to build clam shell.
#2 Umm...See below
#3 Yep, Hornresponse will show the changes I saw.
#4 Yep, for testing at least. (The pressure didn't rip any boxes apart, and the panels did not vibrate too much)
Cheers
Wiliam Cowan |
|
|
| paulspencer |
| quote: | | One path up, two back or the other way around will give you an easy to build clam shell. |
Could you elaborate, I don't quite follow! |
|
|
| cowanaudio |
| The bottom left and right corners are the mouths. You could reverse the whole thing for a mouth in the middle. This arrangement reduces even order distortion products and cancels mechanical vibration. |
|
|
| paulspencer |
Ahhhhh, I get it now. Yours is perhaps a little simpler than my idea of how to do it:
I think I like your version better. |
|
|
| cowanaudio |
Paul
Your drawing is better.
Since this is a new idea, perhaps I should patent it!?$
My design just requires one more internal baffle, which adds bracing. External size is only increased by the thickness of that baffle (25mm). A horizontal brace between the two drivers would probably help.
During the free air testing of the XXLS driver I thought had failed, I had it sitting on it's magnet and playing high level sinewave through it. The driver was almost bouncing off the floor! :bigeyes: The carpet has a 9mm or so pile and the driver frame was moving up and down about 6mm at times. Cancelling vibration between opposing drivers has got to help.
Have a good weekend.
Cheers
William Cowan |
|
|
| paulspencer |
New version:
Part of mouth would have to be covered (too big) but as shown it will accomodate the 15" drivers and a 430 x 430 mounting plate to be slid in to one side.
Mounting plate shown hatched. |
|
|
| rick57 |
(I tried to attach moray's refenced patent, but the file was too big) :(
Hi Paul
So I don’t build up expectations and later find I can’t fit one in, what would be the external dimensions of the current config, if built as a settled design rather than a prototype, for 10/ 12/ 15 inch drivers?
Or even to see if I could fit one in my lounge room as well :D
Thanks |
|
|
| paulspencer |
Top drawing is 2250 x 515 and bottom drawing is 2250 x 436. Mounting plate shown hatched is 430mm square and the opening in which it fits is 450 x 450 - enough to fit any 15" driver. Hatched blue circle shows the cutout required for the PHL driver, hence the baffle hole shown there will be cut for this or for the biggest 15" that anyone is bringing. So in short, any 15" driver should fit.
I'm going to have to build this thing out of 3 sheets, no way around it. I had hoped for 2, but it's obviously not going to happen.
Are any of your drivers in pairs? I'll probably make the second baffle to suit the AV12 without mounting plate.
I'll be interested to see the impact of one vs two drivers. In the next 1-2 weeks my Rythmik Direct Servo kit is coming, not only would I like to try it tapped, but it could also be a good reference sub to compare them all against, being a servo sub. One thing I want to test is if a horn can actually get a conventional driver to compare with a servo sub. Rob Wells commented that his lab horns are very similar to servo subs, except at much higher output. This is what first really got my attention about bass horns. |
|
|
| AJinFLA |
I would strongly recommend against using the Rythmik in a TH without consulting with Brian. My guess is that he will not recommend it either. The TH will certainly give much greater output, but there is also evidence that points towards phase distortion possibly being audible a very low frequencies and mainly at very low frequecies. I have not yet seen Gedlee AES paper so I suppose the jury is still out. I however, cannot imagine how a pair of sealed Rythmik Servo 15's would not supply sufficient low frequency output for music (or even HT for that matter) unless the room was extremely large.
The SQ is superb for music with the sealed (recommended by Brian) enclosures (low GD). I'm not saying it is because of the low GD, but having low amounts certainly can't be hurting the performance.
Keep in mind this GD is inherent in the system before XO.
Now of course if you need 120db @ 25hz at the listening position, TH's do have their appeal ;) .
cheers,
AJ |
|
|
| paulspencer |
Why ask for an opinion on whether something will work when you can simply try it and find out for yourself?
I doubt I'll keep the Rythmik drivers in a tapped horn, more likely dipole or a large horn, or perhaps more short term a sealed box. This is about experimentation first and foremost. It won't be a failure unless we learn nothing. If I come out th e other end wanting to actually use a tapped horn, then that's a bonus. If not I'll re-use the MDF for something better.
If I do use the Rythmik, then I'll see if I can get a custom made servo board. Apparently only the frequency response is required.
