[gamgee]

Hello AudioCircle!
Could not find some introductory thread so I guess it's ok to say hello over here, since OB is my prime interest...
I'm from Croatia, I looove OB speakers, and I really need to get out of lurking mode, 'cause it's driving me crazy...so much to read...

Ok, here it goes:
After a lot of reading, some experiments with "on hand" drivers, gypsumboard, cardboard and "old-wardrobe-from-the-attick" boards, and one "real" experimental "OSB/narrow U baffle+ripol bass" speaker, I want to do proper full OB loudspeaker. I'll add info and my reasoning behind doing things in next few posts, for the time being, here are some sketches...

something like this





or this



or even this

[Russell Dawkins]

Welcome, gamgee.

I would say you are off to a good start. I would question the need for that flat panel behind the woofer magnets - the vertical in the last picture looks like it would interfere with the sound less and could still be made to perform the same function - unless I misinterpret its function.

Russell

[Nate Hansen]

I would question the need for that flat panel behind the woofer magnets - the vertical in the last picture looks like it would interfere with the sound less and could still be made to perform the same function - unless I misinterpret its function.

Russell

Magnet mounting I'd say.........although it'd need to be beefier than the drawing to be really effective.

Is that a fullranger on top? I would try to isolate the fullranger from the woofers as best as you can........full separation is best, to keep the vibrations from the woofer from affecting fullrange.  I'm doing that with my dipoles and its one of the biggest improvements I've made.  Magnet mounting the woofers also seemed to "tighten" up the bass quite a bit.

I like the "naked" fullrange version the best! Although you would for sure need some sort of active eq. Better yet, hang it from swings and get it isolated from the rest of the speaker and you're onto something.......

[gamgee]

Welcome, gamgee.

I would say you are off to a good start. I would question the need for that flat panel behind the woofer magnets - the vertical in the last picture looks like it would interfere with the sound less and could still be made to perform the same function - unless I misinterpret its function.

Russell

Hello Russell,
flat panel behind woofers is magnet mounting so that woofers are only touching the baffle, without strong mechanical connection. The vertical is same thing but for the mid - and its not connected with woofer m. mount (well, actually it is - they both are connected to the same base, but vertically, there is a gap between them).

[gamgee]

Magnet mounting I'd say.........although it'd need to be beefier than the drawing to be really effective.

Is that a fullranger on top? I would try to isolate the fullranger from the woofers as best as you can........full separation is best, to keep the vibrations from the woofer from affecting fullrange.  I'm doing that with my dipoles and its one of the biggest improvements I've made.  Magnet mounting the woofers also seemed to "tighten" up the bass quite a bit.

I like the "naked" fullrange version the best! Although you would for sure need some sort of active eq. Better yet, hang it from swings and get it isolated from the rest of the speaker and you're onto something.......

Hi Nate, that's right, I'm also worried about the woffers mount. I'm trying to make sturdy and as simple and good looking construction as I can. For the "naked" version - I like it too, but...Edge simulation does not look too good (if I'm doing it right).
Here it is, Edge simulated circular baffle only few mm larger than the driver itself:



Here is same thing for the baffle from the first picture:

[gamgee]

But I still haven't given any details about the project...so here it is.

Drivers are:

Tweeter: Beston RT002A Ribbon Tweeter modified for dipole operation
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=277-112

After finding this driver on part express, I contacted Beston and found they are really cooperative, fast and allround pleasure to work with - they promptly accepted the idea of open back and sent me my pair without extra charge - I have the first such pair they made!

Mid: Visaton BG20
http://www.visaton.com/en/chassis_zubehoer/breitband/bg20_8.html

Actually, it's a fullranger and at the moment I'm using it as such - without crossover.

Woofers: There are two pairs of slightly different woofers - reason for this is more a matter of financing the project (I got them for cheap).
First one is really great vintage driver, Tesla ARN 8608 (which I suspect could be used higher than usual woofer range, if need be)
http://www.repromania.net/tesla/tesla-basove-arn-8608/tesla-basove-arn-8608.php
The other one is SAL AHR 3040, for which I don't have a link, it's dicontinued driver but not vintage.
TS parameters of those two pairs are quite similar but not the same - I read somewhere that is a good thing since they will probably end up connected in parallel. Anyway, woofers I'd like to use are out of my price range, and these don't look bad at all, so that's what I'll use.

