|
|
Post by karatestu on Feb 12, 2022 11:40:26 GMT
I thought I would start a new thread about my speakers as the old one doesn't seem relevant anymore. I have learned a lot over the years having tried all sorts of things that either worked, didn't work or I thought they worked and hoped but ultimately were wrong on many levels. Oh well, the journey is as important s the destination. My goals include nothing between amplifiers and drive unit(s) other than wire, up firing mid bass, aerodynamically shaped enclosures for smooth baffle step transition and low re radiation, suitable for a small room at low listening levels. My rooms have a lot of room gain especially in the bass so even a JBL Bluetooth speaker can deliver what I want. Given the above you might think my dream speaker would be a single full range driver or coaxial but as I like the semi omni design then a forward firing tweeter is essential even if is married to an up firing full range. Being obsessed with curves I have ended up with a spherical miid bass enclosure with 2" full range acting as a tweeter in a spherical enclosure directly above the up firing mid bass. It doubles as a sort of diffuser for the mid bass. There are no speaker level components, just wire. The high pass at 7 KHz us done at line level by just changing the value of a capacitor in the power amp. I found some hourglass shaped plastic stools which when modified a bit will be the stands. Below is not the real thing but is a mock up just to give an idea of the aesthetics  It needs some more work yet but but the basics are there. It will either be all black or two tone with the big 3mm steel sphere being a different colour to the rest. |
|
|
|
Post by MartinT on Feb 12, 2022 11:57:33 GMT
Nice improvisation. Have you experimented with an inverted cone compared with the ball as a diffuser?
|
|
|
|
Post by karatestu on Feb 12, 2022 12:52:37 GMT
Nice improvisation. Have you experimented with an inverted cone compared with the ball as a diffuser? No not really. I know an exponential cone is supposed to be better but whatever is directly above the mid bass also has to house the tweeter. So I think a small 15cm sphere is the best compromise for doing both jobs. I am going to try and add a sort of cone to the bottom of the tweeter sphere to try and distribute the mid bass' beaming higher frequencies better and reduce any resonances as sound bounces about in the gap. I wanted to keep thd tweeter as close to the mid bass as possible. It will be a bit like this but with the mid bass transferred to the spherical steel enclosure rather than this crappy chipboard cube.  I had wondered about if I should stick with the plastic sphere or use a steel one hut I think I will keep it plastic. It's not like there is any bass pressure to contain as the high pass is at 7 KHz. |
|
|
|
Post by jandl100 on Feb 12, 2022 12:53:42 GMT
|
|
|
|
Post by karatestu on Feb 12, 2022 13:45:11 GMT
Hi Jerry. Thanks for your comments. I have tried tweeters firing up into cones and the tweeters I have are no good for that. The 2" full range I use as tweeter now (used to use 1" domes) is amazingly good so I want to stick with that. I gather you need a compression tweeter with big rising response in to the upper high frequencies to make it work. To be honest this speaker saga has been going on five years and I just want to get something finished. |
|
|
|
Post by speedysteve on Feb 12, 2022 20:19:41 GMT
Interesting idea.
With the right drivers and resonance damping it should work. Depends on how low freq wise you need to go.
If you were using the stools as a horn...
Frequency response wise it's interesting to model the intended profile in Hornresp (David Bean created amazing horn computer modelling simulator). You can find it over on DIYAUDIO forum. David is a great guy, very helpful and really knows his stuff.
You can see any nasty peaks and troughs and play around with mouth/throat size, length, flare etc.
Also the intended driver T&S are input and taken into consideration.
I've modelled many issues of mid bass horns and tapped horns for folk using this tool.
On building they perform near as damn it, when measured, as the simulator showed they would.
My mid bass speakers are exponential. 1.2m long (throat to mouth), 13cm diameter throats, 70cm diameter mouths - from memory.
I use a 12" driver in a compression format.
