Baffle Step Compensation Adjustment

I'm confused... I've been reading about BSC from the following sources:

http://trueaudio.com/st_diff1.htm

http://sound.westhost.com/bafflestep.htm

And I've come to the conclusion that a BSC passive filter looks something like this:

diff_fix.gif


Well, I'm attempting to adjust the BSC filter from the following crossover network to accommodate a wider baffle that the designer intended. I already have the cabinets built with the same driver compliment and would like to apply his crossover network to it.

http://www.htguide.com/forum/showthread.php4?t=18946&page=1&pp=35

[IMGDEAD]http://www.whatisrazar.com/crossover.jpg[/IMGDEAD]

The design was originally for a 7.5" wide baffle. My baffle is 11.5" wide. I've calculated that this should drop the f3 of the diffraction loss from 608Hz to 397Hz.

My QUESTION (finally =) is....
How do I alter the crossover network shown to reflect this? I don't even really see the BSC filter clearly in the schematic. Help?
 

augerpro

Member
Paid Member
2006-08-27 11:02 am
www.somasonus.net
BSC is generally not a discrete circuit. The XO as a whole incorporates it, mostly at the first inductor of the woofer.

Changing this value substantially obviously affects the XO to the tweeter, so it's not something that can usually be changed much without adjusting everything else. Small tweaks will be possible, but without modelling it, I'm guessing the diiference caused by going from 7.5" to 11.5" will be substantial.

I would contact cjd at htguide to get his advice since that is his design.
 
augerpro said:
BSC is generally not a discrete circuit. The XO as a whole incorporates it, mostly at the first inductor of the woofer.

Changing this value substantially obviously affects the XO to the tweeter, so it's not something that can usually be changed much without adjusting everything else. Small tweaks will be possible, but without modelling it, I'm guessing the diiference caused by going from 7.5" to 11.5" will be substantial.

I would contact cjd at htguide to get his advice since that is his design.


I did, he respectfully indicated that he was not interested in making the modifications. I can understand that. I was hoping that here was some sort of mathematical trickery I could use to lower the f3 of the given BSC. Actually, I'm still fairly certain there must be. Crossover software modeling programs are able to calculate the values somehow. I just haven't been able to figure out which components in the given crossover are involved in the BSC filter and what their mathematical relationships are.
 
As Thawach has mentioned there's no full blown BSC in the filter such as the one shown here.

http://www.quarter-wave.com/General/BSC_Calculator.xls

There is also some other reading here as well:
http://www.quarter-wave.com/General/General_Articles.html

There are some tricks in the xo to help with BSC such as the first inductor for the woofer and probably some extra padding for the tweeter.

Just build it and tweak. The tweeter should have no problems as it already cooked to suit. It's only the woofer that has a new BS F3 which may just need a few less turns on that first inductor.

Working this out with software could be more frustrating and time consuming than tweaking in real life which is more fun and educational any way.

Then comes the question.... how close to the rear boundary are the speakers going to be used. Near the wall, no BSC required, 1.2m into the room maybe 4-6dB BSC.
 

augerpro

Member
Paid Member
2006-08-27 11:02 am
www.somasonus.net
You can model it if you had his measurements. Having said that I do have RS150 measurements on 7.5" baffle and 1.8mh seems quite a bit less than full BSC, particularly with an 18g inductor as advised in that thread. My own design used 2.5mh for a full 6 dB of BSC. I would build it as is and see how it sounds.

If you find you must reduce BSC try this: 1.6mh on the woofer and change the tweeters series 6 ohm resistor to 5 ohm. This will result in a slightly higher woofer and tweeter response *away* from the crossover frequency, which will tend to remain the same level. Phase integration will stay about the same this way. If you want to risk a little bit of phase deviation in return for basically the same response as the original, do the above but also bypass the tweeter's series cap with a 0.5-1.0uf cap, raising the total value to around 8-8.5 uf.
 
Re: bsc freq less - confused?

rj45 said:
Dudes,

I'm not a guru, but since the OP's question was what to do since
his baffle was wider than the original plans, so that the BSC
frequency went from 600 Hz to 370Hz.

So wouldn't he need more inductance, not less?

-Don


Yeah my original feeling was to raise the 1.8mH inductor's value, but I wasn't sure by how much and if anything else needed to change...
 

augerpro

Member
Paid Member
2006-08-27 11:02 am
www.somasonus.net
*In the circuit posted* if you raise the inductor value the woofer's higher frequencies will drop further, no? So bass is higher in relation (disregard DCR loss for the moment), thus it has more BSC correction. If we were to lower the value the high frequencies would rise in relation to the bass, thus less BSC. People really need to get out of their heads that BSC correction is a discrete circuit, and they certainly don't used generic textbook calculators for values.You'll see there is no resistor bypassing that inductor in the design posted, like the theoretical circuit shows, makes all the difference here...
 
augerpro said:
*In the circuit posted* if you raise the inductor value the woofer's higher frequencies will drop further, no? So bass is higher in relation (disregard DCR loss for the moment), thus it has more BSC correction. If we were to lower the value the high frequencies would rise in relation to the bass, thus less BSC.


