I re-did the zma file again with the woofers mounted on the baffle... it worked this time.
I added an extra inductor, and it looks good. I've built the one with only one 1mH and it measures well and sounds ok too. Will probably be better when I get another extra 1mH inductors.
I added an extra inductor, and it looks good. I've built the one with only one 1mH and it measures well and sounds ok too. Will probably be better when I get another extra 1mH inductors.
Attachments
Beware using published response curves in Xsim
I recently completed a pair of 3-way TLs using Volt 15" bass and 3" dome mid plus a Scanspeak tweeter using Xsim for the crossover design. What a brilliant software application! My first crossover was based on the manufacturers' published curves converted to frd and zma using FP Graph Tracer. The result was a ruler-flat response - see attached picture. Too good to be true? YES
Then I measured the drivers in cabinet with Omnimic and DATS. to create real frd and zma data. When I applied these to the original crossover design the results were not what I wanted - see attached picture.
I tried various crossover designs and the best I could produce gave the results in the third picture (the mid was measured inverted). The final crossover is very different from the first attempt. I constructed it and listened. The completed speakers sound very natural and revealing but the treble was a bit strong so I deviated with a 1ohm series resistor to the tweeter, which produced a more natural sound (mainly orchestral and choral music).
My purpose for posting this is to point out how important it is to measure the drivers rather than rely on published date from the manufacturers. I guess most of you know that already.
I recently completed a pair of 3-way TLs using Volt 15" bass and 3" dome mid plus a Scanspeak tweeter using Xsim for the crossover design. What a brilliant software application! My first crossover was based on the manufacturers' published curves converted to frd and zma using FP Graph Tracer. The result was a ruler-flat response - see attached picture. Too good to be true? YES
Then I measured the drivers in cabinet with Omnimic and DATS. to create real frd and zma data. When I applied these to the original crossover design the results were not what I wanted - see attached picture.
I tried various crossover designs and the best I could produce gave the results in the third picture (the mid was measured inverted). The final crossover is very different from the first attempt. I constructed it and listened. The completed speakers sound very natural and revealing but the treble was a bit strong so I deviated with a 1ohm series resistor to the tweeter, which produced a more natural sound (mainly orchestral and choral music).
My purpose for posting this is to point out how important it is to measure the drivers rather than rely on published date from the manufacturers. I guess most of you know that already.
Attachments
anyone know where i can get a newer version or something? i have installed it 20 times even as ADMIN in WIN10 and it wont load. It just says
"The code execution cannot proceed because hhctrl.ocx was not found. Reinstalling the program may fix this problem."
Again i have tried installing it 20 times already
"The code execution cannot proceed because hhctrl.ocx was not found. Reinstalling the program may fix this problem."
Again i have tried installing it 20 times already
Desciption in Win7 for that hhctrl.ocx file is "Microsoft HTML Help Control" that could point to maybe investigate if there are any free feature package for Win10 that make it more compatible with older program codes help system or go to installed Windows objects (probably something as this path: Controlpanel/Programs/Windows objects) and see if anyone in this regard is missing there.
Other suggestion could be manual force Win10 to look for updates and see if anything is missing, and also one can try install in combability mode where you right click installation file and pick its policy to run as for example Win7 or older OS.
Other suggestion could be manual force Win10 to look for updates and see if anything is missing, and also one can try install in combability mode where you right click installation file and pick its policy to run as for example Win7 or older OS.
Sounds like a file missing (or more likely corrupted) in your win10 installation. Funny thing is that Xsim doesn't even use HTML help, but it might have been referenced in a dependencies list somewhere.
Try searching your hard drive for that file (hhctrl.ocx) and see if it's there. If it is, then try getting a copy of that file from another win10 computer and replacing yours with that.
I read that the early released October update for Window10 was pulled back because it was accidently deleting some of the files on the computer. Might have been caused by that? Or, another possibilities is that it might have gotten eaten by a computer virus -- you might want to run a virus scan.
Try searching your hard drive for that file (hhctrl.ocx) and see if it's there. If it is, then try getting a copy of that file from another win10 computer and replacing yours with that.
I read that the early released October update for Window10 was pulled back because it was accidently deleting some of the files on the computer. Might have been caused by that? Or, another possibilities is that it might have gotten eaten by a computer virus -- you might want to run a virus scan.
WHAT?
I'd laugh but it's not even funny.
One of the reasons why I never update right away. But, I heard that you have no choice with Win10. You can't turn off automatic updates... is that true?
I'm on a Mac, but two OS versions behind, no intention of updating any time soon.
Also have a PC box, and that one was recently updated to Win 8.1 ... from XP.
Yes.
There are some tricks to prevent automatic updates, but nothing really works 100%. To make things worse it forces automatic reboot to install updates. And it is really slow when installing updates.
October update broke many things.
Hello,
I just installed XSim 1.2.00 on my new WIN10 system (4K screen with taskbar at the top of the screen). Unfortunately, it is one of those apps that always open with window borders behind task bar. Can't move or resize the windows of XSim, unless I put the taskbar to the bottom. Any fix?
I just installed XSim 1.2.00 on my new WIN10 system (4K screen with taskbar at the top of the screen). Unfortunately, it is one of those apps that always open with window borders behind task bar. Can't move or resize the windows of XSim, unless I put the taskbar to the bottom. Any fix?
^Frequency data export of MLS measurement for example with 48k sample rate, 64k-256k length and 2048 pts. export is problem for XSim due to duplicate frequency points.
Log chirp stimuli is recommended anyway so this shouldn't be a problem in practice. 3rd decimal to frequency value might be added to next Clio 12 revision.
Log chirp stimuli is recommended anyway so this shouldn't be a problem in practice. 3rd decimal to frequency value might be added to next Clio 12 revision.
Last edited:
Here's another way...
Go online to calculate the parts for the XO's order and frequency.
Add a driver in XSim. If it's 1st or 2nd order, XSim already has circuit blocks for it.
For a 4th order LR, just add the capacitors and inductors and you'll have your curve showing up in XSim.
Now, add your driver and try to match it. Adjust level to match your real-life driver.
For example, here's a 4th order LR at 500Hz.
Go online to calculate the parts for the XO's order and frequency.
Add a driver in XSim. If it's 1st or 2nd order, XSim already has circuit blocks for it.
For a 4th order LR, just add the capacitors and inductors and you'll have your curve showing up in XSim.
Now, add your driver and try to match it. Adjust level to match your real-life driver.
For example, here's a 4th order LR at 500Hz.
Attachments
Bill,
I wonder if this is easy, hard or totally impossible to implement in XSim, but...
Would it be possible to have XSim export a convolution file based on the parts in the XO for each driver?
For fullrange notches, that would be 2 files, for a 2-way, that'd be 4 files, etc...
That way we could actually hear the way the XO would sound in the real world.
I remember one of the more expensive XO software being able to generate an audio out with the applied XO parts. Forgot which one... sorry. Maybe LEAP or LspCAD?
Anyway, I thought this would be a killer addition to XSim.
I wonder if this is easy, hard or totally impossible to implement in XSim, but...
Would it be possible to have XSim export a convolution file based on the parts in the XO for each driver?
For fullrange notches, that would be 2 files, for a 2-way, that'd be 4 files, etc...
That way we could actually hear the way the XO would sound in the real world.
I remember one of the more expensive XO software being able to generate an audio out with the applied XO parts. Forgot which one... sorry. Maybe LEAP or LspCAD?
Anyway, I thought this would be a killer addition to XSim.