The paper does look interesting. Geddes has done some very interesting work in the field of perception of distortion. |
|
|
| rick57 |
Hi Paul,
Finally got some time to read and think about this a little.
Not attempting to pour any “cold water”, just to understand better:
Post 1 –
“With the correct driver, the result is that the response is flat and the dips are filled in”
How confident can we be of this (or if the measured response ended up close to the sims), how easy would it be to flatten the sharp notches?
Hi William,
On that note, what distance from the mouth were your measurements taken at?
Post 15 –
The PHL sims a smoother response than the AV-12, so curious why its suspected the AV12 will work out best?
“The kappa is very similar to the PHL except that it has a peak at the bottom end due to the much stronger magnet”
The BL of the Kappa Pro 12 is 17.3, the XLS 17.6, JBL 2035 and the AV-12 BL is 16.
The PHL 5320 is an ‘outlier’ in this group at 23.4. Not clear on this possible correlation?
Mounting
Is it possible to mount the drivers per Planet 10’s Push-push in post 2 www.diyaudio.com/forums/showthread....6902&highlight= to cancel the drivers’ Newtonian motion, and still be a tapped horn?
“Top drawing is 2250 x 515 and bottom drawing is 2250 x 436”.
Both liveable, IMHO ;)
All my drivers are pairs, though how many will we have time to test, maybe we need a weekend? :D
Cheers |
|
|
| paulspencer |
G'day Rick, my latest version is like version 3 linked in the other thread. It's not practical to face the magnets together in this example, and I think it's best to get distortion reduction as well.
About how confident, the unknown is the drivers in question, not the concept of the tapped horn itself. William shows measurements to demonstrate it works.
I'm just speculating about which will work best. Kind of like guessing which horse will win melbourne cup.
What I am looking for though is two peaks and a dip - if there is no dip then it's not going to get filled in and then you have early rolloff and less efficiency down at the tuning point.
But we will find out when we actually measure it.
I'll allow for two drivers of each, but I just wonder if two 15s will be too much SD, but who knows. I'll allow for it anyway.
Comparing the drivers I was talking about the hornresp SPL chart, not their parameters.
If the response in reality works out like those charts, then no chance of eq, we are counting on it working out much better as hornresp isn't really set up to model it all. |
|
|
| rick57 |
Are any of your drivers in pairs?
Except the CSX 10", of which the viewpont of a d-s-t engineeer is they're capable of "excellent transient response", I have four of . . |
|
|
| paulspencer |
| All in pairs - 2 x AV12 and 2 x Rythmik (except 2nd won't have arrived in 2 months I don't think) |
|
|
| rick57 |
G'day Paul,
Thanks for your clarifications
G'day William,
Also, horns as a rough rule of thumb if I understand correctly add 5 – 7 dB ~ don't know if this applies to bass horns.
What about tapped horns; what is the db or watts scale on your graphs?
Cheers |
|
|
| rick57 |
Of pre-existing designs tapped horns are probably closest to Tapered quarter-wave pipes. I don’t know enough about each to say how they differ.
What are any differences?
Thanks |
|
|
| paulspencer |
I believe the efficiency is more if I understand correctly.
If I'm not mistaken, it's more like an undersized conical horn where a trick is used (the tap) to overcome the peaky response that would normally result. I believe a transmission line achieves a moderate improvement over a vented box in accuracy, but comparable output, but a tapped horn should create significantly more oomph. Apparenly the Danley tapped horn has more output than the contrabass to the tune of 10 db. The Contrabass used 2x high excursion 15" with two 18" PRS. This does more with a single 12" driver. |
|
|
| AJinFLA |
| quote: | Originally posted by paulspencer
Why ask for an opinion on whether something will work when you can simply try it and find out for yourself?
|
Because that (manufacturer) opinion might prevent damaging the system?
| quote: | Originally posted by paulspencer
I doubt I'll keep the Rythmik drivers in a tapped horn, more likely dipole
|
Unless much has changed since I asked (Brian) years ago, it's unsuitable for dipole use also.
I'm not trying to discourage you from trying it all, you paid for them, so they are yours to do with as you wish. I'm simply letting you know that the manufacturer has advised me previously against this. It might be based on his knowledge of his product or perhaps he was just being cautious.
Best of luck with your experiments, I'll keep an eye on them to see how they turned out.
