Depth of baffle step

[tvrgeek]

I was playing with several cabinet designs. Trying a very narrow tower (6 inch) which requires sharp edges as the driver is 5 3/4 inches. The baffle step was quite a challenge being almost 5 dB. When I added some width with a 3/4 radius round-over, of course the freq dropped as expected, but the step decreased several dB. I usually just am dealing with what is given, never played with it. Is this normal behavior? I have not seen anything talking about this is the texts.

[fastbike1]

This is often a stated reasons for going to wider baffles > > less to no BSC needed.

One could extrapolate this result from the fact that infinite baffles have no baffle step.

Perhaps this will help. Understanding Cabinet Diffraction ? Audioblog

[sreten]

Hi,

Physics indicates that baffle step is always 6dB. Real baffles also have
diffraction ripples as well, probably its these that you are reducing.

rgds, sreten.

[tvrgeek]

Guess those expensive 3/4 Radius and 1 inch radius bits I bought were not for naught! Theory says 6 dB, but I have never measured that much at 1 meter.

You can drive a screw with a hammer, but it is darn hard to pull out. me
If you think one size fits's all, go buy some shoes. me again

[planet10]

Originally Posted by sreten

Physics indicates that baffle step is always 6dB. Real baffles also have
diffraction ripples as well, probably its these that you are reducing.

But is countered by room gain. That, along with the ripple, make BSC a challenge.

dave

[sreten]

Hi,

Room gain issues are not relevant to the gist of the original post.

The bigger the baffle the further you need to measure to see 6dB.

rgds, sreten.

[AllenB]

tvrgeek, how are you making your measurements? By the way, the roundover is a good idea. Go bigger if you can.

[tvrgeek]

I measure several ways. Nearfield and 1 meter. If I know where it is going to be used, I will measure in the target location at the listening position. I use a "quick sweep" from True RTA, a pink, and an MLS using Sound easy. Every one gives somewhat different results as every one interacts with my office differently.

I finally built a router table so I can spin that 1 inch bit. It is high on my to do list after I work out my tweeter issues.

[AllenB]

Are any of these measurements gated? I ask because if the room is involved with your measurements you'll find significant 'goings on' in the lower mids and bass. You won't be able to pull your baffle out of the room, so to speak.

[tvrgeek]

Allen,
You are quite right. None are gated to the extent I would wish them to be. Hence, the differing results. Gated measurements have their own issues. You may put in a narrow pulse and gate the recording so as to eliminate boundary issues but, even if your drive signal is a nice pretty mathematically pure pulse, drivers don't behave that way, so you get skewed measurements do to the acceleration delays of the driver. This is one of the properties that led me to getting a beginning understanding of amplifier design as the envelop varied with different amps.

My measurement system is not exactly portable, so setting up to measure in the driveway, ( single near boundary) or hanging from a tree is most inconvenient. Parts of the results are better, but the ambient noise where I live makes other parts worse. If Creative can'r gent my e-mu 1616 drivers to quit crashing, I will switch to a M-audio 610 and when I get around to a new laptop, I will be sure it has a firewire port. I don't have as basement to fill with old army surplus pads.

If I had the ability do build such a device: a fast sweep with a tracking notch filter and a control for distance/delay could possibly work. If someone would like to write that code, they could be a hero. I can do the system level design, but I don't code.

The end result is we still have to voice by hand in the intended listening position. All these measurements just help us to get close to start with, or to identify what may be wrong when we listen.

Wonder what those push-button systems do that show up in AVR's do? I don't like the results, but the measurement could be useful.