They play down to 90Hz flat and would go up to about 600Hz but I cross over at 350Hz.
|
|
|
|
Post by karatestu on Feb 13, 2022 10:06:03 GMT
Interesting idea. With the right drivers and resonance damping it should work. Depends on how low freq wise you need to go. If you were using the stools as a horn... Frequency response wise it's interesting to model the intended profile in Hornresp (David Bean created amazing horn computer modelling simulator). You can find it over on DIYAUDIO forum. David is a great guy, very helpful and really knows his stuff. You can see any nasty peaks and troughs and play around with mouth/throat size, length, flare etc. Also the intended driver T&S are input and taken into consideration. I've modelled many issues of mid bass horns and tapped horns for folk using this tool. On building they perform near as damn it, when measured, as the simulator showed they would. My mid bass speakers are exponential. 1.2m long (throat to mouth), 13cm diameter throats, 70cm diameter mouths - from memory. I use a 12" driver in a compression format. They play down to 90Hz flat and would go up to about 600Hz but I cross over at 350Hz. Winisd is as far as it goes for me. Hornresp would probably take me a year just to figure out how to use it. I have considered using the hourglass stool as part of the speakers internal volume but it would need strengthening on the inside with fibre glass or carbon fibre. I have never thought about using it as a horn. Wouldn't know where to start with that. |
|
|
|
Post by karatestu on Feb 13, 2022 11:02:19 GMT
I stopped using the cheap Chinese drivers a while back. I am now using some B&W 6.5" mid bass. I added a worm of blutack at the cone/ dust cap junction to get rid of a break up resonance. It has the added effect of lowering driver Fs, increasing Qts and lowering the sensitivity a smidge to match the 2" full range driver I am using as a tweeter.
The B&W drivers sound much better than the cheap doped ones and I already had them so cost me nowt. I am through with doping as I have realised it gets rid of too much of a driver's get up and go. I am also through with using small dome tweeters. They just don't work with a first order high pass even when crossing over at 7KHz. Too much distortion as I proved to myself by adding more tweeters to share the work - it reduced distortion a lot. The 2" full range when used as a tweeter is vastly superior with a 1st order high pass. No ferrofluid to dry up and they can handle a bit of abuse unlike 1" domes.
|
|
|
|
Post by speedysteve on Feb 13, 2022 11:53:11 GMT
Yeah, there is a learning curve with Hornresp, keeps my grey matter active enough using it from time to time🙂
You could use self adhesive bitumen stick on panels (from automotive panel damping applications), as a cheap and effective damping material. Can he applied inside and out for extra effect.
I have also used a mix of black car body under seal and fine sand. That really adds mass and damping.
With the right consistency it's brushable. (in a thick sort of way).
No substitute to quality drivers.
I really like the Eminence Pro range for bass applications. Their T&S parameters fit certain applications really well.
Used these in many customer tapped and mid bass horn applications.
I've found the higher the frequency the more $$$s or even $$$$s you need to splash.
|
|
|
|
Post by karatestu on Feb 16, 2022 19:47:15 GMT
I isolated the up firing mid bass sitting some open cell foam on top off three Litres of sand I poured in the bottom of the 3mm steel sphere. The driver on top of open cell foam and grit sand didn't work out for a number of reasons. The bass was super tight but not high enough in level compared to the rest. It was rather enlightening about how much bass information I had been missing. It didn't really bother me that the day after the driver had sunk into the foam and the seal I had made with gaffa tape had come unstuck. I was going to start again anyway because I think the springiness of the foam and the thinness of the gaffa tape were robbing some bass by letting the driver move too much and I don't think the seal around the driver was very good. So moving on the next idea involved a long length of 60mm round brass bar hot glued centrally on to the back of the mid bass' magnet. On the other end of the bar I put a 12cm diameter circular piece of aluminum plate. The three litres of grit sand in the bottom of the sphere was retained. Once I had put in enough to find where the aluminum plate should go I poured the rest of the sand in so it went part way up the brass bar. I never took any pics of that but here is a diagram that explains it better.  Instead of the gaffa tape to seal around the gap between the driver's mounting flange and the steel sphere enclosure I used some bitumen flashing tape and cut a ring out of that. This is still just temporary so I can prove if there is any mileage in this design. It looked like this.  |
|
|
|
Post by karatestu on Feb 16, 2022 19:48:34 GMT
So, goodbye crappy chipboard cubes with steel lining and hello steel sphere :dance: I am glad to get rid of them and one step closer to a design that works visually and sonically.
