Open Baffle Bass Boost: +4 to +7dB w/ Passive Xover, No DSP

...I need to be clear that the Marshall Bass Boost LC circuit is VERY different from

1) a series inductor, which attempts to cancel out the 6dB slope of the OB cancellation in combination with a very high efficiency driver, or

2) a series capacitor, which KEF and others have done. This can be optimized in combination with T/S parameters to protect an acoustic suspension system with specified 3rd order slope.

Neither of these can boost the output, can only cut.

Not strictly true - the "KEF series capacitor" operates in conjunction with the compliance of the driver's suspension, which looks like an inductor at the driver's terminals:

https://audiojudgement.com/speaker-equivalent-circuit/

So there will be a degree of "boost" - albeit small.

Edit: Oops - a bit late with my post!
 
...Paralleling a film cap with NPE caps to reduce the loss by lowering ESR is not effective enough below the Fc of the film cap used, so the film caps have to be large to be effective too. Electronic Concepts UL30 series are as they call them Unlytic, and are in the 3" diameter standard computer style size and have low ESR. They are able to stretch the film to a thinner insulator so that they can make them in this size format. I have some to test in use, but have not as yet. Ebay seems to have them at good prices.
https://www.ecicaps.com/film-capacitors/ul30-series/
I need 500uF which would be $$$$! Any idea of what combination of NPE and PP might work for this to be beneficial?
 
Coming to this thread a little late, but an analog crossover (or DSP) and two power amps per side gives much more flexibility than a passive crossover.

Have a look at the late, great Siegfried Linkwitz's site for chapter and verse on this https://www.linkwitzlab.com/filters.htm . That is just one page of a comprehensive site devoted to open baffle speakers.
I am a huge fan of DSP and if you click the links at the top of the article, you’ll see several open baffle designs I did using DSP. But the constraint on the Walnut Dipoles project was specifically old-school passive crossovers, resulting in a traditional speaker you can plug into any system anywhere.

Most people don’t know that you can in fact add 4 to 7 dB bass boost, using only passive components.

“No nuclear weapons, just bow and arrow.”

What I like about the Walnut Dipoles and the Passive Bass Boost circuit is they are a self-contained unit, no outboard electronics, they plug into any conventional system and unlike almost all dipoles 1) they have plump bass with a modest bump just above 40Hz for a "comfort food" sound, 2) they're not very big, 28" high and only 35lb/16kg, 3) they play plenty loud and don't overload easily even with heavy bass content [because of the natural subsonic filtering], and 4) they look fantastic.

My active DSP dipole designs are better but they're unwieldy. I can throw these in a car and take them to a friend's house and hook them up to his stereo.

The 5dB bass boost at 45Hz gives them a bass-rich EQ and SPL capability you might normally associate with a 10" bass reflex, but spacious expansive Open Baffle sound instead of boxy bass reflex sound. And approximately Constant Directivity across the entire range which reflex cannot possibly achieve.
 
Last edited:
Rod Elliot makes a pcb for the Linkwitz transform and I can confirm that it works. I built a closed box sub using two Kef B139s downward pointing in a box that fitted under a large coffee table.

https://sound-au.com/index.html project no. 71

The bass shook the house and was almost frightening when reproducing the lowest organ pedal notes and military band bass drums.
The LT can create boost and is more flexible because you can tailor that instead of just living with the narrow band of frequencies where the peaking (high Q) filter has boost, but the LT does not ALSO provide a HP function to limit the lower frequencies, which I think is one attractive feature to using Perry's approach. The only downside is that it is not a general equalization method for dipoles (for the 6dB per octave dipole loss curve). Perry's speaker cabinet design allows him to use only the peaking HP filter, and that is clever.
 
Perry, you say this could work well with 0.5 QTS drivers. What would the SB Audience Bianco-15OB350 look like if you have the spare time?

http://www.sbaudience.com/index.php/products/open-baffle-drivers/bianco-15ob350/

These are significantly cheaper in Aus than the delta lites and interestingly have a impedance peak lower and also have a peak in response just above the impedance peak.

The thing that annoys me with these drivers is the dip at 50hz and the earlier break up. But could be a nice trade off.
 
