L value in BSC

**BoomieMCT** · **04-02-08, 10:35 PM**

What does everyone use to calculate the value of the inductor in a baffle step compensation circuit? I have a 6" wide baffle and a 6.5 ohm driver. Edge suggests adding a 2.7mH inductor but my own calculations come to 0.97mH. Since these are not cheap parts I'm wondering what y'all use.

**Anthony** · **04-03-08, 08:31 AM**

Edge's calculations are probably for an enclosed speaker. I didn't think you used a BSC on an open baffle speaker (because of the hump and steeper slope rather than just the rolloff). I could be wrong, though.

**BoomieMCT** · **04-03-08, 08:53 AM**

Both calculations are for enclosed speakers.

I imagine it shouldn't matter though as the BSC comes from the transition from 2pi to 4pi space. I always thought dipoles had that effect as well, it was just less important as they are dominated by the destructive intereference effects from the back wave.

**Anthony** · **04-03-08, 08:56 AM**

I guess that's my point. Why put something in to potentially distort the signal when it will have only a minor effect in the overall performance.

Why not just remodel without the BSC and pick different crossover points?

**BoomieMCT** · **04-03-08, 09:01 AM**

Because the final product (which won't be a dipole) will have a BSC. Even though that is mainly to boost lower frequencies then I'm really paying attention to there is some effect at the x-over point. That's why I need to add it in.

**Anthony** · **04-03-08, 09:06 AM**

Oh, I thought you were making dipoles. I'm sorry I misunderstood

In that case, you should build a much bigger baffle for individual testing, then determine enclosure size, build the enclosure, and then remeasure for crossover design.

The only way to properly design a crossover is to measure the speakers in their final baffle (which of course makes for a lot of trial and error if you have to change the baffle or cabinet volume).

**BoomieMCT** · **04-03-08, 09:13 AM**

Instead of making a whole enclosure I wanted to make sure I even had drivers I liked first. There is a lot of information that can be gathered from just a baffle even if the final product will be enclosed. For example, if you look at the same baffle in Edge and switch back and forth between OB and not (the little check box) the difference at the x-over point is less then half a dB.

Also, looking at Edge simulations there are more induced effects around the x-over point from a very large baffle then there are on a small baffle. This is why I used a small baffle.

But this is all off topic. My issue is that LDC, Edge, Lalena and other BSC calculators are not giving consistant values of L for the BSC I will eventually need. I'm getting values ranging from 0.8mH to 2.7mH. THIS is the question I need people's insight into.

**JCD** · **04-03-08, 11:54 AM**

Since I haven't jumped in the pool this deep yet, I can't offer personal experience -- however, I've been talked out of doing BSC circuits by someone I've got a certain amount of respect for. His comments mirrored a lot of what Anthony mentions.

That being said, this website has a simple excel spreadsheet that calcs both the BSC and Zobel networks that is based on a paper my Martin King.

JCD

**BoomieMCT** · **04-03-08, 12:00 PM**

I assume you are talking about Anthony's suggestion of "picking different x-over points"? Unless I can find a tweeter that plays down to 800Hz that isn't an option.

I'm interested in what argument made you decide that BSC's were bad. They are very commonly used in almost every design I've seen. I'm not challenging your statement - I just want to learn more about your viewpoint. Outside of active EQing I know of no other way to address this.

I'll check out that spreadsheet when I get home.

**Anthony** · **04-03-08, 12:08 PM**

I was only questioning BSC's for dipole. Otherwise, they are almost a requirement, unless you want really wide speakers.

I still don't think you will get a good feel for bass extension or overall flatness unless you build a really wide baffle for testing. The reason Zaph and others use 4' wide baffles is so BSC never becomes an issue in raw driver testing.

If you are wanting to test the width of your baffle and it's effects, you need to get rid of that backwave or the hump and steep rolloff will color any results you measure. Just toggle the dipole/open-baffle checkbox in Edge to see how drastic it changes.

As for you question on the calculators: I wasn't using the BSC calculator, but I found a large disparity between Speaker Workshop, Lalena, 12 volt, and PCD Calculator for things like Zobel, lpad, and notch filters. They would all show different values, some way off of others. Since none of them show the underlying math (and assumptions) I never knew which to believe. I usually defaulted to Speaker Workshop, but it does not have a calculator for all of those things.

**BoomieMCT** · **04-03-08, 01:01 PM**

Quote:

Anthony wrote:

I still don't think you will get a good feel for bass extension or overall flatness unless you build a really wide baffle for testing. The reason Zaph and others use 4' wide baffles is so BSC never becomes an issue in raw driver testing.

True, but as I stated I'm not doing this test for the bass. I'm only looking at dispersion and the x-over point. At the frequencies I'm currently looking at it shouldn't be an issue AND I should be able to get more useful data with a real sized (small) baffle.

Quote:

Anthony wrote:

As for you question on the calculators: I wasn't using the BSC calculator, but I found a large disparity between Speaker Workshop, Lalena, 12 volt, and PCD Calculator for things like Zobel, lpad, and notch filters. They would all show different values, some way off of others. Since none of them show the underlying math (and assumptions) I never knew which to believe. I usually defaulted to Speaker Workshop, but it does not have a calculator for all of those things.

Yeah, I thought as much. My 0.97mH number is based off using only the thinnest dimension of the baffle - in this case 6 inches - and doing the math on that (I made my own spreadsheet a while back). I'm sure Edge uses the plot generated from the whole simulation. I do understand how even using similar techiniques you can get some wiggle room but a factor of three is odd.

**JCD** · **04-03-08, 07:36 PM**

Quote:

BoomieMCT wrote:

I assume you are talking about Anthony's suggestion of "picking different x-over points"? Unless I can find a tweeter that plays down to 800Hz that isn't an option.

I'm interested in what argument made you decide that BSC's were bad. They are very commonly used in almost every design I've seen. I'm not challenging your statement - I just want to learn more about your viewpoint. Outside of active EQing I know of no other way to address this.

I'll check out that spreadsheet when I get home.

No harm, and I'm not offended if you totally disagree..

This is a quote lifted from where he posted the bulk of his opinion:

Quote:

I really don't like BSC filters, in general, because they are extremely difficult to design from scratch without measuring and testing the output of the speaker in the room you are putting it in. If you design the BSC filter based on theory and modeling, the results are almost always pretty poor. Also, EVERY series component will be passing all audio through it all the time, so any performance concerns about the quality of the components are very important to consider. In my experience, simply designing the system appropriately and calibrating the tweeter levels beased on measured performance on the finsihed baffle is more than enough to get an acceptably flat response and a sweet sounding output. Also, filters like the Zobel network can also reduce the high frequency output at the higher end of the midrange response, which is similar the impact of a BSC.

JCD

**BoomieMCT** · **04-03-08, 10:02 PM**

I understand some of these points but if I'm crossing to the tweeter at 2200Hz and the baffle step effects are at 800Hz (ish) and below chaning the tweeter level won't do much for me. If I was making a 3-way that happened to cross to the woofer at the same point baffle roll-off started that would be one thing but it doesn't apply here.

**JCD** · **04-04-08, 03:40 PM**

I won't pretend to have as much knowledge/experience as either of you two, but this was in reference to a mtm design with a sub to fill out the bottom, so the crossover was supposed to be set ~2200 as well.

And I had the same reaction to his "anti"-BSC -- i.e., seemed like a prudent thing to do.

I'm slightly torn here myself, he is someone who I've got a lot of respect for, but BSC are kind of the accepted thing. I'll probably end up trying with and without at some point.

JCD

Advertisement