The brass bar with aluminum plate idea has worked a treat, I'm well pleased with it. It obviously makes the magnet seem heavier and the plate on the end further prevents movement . The driver doesn't touch the enclosure at it's mounting flange (only via the bitumen flahing tape) and the brass bar with aluminum foot plate doesn't touch the enclosure either due to it resting on sand.
The sand is there for several other reasons. This driver doesn't work in 13.5 litres even with a worm of blutack lowering Fs and increasing Q. So the sand eats up the volume required to get the bass response where it needs to be. I think it also provides some vibration control. With the sand being in the bottom of the sphere it helps keep the centre of gravity low.
Furthermore, the sand makes the inside of the enclosure not spherical so helps avoid the possible strong internal resonance from every dimension being the same. The sand also helps to support the brass bar with driver glued on the top. So this sand is a quite essential part of this design.
There are a couple of drawbacks. The speaker must never be allowed to go on it's side otherwise I will be tearing it apart to reset everything. Don't want sand in the drivers workings so these things must always stay upright
|
|
|
|
Post by karatestu on Feb 16, 2022 19:51:56 GMT
Some observations on vibration. There is still very little I can feel in the steel sphere enclosures. I am over the moon with how well the sand is damping it. It is an order of magnitude less than the vibration I felt when using the steel lined chipboard cubes. It's ridiculous really how much of a difference there is.
I have listened to a lot of music - can't stop :guiness; . The main thing I can say with where I have got to now is the amazing clarity across the board. The bass is now so detailed and clear that I realise how much mud is produced by an enclosure no matter what it is made of. I am hearing parts to bass lines I never knew existed - previously masked by floor and enclosure mush (no better word for it.)
I can't be sure how much this is due to the much reduced cabinet vibration because I have changed to steel spheres at the same time but my hand on the enclosure detects a massive difference so it must be partly due to that. A 3mm steel sphere is a very strong shape so there will be little flexing going on and for a given volume a sphere has the lowest surface area.
What effect the spherical enclosure shape has had on things I am not completely sure but I can say the sound staging and imaging is better than anything I have ever heard. There will be less ripple in the baffle step transition region as was proven by harry Olson inhis experiments with different shaped enclosures.
I am still experiencing all the dynamics I was previously. These latest changes have only added positively to everything. Transparency has got to a stage where instead of cleaning the window somebody has bloody removed it :dance: Cleaning up the bass and removal of reradiation from sharp 90 degree corners has unearthed a bloody treasure trove of extra detail in the music which doesn't sound false.
There is an ease to the speakers that is just wonderful. Everything has now fallen in to place. Yes it's highly detailed but I am enjoying the music even more now, the dynamic and emotion is all still there. No added music robbing filters were needed to do this or new devilry (dsp) - just attention to cabinet shape, positioning or drivers and vibration control. I know there are no 12" drivers or high sensitivity drivers and what they bring to the party but I can live with that. I feel like I am listening to monitors but ones that have loads of life (no BBC here) and which sound bigger than they have any right to without throwing it all at me (point & squirt).
Laters.
