Hi Djim,
The opening that you proposed in Post #482 of your link http://www.diyaudio.com/forums/subwoofers/190635-th-18-flat-35hz-xoc1s-design-49.html is offset completely towards the S1 side of the driver location. The photo of the B52 posted by djk Post #488 is not. The B52 opening looks approximately centered on the driver.
The FEA picture from that same thread, Post #480 is quite interesting, but does it really reflect the behaviour of a pressure wave in the sub 100Hz range?
How can one come to any conclusions without some very extensive measurements?
Regards,
The opening that you proposed in Post #482 of your link http://www.diyaudio.com/forums/subwoofers/190635-th-18-flat-35hz-xoc1s-design-49.html is offset completely towards the S1 side of the driver location. The photo of the B52 posted by djk Post #488 is not. The B52 opening looks approximately centered on the driver.
The FEA picture from that same thread, Post #480 is quite interesting, but does it really reflect the behaviour of a pressure wave in the sub 100Hz range?
How can one come to any conclusions without some very extensive measurements?
Regards,
Hi Oliver,
We seem to have a different explanation about the changing position of the high pressure zone since in my view it will not be transferred to S1. In my view it works in three ways;
1.) Transferring the high pressure zone from the left side of the cone back to S2 (and not S1), where it suppose to be.
2.) Minimising the differences in distance from any position of the cone surface to S2 (that's why it is wave shaped).
3.) To equalise the pressure differences within the cone (at what degree I don't know yet)
While the driver side (right side) near S1 in the original concept has the lowest pressure, with my suggestion the narrowing of the left side of the driver is similar. Both sides still have a difference in pressure but the difference should become smaller.
I agree that actual measuring of the real thing will tell, till what degree this system works and if it is sufficient enough. Danley’s "V" shape method isn’t perfect either although I do like the fact the distance between S2 and S1 is enlarged. Driver positions that are located more downwards in the horn/pipe do have less difference in loading between the two sides, left and right side of the cone and front and backside of the cone. Therefore I would use such position especially for drivers with low BL forces and less sturdy cone/suspension systems.
Personally I believe that symmetric centre loads can give the best results if worked out properly (with cone volume make up) and the right driver characteristics. Another advantage of the symmetric centre load is you can use the long frontal path to create a resonance in the area of the bandpass where Dynamic Compression is located. With the horn expanders I presented elsewhere you can also do such thing and is based on normal 1/4WL behaviour.
In my view FEA modelling can give better inside in locating pressure zones, although I have questions with the model you are referring too. As I already explained I don’t expect the highest pressure point to be somewhere inside the horn path, where the horn/pipe is already expanding! I am also not sure how it is modelled but it looks like it uses constant fluid streams. Sound waves are build up by air molecules that collide against each other and pass there energy instead of streaming out the horn in a constant direction.
We seem to have a different explanation about the changing position of the high pressure zone since in my view it will not be transferred to S1. In my view it works in three ways;
1.) Transferring the high pressure zone from the left side of the cone back to S2 (and not S1), where it suppose to be.
2.) Minimising the differences in distance from any position of the cone surface to S2 (that's why it is wave shaped).
3.) To equalise the pressure differences within the cone (at what degree I don't know yet)
While the driver side (right side) near S1 in the original concept has the lowest pressure, with my suggestion the narrowing of the left side of the driver is similar. Both sides still have a difference in pressure but the difference should become smaller.
I agree that actual measuring of the real thing will tell, till what degree this system works and if it is sufficient enough. Danley’s "V" shape method isn’t perfect either although I do like the fact the distance between S2 and S1 is enlarged. Driver positions that are located more downwards in the horn/pipe do have less difference in loading between the two sides, left and right side of the cone and front and backside of the cone. Therefore I would use such position especially for drivers with low BL forces and less sturdy cone/suspension systems.
Personally I believe that symmetric centre loads can give the best results if worked out properly (with cone volume make up) and the right driver characteristics. Another advantage of the symmetric centre load is you can use the long frontal path to create a resonance in the area of the bandpass where Dynamic Compression is located. With the horn expanders I presented elsewhere you can also do such thing and is based on normal 1/4WL behaviour.
In my view FEA modelling can give better inside in locating pressure zones, although I have questions with the model you are referring too. As I already explained I don’t expect the highest pressure point to be somewhere inside the horn path, where the horn/pipe is already expanding! I am also not sure how it is modelled but it looks like it uses constant fluid streams. Sound waves are build up by air molecules that collide against each other and pass there energy instead of streaming out the horn in a constant direction.
I quite agree that it will be impossible to come to any conclusions without measurement.
The 6 fold design was firstly tweeked as your suggestion, with the extra ramp and the small flare at the mouth.
The discussion on the compression plates started with PASC's findings with the TH18 design. The TH18 seems fine pushed to 1KW plus but ultimately does not seem to keep good control over the cone when pushed into the 2KW plus power levels that are possible with the best 18" drivers.
The 2:1 compression plate is an established way of acheiving more control of the speaker cone.
The drawing of the compression plate was done so that it could be incorporated into Crescendos 6 fold test box if wanted. Unfortunately after setting it out, it it became apparent that it would be difficult to retofit the compression plate once the baffle had been cut. It would need to be fitted to the test box when bullt.
Ultimately to prove if it has any real worth it would need to be cut out of the test box and the cabinet re-tested with a fully cut out baffle.