BTW, which Rythmik kit did you get? He seems to have expanded the choices quite a bit.
cheers,
AJ |
|
|
| cowanaudio |
G'day Rick
>On that note, what distance from the mouth were your measurements taken at?
The published graphs were taken with the microphone in the center of the plane of the mouth. The plots were taken indoors, and this was done to eliminate the effects of the room.
>Also, horns as a rough rule of thumb if I understand correctly add 5 – 7 dB ~ don't know if this applies to bass horns.
Yep, 5-10 dB gain seems reasonable.
>What about tapped horns; what is the db or watts scale on your graphs?
All the horns I measured measured between 95dB/2.8V/M and about 102dB/2.8V/M. As the mouth size goes up the efficiency goes up too. You get less efficiency than Hornresponse predicts.
With the right driver, the dip between the first two peaks will be eliminated.
Cheers
William Cowan |
|
|
| bear |
An interesting idea...
I wonder if there is any real gain here... ?
gain over a 1/4 pipe? (for the freqs the pipe works)
Also, I wonder about the pressures on the drivers - especially the one that happens to be out of phase with the other at any given freq... thinking about the relationship between the peaks of the "main driver" being filled in by the "other driver"...
Also, it's a big box for 30Hz.
What's the relative differential with a 15" driver with the right specs in a similarly sized volume ported cabinet?
How about compared to one of the many bandpass cabinets?
...thinking out loud.
_-_-bear |
|
|
| paulspencer |
AJ are you suggesting that this might destroy the Rythmik driver or amp or servo board?
I'll have to see what Brian has to say. As I understand, the main issue is getting the servo board to understand the frequency response so it can get it right in the nearfield. It may be some time til I settle on something that I'm happy to stick with for a while. A lot of "try things out" urges to use up first.
Bear, this is more like an 18 Hz horn with the right driver. The output should be more than a TL and distortion lower due to efficiency gain. Also the second driver is not what fills in the gaps, it is the fact that the rear output from each driver which is near the mouth loads the horn as a 1/2 wave horn - it does this by virtue of the fact that radiates from the mouth back into the throat, then back again. Effectively this allows the driver to behave differently across the frequency range in such a way that it overcomes the shortcomings of a too-small horn. I've tried sims of bandpass subs with my drivers and I haven't been able to model any significant advantage, even with a size-no-object box.
Will be very interesting to see how things turn out. The results here could either inspire or deter many other attempts. |
|
|
| GM |
Greets!
FWIW, I visited DSL the other day to see the 'container' tapped horn, but unfortunately they were behind schedule due to stuff beyond their control, so didn't get a chance to experience forty horn loaded sub drivers, though a quickie audition of some their products using action scenes from 'Spider-man 2' left no doubt that the T-O-P delivers eyeball flattening bass in spades..........
GM |
|
|
| paulspencer |
GM, what can I say but I'm envious!
| quote: | | no doubt that the T-O-P delivers eyeball flattening bass in spades.......... |
By tower of power are you referring to the DTS-20? |
|
|
| GM |
Greets!
Right........
GM |
|
|
| paulspencer |
| What were your thoughts on the TOP vs contrabass? |
|
|
| GM |
Greets!
Didn't have any, I mean I haven't heard the CB in years and my acoustic space is maybe 1/20 what the TOPs are in. WRT SQ, I was very disappointed, but I'm not familiar with either of the pieces used to demo plus there were lots of other speakers, tapped horns scattered around in close proximity to act as various BW passive radiators, so I'm sure the system has much more potential than I heard.
Really, what impressed me most was the Unitys. The movie scene where lasers? zap the pellet were so clear, razor sharp and dynamic I looked down at my chest, it speared me so hard, realistically, like a small caliber gunshot wound. I'm looking forward to getting my system going again to see how the Altecs fare.
GM |
|
|
| AJinFLA |
| quote: | Originally posted by paulspencer
AJ are you suggesting that this might destroy the Rythmik driver or amp or servo board?
|
I think if you use due care you should be fine.
| quote: | Originally posted by paulspencer
I'll have to see what Brian has to say. |
That would always be my advice ;).
Best of luck. I'll be watching. I do have 4 XLS12's laying around.