I'm over the fecking moon. My speaker design journey has ended, just the aesthetics to sort out and then move on to the 12" driver version
|
|
|
|
Post by karatestu on Feb 16, 2022 19:58:57 GMT
I removed the 24 x 24" x 2" concrete slabs under my speakers since now I have isolation they are not required. Bloody fugly things, it looked so much better without them. I noticed a change in sound that I liked as the height dropped by 2" so was curious to lower the spheres even further to hear what would happen. I did this before and liked it but the hourglass stand puts the drivers higher than I have ever had them. So I temporarily removed the hourglass stands and put a slightly shorter 110mm pipe under instead.   The tubes work really well half way under the circular base and half way sticking out. The height had lowered by about 4" and I liked it so not one to stop there I cut some even shorter pipes (6") and put them under.   I really like them low down both visually and sonically |
|
|
|
Post by MartinT on Feb 16, 2022 19:59:35 GMT
I isolated the up firing mid bass sitting some open cell foam on top off three Litres of sand I poured in the bottom of the 3mm steel sphere. So I wasn't far off with draught excluder strip? Really glad you've met with success. Your efforts have won through. Now you can turn your attention to the tweeter - possibly use an idea like loose fitting or the Roksan spring idea? |
|
|
|
Post by karatestu on Feb 16, 2022 20:00:13 GMT
How to explain the difference as the sphere gets closer to the floor ? It must be due to boundary enforcement and changing distance to the ceiling As the mid bass gets closer to the junction between the floor and wall the bass gets boosted a bit. Before I did isolation at speaker and driver level the bass got too muddy at the low level with certain music.
Now with the most articulate, clean and fast bass I have ever had it is not too much. The bass detail and cleanliness stays the same but the level relative to mids gets higher. Hence I am quite liking the added oomph.
So now I have a dilemma. I like the look of the short straight pipe more than the hourglass :shock: I might like the look of the hourglass stand more than the straight but it is going to have to be shortened at both ends considerably. Once that is done there is no going back (without buying more hourglass stools at £18 a pop.
|
|
|
|
Post by MartinT on Feb 16, 2022 20:02:13 GMT
For stability, you might want to cut a centre section out of the stools and join the two parts together. That way you keep the wide top and bottom.
|
|
|
|
Post by karatestu on Feb 16, 2022 20:33:10 GMT
I isolated the up firing mid bass sitting some open cell foam on top off three Litres of sand I poured in the bottom of the 3mm steel sphere. So I wasn't far off with draught excluder strip? Really glad you've met with success. Your efforts have won through. Now you can turn your attention to the tweeter - possibly use an idea like loose fitting or the Roksan spring idea? Thanks Martin. Yeah not far off with the draught excluder. But it isn't wide enough to bridge the gap. The mid bass' mounting flange does not sit on the enclosure but is inside the cut out in the steel sphere with a 2mm gap all around. I had to cut a ring out of bitumen flashing tape. The driver is held up by the brass bar which is held up by the grit sand so the flashing tape doesn't have to support anything. It's only temporary until I can come up with something more aesthetically pleasing. But it has proven to me beyond any doubt that isolating the driver from the enclosure is well worth doing. The effect is quite remarkable. I have already isolated the tweeter Martin. I did it before the mid bass and it too had a massive effect. It's what spurred me on to isolate the mid bass. I did it in a similar way but the gaffa tape is taking the weight of the tweeter and so isn't ideal. The tape keeps coming off and the weight of the tweeter distorts it so I am going to do that with something else. I am gobsmacked at the improvement isolating the drivers has brought. And it all started off with an inner tube under the whole speaker. The tricky bit now is getting it to look pleasing to the eye. |
|
|
|
Post by karatestu on Feb 17, 2022 5:28:47 GMT
For stability, you might want to cut a centre section out of the stools and join the two parts together. That way you keep the wide top and bottom. Hi Martin, I am thinking of running a threaded rod from base plate through the hole in the bottom of the sphere ( they come with a 8mm hole in the pole of the hemisphere) and putting a washer and nut on both ends to clamp it all together - that should be stable. I might even fill the stand with something. Not sure yet. These speakers weigh quite a bit already and I don't want to go overboard as hopefully I will be using these in my old age (not planning on moving them too much though). The wiring is going to go down the inside of the stand and terminate at four binding posts but not sure exactly where yet - base plate or bottom of the upright. |
|
|
|
Post by karatestu on Feb 17, 2022 5:43:22 GMT
Yesterday I posted loads ov stuff here about my speakers. Sorry if it was too much to read in one go but I copied posts over from hfs.
|
|
|
|
Post by MartinT on Feb 17, 2022 6:15:50 GMT
No problem, Stu. Your ongoing project is interesting to follow.
|
|
