[Piek]

Hello I am Micha and this is my first contribution to this forum!

I want to share my experiences with this open baffle which is not yet finished.

The idea was to create a passive speaker and to do it as quick as possible, because I had problems with my active setup and needed a less complicated speaker to just " plug on and listen to" without digital clipping and alike.
So I used all the parts I already had from former projects, 2x Fostex Fe167e, 2x Monacor SP-382PA out of my Ripol'es, and some sheets of wood that would have been thrown away.

The sheets were 1000x400x18mm so the baffle size was given. Simulating the FE167e in "Edge" gave me the position 890x80mm on baffle, this position would compensate a 7db dip in his response at 1,2kHz and give a flat response above this.
In Order to get some bass, I placed the Monacor 15" on 200x200mm down at the bottom.

Bass cutouts are cut in half with a slight angle and used as stands.

Since the OB roll off with Fe167e is around 200Hz I put a 10mH coil as a 6db crossover to roll off the bass, in order to prevent Fe167e's excursion there is a huge 110uF Capacitor. The Fostex needs a 'brake' too it is attentuated by 4,7Ohm+15Ohm.





Quick in room-measurement at 1m distance to walls and microphone:


Bass needs to be attentuated below 100Hz in order to have the full experience at listening position.

The sound is very open, tight, relaxed with no listening-fatigue, the "Bretthart" shines at high volumes, she will give you a live sound and is not designed for listening at low volumes because there should be some air moved!

This is just an example for an (fun-) open baffle, if you want to build one, keep in mind that you could use a fullranger that has the same efficiency as the woofer or vice versa to prevent using a voltage divider (sp?)

Regards, Micha






 

[Graham Maynard]

Hi Micha,

Simple, neat and effective.
HF just a little down.  Could maybe do with something like 1uF between amp and top driver to bypass the 'brake' at HF.

Cheers ....... Graham.

[Piek]

Thanks,

this will probably solve the falling highs, even though I don't really 'hear' it as a problem.
[edit: Pic deleted]
SPL is not correct in the simulation, it should be around 90db.

regards, Micha

[Rudolf]

Hello Micha,
nice to see you on this forum too. Your wallpaper tells the story even without giving your name.
What about the fostex phase plug? Looks like you have been to the curtain shop lately.

BTW: I don´t believe in that "linear response on axis" any longer. Did you look in Boxsim (your simulation program) what the Fostex driver position does to the 30° and 60° response? You could try in Boxsim to put the Fostex close to the upper edge of the baffle but horizontally centered.

If you don´t mind I would like to play with those Boxsim data. Would you mail your .BPJ file to me?

Thanks
Rudolf

[Ray Theon]

This looks interesting. Could anyone comment on the advantage of using two drivers taht were similar in sensetivity? would it matter if you were using 2 amps, a sub amp on the bottom, something nice on top?

[Piek]

Looks like you have been to the curtain shop lately.

   bought 3 different ones for 1€ at the local "Karacho" sale

I know what you are talking about with the off axis listening and crossoverdesign, but its not really an issue for this OB and its purpose for me because listening position is absolutely on axis.
When I need constant directivity for free anywhere-in-room-listening I can switch to my mains in the background.
Will send you the boxsim-file later.

advantage of using two drivers that were similar in sensitivity?

Passive, you wont need the 2 Resistors at the fullranger and probably end up with a higher overal-sensitivity of the speaker.

would it matter if you were using 2 amps, a sub amp on the bottom, something nice on top?

This would be cool and surely more flexible you wont need to match drivers in sensitivity then, because you can match your chassis via your amps individual volume control .
You  should find an active crossover that is capable of crossing over in the 200-300Hz range and boost up the bass or you could build passive line level crossovers (pllxo) yourself, but thats another story.

-Micha

[Piek]

Hi,

I am considering this Projekt as finished so far (measurement at 2m listening distance):

(red=without correction circuit, black=with circuit[edited])

Bass is plenty there now, I've changed the room and amps, they were a little "bass shy" with my DIY Amps (myrevc) I think its the low impedance of 4Ohm they cannot drive. Now with a Nikko TRM-750 oldie everything is very solid at the bottom end.

Going to try a T-circuit if I get cheap transformers, we'll see.

-Micha

[D OB G]

Hi Rudolf

"I don´t believe in that "linear response on axis" any longer."

What approach do you take now?

David

[Piek]

Hi Rudolf

"I don´t believe in that "linear response on axis" any longer."