TH sort of SPL might get me thrown out of the neighborhood however :) .
cheers,
AJ |
|
|
| freddi |
| what's a good single 12" and flare for TH which will just comfortably fit 7'10" ceiling w/o scraping? is it safest to go w. William's 30Hz TH and that Peerless? |
|
|
| paulspencer |
Well your ceiling height is spot on the height of the TOP, so I'd imagine you could just make it just shy of those dimensions to fit your room if standing it up. Should do better than 30 Hz. I can't recommend any driver yet, except point to William's results in which the car version of the old XLS performs well. Richard found driver recommendations I believe - not sure of the source - he might like to chime in. I'll leave other advice to William since he has runs on the board on this one. So far I just have ideas that I haven't yet tested.
Actually I can't really recommend this type of sub yet, not having heard it myself, and not having tried it yet. |
|
|
| rick57 |
“Richard found driver recommendations I believe - not sure of the source - he might like to chime in.”
I’m sure I saw these posted by William, in a thread here
William am I right?
Will the benefit of two XLSs (apart from, if wired in parallel a likely + 6dB) be how much less distortion do you think? |
|
|
| cowanaudio |
G'day Paul
The 830515 car XLS driver is a VERY POOR performer in my 30Hz tapped horn. The graphs on my webpage were labeled incorrectly. This was pointed out and rectified last week. The 830500 worked much better.
Rick
Running a pair of drivers, facing in oposite directions, will yield a very worthwhile improvement in distortion, at a given level. Each driver will have half the excursion, and the even order modes will nearly cancel out. This is good in a final implimentation, but I would not recommend running dual drivers in Paul's test box. This will only complicate the issue.
Cheers
William Cowan |
|
|
| paulspencer |
| quote: | | The 830515 car XLS driver is a VERY POOR performer in my 30Hz tapped horn. The graphs on my webpage were labeled incorrectly. This was pointed out and rectified last week. The 830500 worked much better. |
Oops. Glad to hear it actually! We'll have the one that works well, and since it looks like AV12 in hornresp, that's a good sign.
I'm not sure I understand why you don't recommend dual drivers in the test box? Could you elaborate a little more?
I was hoping to reduce the vibration so bracing is less critical, and be able to then brace the test box and improve it a little and have a useable box, rather than having to scrap it and build from scratch. If I allow for 2 drivers in the test box, I could experiment with one or two drivers driven.
I'd also be interested to hear your subjective thoughts on your tapped horns vs other subs. |
|
|
| cowanaudio |
G'day again Paul
I think the first tapped horn you build should be as simple as possible to maximise your chances of success. If you build your horn according to "plan A" I think your chance of success is very high. Two drivers will need a bigger box.
>I'd also be interested to hear your subjective thoughts on your tapped horns vs other subs.
Lots of output. Very low distortion because of excellent cone control. I was unable to upset the drivers with the full output from an AM1600. The big horn plumbed the depths with more ferocity than any other sub I've experienced. Even an 800L dual 2245H vented box couldn't get close to the big tapped horn for sheer low frequency grunt. Wherever I placed the mouth of the horn the gyprock on the walls would come loose and start banging against the studs.
Tom's fireworks recording was simply jaw dropping with this sub running under my Unity system. The sub integrated seamlessly in my very high resolution system.
It was big, very big, but had a small foot print when standing up. The big tapped horn has been dismantled. The 30Hz horn gets used in the workshop.
Cheers
William Cowan |
|
|
| freddi |
G'day William - -my lack of mobility, tools & help are a hinderance but gotta get something better than Klipschorns - K-H sound kinda like BP plus a bit of non-coherent old-school PA.
Karlson coupler = stubborn pick of midbass over reflex and some horn as cone excursion is controlled at high peak outputs. Drums are great. I briefly ran one Unity passively xovered to an 18"K but may use 15" coupler with K-tube - if can guess a smooth playing front section from previous observations.
I don't know if bass k-couplers are practical although exemplar chose to re-tune and boost K15 with an ALtec 15 and subjectively that's pretty good.
does your little 30Hz Peerless TH play extremely taut once xover is dialed in?
Best,
Freddy |
|
|
| rick57 |
G'day Freddy
I’m also very big on drums with fast response & minimal overhang.
However, THs normally only cover 2 octaves up to about 60-70 Hz, below the fundamentals of even bass drums.
Still a good question ~
G'day William,
What’s your experience of the response speed & overhang of THs vs other horns or conventional alignments?
And on bandwidth, has anyone looked at the patent 3047090 which it was suggested extracts a further half an octave? http://free.patentfetcher.com/Paten...ch&PN=3+047+090
It’s of a kind of folded horn, so I haven’t been able to see it’s relevance.
