How do you calculate and implement BSC?

[bigg]

How do you calculate and implement BSC? Also, how do you know if a design; a published design you are building, or one you purchased, has BSC?

[MSaturn]

There's a simple equation - 4560/baffle width in inches - that will give you the -3dB down point of baffle step (roughly.) I used this before I got Jeff B's modelling software.

Check out his stuff - he's compiled spreadsheets that will calulate not only step but complex diffraction effects, and save the results to an original .frd file (which you're hopefully using to model!)

Usually a speaker in te commercial arena will have minimal baffle step or up to 3dB. Most speakers I build have about 4.5dB - a stepping point between 3dB, which I consider minimal, and 6dB, which is full compensation.

Perhaps Jeff will post here. He's got a link to his software as his signature.

[Jeff B.]

There's a simple equation - 4560/baffle width in inches - that will give you the -3dB down point of baffle step (roughly.) I used this before I got Jeff B's modelling software.

Check out his stuff - he's compiled spreadsheets that will calulate not only step but complex diffraction effects, and save the results to an original .frd file (which you're hopefully using to model!)

Usually a speaker in te commercial arena will have minimal baffle step or up to 3dB. Most speakers I build have about 4.5dB - a stepping point between 3dB, which I consider minimal, and 6dB, which is full compensation.

Perhaps Jeff will post here. He's got a link to his software as his signature.

The link to my software is below.

The easiest way to tell at a glance if an existing design incorporates baffle step compensation is to look at the size of the inductors used in the low pass crossover. If the inductors are under 1.0mH for an 8 ohm driver, then there is little to no compensation, if the value if closer to 1.8-2.2 or higher then there is BSC. Just at a glance you can see this much. Of course, there are other variables that change some of this, so it is just a rule of thumb.

Jeff

[bigg]

Thanks MSaturn and Jeff Bagby. That will be a huge help!

[Ryan_M]

Good question.

I've seen it mentioned before in a write up that "...you can see there's about 4db of BSC..." and I wonder how you can see that? Looking at a the FR plot, if the response is ruler flat out to where the speaker rolls off naturally does that mean full BSC (6db) has been added or none? And if one wanted to apply 4.5db of BSC should I see a 1.5db dip in the FR response from ~500Hz (or what ever the BSC frequency happens to be) on down?

[bigg]

This damn forum! I typed out a long, detailed response and it dumped it! Said I wasn't logged in! I was too!!!!!

[MSaturn]

It auto-logs you out after a certain period of time. It does sort of suck, but at least usurpers can't post things you don't want them to if you're AFK!

[Ludo]

This damn forum! I typed out a long, detailed response and it dumped it! Said I wasn't logged in! I was too!!!!!

When you type a really long message, select all the text and copy it to your Windows buffer before you hit the submit button. Then if it turns out you are logged out, you can relog back in, reply to the thread again, and paste the text back in and submit.

It stinks and I hate it when it happens.

Louis

[bigg]

I want to thank all of you for your input. I was hoping to get a simple, user friendly kind of equation that would be intuitive to use while designing crossovers. I love building speakers, but I'm not a crossover geek. I have designed and built a couple of crossovers for some simple, small 2-way speakers that I built and they sound great. I used no BSC. But, I would be glad to use BSC if I could do it easily and with the confidence that I was doing it right and not just screwing up something else. I've tried other filter ideas before that I had to rip out of the crossover because they sounded terrible. But I know that the designers here know a lot more than I do, so I'm glad they're around so I can build their designs.

[brianpowers27]

Good question.

I've seen it mentioned before in a write up that "...you can see there's about 4db of BSC..." and I wonder how you can see that? Looking at a the FR plot, if the response is ruler flat out to where the speaker rolls off naturally does that mean full BSC (6db) has been added or none? And if one wanted to apply 4.5db of BSC should I see a 1.5db dip in the FR response from ~500Hz (or what ever the BSC frequency happens to be) on down?

It depends on how the plot was built/measured. If you took a regular FR plot from a manufacturer you would need to subtract the effect of BSC out and then use an inductor to shape the response back to flat.

If you used a measured response (on baffle), the FR would show the loss of BSC according to the space it was in. You would then use BSC shaping to get back to normal.