What approach do you take now?

David

Probably listening off-axis where response is flat and leave on-axis-response as is.

-Micha

[Rudolf]

Hi Rudolf
"I don´t believe in that "linear response on axis" any longer."
What approach do you take now?

David,
I could not get decent data for the FE167e. So I simulated the Visaton FRS 8 on a open baffle of 100x20 cm.
First with the driver center at 89 cm height and 5 cm from the right edge:



While the response on axis (blue) is quite linear, the response at 30° and 60° off axis is very different for the left and right side.  At 1000 Hz the figure "8" radiation pattern looks like the diagram below. Keep in mind that even at 15° the response will look more like the 30° response than the 0° response. That´s what my own measurements tend to show.

Second simulation (left) is with the driver at the same height but horizontally centered on the baffle:



Obviously we see a nasty dip on axis at 3000 Hz. But the 30° and 60° patterns are very similar. So I would prefer to equalise the response for the 30° angle and listen to the speaker at that angle too.

Keep in mind that 0° radiation is just a narrow beam in space, while 30° radiation is along a cone surface and 60° is along an even wider cone. So the 30° and 60° radiation is of much more significance for the summed response in the room than the 0° radiation.

To keep things in perspective: If your OBs are far from the side walls, if those walls don´t reflect much and if you are listening close to the OBs - then you should forget about 30° and 60° and concentrate on 0° linearity.
But if reflected sound is a prominent part of your listening experience, you better look for off axis linearity and ignore that 0° dip.

Micha already knew what my strategy is.

[D OB G]

Thanks Rudolf,

Interesting you should say that.

Since I have got DEQX, I can achieve a perfectly linear response at whatever off-axis angle I want (or a modified response if I want to).

I have found that I am getting my best results at the moment, with I suppose "normal" distances to the walls, by toeing the speakers out!! 20 degrees, and then producing a linear response to the listening chair (i.e. the power response in the room is more even, with the rising response off-axis- I also run a tweeter through a 10 ohm resistor pointing behind the speaker- inspired by Linkwitz).

The sound stage is bigger, more defined front to back and side to side, with more air, and the phase modified recordings of instruments outside the width of the speakers are producing very good images.

There is obviously plenty of scope for further experimentation.

David

[Graham Maynard]

Hi Rudolf,

Very interesting measuements.

Yet I wonder, if the straight-on response did not have that suck-out, would reproduction not be better on the driver axis ?

My feeling is that the suck-out is due to symmetrical sharp baffle edge diffraction, and I wonder if you can test that arrangement again with foam pipe insulation clipped over and around the circumference of your baffle.  This is about 2" dia and cheaply available. 

I have found that rounded edges significantly 'quieten' baffles, and on dipoles that roll shapes are best.  Even a roll of soft cloth or carpet might well take out that centre position dip.

I tried the FRS8 on a 4" wide baffle between two vertical 12" dia rolls of carpet which were as close as 1" separation behind the driver.  This 'baffle' gave best reproduction for that Visaton; no notable resonances or reflections, and excellent axial reproduction.

Cheers ......... Graham.

[Rudolf]

I tried the FRS8 on a 4" wide baffle between two vertical 12" dia rolls of carpet which were as close as 1" separation behind the driver.  This 'baffle' gave best reproduction for that Visaton; no notable resonances or reflections, and excellent axial reproduction.

Graham,
those diagrams were no measurements but simulations. But your assumption is perfectly right. And Boxsim is a nice toy. I simulated the FRS8 for the same baffle and symmetrical position as before. Black is without any taper at the 1,8 cm thick baffle. green is for a 2 cm taper and red for a 5 cm taper (but don´t ask me what a 5 cm taper on a 2 cm baffle is supposed to look like in real life )



Honestly I would not have believed that tapering or rounding edges could completely level that dip. But obviously it does. And the combined response looks much nicer now too:



Thank you very much for bringing up this issue!

Rudolf

[weidok]

I have some boxsim data from the Fostex FE 167 E
did it with spltrace ... but its better then nothing

if anybody wants the boxsim files just let me know

..
I am building a OB with Peerless SLS 315 and Fostex FE 167 E
just testing a tappered NOBOX with active X over
in a T amp Omnes Audio 2.1
mk II

NOBOX


T amp Omnes Audio

[Graham Maynard]

Hi Rudolf,

Thanks for the follow-up.

Interesting ........ Graham.