Half an octave of itself is not a lot, but it would make integration with the driver ‘above it’ a lot easier. It would be good to know why bandwidth is only 2 octaves, and if this could be stretched a little, maybe by sacrificing a modest amount of the generous efficiency.
Maybe an understanding of transmission lines (something I’ve never really looked into) esp tapered quarter-wave pipes would help. Any TL folk eg dave? MJK? reading this?
Cheers |
|
|
| cowanaudio |
G'day Freddi and Rick
Firstly, the quality of bass from a correctly implemented tapped horn is excellent. This is also the case for low Q sealed, dipole and well designed bass horns. The design itself is not floored.
The big dip in the top end cannot be removed. We tried some damping strategies (Which partly worked but cost us sensitivity) and we also tried spacing two drivers along the horn, again which only partly worked. If you think about the four acoustic paths in a tapped horn, it's quite easy to see where the cancellation related dip comes from. The reflection off the throat of the horn from the mouth side of the driver is nearly perfect. We were getting nulls in the order of 20 to 30dB. Essentially the entire signal is being reflected back to the mouth and destructively interfering with the direct output. Tom's horns are essentially two octave devices as well. (Give or take a bit for driver choice and horn layout.)
These four paths are what makes modelling this beast so difficult. Hornresponse will only model the dominant full path. A complex sum has to be made on all four paths to generate an accurate model. I guess this is a topic for discussion elsewhere.
Cheers
William Cowan |
|
|
| rick57 |
G'day William,
I see what you mean about cancellation causing dips and modelling being difficult.
But I’m still interested in reducing the dip in the top end, and willing to sacrifice a little of the generous efficiency.
If you get a time to post here, or a new thread, I’d love to here about the damping strategies, and how much they gained (more so if they would also work with the 830500).
Cheers
Rick |
|
|
| cowanaudio |
G'day again rick
This plot gives you an idea of what some acoustically absorbant material will do at the velocity maxima at 75Hz. Red is undamped. You reduce the dip to just 8dB, but lose 6dB of sensitivity in the process. My conclusion was that it is not worth doing. |
|
|
| cowanaudio |
| This would be the green trace above. Damping is polyester, 35Kg/M^3, compressed to about 50Kg/M^3 |
|
|
| cowanaudio |
| And this would be the blue trace above. Damping is polyester, 35Kg/M^3, uncompressed. |
|
|
| djk |
| Knowing what doesn't work is very helpful, thanks. |
|
|
| freddi |
| thanks William= cool stuff - guess a shelf like on K15 wouldn't do anything (good?) |
|
|
| cowanaudio |
| It might look cool! :cool: But I think that's about all. |
|
|
| djk |
Fred, I have a drawing of a taped horn with a cavity in front of the woofer with a K-slot in it.
If you look at the 15" model from Danley it has a deep dip at around 140hz, then comes back and is usable to about 250hz. A K-box front chamber resonates about 160hz. The same idea may help out in the taped horn (or it may not). Have to make some sawdust to find out. |
|
|
| rick57 |
Thanks William, very helpful
Wonder if locating the polyester in a different location might give a less unfavourable trade-off?
If we get time at Paul's, we could try, meantime anyone (with TL experience?) have an opinion?
Cheers |
|
|
| cowanaudio |
G'day Rick
Tried that too, only gets worse from there. |
|
|
| paulspencer |
Rythmik Direct Servo kit arrived yesterday. Very very impressive. It will be very good as a reference sub. I'm impressed by how clean it is, and how dynamic. It's actually better than I expected.
I hate to do this, but I think we'll have to go back to the first design. If the second needs to be twice as big, then I'll stick with the first. Or if I do build the 2nd design, probably end up just using one driver in it. Any bigger will probably be too big.
Richard, as it stands I think you have two good placement options if actually used permanently. Standing in a corner, or sitting along a wall on it's side. With the latter, you could put a TV on top and make it look just like an entertainment unit. |
|
|
| rick57 |
Hi Paul
"I think we'll have to go back to the first design."
The bolt holes, outside to outside on the 15s are 374, 385 & 394 mm, allow a little for the bolt heads. Will all three fit?