Good ! It can take a pretty stout amp and would assume will not work with things such as single ended tube amplifiers with relatively high output Z - I've done it with Karlson boxes and little laminated steel core inductors such as those sold under the Dayton name.
One amp seemed to "ratchet" into protection on a heavy drum beat but the 200w rated Samson amp may have clipped. The inductor's filed shook a pair of pliers

Hornresp Filter Wizard can simulate the filter with whatever.
 
might be a silly question but where does this approach differ from using for example a graphic eq and adding the boost at the desired frequency and a cut at the frequency bellow for the protection . I mean electrically hoe does this differ hoe would the demands placed on the amp and driver differ .
not bashing I think this is really interesting just wondered which would be the preferred method be taking costs into account and any added distortion and noise from the eq .
could one build a simple circuit to boost and cut a specific frequency like a 2 band eq for a similar cost to this approach
 
might be a silly question but where does this approach differ from using for example a graphic eq and adding the boost at the desired frequency and a cut at the frequency bellow for the protection . I mean electrically hoe does this differ hoe would the demands placed on the amp and driver differ .
not bashing I think this is really interesting just wondered which would be the preferred method be taking costs into account and any added distortion and noise from the eq .
could one build a simple circuit to boost and cut a specific frequency like a 2 band eq for a similar cost to this approach
The difference is adding external electronic EQ devices. The goal of the Walnut Dipole project and this circuit is too boost the bass using only passive components inside the speaker system. Of course you can use external EQ and DSP any way you like.
 
Perry, you say this could work well with 0.5 QTS drivers. What would the SB Audience Bianco-15OB350 look like if you have the spare time?

http://www.sbaudience.com/index.php/products/open-baffle-drivers/bianco-15ob350/

These are significantly cheaper in Aus than the delta lites and interestingly have a impedance peak lower and also have a peak in response just above the impedance peak.

The thing that annoys me with these drivers is the dip at 50hz and the earlier break up. But could be a nice trade off.
Mitch,

My Bitches Brew speakers use a 15OB350 woofer so I disconnected the middle woofer, measured frequency response and impedance of the middle woofer only and put that data into VituixCad. Below: There is no EQ applied to this woofer. This is crude 1/3 octave data, in my room about 1.5M away, in the Bitches Brew Open Baffle enclosure:

sb acoustics 150B350 farfield vituixcad no filter CROPPED.jpg

The main point of this graph is the dipole roll off below 70Hz. The measurement is not 2.83V/1M, so the dB SPL numbers are relative.

BELOW:

With 330uF capacitor and 25mH inductor.
sb acoustics 150B350 farfield vituixcad MBB filter.png

This filter gives +6dB boost at 45Hz, impedance minimum of 4 ohms, and f3 of around 40Hz. The 15OB350 looks like an excellent candidate for the Passive Bass Boost circuit.

These measurements aren't exactly an apples to apples comparison with the Walnut Dipoles but they are useful. I'd have to do a lot more experimenting to know whether the 450Hz peak is "real" or if it's just an artifact of my quick and dirty measurements. In any case the SB is a very good woofer with nicely behaved midrange and no breakup problems.
 
Last edited:
Good ! It can take a pretty stout amp and would assume will not work with things such as single ended tube amplifiers with relatively high output Z - I've done it with Karlson boxes and little laminated steel core inductors such as those sold under the Dayton name.
One amp seemed to "ratchet" into protection on a heavy drum beat but the 200w rated Samson amp may have clipped. The inductor's filed shook a pair of pliers

Hornresp Filter Wizard can simulate the filter with whatever.
The shape of the impedance curves gets imposed on the frequency response curve for a speaker being driven by a tube amp.

I don't know that the impedance of a dipole with Passive Bass Boost is any "worse" for a tube amp than no filter at all, it's just different. The passive circuit exhibits an impedance dip (3 ohms for the Walnut Dipoles) instead of a peak (25ohms if I didn't have the LC filter).

So instead of a tube amp giving "fatter" bass at resonance with this circuit, it delivers a little less.
 
@Mitch311 The 15OB350 has a higher mechanical Q (Qm=12) which means less natural damping. Higher Qm means larger peak boost in the Marshall Bass Boost circuit and lower dips in impedance. It's not hard to get an 8-10dB peak with this particular woofer, and impedances that dip down to 2 ohms and below.

You run into that problem/opportunity with the 15OB350 when you try to push the boost frequency below 45Hz.

That's fine if your amp is OK with 2 ohm loads, but with this woofer I tried increasing the DCR of the inductor from 0.8 ohm to 1.8 ohm (which means you can buy a cheaper inductor, US Coils sells all different grades) so with a 1.8 ohm DCR I can push the boost frequency down to 40Hz with a Zminimum of about 3 ohms:

sb acoustics 150B350 farfield 2ohm DCR lower F3.png

Here, the filter boosts response by 7dB. The cutoff point has moved down from 40Hz to 35Hz, and the SPL at the cutoff point drops a bit as well. Goes a little lower, and a bit leaner.

The point being that you can play with driver Qm and inductor DCR to find other useful compromises.
 
Last edited: