Info on Open Baffles

[EddieT]

Hi Y'all.

Been searching for some good info on OB and found this to share if others are interested:

https://jelabs.blogspot.com/2012/06/je-labs-open-baffle.html

It's a pretty good source for lots of hifi stuff.  The owner is really knowledgable and experienced.

Enjoy!

[Cloth Ears]

16 hours ago, EddieT said:

Hi Y'all.

Been searching for some good info on OB and found this to share if others are interested:

https://jelabs.blogspot.com/2012/06/je-labs-open-baffle.html

It's a pretty good source for lots of hifi stuff.  The owner is really knowledgable and experienced.

Enjoy!

You'll find this much more informative - http://www.quarter-wave.com/

[davewantsmoore]

7 minutes ago, Cloth Ears said:

You'll find this much more informative - http://www.quarter-wave.com/

Martin pages are very good, and build up from fundamental physics/maths like you should, to hopefully avoid the garbage in garbage out issues of both measurement and listening.

 

... but I think the big issue is that directivity is not considered.... which is one of the big draw cards of an open baffle (dipole) speaker (ie. you have the opportunity to make a speaker which has improved directivity at lower frequencies (eg. << 2khz) over a monopole.

 

It all depends on what sort of OB you want to make though.   If it's more old school one (eg. wide baffles and/or single drivers) .... then Martins pages are even perhaps a bit too "in depth" for the casual enthusiast.

[Cloth Ears]

9 minutes ago, davewantsmoore said:

Martin pages are very good, and build up from fundamental physics/maths like you should, to hopefully avoid the garbage in garbage out issues of both measurement and listening.

 

... but I think the big issue is that directivity is not considered.... which is one of the big draw cards of an open baffle (dipole) speaker (ie. you have the opportunity to make a speaker which has improved directivity at lower frequencies (eg. << 2khz) over a monopole.

 

It all depends on what sort of OB you want to make though.   If it's more old school one (eg. wide baffles and/or single drivers) .... then Martins pages are even perhaps a bit too "in depth" for the casual enthusiast.

I'm not sure that directivity can be considered - other than to note that there's not much you can do about it. Cancellation around the edges of the baffle is one of the reasons that speaker enclosures are so popular - because you don't have to worry about the size of the speakers and that the sound disappears when you walk past them.

And yes, there's tricks to ameliorate the effect, but essentially, you have the same signal (out of phase) coming from the back and the front of the speaker.

Luckily, I sit in front of my speakers, so even if I did have OB, I wouldn't be too worried about what happens next (to the speakers)

[davewantsmoore]

34 minutes ago, Cloth Ears said:

I'm not sure that directivity can be considered - other than to note that there's not much you can do about it.

The dimensions of the driver and the baffle are completely responsible for the directivity.

 

38 minutes ago, Cloth Ears said:

Cancellation around the edges of the baffle

Can be used to optimise the directivity, if designed well.

This is why for example the Linkwitz Orion speaker moved from a constant 35cm wide baffle with 1" + 8" + 10" .... to a narrower funny shaped baffle with a 1" + 4" + 8" + 10" driver layout.   This improved the directivity.

 

Martin only considers the driver size and baffle width in regards to the bass, and a single axis frequency response..... were as in a high quality speaker it is actually more important to optimise the driver size, layout, and baffle size in regard to the frequency response vs listening angle (aka. directivity) as this cannot be "fixed" after the fact.   Where as the bass level, and axial frequency response can be changed by things like EQ, or using extra drivers, or choosing a specific listening angle, etc.

 

But all that being said... his pages are very good, and he's a great contributor to the community (actually helping people, rather than just being a 'naysayer' is lots of work).

 

[EddieT]

On 26/9/2023 at 8:22 AM, davewantsmoore said:

 

 

On 26/9/2023 at 8:12 AM, Cloth Ears said:

You'll find this much more informative - http://www.quarter-wave.com/

thanks guys1

[Ian McP]

How’s your German? Plenty of dipole info in here, the graphs may interest you.

http://www.dipolplus.de/thema1.html

 

 

[RockRolley]

Hi all,

I have a pair of Goodmans 12" Axiom 15/16 ohm drivers which I aim to pair with coles super tweeters. I'm wanting to experiment with OB. I know that a good OB can produce a lovely soundstage, but I also value nice bass. 

Am I barking up the wrong tree with OB? I know that even 10" drivers can produce nice bass in a ported enclosure.

Anyway, I would like to try it all things being equal. I don't understand too much technical talk, so what I'd love, is if anyone can say with confidence, 'if you make baffles in these dimensions with the 'full range' driver placed here and the super tweeter placed here, it will produce ample bass and sound fantastic'. 