Thanks |
|
|
| paulspencer |
| I'm allowing for a maximum dimension of 400mm so yes they should fit. |
|
|
| rick57 |
. . maximum dimension of 400 mm
by the skin of their chinny chinny chin ;) |
|
|
| paulspencer |
| If you look closely at the dimensions for the PHL, there is more room than you think. The hexagonal shape means it will have around 10mm clearance. The JBL may be another matter. |
|
|
| rick57 |
| The PHL is the biggest, the JBL the smallest. |
|
|
| rick57 |
Maybe our militaries could use this in Iraq, or even on try em on their opposition parties. ;)
Go George, Tony & Johnny go! |
|
|
| Patrick Bateman |
| quote: | Originally posted by GM
Greets!
Didn't have any, I mean I haven't heard the CB in years and my acoustic space is maybe 1/20 what the TOPs are in. WRT SQ, I was very disappointed, but I'm not familiar with either of the pieces used to demo plus there were lots of other speakers, tapped horns scattered around in close proximity to act as various BW passive radiators, so I'm sure the system has much more potential than I heard. |
Isn't anyone going to comment on this?
In my personal experience, dual reflex bandpass is very VERY hard to get correct. And the tapped horn has a lot in common with dual reflex bandpass. Subjectively I've found that dual-reflex designs sound 'sluggish' and 'out-of-step' with the mains. Whereas a single reflex design can be integrated without much fuss, as long as the levels, the response curve, and the timing is fairly accurate.
Has anyone heard a lab horn AND a tapped horn? Is a lab horn more 'musical?' |
|
|
| Patrick Bateman |
| quote: | Originally posted by cowanaudio
G'day again rick
This plot gives you an idea of what some acoustically absorbant material will do at the velocity maxima at 75Hz. Red is undamped. You reduce the dip to just 8dB, but lose 6dB of sensitivity in the process. My conclusion was that it is not worth doing. |
Thanks for posting graphs! I'm personally interested in using the "tapped horn" concept for the midranges in a small unity horn. So your experiments in damping and controlling the dip will save me some time :) |
|
|
| rick57 |
"tapped horn has a lot in common with dual reflex bandpass"
Agreed bandpass can be out of time, didn’t like the one I heard
I thought tapped horns are more like a Tapered quarter-wave pipe? |
|
|
| Patrick Bateman |
Both a dual reflex bandpass and a tapped horn utilize the rear wave to augment the overall response. A front loaded horn, like the lab horn, does not.
A transmission line uses the rear wave to augment the front wave, but the rear wave is significantly damped. So it's contribution to the overall response is much less dramatic.
In a tapped horn, it appears that the rear wave is just as prominent as the front wave, which is why the efficiency is so high. |
|
|
| paulspencer |
Patrick, it does concern me a little that GM wasn't impressed with the SQ, but not being there or seeing the setup and the source material, it's hard to make a secondhand judgement. What concerns me more is that there is a dip on the midbass that seems unavoidable, although Danley seems to have pulled it off.
Could it be that there is some special trick to do it that William didn't find? Or could it be some unobtainium driver?!
GM, wow! :bigeyes:
40kw x 40 drivers and 105 db @ 250m!
I wonder what they actually use it for!
One thing I've noticed is that the midbass dip actually moves higher as you reduce the length. One possibility would be to make two tapped horns of different tunings and attempt to get them to further balance out the other dips. I wonder if this could work.
Another idea - what if a second driver was used and placed higher up - it would be like a second tapped horn tuned higher.
William did you try anything like this and do you think it could prove an answer?
Ideally I'd like to get this thing to be useable up to 80 Hz.
Ok so here's the idea:
We have a 1.8m and a 2.4m tapped horn - this is what it looks like:
http://farm1.static.flickr.com/168/428888068_bc5de3b6cb.jpg?v=0
Notice that the upper bass dips are pretty well cancelled. |
|
|
| rick57 |
> it does concern me a little that GM wasn't impressed with the SQ
That was the opposite to WC’s more extensive experience . .
> I'd like to get this thing to be useable up to 80 Hz.
Me too originally, not rally now; what are you likely to want to integrate it with, that 15 Hz is significant?
OTOH, it was suggested to me that
“as there is no damping material, I would expect midrange sound waves from the back of the cone to reflect back from the untreated and parallel surface of the box onto the cone and be heard at the front of the cone.
At high volume levels one might experience standing waves.
Also, the driver being buried, as it were, inside the box would give rise to a beaming effect as the frequencies increase.”
Operating below 100 Hz, I think these are both non-issues, anyone think they are?
Cheers |
|
|
| cowanaudio |
G'day Paul
>Another idea - what if a second driver was used and placed higher up - it would be like a second tapped horn tuned higher.