I really appreciate your insight, but I want to be clear about what you stated. I believe your instructions would be correct for raising and lowering the magnitude of the BSC, but I'm interested in lowering the FREQUENCY at which the BSC occurs. Are we talking about the same thing?
 

augerpro

Member
Paid Member
2006-08-27 11:02 am
www.somasonus.net
m0tion said:



I really appreciate your insight, but I want to be clear about what you stated. I believe your instructions would be correct for raising and lowering the magnitude of the BSC, but I'm interested in lowering the FREQUENCY at which the BSC occurs. Are we talking about the same thing?

We would be IF the crossover posted were using a textbook BSC circuit, but it's not. If you have access to modeling software and real driver measurements, construct a crossover like the one posted. Then change the inductor value and see what happens. Again there is no resistor bypassing the inductor, this changes how each circuit behaves.


marchel said:
If I were the OP, I would change the 1.8mh inductor into 2.5mh And then tweak the tweeter L pad if necessary.

Just so we are understanding what this will result in, do you agree that going to a 2.5mh inductor will drop the higher frequencies of the woofer? This then requiring the tweeter to be dropped in level to match? If so, do you agree that you have now increased the amount of BSC, as the bass is now at a relativeley higher level than the midrange and treble?

Another mental example: if you placed a woofer on a baffle and made that baffle larger and larger, the inductor would get smaller and smaller until the point that the baffle became infinitely large, at which point the inductor would have 0 value. Do you agree?

I think perhaps the key here is understanding the difference between textbook BSC circuits and values and what we use in the real world. Anyone who's used crossover sim software like LSPcad or SE or SW knows intuitively what I'm talking about.
I'll sim some examples and show you. The main advantage to a textbook filter is that you can smooth out the response right at the baffle step, imposing a knee right at the baffle step, while keeping the higher frequency response flat because of the resistor. Using an ordinary inductor drops the response linearily, so you may end up with a small dip just before the baffle step, but I've only seen this effect on very small baffles. Larger ones tend to even it out.

BTW if you see diy designs on the net with a reduced BSC version, does this version's inductor usually use a larger or smaller value? I think you'll find that it is smaller with out exception.
 

augerpro

Member
Paid Member
2006-08-27 11:02 am
www.somasonus.net
Ok here is a quick hack and slash sim using the RS150 and 810921 measurements from my bookshelf speaker i built, which use 7.5" baffle. The network layout is nearly identical to cjd's. The blue is the original near full BSC and uses a 2.5mh inductor on the woofer. The red is with the inductor dropped to 1.8mh, which required dropping the tweeter series resistor from 6 ohm to 3 ohm, and raising it's series cap about .8uf to 5uf total. Now comparing the bass response with the response higher in frequency, it should be clear the blue response has much more BSC by about 3-4 dB and uses the larger 2.5mh inductor. While the red has less BSC and uses a smaller 1.8mh inductor.

BSC_comparo.jpg




Now this only for a given baffle width, but you get the idea of BSC works just using an inductor and a typical baffle.

I see in that htguide thread that cjd means 4 dB when he says it is full BSC, which is probably realistic in most rooms. Given that, I would again suggest teh OP build the design as is and see if he feels it needs less BSC. My guess it will sound nuetral to slightly warmish, but otherwise fine. The main problem changing the baffle width I see isn't BSC, but phase integration of the tweeter and woofer adn smooth response under the crossover frequency. These are somewhat unpredictable without having measurements on the actual baffle.
 

augerpro

Member
Paid Member
2006-08-27 11:02 am
www.somasonus.net
augerpro said:
Now this only for a given baffle width, but you get the idea of BSC works just using an inductor and a typical baffle.

I want to make sure this comment was not missed. Especially since I don't want to go off on a tangent myself! Which I probably did :dead: ...Changing baffle width is not corrected by simply changing baffle step correction, but you could raise the inductor value to lower the woofer response back down. But the response is slightly different around the crossover frequency, so again I think the concern here is not BSC changes, but maintaining smooth integration with the tweeter, whether or not the woofer's inductor is changed or not. The difference between 7.5" and 11" baffle is only about 1 dB or so under 800Hz.

Bleh, I need my morning coffee.
 
I'd take Brandon's advice. Go first with a 2.0 mH inductor with a bit less padding on the tweeter. If you need less BSC after extensive listening, you can easily modify it by unwinding the coil and adjusting the padding resistance. But this simple method will only work within a certain limit. I'd not go outside of 1.5 to 2.0 mH range. Within this range, the designed phase tracking won't be ruined. And I think this range should be sufficient for your purpose.

To unwind an inductor, an online calculator:

http://www.lalena.com/Audio/Calculator/Inductor/

is pretty accurate if you know how to use it. I've actually compared a result (i.e., number of turns or wire length that needs to be taken out to reach a certain value) obtained from the calculator to an actual result obtained using my LC meter, and found excellent accuracy.