Bottom line, don't assume much. The response can look flat because BSC has been compensated for. At what level you ask? Only the designer knows.

[bigg]

Doesn't it matter how big the driver is in relation to the baffle? Or how the speaker is placed/used in a room?

[JustinG]

Doesn't it matter how big the driver is in relation to the baffle? Or how the speaker is placed/used in a room?

The first half, I don't believe it does, it will affect the defraction effects but as far as BSC it doesn't matter.

To answer the other part of your question, the placement doesn't have an effect on the point where BSC happens but it will affect the amount you have to compensate. That is, if you have the speaker right next to a wall you might only need a few dB of BSC but if it is in the middle of a room you would need all 6dB
I hope this helps
-JustinG

[curt_c]

Doesn't it matter how big the driver is in relation to the baffle? Or how the speaker is placed/used in a room?

Driver size will have an effect on the diffraction effects, including BSC, but only at the upper end of the BSC ‘curve’. Baffle step is not a smooth transition from 4pi to 2pi space as one might think, but has a peak in response as it approaches the 2pi response. This peak can be as much as 2 to 3 dB above the nominal response.
If baffle size is held constant and a large diameter driver is compared to a point source. This peak in response, as well as the diffraction ripples at higher frequencies will be ameliorated. This is due to the wave launching from various positions on the cone of the larger driver.

BSC correction is dependent on room position, mainly how close the speaker is positioned to the front wall. At low frequencies the closer proximity of the wall makes the baffle appear larger to the driver. (as large as the wall) In wall speakers do not require BSC, and consequently will be up to 6 dB more sensitive than conventional box speakers. The disadvantage of in walls is they will not have the image depth that speakers placed well out into the room can provide. This is probably not a big deal for HT setups, but IMO is very important to the 2 channel experience.

C

[fbov]

Can one simplify the crossover design exercise by measuring frequency response of the driver in the indented enclosure placed at the intended listening location?

I've generally found that the fewer steps required in a new process, the less likely I am to hit a roadblock. If I have the drivers and the box, as well as a mic/mixer and software, a simple in-room measurement will show me what BSC I'll need for that combination, will it not?

Thanks,
Frank

[JustinG]

Can one simplify the crossover design exercise by measuring frequency response of the driver in the indented enclosure placed at the intended listening location?

I've generally found that the fewer steps required in a new process, the less likely I am to hit a roadblock. If I have the drivers and the box, as well as a mic/mixer and software, a simple in-room measurement will show me what BSC I'll need for that combination, will it not?

Thanks,
Frank

That would be correct.

[brianpowers27]

Can one simplify the crossover design exercise by measuring frequency response of the driver in the indented enclosure placed at the intended listening location?

I've generally found that the fewer steps required in a new process, the less likely I am to hit a roadblock. If I have the drivers and the box, as well as a mic/mixer and software, a simple in-room measurement will show me what BSC I'll need for that combination, will it not?

Thanks,
Frank

I will run a few sims ahead of time beforee I build anything but when it comes time to purchase xo parts I will use real measurements. You are correct about your assumptions regarding on cabinet response as it pertains to BSC.

[bigg]

What does a BSC circuit look like? I haven't found it in my books.

[brianpowers27]

What does a BSC circuit look like? I haven't found it in my books.

I use the term loosely....

A BSC circuit can look like a RL circuit or incorporated into the woofer's inductor. Either way it usually requires padding the rest of the system.

[bigg]

Hi,
Yeah I found it. Thanks.

[brianp]

I'm surprised no one has mentioned the 2.5 way approach to BSC. Assuming that you can accomodate the larger box, lower impedance in the bass (where the two woofers operate in parallel), and the cost of an extra woofer and large inductor per side, this approach offers several advantages.

Most notably, higher sensitivity and the advantages of higher dynamic headroom that accrue therefrom. When you apply BSC in the crossover to a single woofer, you throw away up to 6dB of on-axis sensitivity, and have to pad the tweeter down to match the reduced level of the woofer in its upper range. A lot of your amplifier power then goes into heating the padding resistors. With a 2.5 way, you run the "upper" woofer up to crosover with the tweeter, and use the lower one to fill in the on-axis response at the lower frequencies. The tweeter requires less padding, and the output of the entire system is improved.