Support for the DIY speaker builder

Topic Author
Posts: 6
Joined: Tue Jun 14, 2016 9:43 pm

Another ZMA Question

Wed Jul 06, 2016 7:43 pm

I should know this, but have managed to confuse myself.

For the ZMA file to be imported for each driver, in the case of a midrange or woofer, should the ZMA be a "free air" file, or should it reflect the impedance with the speaker in the (modeled) box?

If the former, is their a tool or utility to tell WinPCD some box detail so the impedance can be transformed appropriately?

Thanks and regards,

User avatar
Posts: 90
Joined: Thu Feb 21, 2013 1:16 am
Location: Canton, MA

Re: Another ZMA Question

Wed Jul 06, 2016 9:55 pm

You should use the impedance in the box. The box alters the impedance as seen by the crossover around Fcb (closed box) of the box. This may affect the interaction with the crossover if the crossover Fc is close enough to the box tuning.

WinPCD has no box model capability, at least not yet. I'd like to add this, but have not had time as I continue to add to the the crossover capability when I get the time. I am, however, starting to look into a box section or separate program. The box part shouldn't be nearly as difficult as the rest, but it still will take time. I started back with full time work in October, so my time is limited now.

Topic Author
Posts: 6
Joined: Tue Jun 14, 2016 9:43 pm

Re: Another ZMA Question

Fri Jul 08, 2016 12:05 pm

I appreciate what you've accomplished so far!

For my purposes, I would like a replacement for Response Modeler, but would want to see integrated room, boundary, and diffraction simulation.

Unfortunately, the Excel based RM has never worked right on my computer. Pressing buttons results in errors or operations that simply don't run. I'm going to ask Jeff if there's a way to run it w/o the macros.

I've also looked at Unibox and HornResp. The sim on the former stops at 1,000 Hz, so the higher frequency impedance model is missing. HornResp runs to 20 kHz. Below 1 kHz, HornResp, Unibox, and traced measurement agree within ~10% (sample size = 1). The model for voice coil inductance is different in HornResp and it predicts much higher Z values above 1 kHz than traced measurement graphs show.

In the short term, it's probably easier to just measure the drivers in-box and use that.

Who is online

Users browsing this forum: No registered users and 1 guest