I think that the 2:1 compression plate could have a positive effect on the 6 fold design. The 6 fold has a parallel first section that makes it slightly unconventional. Any ramps added to the first section reduce the internal volume of the speaker. This is unlike the TH18 & SS15 fold, where the right shape ramp might make it possible to actualy increase the overall path length of the speaker!
As to the FEA sim, this shows the flow thru the speaker of constant pressure flow. As such its frequency is DC.
What the Sim demonstrates that the S2 point where Hornresp reports the compression ratio is not strictly true. There are 2 points to mention.
Firstly that at the S2 point, half the speaker driver is past that point in the horn. To my mind that means that at the actual S2 point, on an offset style of TH horn, the compression is reduced by 50%. The full compression point in the horn is only reached at the edge of the driver - As such the compression ratio reported in Hornresp still has relevance, it's just that the actual point in the horn where the where the full compression exists is offset!
The second point to mention is that the pressure flow across the speaker throat does not just exist above the baffle plate. The baffle cut out AND the shape of the speaker cone are relevant, and make a difference to the cross sectional area of the throat. This reduces the supposed compression ratio at the S2 point even more
The 2:1 compression plate adresses these points. Who knows, it might mean a design that delivers everything that it promises in a sim!
Regards Martin
Hi Djim and Xoc1,
I can see where the idea for the B52 throat opening comes from. And I have no problem with the compression ratio.
The following only applies to this particular type throat chamber and horn coupling:
I just want say that having the throat opening offset completely to one side does not seem right to me. It also does not provide for "minimising the differences in distance from any position of the cone surface to S2". It in effect creates a new position for S2 somewhere in the middle of the opening, and the pressure distribution-and particle flow-inside the throat chamber and out off the throat chamber cannot be anything like evenly distributed(?).
I'm just guessing, but I'd say keep the alignment of the throat opening centered with the driver no matter what shape the opening takes.
Regards,
I can see where the idea for the B52 throat opening comes from. And I have no problem with the compression ratio.
The following only applies to this particular type throat chamber and horn coupling:
I just want say that having the throat opening offset completely to one side does not seem right to me. It also does not provide for "minimising the differences in distance from any position of the cone surface to S2". It in effect creates a new position for S2 somewhere in the middle of the opening, and the pressure distribution-and particle flow-inside the throat chamber and out off the throat chamber cannot be anything like evenly distributed(?).
I'm just guessing, but I'd say keep the alignment of the throat opening centered with the driver no matter what shape the opening takes.
Regards,
Last edited:
I take it ramps are the only solution, aside from redesigning the box, to change the first section from being parallel. What type of 'problems' do you expect one to see/hear from this?
How many of you glue your test boxes? I was thinking of using gasket tape on the mating surfaces and screwing them together. This way, I can take a side off and replace the compression plate baffle with an open baffle. How bout glueing all but the baffle and the one side? Might be quicker, but would prevent any other easy box changes..
(on second thought, glueing probably won't be much, if any, quicker)
How many of you glue your test boxes? I was thinking of using gasket tape on the mating surfaces and screwing them together. This way, I can take a side off and replace the compression plate baffle with an open baffle. How bout glueing all but the baffle and the one side? Might be quicker, but would prevent any other easy box changes..
(on second thought, glueing probably won't be much, if any, quicker)
Last edited:
Hi Oliver,
Your guising makes me think again and there might be a solution. If the line of the wave edge uses stronger bended curves (bigger waves) it might enhance the impedance over a wider area cq frequency.
I have another alternative but I have to draw it in paint because with my English nobody would understand it ;-)
(Ps Justin I use rubber tape for testing and it works, you can measure impedance to see if anything strange occurs within the plot that shows evidence of leaking)
Your guising makes me think again and there might be a solution. If the line of the wave edge uses stronger bended curves (bigger waves) it might enhance the impedance over a wider area cq frequency.
I have another alternative but I have to draw it in paint because with my English nobody would understand it ;-)
(Ps Justin I use rubber tape for testing and it works, you can measure impedance
to see if anything strange occurs within the plot that shows evidence of leaking)The following example uses a wider impedance throat similar to the function of the "V"shape of Danley. Meanwhile it still keeps the baseline more towards S2 in order to move the high pressure zone to S2.
The best function is achieved when the top opening (light blue distance) doesn't exceed the diameter of the VC. Of course the baseline of the throat opening can be made wave shape for similar function as suggested before.
The best function is achieved when the top opening (light blue distance) doesn't exceed the diameter of the VC. Of course the baseline of the throat opening can be made wave shape for similar function as suggested before.
Hi Justin,
That tool seem to measure impedance only at 1Khz. For less money you can buy the Dayton Audio WT3P
I've got 3 panels cut - 2 of which I'll need to trim due to the top board being the full depth of the cab after all. I'm planning on cutting more tomorrow (technically, today). What's your progress looking like?
I'm reading several posts elsewhere regarding the dbx RTA-M mic I have, not being uber accurate (expected at its price point, I suppose). I'm not sure how 'off' my measurements will be. This has me pondering an SPL meter purchase..
I have everything ready just waiting for Xoc1 to arrange the panels on the 6 fold plan , then I will start building.
Last edited:
The dbx RTA-M mic should be reasonably accurate in the range of the cabinet.
I have an RTA-420, IIRC has the same element, it measures a few dB less at 30 Hz than it should, compared to my B&K 4004 mics.
An SPL meter with better LF response will cost big $$.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.