>William did you try anything like this and do you think it could prove an answer?
We tried this with two drivers in the 30 Hz horn, and it didn't help. Here's a picture of one of the spaced driver test boxes, in the other the top driver was 1/2 way up the line. I'll have to look for the plots. They were nothing special.
Cheers
William Cowan |
|
|
| paulspencer |
Oops, here's the pic properly linked:
Beaming?!!!!!!!
This is a subwoofer! Now if we could get more directivity in its bandwidth, that could in fact be an advantage rather than a problem. Of course, dipole is the only way this is going to happen.
The other issue is also irrelevant to a subwoofer. This would be no different to any other subwoofer. Sound inside any box speaker will get out through the cone. In most speakers this is undesirable. In this one, the sound getting through the cone is a much lesser proportion. We are talking about a subwoofer here, which makes it pretty much irrelevant anyway.
| quote: | | In my personal experience, dual reflex bandpass is very VERY hard to get correct. And the tapped horn has a lot in common with dual reflex bandpass. Subjectively I've found that dual-reflex designs sound 'sluggish' and 'out-of-step' with the mains. Whereas a single reflex design can be integrated without much fuss, as long as the levels, the response curve, and the timing is fairly accurate. |
I don't think there is a lot of similarity with an 6th or 8th order bandpass. Tuning is based on line length like a TL, hence it should not be difficult to get the right tuning. I believe it isn't finicky in that way.
Some argue that undersized horns behave like bandpass more so than true horns. Still, from a design point of view I think a tapped horn is very different.
| quote: | | Has anyone heard a lab horn AND a tapped horn? Is a lab horn more 'musical?' |
I'd also like to know if anyone has ...
| quote: | | A transmission line uses the rear wave to augment the front wave, but the rear wave is significantly damped. So it's contribution to the overall response is much less dramatic. |
The damping is to tame the resonances and TLs are normally used for midrange as well. This damping should not reduce bass output unless overdone. As I recall from playing around with Martin Kings simulations, if the density isn't too high, the resonances will be damped without reducing bass reinforcement from the line.
| quote: | | In a tapped horn, it appears that the rear wave is just as prominent as the front wave, which is why the efficiency is so high. |
I believe it also allows it to behave as a true horn with a much smaller size by operating as a 1/2 wave horn one octave higher than its 1/4 wave operation which determines low frequency extension and output.
I think the reason for the upper/mid bass dip is the same reason you get this problem with a BLH. I wonder if I should actually be modelling these as a BLH instead.
William ....
Damn, it didn't work!!!!! :(
Ok but could that be because of two drivers in one box creating other problems? I'm curious what would happen if you were driving the offset driver with the bigger tapped horn both measured at the same time. My chart does suggest that if the tuning is different enough that the other peaks and dips should cancel.
One more thought. In the Danley version, does anyone know how the driver is actually mounted? It has an access panel in the middle - could it be that the driver is actually in there? It would seem a strange thing to do ... or does that panel serve another function? |
|
|
| GM |
| quote: | Originally posted by paulspencer
One thing I've noticed is that the midbass dip actually moves higher as you reduce the length. |
Greets!
Right, this is the pipe's acoustic 3rd harmonic, so as the acoustic path-length is shortened, the dip must rise in frequency and if I understand TD's white paper correctly, it's what gets filled in by the driver's anti-phase output, putting the first major dip at the 5th harmonic somewhere above a sub's passband.
GM |
|
|
| rick57 |
Originally posted by paul
One thing I've noticed is that the midbass dip actually moves higher as you reduce the length.
Originally posted by GM
as the acoustic path-length is shortened, the dip must rise in frequency and if I understand TD's white paper correctly, it's what gets filled in by the driver's anti-phase output, putting the first major dip at the 5th harmonic somewhere above a sub's passband.
So the trade-off is upper vs lower cutoff?
GM
IIRC you said the ToP is effectively a 6th order BP.
Would you also sat that with Patrick’s view that it’s *dual reflex bandpass; and that
“dual-reflex designs sound 'sluggish' and 'out-of-step' with the mains.
Whereas a single reflex design can be integrated without much fuss, as long as the levels, the response curve, and the timing is fairly accurate.”
Cheers |
|
|
| pinkmouse |
| quote: | Originally posted by paulspencer
One more thought. In the Danley version, does anyone know how the driver is actually mounted? It has an access panel in the middle - could it be that the driver is actually in there? |
I assume you have seen this thread? |
|
|
| GM |
Greets!
