diyAudio (
-   Software Tools (
-   -   FREE Passive Crossover Design 7 software released (

CharlieLaub 3rd June 2010 08:44 PM

FREE Passive Crossover Design 7 software released
NOTE: I am posting this on behalf of Jeff Bagby, the programmer and guru behind this excellent free tool for crossover design (it also does active crossovers!).


Passive Crossover Designer 7 Released!

Good afternoon! I am pleased to announce that Passive Crossover Designer 7.0 (PCD 7) is now available.

I did not change the electrical voltage divider and circuit within the program this time, but dramatically changed how the program functions in Excel.

I got perturbed at Microsoft for the way they changed things in Excel 2007, so in order to make the program function more smoothly I decided to make it work more like an Excel file and less like a "program", although it still does all that is did before.

Some of you will love the fact that you can see sheet tabs, column headers, and be able to unprotect the main page and change chart settings on your own (but be careful not to mess something else up, you're on your own there).

Some will go gaa-gaa over the fact that I am now allowing you to save the file under a new name with all data, frd, and zma file intact. My personal version have worked this way for a long time, but now I will allow that for other users too. The files are about 5mB, but hard drives are pretty big these days. . The old way was a security feature, but I decided to relax some of that for version 7.

Below is a list of all of the changes or version 7. Please read them carefully so you know what you are dealing with. Please make special note of the Z axis change in blue. Missing this will change your results considerably.

As before, there are versions optimized for Excel 97-2003 and for Excel 2007-2010 available.

I would really appreciate it if those of you who post on other forums would link this announcement for me.

Special thanks to Charlie, as always, for hosting the files.

Jeff B.

Changes for Passive Crossover Designer Version 7.0

The new open view of the file will show column and row headers, sheet tabs, and formula bar. The ability to switch to a full-screen view with these features hidden is still present by clicking the "Full View" button under the Chart Manager.

This version offers the ability to save the file with all data loaded under a new name and be able to reopen the file with the project intact. Simply click on "save as" in Excel, and give it a new name for your project. The spreadsheet will compress and save containing all data including frd and zma file data. When the file is reopened, the project will be exactly as it was when filed. (Files saved in this way will about 5mB in size)

Closing the file will no longer shut down the Excel application. All other open files will remain open.

Removed the System Description page. This page was causing too many issues with screen updating in Excel 2007 and was being used by too few users to justify maintaining it.

Built all of the old schematics and crossover layout pictures into the program rather than external and accessible through hyperlinks. There are no longer additional file folders that need to accompany the program to view these layouts. The buttons to access these layouts are in each driver section button grouping.

Added the ability to import second woofer and second midrange frequency response and impedance data files, so five different drivers can be imported if desired. (Note: If, however, you wish to model two woofers or two midranges and do not want to import a second set of data, that is OK, as the program will automatically use the original data as it did before as a default. The ability to import the second set of data is under the selection for two woofers and two midranges and is only an option to be used as desired. This allows you to import two different diffraction responses for drivers in an MTM, for example. Or you can import two completely different drivers with different frd and zma file if desired as well.)

Changed the Z-offset convention to be negative if the acoustic center is "behind" the zero plane, as in the case of the woofer's acoustic center being behind the tweeter. (Note: All .csp or saved project files earlier than version 7 will be automatically be changed to the negative of the stored value when imported. Version 7 files will not be changed but will remain exactly as saved.) (Note: dlr, this works the same as CALSOD now with the same geometric entries )

Corrected a problem in the vertical axis adjustment. It was originally upside down. This has been fixed. Now, if you set the woofer to be -.150 for a Y offset and set the vertical axis to -.150, for example, you will be aligned with the minimum phase axis of the woofer again and the phase will show no excess delay, as it should.

Refined the driver directivity and power response approximation models for greater accuracy.

Removed the separate hidden phase graph (used in earlier versions) since a dedicated phase chart can be printed and saved from the main graph's settings now.

Added buttons to remove the woofer or the tweeter (or both) from the simulation if desired for greater versatility. This can allow for modeling of single driver systems on the main graph, or to allow for simply modeling a woofer or tweeter's response individually on the main graph if desired.

Added special buttons to compress all calculations or to re-expand the calculations, if desired. (Note: when saving the file, it will compress the calculation automatically. If you wish to continue working in the same project you must click the "Expand" button to restore the program.)

