I have a pair of midrange paging drivers that will allegedly do 100 Hz on the low end (yeah, right).
1 3/8 X 18 Thread 40W Rms Midrange Driver Threaded | 54-050 (54050) | Distributed By MCM
Pondering if there's any potential used as a midrange K-tube with some PVC tube stuck on the end.
1 3/8 X 18 Thread 40W Rms Midrange Driver Threaded | 54-050 (54050) | Distributed By MCM
Pondering if there's any potential used as a midrange K-tube with some PVC tube stuck on the end.
you may get a surprise and find they can play that low the question will be how loud? they can't push a lot of air given the diaphragm size. I have found that a 1/4 wave line is all you need. Pick either the given Fs or look at the peak on the impedance graph if one is available. I like to use the double cutaway K-Tube which looks rather like a bishop's hat see the one on the left side. Best regards Moray James.
Attachments
That is a really interesting driver - how does it sound and does it really reach 100 Hz? You can so a quick and dirty test with just stiff card stock.
Is that a copy of the Atlas Sound driver purported to be "just like" a WE555?
Assuming a fairly standard profile, that is, a single half-elliptical cut in the tube, the effective length is roughly 90% that of the slotted length. An initial un-slotted section changes things, but I have not studied that much myself. I base this upon my measurement of BMS 4550 on a Transylvania Power Company "The Tube". It is 5.3" long and slotted on its full length. Versus the "naked" driver, i.e. no horn/wg, there is a reduction of distortion starting at 700Hz. This suggests the need for a 37" or so tube to work down to 100Hz. The proportions will then be much different from my tube that my assumptions might not hold up though.
The Faerber Equilibrium speaker uses a super-long unslotted section, I'd sure like to see raw SPL and impedance measurements for that one. Freddi experimented with something similar.
Assuming a fairly standard profile, that is, a single half-elliptical cut in the tube, the effective length is roughly 90% that of the slotted length. An initial un-slotted section changes things, but I have not studied that much myself. I base this upon my measurement of BMS 4550 on a Transylvania Power Company "The Tube". It is 5.3" long and slotted on its full length. Versus the "naked" driver, i.e. no horn/wg, there is a reduction of distortion starting at 700Hz. This suggests the need for a 37" or so tube to work down to 100Hz. The proportions will then be much different from my tube that my assumptions might not hold up though.
The Faerber Equilibrium speaker uses a super-long unslotted section, I'd sure like to see raw SPL and impedance measurements for that one. Freddi experimented with something similar.
Quite possible - I'm not all that versed in WE stuff, but remember reading something to that effect some time back. I think the diaphragm and/or phase plug design was supposed to be almost unchanged.
GM, how would you analyze the following K-Tube measurement? This is the Transylvania 1" tube, 5.3" long, fully-slotted, on a BMS 4550 1" driver. Mic was "nearfield", which probably means a few inches away IIRC.
Blue Trace - "naked" driver and its distortion
Red Trace - loaded with The tube and its distortion
An externally hosted image should be here but it was not working when we last tested it.
We can see distortion is reduced above 700Hz, up to ~1600Hz. Output is stronger below 1600Hz down to driver cutoff. The 2kHz dip and distortion peak may be throat mismatch/reflection? Response is often smoothed-out by the K-Tube in the upper registers.
Here is Emilar EC-175:
Blue Trace - "naked"
Red Trace - loaded with The Tube
An externally hosted image should be here but it was not working when we last tested it.
For some reason, the 2kHz dip is worse on the Emilar. I suspect its exit angle to be greater the that of the BMS, which is 14°. The BMS does not exhibit as strong a dip in the far-field, but the Emilar still does.
Last edited:
This is the crux of the matter and since a 1" tube is 'done' at around 4316 Hz, it realistically will be 'done' ~3 octaves away or ~540 Hz, so figure only usable from ~1080 Hz, just as its response indicates. Increasing the tube's diameter to 2" will help down low at the expense of the HF output, but now power handling has been severely compromised without increasing the driver's Fs.
No actual experience with loading any kind of tubes with compression drivers though, so curious how this pans out.
GM
Don’t know, Altec-Lansing didn’t carry over any part of the 722 design AFAIK, only the Lansing drivers and no clue who might have cloned [parts of] WE designs after the patents ran out.