As TD states and shows in an excursion plot sim in his Tapped Horn WP, there's actually less gain at tuning than in a BR, only the gain due to pipe loading, so yes, you're 'robbing Peter to pay Paul', but it's a good one for a prosound sound app if the TP is tuned low since they're only really interested in the ~60 Hz - up BW and particularly in the ~80 Hz - up BW. For a typical music/HT app, the TP makes no sense to me since it's going to be XO'd <80 Hz and typical high excursion drivers combined with relatively cheap mega power in simple pipe horns can damage houses at low distortion levels, not to mention these can easily be ~accurately simmed using any TL or BLH software.
WRT the TOP's SQ, as I noted, until proven otherwise, I believe it was all the other horn loaded cabs in close proximity 'singing along'. WRT PB's opinion, I've auditioned 8th order BPs (Karlson K-15) that if BW limited and combined with mids/HF compression horns (Altec 805 multi-cell/288 driver in this case) was heart attack 'fast' enough to make you believe you were standing next to a Nick Mason drum solo, so IMO it's dependent more on the design/implementation than the number of orders of resonance.
As always though, YMMV.
GM |
|
|
| pinkmouse |
| quote: | Originally posted by GM
... was heart attack 'fast' enough to make you believe you were standing next to a Nick Mason drum solo... |
Been there, had the t-shirt, (last seen polishing the car). He's actually fairly quiet as drummers go... ;) |
|
|
| GM |
Greets!
True, but I can't remember the names of any of the heavy metal band's drummers. :( Regardless, it takes a pretty serious wide BW HE system to reproduce any drum set solo at ~accurately at ~live levels.
GM |
|
|
| Ray Collins |
Try Joe Morello.
Ray |
|
|
| pinkmouse |
| Nah, he's a wuss. Now this guy is loud... :) |
|
|
| Ray Collins |
Agreed.....I was going for precision rather than volume.
Ace of Spades...this guy is a beast!!! |
|
|
| rick57 |
GM
Seems to me for music like Nick Mason etc, 80 - 400 Hz is most critical.
I’ve considered and dropped Onkens, am deferring bass horns till I have more time & skill, and thinking of a simple vented mid bass 15”. I have a pair of JBL 2035HPL’s with Fs 43 Hz, 7 mm Xmax, low Le of 0.25, BL of 17 etc. (I don’t seek to get to live levels, say 105-110 dB will do) ~ not Altecs, but within my budget.
Other than horns, in your opinion would a Karlson or some other rig do transients/ drums any better than a good driver in a good vented box?
Cheers |
|
|
| GM |
Greets!
Freddy and a few others don't mind XOing Ks as high as 500 Hz, but going from a near 50 yr distant memory, I do, so something like what RCA-fan calls a 1/2 horn seems a reasonable solution, though I prefer the better damped ~max flat impedance TL variant.
Got the rest of the HPL's specs?
GM |
|
|
| rick57 |
Hi GM
I thought a concern of all or most Ks, is a peak at c. 200 Hz, often about 2 dB. Guess it might be notched out, but maybe it’s only the peak giving an illusion of better transient response?
A constraint, and a workaround: I don't want a box bigger than about 70 L (2.5 cubic ft), *but I will use an active “sub” underneath which goes deep, crossing about 60 – 80 Hz
The tapped horn or a TL would be an alternative to the sub.
I apologize for digressing to the next ~3 octaves above what a TH can cover.
The HPL's specs are:
Fs 43
Re 4.7
Qms 4.9
Qes 0.36
Sd 880 cm2
Vas 192 L (6.8 cubic ft)
Xmax 7.0 mm (0.28”)
Le 0.25
With the low Le, I had one steam of thought that the 2035 could be used even to say 1200 Hz without too much distortion(?), avoiding the need for a mid driver. The idea of avoiding a mid driver is one less crossover, less induction losses, and allow a sweet low powered amp (Gordon Rankin’s 2 watt 45 SE Bugle).
Now beginning to think that’s the horse driving the cart: I should use a mid (have some good candidates of similar efficiency) and only need to take the mid bass up to eg 300-400 Hz, ie the piston range of a 15".
Rather than depth, I was asking after optimizing midbass –lower mid transient response & impact. Maybe within those parameters it comes down to the quality of the driver and the box construction?
Thanks
|
|
|
|
