Ryan,
There is a great resource maintained here dedicated to the Karlson speaker: The Karlson Homepage
In the theory section, it says to make the cut up to 2/3rds the length of the TL and make the profile an exponential. I think a hand drawn smoothly curving arc that goes from a tangent slope of a few deg at the cusp to maybe 30 deg at the end (terminus lip) should be good. The point is to not have a distinct boundary of where the end length is - to fool the quarter wave into not attaching at a precise location, thus smoothing out the peaks and dips. So for a 60 in line, the cut goes from the end to 40 in down the pipe.
There is a great resource maintained here dedicated to the Karlson speaker: The Karlson Homepage
In the theory section, it says to make the cut up to 2/3rds the length of the TL and make the profile an exponential. I think a hand drawn smoothly curving arc that goes from a tangent slope of a few deg at the cusp to maybe 30 deg at the end (terminus lip) should be good. The point is to not have a distinct boundary of where the end length is - to fool the quarter wave into not attaching at a precise location, thus smoothing out the peaks and dips. So for a 60 in line, the cut goes from the end to 40 in down the pipe.
You can just use a half-ellipse whose minor axis is the pipe's circumference and major axis it length. Short unslotted sections at the start can be used as well. For small 1" K-Tubes, I draw them in illustration programs, print and cut the pattern out. I wrap it around the tube, mark the edge and cut it out with a Dremel tool.
IG
IG
one of the most successful ways imo is to follow Karlson's X15 system (appeared first in 1965) where the slotted waveguide tweeter is mounted inside the bass coupler
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
I have never played with such large K-Tubes and have not seen many measurements for them, but smaller K-Tubes don't seem to truly have a tuning as much as they are acting as an anti-horn of sorts. I think they do act as impedance transformers and do it quite cleanly when looking at impedance charts, one hallmark of a great waveguide. I worked out some maths a while ago to figure out the effective lenght of a regular semi-elliptical slotted tube. The driver is likely close to unloaded, if not so, by the time ¼-WL of its absolute lenght is reached. Anecdotal evidence, though based on K-Tube use by the team who developed it, suggest it can probably be used that low though.
IG
IG81 So if I used the longer pipe more for visual effect and its waveguide properties and tuned the lower enclosure the driver is in would the be more effective for getting the lower frequencies? (fyi I'm very new to all these enclosure types and just brainstorming if you can call it that.)
Also what would make a good speaker it terms of specs to try?
Bravox CX50CF Carbon Fiber Series 5-1/4" 2-Way Coaxial Speak 269-508 I found these in the clearance bin.
I could use the driver specs and make the lower enclosure a BR tuned low and then use approx a 60" x 4" tube with a 40" taper slot.
Or should I use a true full range single driver?
Also what would make a good speaker it terms of specs to try?
Bravox CX50CF Carbon Fiber Series 5-1/4" 2-Way Coaxial Speak 269-508 I found these in the clearance bin.
I could use the driver specs and make the lower enclosure a BR tuned low and then use approx a 60" x 4" tube with a 40" taper slot.
Or should I use a true full range single driver?
I think I'd keep the pipe length around 20-24" initially in this case and use the bottom enclosure for tuning the low-end. A weak-motor driver in a sealed box with Qtc~1.0 might be cheerful. A longer pipe might make for a tad more upper-bass if needed. Whether using a fullrange or coaxial, you probably want a tilt towards more HF output, a bright-sounding driver IOW. You are already listening off-axis and there will probably be some losses above 3-4kHz or so in a 4" pipe.
IG
This is a good thread. I have been very interested in the Karlson aperture speakers for some time now and I think I have a way to model this in AkAbak. I can approximate the K-slot as a series of ports (vents) in successive sections with the vent CSA increasing exponentially. I will see how this goes and report back once I get something working...
