Softening edges is a necessity IMHO, 3/4 inch radius is the least I now do, but if you can only do 1/2 inch, it is far better than nothing.
You are fighting a round baffle. That is the absolutely worst mechanical arrangement for a baffle with regard to ripple in the baffle step. As I said, if you are building an integrated system so you can apply eq where needed, you can probably do OK. Just expect it. As an example: assuming a 130 mm round baffle with an 80mm driver in it. Expect a 2 dB hump at about 2100 Hz and a 2 dB dip at 500 with about a 6 dB step with pole-zeros at about 250 and 450Hz. It will make the speaker " shout" in a most unpleasant way. This is just from a model, so the reality in measurement will be different. Using a full range, or more correctly a wide range, you will have breakup modes to deal with. This can be nasty. Nasty enough I abandoned my FR efforts and went back to tweeters.
Part of engineering is problem solving. Totally agree. Part of engineering is also making sound basis decisions to minimize the problems that have to be fixed later.
I assume you have read Olson.
You are fighting a round baffle. That is the absolutely worst mechanical arrangement for a baffle with regard to ripple in the baffle step. As I said, if you are building an integrated system so you can apply eq where needed, you can probably do OK. Just expect it. As an example: assuming a 130 mm round baffle with an 80mm driver in it. Expect a 2 dB hump at about 2100 Hz and a 2 dB dip at 500 with about a 6 dB step with pole-zeros at about 250 and 450Hz. It will make the speaker " shout" in a most unpleasant way. This is just from a model, so the reality in measurement will be different. Using a full range, or more correctly a wide range, you will have breakup modes to deal with. This can be nasty. Nasty enough I abandoned my FR efforts and went back to tweeters.
Part of engineering is problem solving. Totally agree. Part of engineering is also making sound basis decisions to minimize the problems that have to be fixed later.
I assume you have read Olson.
My understanding is that curved/round speaker shapes get around the inherent structural and acoustic problems of box speakers. A sphere is the ideal exterior shape. Internally, a sphere is not ideal acoustically, so kind of by happy accident and because its a floor standing design, I've ended up with a teardrop shaped top section with no real baffle per se, and a 'tail' like stand which will absorb standing waves. The top diameter where the driver sits is 93mm across and sharp edges in the render will be softened.
If I can recommend a great FR, try a pair of Omnes 3.01; inexpensive (30euros each), smooth and detailed treble with no breakup issues imo.
If I can recommend a great FR, try a pair of Omnes 3.01; inexpensive (30euros each), smooth and detailed treble with no breakup issues imo.
Last edited:
TVRgeek
Considering my decades long use of technical papers from the organizations you cite as well as many others ranging from consumer to international telecom to military as part of my job function as an electronics and mechanical designer, your attempt to condescend to me on a technical basis is poorly founded.
It may be a good idea for you to take up any issues you have with KEF's technical rigor in their publications with them before attempting to disparage their ACE technology by inference on that basis.
You have in fact presented nothing whatsoever that mitigates against the efficacy of this approach of enclosure compliance enhancement. Until you bring something to the table technically, it will be understood that you are merely expressing a casual personal opinion.
Considering my decades long use of technical papers from the organizations you cite as well as many others ranging from consumer to international telecom to military as part of my job function as an electronics and mechanical designer, your attempt to condescend to me on a technical basis is poorly founded.
It may be a good idea for you to take up any issues you have with KEF's technical rigor in their publications with them before attempting to disparage their ACE technology by inference on that basis.
You have in fact presented nothing whatsoever that mitigates against the efficacy of this approach of enclosure compliance enhancement. Until you bring something to the table technically, it will be understood that you are merely expressing a casual personal opinion.
Last edited:
Tested fish tank activated carbon. It raised Fs slightly (2 Hz) and only lowered Qts by about .05. Hump down by about half a dB. Subjective, but I think it did not do as well on controlling reflections as wool. I assume it is coal based, not coconut as it did not say. If it is the micro pores, I am not impressed. $35 going into the flower bed. It is going to cost about $150 to get a pail of the clean coconut delivered, so I may not bother without much better evidence. Right now, I'll still choose glass for big sealed subs and wool for mid-bass.
Bill, back on subject. You have no extended baffle, but you still have a baffle even if it no more than the mounting flange. Measurement will show what you will need to deal with on the eq. You can get a hint if you download The Edge and spend 5 minutes looking at the simulation. Only a sim; reality will be a bit off. The good news is the larger the radius, the less of a problem. When the radius is large enough, it becomes an effective sphere which is the smoothest of all possible shapes. As you are doing a FR, you won't have any problem with the BSC overlapping the crossover.
I don't know what problems you think rectangular boxes have a curved box solves. A curved box may be easier to push the resonances up higher where they are less problematic. No magic, just different. Should be fine. Square boxes do wind up looking like just square boxes which does not meet your design goal.
I have not heard of the FR you mention. Got a link? Always looking for a well behaved driver. My efforts with a Fostex and a Fountek were not very happy. You say no breakup issues. I think you mean no serious issues that may need eq, as all drivers have breakup. Physics. If it is high enough, and low enough Q, it may not be a problem.