Added a button to unprotect the main page of the file for better chart management. This was a tough decision to make because when the page is unprotected it is possible to inadvertently make other changes that would break or alter the program's function. Because of this nearly all other function buttons will reset the protection automatically. However, while unprotected you can select a chart, view the value for a given point, and even change the scale, line weight, or line color, etc. (Note: The end-user is responsible for all changes made to the program while the main page is unprotected.)

Made the textbook calculators more accessible by adding button to each to the button group in the Woofer, Tweeter, and Midrange sections.

Added Pop-up warning boxes for you to confirm your actions when clearing the crossovers or resetting the project. This ensures that you will not clear what you are working on accidentally.

A "Quick Guide" instruction manual is now built into Version 7 accessible via a worksheet tab at the bottom, so there is no additional download for that.

There is an additional download available that includes a 35 page User Guide in pdf file format and several sample files that can be imported to demonstrate the programs features. This User Guide does not include Version 7 updates, but it is still helpful in understanding the program and providing sample projects to experiment with.

Files are available here:
Jeff Bagby's Loudspeaker Design Software

Jeff B.


jayam000 15th June 2010 03:16 AM

is it possible to design only HF crossover for tweeters?

stewin 25th May 2012 11:48 PM

thanks jeff . is your design 6db 12db 18db or 24db

CharlieLaub 26th May 2012 03:33 PM

I noticed your question - I don't think that Jeff visits this forum. You might go to the PCD software page - his EMail is posted there. If that doesn't reach Jeff, then you can find him on the Parts-Express TechTalk forum. He posts there frequently.

Here's my thoughts: the slope that you achieve depends on the components you use to create the electrical filter, plus the slope of the driver response. These combine to give you an overall "acoustic slope". I can't recall what the maximum filter order for the electrical filter is. Just download the software and test it out for yourself (read the manual first!). Many DIYers on the techtalk forum use this software for designing their speakers, and could answer your questions better than I could.

Have fun,


nannoo 18th February 2014 04:01 PM

To avoid starting a new thread, how accurate is this tool?
Out of interest sake, I tried to model an existing known crossover design and came up with a very bizarre looking graph.

The crossover I modelled was that from the B&W XT4 3-way.

The main eyebrow-raiser is the midrange xo unit, which has a first order low pass (1.2mH inductor) and second order high pass (47mF cap, 1.1mH inductor). According to the software designer this results in a massive resonance in the response of the filter circuit at around 850Hz (based on a load of 8ohms, which these drivers approximate to).

Any ideas?

Should I start a new thread to discuss this?

Neil Davis 18th February 2014 09:56 PM


Originally Posted by nannoo (
To avoid starting a new thread, how accurate is this tool?

See my reply to jake9115 in a previous thread:

The PSD-Lite program is a different implementation than Jeff's PCD program, but it does the same thing--it solves the crossover network equations to calculate the voltage at the driver terminals for a given input. Then it uses the driver transfer function (from the "FRD" measured response) to determine the system response. If you have good measurements of the driver impedance ("ZMA" file) and driver response, these modelers are very accurate, and they have been validated many times. However, if you don't have good data on the driver impedance and driver response, these programs won't be useful.

nannoo 19th February 2014 12:48 AM

D/loaded, I' tried to model with the impedance measurements for the drivers taken with ARTA, (in PCD 7)
No matter what I do, same res. peak.
Tried in this new recommended software, same issue.
Not sure how to resolve, no matter what impedance value I use for the mid (have tried using various fixed as well as the measured values) still has the same peak, just different variations of.

nannoo 19th February 2014 01:05 AM

>>> Mystery solved, was simply a case of the topology, where B&W for some reason spec the series components first then parallel in there x/o designs... bizarre.

wintermute 19th February 2014 01:09 AM

Try adding a little series resistance to any shunt caps and see whether that removes your peak. If you model with perfect caps you will likely get resonant peaks :) A real world implementation may not need the extra resistance (but sometimes it will if it was larger, say 1 ohm or more).


Bill poster 15th April 2014 08:34 PM

Total Noob question, i've just downloaded your passive crossover program and reading the instructions - how to 'unprotect' the excel sheet and start a project? Its asking for a password i don't have.

All times are GMT. The time now is 10:23 AM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 18.75%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio