W5-704s Design Help Needed

[69stingray]

Hello everyone, I am trying to design a small bookshelf speaker using Tang Bang's W5-704s speaker and Dayton Planar. I am using the FRD tools (http://www.pvconsultants.com/audio/frdgroup.htm )

I already have the planar, and therefore, would like to crossover as high as possible, I am thinking around 4,000 Hz.

Zaph was nice enough to post the T&S parameters and freq. response from a set a he has.

Tang Band w5-704s

I traced his response and was about to get a nice 4,000 Hz crossover using a parallel trap filter and a 3rd-order Butterworth crossover.

The problem I run into is when I add in the Baffle Step. My baffle is 11 in x 17 in and the hump in the baffle step matches perfectly with the hump in the frequency response, giving me 10 dB hump to deal with.

I have tried a RL contour network, RCL trap filter and even the "baffle step" network, all with no luck. The best I can to is get the hump down to 4-5 dB. I was hoping the slope of the crossover would smooth it out, bit it didn't.

My baffle is just about the same size as Part Express's pre-fab baffle and a lot of people use them without issues. I am doing something wrong here? If not, any crossover suggestions?

Baffle Sim:


Combined Response:


Combined Response Imported into Passive Crossover Designer:

[MarkMcK]

I have no idea why your simulation is generating these predictions.

I do have experience with the transducer and would not expect you to be able to experimentally verify the predictions.

I might suggest playing with other configurations to see what can change the prediction.

Mark

[Zaph]

Hmmm, something is wrong with your diffraction simulation. Check your numbers. If your baffle is 11" x 17", the first diffraction ripple peak should be below 1kHz. You show it at 2kHz. I don't think PE has any boxes of that size. 11" x 17" seems kinda big for a 5.5" woofer.

I find this woofer's natural response to integrate nicely with most normal size boxes for *very* smooth response.

You might consider the .5 cu ft box (9" x 14") vented or the .38 box sealed. If that's your simmed-in bass tuning, it looks a little peaky also.

I haven't used Paul V's program in a while, but try making it look like this:

[69stingray]

Thanks everyone- I think I may have figured it out....real stupid...I may have done the baffle sim in metric... In the back of my mind I knew it had to do something with the baffle sim. I should of done the back-calculations sooner to verify the width vs. freq. increase.

John- that is the predicted enclosure. Right now it has +3 dB @ tuning just because I know my brother-in-law like the bass. I think I am going to move the port to the back, if I do, I will remove the +3 dB because he should get reinforcement from the walls/room.