If there are any further ideas on top of that as far as making a stand or any other design ideas/features, I'm all ears too.

Thanks, brains trust.

Simon

NB: PICS NOT MINE-FOR ILLUSTRATION PURPOSES ONLY

[Ittaku]

OB can produce great bass, you just need an awful lot of driver surface area and/or excursion. Depends greatly on what your threshold is for good bass, but one 12" is not enough IMO.

[RockRolley]

7 minutes ago, Ittaku said:

OB can produce great bass, you just need an awful lot of driver surface area and/or excursion. Depends greatly on what your threshold is for good bass, but one 12" is not enough IMO.

Thanks, well, I have IMF RSPM transmission lines and they produce incredible bass, but I suppose there's no point trying to create that with OB, Essentially, I liten to a lot of 60's-70's pop/rock, Beach Boys, Beatles, Doors, Hendrix, Grand Funk, etc. so if it can reproduce bass guitar well, that would be great, but not necessarily deep tuba...

One thing that attracted me to OB was the idea of full range speakers and I don't have to worry too much about crossovers (aside from the super tweeter in this case), so if I start adding more drivers, it may start to defeat my ends...

[Ittaku]

2 minutes ago, RockRolley said:

Thanks, well, I have IMF RSPM transmission lines and they produce incredible bass, but I suppose there's no point trying to create that with OB, Essentially, I liten to a lot of 60's-70's pop/rock, Beach Boys, Beatles, Doors, Hendrix, Grand Funk, etc. so if it can reproduce bass guitar well, that would be great, but not necessarily deep tuba...

One thing that attracted me to OB was the idea of full range speakers and I don't have to worry too much about crossovers (aside from the super tweeter in this case), so if I start adding more drivers, it may start to defeat my ends...

A bass guitar goes down to about 40Hz. You could just squeeze in good bass at a respectable loudness out of 12" OB drivers at that frequency but not if you want it to get particularly loud. Other instruments do go a lot lower though.

[RockRolley]

11 minutes ago, Ittaku said:

A bass guitar goes down to about 40Hz. You could just squeeze in good bass at a respectable loudness out of 12" OB drivers at that frequency but not if you want it to get particularly loud. Other instruments do go a lot lower though.

Thanks. Any tips for baffle dimensions or designs? 

[Ittaku]

4 minutes ago, RockRolley said:

Thanks. Any tips for baffle dimensions or designs? 

Not offhand sorry. Everything is a compromise - wider baffle, lower frequency before the baffle step loss, but less "open" sound.  Knowing the characteristics of the driver and modelling them with software is the only way to know even vaguely what you might end up with. Blind placement in a baffle of some empirical legacy design is from the dark ages and I wouldn't do it unless you were spending only spare change on it and were happy to experiment for the sake of experimentation.

Edited December 14, 2023 by Ittaku

[RockRolley]

23 minutes ago, Ittaku said:

Not offhand sorry. Everything is a compromise - wider baffle, lower frequency before the baffle step loss, but less "open" sound.  Knowing the characteristics of the driver and modelling them with software is the only way to know even vaguely what you might end up with. Blind placement in a baffle of some empirical legacy design is from the dark ages and I wouldn't do it unless you were spending only spare change on it and were happy to experiment for the sake of experimentation.

Ahhh… well, if I can find someone with the modeling software, I’d certainly go for that, however, I was thinking of just getting some used ply to experiment with before getting some nice veneered birch ply or even actual timber. 
 @andyr had offered to assist me to figure out crossover with his software, but not sure if that software would also model baffle response 

[andyr]

2 hours ago, RockRolley said:

 @andyr had offered to assist me to figure out crossover with his software, but not sure if that software would also model baffle response 

 

No sorry, Simon ... it can't do that.  

 

To help you confirm your ideas ... you might like to come over and have a listen to my zero baffle spkrs and see what a heavy dose of DSP - and an active setup - can accomplish!  

 

[davewantsmoore]

3 hours ago, Ittaku said:

you just need an awful lot of driver surface area and/or excursion

"Just" double the same (size, excursion) monopole.

 

2 hours ago, RockRolley said:

was thinking of just getting some used ply to experiment with before getting some nice veneered birch ply or even actual timber. 

A driver on an open baffle will have less bass than the equivalent driver(s) in a box, due to the "dipole" effect...... so it will sound "bass shy" without the right correction filter to make the response the same as the equivalent monopole (ie. put the bass back).