Baffle:

Baltic birch ply, 3x12mm sandwich (which allows for some things as sand filling the baffle), asymetrical wing construction to avoid resonant cavities behind the speaker and to have asymetrical placement of drivers on baffle without baffle becoming to large, disconnecting lower and upper baffle (as much as posible) to minimise vibration transfer from woofers and to create an empty space around tweeter (no baffle reflection, at least in horizontal plane).

Ok, I'll fill in other data later, some measurements I did (newbie there too), and experiments I tried...
I hope you'll find this project interesting 'cause I need all the help I can get 

[gamgee]

I apologise for the delay, that's because I'm new on the forum...should have posted few times before starting my own thread...

[D OB G]

Hi gamgee.

IMO you can’t do better than a sand filled woofer baffle.
In one location I have a true infinite baffle for the bass.  6 X Kef B139s in a concrete floor 150mm thick over a large garage.
I have also used a baffle about 2m X 2.5m with a 100mm thick filling of sand.
The concrete mounted woofers, with low and powerful enough signals, can vibrate the whole 6m X 8m  floor.  I believe that concrete may be inherently underdamped.  It can ring.
The sand filled baffle never vibrated under any conditions… not at all…not even the slightest.
(I followed Gilbert Briggs’ Wharfedale OB idea of course).
And that was without magnet mounting, which might even be better.

Regards,

David

[Nate Hansen]

Tweeter: Beston RT002A Ribbon Tweeter modified for dipole operation
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=277-112

After finding this driver on part express, I contacted Beston and found they are really cooperative, fast and allround pleasure to work with - they promptly accepted the idea of open back and sent me my pair without extra charge - I have the first such pair they made!

Now this I am excited to see! Could the little "horn" be cut away (or removed) to make the assembly smaller?

[gamgee]

Hi gamgee.

IMO you can’t do better than a sand filled woofer baffle.
In one location I have a true infinite baffle for the bass.  6 X Kef B139s in a concrete floor 150mm thick over a large garage.
I have also used a baffle about 2m X 2.5m with a 100mm thick filling of sand.
The concrete mounted woofers, with low and powerful enough signals, can vibrate the whole 6m X 8m  floor.  I believe that concrete may be inherently underdamped.  It can ring.
The sand filled baffle never vibrated under any conditions… not at all…not even the slightest.
(I followed Gilbert Briggs’ Wharfedale OB idea of course).
And that was without magnet mounting, which might even be better.

Regards,

David

6 X Kef B139s in a concrete floor 150mm thick... wow...
sometimes woofers vibrate the floor at my place and they are all on the baffle...what happens at your place where they are in the floor I can only imagine, lol...
So, if I understand correctly, you made Wharfedale OB replica or similar loudspeaker? Is that speaker working together with those KEFs in IB concrete floor?

[gamgee]

Now this I am excited to see! Could the little "horn" be cut away (or removed) to make the assembly smaller?

Waveguide is one of three main parts that housing of the tweeter is made of. It is connected to the rest with 4 screws and I supose you could remove it without any trouble. I did not try to do that, though, but it seems easy enough...
...
ok, after going through my email corespondence with Peter from Beston I found this in one of his first mails!


so it is like I thought - just unscrew and remove.

Now I'm interested why you want to remove the faceplate/waveguide (since it improves dispersion)?
Is it because you want to reduce the area around tweeter some more? Without waveguide you will get aproximately 1 cm off around the tweeter...

[Nate Hansen]

Now I'm interested why you want to remove the faceplate/waveguide (since it improves dispersion)?
Is it because you want to reduce the area around tweeter some more? Without waveguide you will get aproximately 1 cm off around the tweeter...

To reduce the area, yeah. In my case I run the drivers 'naked', suspended from wires (with a Neo3 tweeter). The idea is to push the dipole directivity as high as you can, so smaller is better here.  With a tiny 'baffle' the dipole peak is pushed higher, dictating when the driver starts to beam.  The horn isn't need for controlling the directivity, as the behavior of the dipole does this itself.

[gamgee]

To reduce the area, yeah. In my case I run the drivers 'naked', suspended from wires (with a Neo3 tweeter). The idea is to push the dipole directivity as high as you can, so smaller is better here.  With a tiny 'baffle' the dipole peak is pushed higher, dictating when the driver starts to beam.  The horn isn't need for controlling the directivity, as the behavior of the dipole does this itself.

