need help with passive notch filter design

[xjr100]

Originally Posted by tvrgeek

The graphs above show all the symptoms of baffle step hump at about 650. A notch is not the best way to deal with it.

More than agree with that.
I had one project, where was about same picture ~600Hz. I've put notch there, got almost flat response, but after listening I removed it.

[jimbones]

Originally Posted by xjr100

More than agree with that.
I had one project, where was about same picture ~600Hz. I've put notch there, got almost flat response, but after listening I removed it.

I have to run the edge again. But based on that graph I also agree with you. It needs BSC not a notch. However, when I ran a quick sweep test and from what I remember the real world measurements and graph (edge) are different. I will redo the actual measuremenmts.

BTW, I am new at this so your help is really appreciated and if I don't quite understand somethings you'll understand why .

My measurement technique is to use a sine sweep and behringer RTA. I also have the ability to use pink noise but the measurement jumps around a bit so I prefer a sweep. My calibrated mic is about 5 ft from the speaker about 3 ft off the ground. Is this correct way of doing it? Thanks Much!~!~~~

[sreten]

Hi,

Here is an example of a parallel notch filter in series with the drivers,
to deal with the "dipole peak", which is not the same as the baffle
hump you get with a closed back baffle.

https://sites.google.com/site/undefi...diy-sunflowers

rgds, sreten.

[tvrgeek]

Originally Posted by jimbones

I have to run the edge again. But based on that graph I also agree with you. It needs BSC not a notch. However, when I ran a quick sweep test and from what I remember the real world measurements and graph (edge) are different. I will redo the actual measuremenmts.

BTW, I am new at this so your help is really appreciated and if I don't quite understand somethings you'll understand why .

My measurement technique is to use a sine sweep and behringer RTA. I also have the ability to use pink noise but the measurement jumps around a bit so I prefer a sweep. My calibrated mic is about 5 ft from the speaker about 3 ft off the ground. Is this correct way of doing it? Thanks Much!~!~~~

If you have a calibrated mic, you are 99% there. Now download ARTA so you can do gated MLS measurements. Sweep (I use TrueRTA quick sweep) is fine for low bass, but everything else you need a gated measurement so you are not measuring the room, but the speaker. Of course, if yo can put your speaker on a 50 foot pool in a quiet valley, then sweep works fine.

It is an important discovery that the simulations do not always agree with the measurements. Two reasons. Measurements are not easy and simulations are only so good. Both are informative. The craft in speaker building is learning to interpret both and match that with your ears.

The best resource to really understand speaker measurements is D'Appolito Measuring Loudspeakers. Technical for beginners, advanced users know it. It is how you get from beginner to advanced. It really helps to understand what the measurements mean. Otherwise, you won't know when the model or measurement makes no sense.

I measure nearfield, one foot, 1M and listening position. I also measure on axis and at 15 degrees. I am voicing for on-axis, but looking to "fix" off axis. Subs I just measure at 1M. No reason to play around with them.

This forum is fantastic at helping those who put in effort and listen. A few of us do get a little rough on people who won't put in their own effort ( That big Y in DIY) or who are convinced their bright idea that violates the laws of physics has to be correct even though they have not tried it. I am a bit of a newbee, as I have only been at it for about 35 years off and on. I do have an advantage of a degree in electronics so struggling with the technical papers from those who really do know is not quite as tough as it once was. There are quite a few folks here who make their living at this. Somehow accountants seem to have difficulty with imaginary numbers and "S" planes.

[xjr100]

Originally Posted by jimbones

My calibrated mic is about 5 ft from the speaker about 3 ft off the ground. Is this correct way of doing it?

I never had experience with open baffle projects, but supppose you must use 2 type of measurements: one is MLS as tvergeek offered, another similar to yours, as open baffle interacts with the room. Which combination in "flatnes" is best will be found after many listeting tests.

MLS will show you errors in crossover design as well baffle reflections - at the beginning you can understand, what is speaker response in "simulated anechoic camera". But as it is OB, you must controll, how it interacts with the room.

In your measurements - nearfield is mic directly to cone - this is for you woofer, far field - about 1m from speaker on-axis. Recommend you to adjust height, that any part which can reflect (floor, ceiling, walls) must be on longer signal way. If you measure speaker at 1.6m and mic is 1m from floor you lower relable MLS measurmet will start from ~360Hz, but if you rize the speaker to 1.3m and measure from 1m it will be ~200Hz.

[jimbones]

TVR thanks for the honest (and sometimes humorous) response. I downloaded arta. Now I should give a complete picture of my equipment: I use a laptop. My mic is one that came with a Behringer DSP8024.

I looked at the Behringer manual I don't think it does gated measurements. Soooo, from doing a quick read of the arta help screen it looks like the mic gets connected to the laptop soundcard? Well my mic has balanced connectors. I will need to take more time to read the help (the Y in DIY ) or of course you can tell me if I am on the right track . My first thought is i am going to have to find a way to go from balanced to mini plug? or just buy another mic?

Just curious, can I use the arta as a gated generator and do measurements with the Behringer?

[jimbones]

Opps never mind the arta help screen says NG with the laptop