I don't know what problems you think rectangular boxes have a curved box solves. A curved box may be easier to push the resonances up higher where they are less problematic. No magic, just different. Should be fine. Square boxes do wind up looking like just square boxes which does not meet your design goal.
I have not heard of the FR you mention. Got a link? Always looking for a well behaved driver. My efforts with a Fostex and a Fountek were not very happy. You say no breakup issues. I think you mean no serious issues that may need eq, as all drivers have breakup. Physics. If it is high enough, and low enough Q, it may not be a problem.
Ah, Tracked it down. I don't believe the frequency response for a second. Not from a 3 inch with a phase plug. 8 or 9K maybe, which may suit your goals fine. It looks like it could be a neo magnet which would be a good thing. I am betting on breakup issues below 8K and some of that may be what makes it look like it has some top end. Typical for a FR. May be a good selection for your goals. Only a prototype will tell.
Even at 220 for a crossover, that is a small driver to fill a room.
Even at 220 for a crossover, that is a small driver to fill a room.
Hi -
Did you take any measures to prevent it from adsorbing any atmospheric moisture? According to the KEF paper, that is important, since from other readings (not just the KEF paper), it appears that coconut charcoal will absorb water vapor preferentially to gases such as oxygen and nitrogen.
For my Iron Lawbreakers, I added conventional packets of silica desiccant to the coconut charcoal (almost certainly not as effective as molecular sieve) and sealed the coconut charcoal in their own airtight plastic bags. This brought down the BR impedance peak to about 32 ohms whereas without the charcoal it would have been about 60 ohms. There was also about a 50% fiberglass fill in the remaining volume behind the 15" bass driver (JBL 2220A) primarily to minimize reflections from the back of the cabinet through the cone. This is BR loaded at ~32 hz in a 100 liter cabinet and gives a flat response to ~50 hz with useable response down to Fb when placed reasonably close to a back wall away from corners with a speaker sensitivity of 99-100db w/m (16 ohm). I've played with these woofers before and never have gotten even close to satisfactory bass extension from them in a reasonable sized cabinet before trying this approach (the 2220A's are 101db sensitivity, Fs=37hz, Qts=0.17). A UREI 811 Altec 604 based single woofer monitor in a 5 cu ft box is rated to within +/- 3db to 80 hz in free space with 4-5 db lower sensitivity (based on UREI's 1V/1m at a nominal 8 ohm rating).
For my HT speakers, I found a product called 'molecular sieve' made by CR Laurence Co. that is commonly used in insulated window units to prevent humidity from leaving moisture or frost on panes when the temperature between panes drops below the dew point. You can get 5 lb of this stuff for $30 or less, depending on where you look. From what I read, it is capable of maintaining a RH of about 10% and will adsorb moisture preferentially from activated charcoal.
To give an example of the results I have gotten with the coconut charcoal and molecular sieve, closed box 9 driver line array speakers using Aura NS3 equivalent 3 1/2" drivers from Parts Express in approximately 0.65 cu ft cabinets with ~40% coconut charcoal fill and molecular sieve measure 3 db down from 40 hz - 60 hz (measured from near the middle of the room depending on exact placement of the speaker in the room - the 40 hz ones are about 2' away from the back wall on the long end and about 4' from the side walls and the 60 hz ones are side channel speakers placed immediately against the side walls) without the use of any LF equalization in my 17 x 25' HT room. After this, Audyssey eq is applied, extending the flat response of these speakers to the 30-40hz range without subwoofers which is how I have been satisfied to operate them so far, although if I find some time/energy, I may add the SWs (presumably with their own charcoal fill).
Last edited:
Btw, I don't imagine the charcoal will be near as absorptive at higher frequencies as wool/fiberglass, etc., and I don't regard it as a replacement for these materials. In fact, above 100hz, its main effect may start to become to reduce the enclosure's effective compliance because of the volume it occupies, so that may cause a dual effect of causing some relative boost at somewhat higher frequencies, similar to that of having an undersized cabinet, as well as increasing the apparent cabinet compliance below 100hz. But this is just speculation.
I don't have the source of the coconut charcoal I got immediately available, but I recall I was getting 2 cu ft delivered for just over $100 a few years back.
I don't have the source of the coconut charcoal I got immediately available, but I recall I was getting 2 cu ft delivered for just over $100 a few years back.
with off topic / unnecessary, unhelpful or vaguely insulting comments. I'd probably best not to get into more discussions until physically start making the design in a few weeks and then hopefully it can be a bit more positive . 
As AC is claimed to act as a desiccant, it is totally unsuitable for this application, so maybe it's proponent will give up or at least move to the other thread where I was posting results of stuffing tests to clear this back to your sub/satellite system build.
I would be very interested in the actual measured response of the FR's you have picked. Have you looked at The Edge or Olson's papers? You never said if you were planning a packaged active eq unit with this system or do you hope to use passive eq?
I would be very interested in the actual measured response of the FR's you have picked. Have you looked at The Edge or Olson's papers? You never said if you were planning a packaged active eq unit with this system or do you hope to use passive eq?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.