Well, the dip is at the tube’s 3rd harmonic, so apparently the K-slot only lowers its Q judging by how broadband it is on both, which makes sense to me. Not knowing the design details, PWT responses or any experience with either driver, I’ve no clue why there’s such a big difference other than the 4550 being rated for an 800 Hz XO and the Emilar’s 500 Hz.
GM
That's @kenpeter 's old (and much despised) clothespin. IIRC, I "might" have that actual piece somewhere in the garage. That 2" diameter waveguide is tuned higher than I want for this experiment and from listening to it many years ago, it wasn't that good sounding on an eminence PSD 2002 compression driver. That said, I'm aiming to go lower for this experiment (maybe as low as 300 - 500 Hz). Back then, we did have much better luck experimenting with K-tubes using some smaller vintage alnico compression drivers (which I should still have, hiding somewhere).
I'll try and generate impedance and response graphs when I get going with this project.
Last edited:
Thanks GM!
If you guys can believe it, the guys who made the Transylvania tube apparently recommended nothing less than a 500Hz active crossover using JBL or Altec 1" drivers. We have this information on a 2nd or 3rd hand account, so take it FWIW. I can't see how it'd be sane to do so and I'm not even talking about pro use here.
If you guys can believe it, the guys who made the Transylvania tube apparently recommended nothing less than a 500Hz active crossover using JBL or Altec 1" drivers. We have this information on a 2nd or 3rd hand account, so take it FWIW.
I can't see how it'd be sane to do so and I'm not even talking about pro use here.
Here is the Transylvania Tube and next to it, a brass tube I made, which is essentially the same, but with a ~1" or so unslotted "stub" section.
Driver is BMS 4550 again. Mic was probably a foot or two away, at a certain angle down the slot, judging from the rolled-off HF. The brass tube apparently introduces a notch, which I can only attribute to the plain stub at the base. This also allowed support for greater LF output
An externally hosted image should be here but it was not working when we last tested it.
Driver is BMS 4550 again. Mic was probably a foot or two away, at a certain angle down the slot, judging from the rolled-off HF. The brass tube apparently introduces a notch, which I can only attribute to the plain stub at the base. This also allowed support for greater LF output
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Any slot fabrication hints that you can offer?
I was thinking of transferring a paper template and using a dremel cutting wheel to define the rough profile and a dremel grinding wheel or sanding disc to finish things up.
You're welcome!
If anything, the driver needs more loading to flatten it out based on its published response, so a 'fast' adapter to a longer 1/2" dia. pipe might be worth experimenting with. This increases the pipe's HF cutoff a full octave higher and if ~25.1" long [1 WL of 540 Hz], it may still be usable to ~1080 Hz if it damps the ‘dip’ good enough without rolling off the top end too much.
GM
Pretty much what I did. I drew my half-ellipse in Illustrator and printed, cut with Xacto knife, wrapped around the tube, transferred the slot area to be cut with a black marker. I did the rough cut with small cut-off wheels and mostly filed the rest by hand, since brass is pretty soft.
http://sd-1.archive-host.com/membres/up/78070497537767987/Acousticaltube.pdf
Here's analysis and measurement of a BMS 4540ND mounted on three types of tube: 14cm straight, 14cm elliptic profile and 7cm K-coupler. I find they short-changed the K-Tube as, not only is it shorter than the other two, but the K-slot itself covers ~2/3 of the already short tube. I can't see it having much effect, save on the shortest wavelengths, which are probably affected by the tube's diameter at this point. They were not not out to prove which one is "better" obviously, it seems like an objective study, but I just don't think they worked with a correct implementation of the K-slotted tube.
Here's analysis and measurement of a BMS 4540ND mounted on three types of tube: 14cm straight, 14cm elliptic profile and 7cm K-coupler. I find they short-changed the K-Tube as, not only is it shorter than the other two, but the K-slot itself covers ~2/3 of the already short tube. I can't see it having much effect, save on the shortest wavelengths, which are probably affected by the tube's diameter at this point. They were not not out to prove which one is "better" obviously, it seems like an objective study, but I just don't think they worked with a correct implementation of the K-slotted tube.
I wonder why they picked a shorter K-tube. Either there's something special about the shorter aspect ratio or they were out to rig the results. I can't believe it was gross oversight.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.