 

[Cloth Ears]

2 hours ago, RockRolley said:

Ahhh… well, if I can find someone with the modeling software, I’d certainly go for that, however, I was thinking of just getting some used ply to experiment with before getting some nice veneered birch ply or even actual timber. 
 @andyr had offered to assist me to figure out crossover with his software, but not sure if that software would also model baffle response 

When I was interested in that, I paid Martin King about $20 for use of his MathCad worksheets (transmission line and open baffle). MathCad itself could be used in it's unlicensed version, as long as you don't need to save anything.

Martin's OB stuff is on http://www.quarter-wave.com/index.html, and he was also contactable through there. I juts cannot see where I asked to be able to use his worksheets.

 

3 minutes ago, davewantsmoore said:

"Just" double the same (size, excursion) monopole.

 

A driver on an open baffle will have less bass than the equivalent driver(s) in a box, due to the "dipole" effect...... so it will sound "bass shy" without the right correction filter to make the response the same as the equivalent monopole (ie. put the bass back).

 

But, for non-bass applications, a simple sheet can work wonders for midrange...

 

 

Edited December 14, 2023 by Cloth Ears

[davewantsmoore]

1 hour ago, Cloth Ears said:

But, for non-bass applications, a simple sheet can work wonders for midrange...

At the risk of diving down a "how to open baffle" hole ....  It depends on what frequency the "dipole peak" is (dictated by the width of the baffle), as this area must be avoided.

 

Below the dipole peak.... the driver acts like a dipole.

Far above the dipole peak.... it pretty much just behaves like a monopole.

 

 

1 hour ago, Cloth Ears said:

Martin's OB stuff

Martin is great (like both SL and JK) as he proceeds from "first principles" and builds up from there.... and very much can be learned and taught by that route.

 

Did you see his "mic drop" last week?   (He's quit the OB scene to focus on TLs)

Edited December 14, 2023 by davewantsmoore

[andyr]

20 minutes ago, davewantsmoore said:

....  It depends on what frequency the "dipole peak" is (dictated by the width of the baffle), as this area must be avoided.

 

Is there an equation which relates baffle width with dipole peak, Dave?

 

20 minutes ago, davewantsmoore said:

Below the dipole peak.... the driver acts like a dipole.

Far above the dipole peak.... it pretty much just behaves like a monopole.

 

Do you mean by this, that a LP filter should be applied to the driver ... so it operates only from the dipole peak, downwards - and so always behaves like a dipole?

 

[Cloth Ears]

12 hours ago, davewantsmoore said:

Below the dipole peak.... the driver acts like a dipole.

Far above the dipole peak.... it pretty much just behaves like a monopole.

 

To a certain extent...the actual behaviour of the driver only behaves like a monopole in these instances where the monopole is in an 'infinite baffle' (i.e. qts is less than 0.5). Otherwise, the behaviour of the driver in an enclosure is somewhat different, being affected by the air in the enclosure.

But I get your point, as far as frequency is concerned.

 

12 hours ago, davewantsmoore said:

Did you see his "mic drop" last week?   (He's quit the OB scene to focus on TLs)

I did not. I'm not surprised, as it was his main interest when he started (and going to wide on a subject can really destroy the enjoyment). And those OB folks are sooo argumentative...

 

Edited December 14, 2023 by Cloth Ears

[RockRolley]

Thanks everyone for your input. I had a brief look at quarter wave, and I'm afraid it seems above my head... I may be reduced to 'dark ages' style experimentation based on other designs...

[davewantsmoore]

5 hours ago, Cloth Ears said:

To a certain extent...the actual behaviour of the driver only behaves like a monopole in these instances where the monopole is in an 'infinite baffle' (i.e. qts is less than 0.5). Otherwise, the behaviour of the driver in an enclosure is somewhat different, being affected by the air in the enclosure.

This only has any effect around the "resonance frequency" of the driver in the box (or baffle)..... so none in the "let's just use it as a midrange", example.

 

At frequencies above this, they behave the same, for the same sized baffles.... aside from the area around the dipole peak.   In this area the frequency response of the open baffle is more distorted than the box.    Although, the box is also distorted (diffraction from the box edges) in this range (the dipole is just worse).

 

5 hours ago, Cloth Ears said:

And those OB folks are sooo argumentative...

Sorry.  ; )

 

3 hours ago, RockRolley said:

based on other designs...

This is the way.... as at least then you have something tangible to compare your results with.

 

Open baffle design trips a lot of people up IMVHO.   At high frequencies it's really not doing much "open baffle" ... and at lower frequencies, the design process is very similar to a box speaker, but more difficult.    Because the "wood working" side is easier, people tend to expect the whole thing is easier...... but to design a high quality speaker the process is actually somewhat harder (than a box speaker).