This modeling of the Karlson is taking longer than I thought. Making progress but there is a lot of thought that I have to put into how to generalize the geometry so that I can use this bit of code on other projects. Don Hills confirmed that my approach should be ok. AkAbak has max 56 nodes per system so I should be ok.
xrk971, I take it you are trying to model a K-Tube type application and not the Karlson bass enclosure, is that correct? I don't know much about the large tubes as built in this here thread, but if you can simulate a generic 1" compression driver along with a 1" diameter, 5.3" long, fully-slotted K-Tube, checking the impedance plot might be indicative that your K-slot definition is good or not. This 1" tube will only exhibit the two expected larger impedance peaks towards the lower-end of the pass-band and no other peaks or anomalies really.
Our late K-buddy Gregg Baker had an Akabak model for the bass enclosure, but I forget what he did to define the slot, it might not even have been fully worked out yet.
As far as the region of performance dominated by T/S parameters goes, a simple band-pass model equivalent to the exact K-enclosure is close enough in some cases, but the K-slot will start to have some unpredictable effect starting at a few 100's Hz.
IG
Our late K-buddy Gregg Baker had an Akabak model for the bass enclosure, but I forget what he did to define the slot, it might not even have been fully worked out yet.
As far as the region of performance dominated by T/S parameters goes, a simple band-pass model equivalent to the exact K-enclosure is close enough in some cases, but the K-slot will start to have some unpredictable effect starting at a few 100's Hz.
IG
my laptop with the archive of Gregg's forum has crashed - I'm not sure if he went far for the slot - - he did try to roughly simulate the tilted baffle (typically 23->30 degrees back from vertical)
Carl has a lot of math on Ks but not shared it. I'm weak with math.
my experience with large tubes is very limited
here is a rough comparison of a six inch ID tube 28 inches long to a 150Hz tractrix horn whose mouth area is considerably larger than the slender tube
the rear chamber for the tube was probably too small - the tube had a rather restrictive exponential slot (that's why it ended in a point) - highest frequencies tend imo to ride down the long axis - lower can emit sideways from the slot
I had this tube tilted - can't remember how high - maybe 20 degrees above horizontal - ? this same tube made a nice sounding backload for a Fostex 164.
Robert Reams who has the Karlson-Hypex paper used a dual slot tube he called a "clothespin" - I think it was meant to run from an EV12L
Carl has a lot of math on Ks but not shared it. I'm weak with math.
my experience with large tubes is very limited
here is a rough comparison of a six inch ID tube 28 inches long to a 150Hz tractrix horn whose mouth area is considerably larger than the slender tube
the rear chamber for the tube was probably too small - the tube had a rather restrictive exponential slot (that's why it ended in a point) - highest frequencies tend imo to ride down the long axis - lower can emit sideways from the slot
I had this tube tilted - can't remember how high - maybe 20 degrees above horizontal - ? this same tube made a nice sounding backload for a Fostex 164.
Robert Reams who has the Karlson-Hypex paper used a dual slot tube he called a "clothespin" - I think it was meant to run from an EV12L
An externally hosted image should be here but it was not working when we last tested it.
Bose owns a patent on a K-tube but gives no reference to K
Patent US8351630 - Passive directional acoustical radiating - Google Patents
Patent US8351630 - Passive directional acoustical radiating - Google Patents
Rip-off. :/
I think the K-slot at its purest is that used on the K-Tubes. It's probably one of the best options to make such a device work. The Bose patent does not show it in the drawings, but vaguely mentions something that possibly covers it in claim 10:
10. An acoustic apparatus in accordance with claim 1, wherein the width of the opening varies along the length of the pipe.
Many of the other described embodiments have already been tried by DIY'ers over time.
IG
This could go in at least three threads. Driver is a Visaton FRS-8. Cabinet is 5mm foamcore. I used a spine to locate the front panel and brace the driver.
Cheers Steve
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
These replace the Spherical IKEA salad bowl speakers for my computer. An externally hosted image should be here but it was not working when we last tested it.
First impressions? Definitely bouncier, and smaller, so a bit of desk real estate is reclaimed.Cheers Steve
There are two slot vents at the top, with an internal shelf. What we have here is a 3/4 scale model of a K4 enclosure from plans posted here, a K3 if you wish. I had the drivers, had the board, and wondered if I could make a Karlson type enclosure. It certainly seems to throw the sound at you - surprisingly good at classical orchestral music.
Cheers Steve
Cheers Steve
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.