I'll try and see later today if I still have FR graph for the driver without waveguide...

[scorpion]

Yes, let us know how the Bestons are, nice find.

I also think that the Visaton BG20 is a find. You have probably read  the German Klang + Ton 'Cheap Trick' where they used the BG20 successfully
in a qouple of variations for an only cone speaker, tweeter, mid and bass all the way. The BG20 was crossed at 300 Hz
and 3.5 kHz with no other correction applied for a revealing mid, that should be about optimal also for you, I think.

/Erling

[gamgee]

Yes, let us know how the Bestons are, nice find.

I also think that the Visaton BG20 is a find. You have probably read  the German Klang + Ton 'Cheap Trick' where they used the BG20 successfully
in a qouple of variations for an only cone speaker, tweeter, mid and bass all the way. The BG20 was crossed at 300 Hz
and 3.5 kHz with no other correction applied for a revealing mid, that should be about optimal also for you, I think.

/Erling

Yes, I think both drivers are nice finds (although I can't really take credit for BG20 - it's quite known over here, a friend suggested it to me). Do you remember which issue of Klang+Ton was that? What do you mean when you say "for an only cone speaker" - loudspeaker with no ribbons and similar or something else?
From what I managed to measure, 300 Hz would be lower limit for BG20 (on a moderately sized baffle), and my plan was to use it in between 300Hz and 2500Hz, as that is where BG20 sounds best. What I really would like is to achieve as constant directivity as I can, without introducing yet another, smaller driver. So, I made experimental phase plug, and when I gather enough courage, I'll "operate" one of the BG20's. Hopefully, phase plug will  extend wide dispersion region to where RT002A takes over.

[gamgee]

To reduce the area, yeah. In my case I run the drivers 'naked', suspended from wires (with a Neo3 tweeter). The idea is to push the dipole directivity as high as you can, so smaller is better here.  With a tiny 'baffle' the dipole peak is pushed higher, dictating when the driver starts to beam.  The horn isn't need for controlling the directivity, as the behavior of the dipole does this itself.

I asked Peter from Beston for permision to publish their measurements...his comment is that they are not made on IEC standard buffer board, and that it is good only for comparison at the same testing condition. So, here is comparison of standard RT002A (closed back) and modified RT002A (open back), without waveguide. I think that's what you are interested in?



I still prefer open back version with waveguide...



What do you think?

[Nate Hansen]

On axis doesn't tell us much, other than that without the faceplate it needs to be crossed higher. Personally I'd like to see off axis measurements of the tweeters without a baffle, although it's about an inch wider than the Neo3, so naked I highly doubt it would be any better than the B&G.

That said, in your application your baffle will determine the directivity. With that in mind it looks like a nice alternative to the Neo3.

I'm looking at the RT002A's little brother, the rt001a. I'd like to see it without a back.......looks like it could work!

[gamgee]

Sorry, I don't have off axis measurements...I'll try and ask Peter if they are available.
I decided on RT002A because of the 92 db sensitivity (same as BG20) and 2500 Hz recomended crossover frequency. RT001A has 88 db sensitivity and RT003C (round shape!) has 93 db but 4000 Hz rec. freq. ... so, RT002A for me.

[Nate Hansen]

I now see from your drawings that you do want to run the tweeter naked.....cool! You also mention maintaining constant directivity. I would say use as small a baffle as you can for your mid to maintain its directivity as high as you can, even then it will be beaming well before the tweeter, causing a narrowing of the directivity pattern before the tweeter takes over. You are going to use active eq right?

I'm also wondering about that horn on the tweeter. To the front of the speaker it will allow you to cross lower, but to the rear I think you might end up with a hole in the response due to the lack of that horn loading to the rear......

[scorpion]

K + T speaker, it is a cone tweeter not a dome or ribbon or planar or anything else, just a plain old cone. The variation with BG20 as a midrange was in Nr 1 2009.

I find the 'backless' Beston measurements very interesting.

Have a look at GR-Research's Super V which has a coax hornloaded compression driver forward firing with the membrane naked at the rear:



It will be more or less the same here. You don't want the backwave to dominate the sound just give a subtle addition.
I agree with Nate to keep midrange baffle as narrow as possible for constant directivity and also let midrange and treble be totally physically separated from the bass baffle. Should be easy to fix judging by your sketch.

This also opens for vertical physical alignment for no phase difference between speaker units.

/Erling