 

Don't let that deter you from slapping a driver on a sheet of wood though... it's all good fun.

 

5 hours ago, Cloth Ears said:

I did not. I'm not surprised

... or at least he is quit(ting) after he releases his set of papers he's been working on.

 

Edited December 15, 2023 by davewantsmoore

[davewantsmoore]

18 hours ago, andyr said:

Is there an equation which relates baffle width with dipole peak, Dave?

Yeah, I'm pretty sure I gave it to you in your speaker thread.

I don't much use it, as you can see the effect clearly in measurements of a driver on a baffle (vs the driver near field or very big baffle, etc).

 

It is roughly = speed of sound / 2 x distance between sound sources

 

So for a 6.5" woofer with zero baffle is 343 / 2 x 0.17 = 1 kHz (distance is centre of front of cone to centre of rear of cone.... so driver width = close)

... or woofers in a big baffle (or smaller baffle with folded wings) where the path from front to back is 1m is 343 / 2 x 1 = 170 Hz

... or two 1" tweeters mounted back to back could be 343 / 2x 0.05 = ~ 7kHz

 

18 hours ago, andyr said:

Do you mean by this, that a LP filter should be applied to the driver ... so it operates only from the dipole peak, downwards - and so always behaves like a dipole?

Yes.... but "should" is a strong word.   If you don't, you will enounter the consequences, but there's always compromises.  It's really just something to be considered from the get-go (ie. at the pencil and paper stage of the design, if not the simulation stage).

 

Obviously many many open baffle designs (especially of old) put a driver on a big baffle and operate the all the way up.    The dipole peak puts a kink in the response, and this cannot be equalised out because the error is different at different angles of observation.   This is the same problem as diffraction from the edges of a box speaker, just that a open baffle has it worse.

 

If we are trying to build a speaker with a good frequency response at all angles ..... a dipole is a good way to try to do that..... but we would only want to use the driver below the dipole peak.

 

Below the dipole peak, you get constant directivity.

Around the dipole peak you get diffraction (just like a box speaker, but somewhat worse).

Above the dipole peak the response settles down to behave just like a monopole/box with the same sized baffle.

 

 

An example of this can be seen in the Linkwitz open baffle speaker(s).

 

In the earlier "Orion" speaker, it used an. 8" woofer + 1" tweeter.     It was a big stretch to get the 8" (with effectively zero sized baffle) up to ~1kHz without distortion from the diffraction ... and it was a big stretch to get the 1" down to 1kHz without distortion from over-driving, and also distortion from the 8" sized baffle being too wide.

 

... so the later version of the speaker (LX521) used an 8" woofer, 4" woofer, and 1" tweeter, with the woofers operating comfortably in their "true" dipole ranges, where the SPL vs frequency is constant with radiation angle.

 

 

Sorry to yank the thread sideways ..... although it is all about what to expect when you put a driver on a sheet of wood.....   vs driver on a tiny (effectively zero sized baffle) .... vs driver in a box.     It's often overestimated how much bass you will lose (it's only 6dB/octave) so it isn't unsurmountable ...... but OTOH, that 6dB is often unaccounted for, as in, if you don't correct for it properly, your speaker will have a pronounced tilt to it, and sound "like something" (insert subjective something here, good bad, could be either.... some might say clear and vivid, some might say shouty, some might say no bass) ...... then then the diffraction also goes unaccounted for (it's no different, really (just worse), to what you have to deal with in a box speaker).

Edited December 15, 2023 by davewantsmoore

[davewantsmoore]

36 minutes ago, davewantsmoore said:

the process is actually somewhat harder (than a box speaker)

 

Part of the big reason is that if you have a moderately sized dipole woofer (say 8", whatever) .... and you are looking to correct its rolloff, and you can't/won't simulate the rolloff (and correction) ... then you have to measure it..... but the ground gets in the way ... or the surfaces of your room gets in the way.   It can be very hard to get accurate measurements <<< 1khz.       It's a bit like how the "baffle step" is the hardest part of getting typically box speaker right..... similar issue in a dipole, but again, dipoles are just a bit harder for a number of reasons.

[GlassAudio]

On 25/09/2023 at 6:27 PM, EddieT said:

Hi Y'all.

Been searching for some good info on OB and found this to share if others are interested:

https://jelabs.blogspot.com/2012/06/je-labs-open-baffle.html

It's a pretty good source for lots of hifi stuff.  The owner is really knowledgable and experienced.

Enjoy!

Archive of the John K's Music and Design site is one of the very best sources for dipole design,
https://musicanddesign.speakerdesign.net/tech.html