Automatic crossover designing with XMachina

Now for angles, I measured vertical -180 to +180 in 10 degree increments,

But for horizontal I only measured 0 to +180 in 10 degree increments, does Xmachina mirror missing angles like VCad?
Yes, but you have to tick appropriate options. In this case
10_mirror.jpg
 
For now, I have made an index of descriptions of functionalities that were implemented after the creation of the first (and only) pdf. I think that I will compile it into a new pdf, so if any of the descriptions need to be clarified, expanded or corrected, please let me know.

Index of features not included in the pdf:
  • Power Response target with off axis measurements LINK
  • System off-axis responses on charts LINK
  • Polar Maps LINK
  • Automatic Curve Reader LINK
  • Parallel processing LINK
  • Force common GND filter architecture LINK
  • Design time manager LINK
  • Impedance phase equalization LINK
  • Impedance calculation up to 200kHz LINK
  • Target follows the woofer at LF LINK
  • Optimise for crossover price LINK
 
Last edited:
  • Like
Reactions: 1 users
Any chance to add a provision for "corner" in power resp area (1/4 sphere or 1/4 cyl)? Say we put in curves to +/- 45 deg horizontal and it can skip all the wider angle calcs, etc. We just would need to know what it thinks the data is / if it lifts +6 or whatever and we can adapt in measurement methods/levels if you don't want to. In any case, thank you again for the great tool!
 
In fact I was hesitant to consider even a 1/2 space approach because even if the design system is to operate close to the baffle, the backward energy will not be absorbed, at least not entirely, so it should not be completely ignored. I have eventually implemented 1/2 space only because these cases were analyzed in the Tylka paper. However if there was any practical use for the 1/4 space approach it would be technically doable, why not. Either way, it would have to wait until the next software update campaign, and this might be not very soon.
 
  • Like
Reactions: 1 user
No problem. Thank you for taking the time to think about it. Sincerely appreciate all you have done. vcad can do some things, but I like your software a great deal and was hoping there could be a way to stop it trying to extrapolate so much (and asking for confirmation in the batch jobs). I realize that corner speakers are not common, but they are what I love :) Much gratitude to you (and to Kimmo, for that matter)!
 
Please note that the only place to download XMachina is https://xmachina-ai.blogspot.com/p/current-version-info.html.

I was notified by one forum user (who is an admin on another forum) that someone who seemed to imply he was the owner/developer of XMachina posted a link redirecting to a suspicious area. The post was deleted because the provided link raised a serious security alarm.

Unfortunately, this may happen again and I don't see any way to prevent it. So be careful. Never download XMachina from a place other than xmachina-ai.blogspot.com. Do check the signatures of the downloaded file before proceeding.

It is worth adding here that it is generally a good practice to run freeware programs from the Internet using the sandbox functionality or to use a dedicated computer for this purpose.

Stay safe.
 
  • Like
Reactions: 4 users
Greetings,

I've had a project in my mind to make a 3Way Speaker, and first thing I did was download this brilliant tool, and enter the Frequency Curves for all 3 Speakers via Images from companies technical data (which took me some time but anyways). And I have three questions (I'm a f*cking newbie hahah) :

Speaker Project.PNG


- From what you can see, the midrange (BliesMa M74B-6) is much louder than the two others (PurIfi PTT8.0X08Ohm & Raven Point Source AMT Driver), so how do I make my target SPL ? (I'd like the crossoverpoints at around 300Hz and 9 000Hz, I guess that's what's best as they would be used as for my HiFi and Studio Monitor setups).

- Is that Target SPL will be the Sensitivity Level (XdB/W/m) or has it nothing to deal with ?

- How do you make a TXT file (which informations and in which which order mostly) for Capacitors, Inductors and Resistors ?

Apologizes for those newbie questions, may be they've had already been answered, and don't hesitate to tell me if I have to make a special thread for that.

Many Thanks.
 
- From what you can see, the midrange (BliesMa M74B-6) is much louder than the two others (PurIfi PTT8.0X08Ohm & Raven Point Source AMT Driver), so how do I make my target SPL ? (I'd like the crossoverpoints at around 300Hz and 9 000Hz, I guess that's what's best as they would be used as for my HiFi and Studio Monitor setups).

- Is that Target SPL will be the Sensitivity Level (XdB/W/m) or has it nothing to deal with ?
Click your system target as you like, for example a flat line. System target domain is dBspl. But remeber not to draw the target above driver's spl (in its usable range). XMachina is a passive crossover designer so it can easily attenuate a driver but it can't amplificate it.

How do you make a TXT file (which informations and in which which order mostly) for Capacitors, Inductors and Resistors
for example, write something like below in notepad and save it as txt. (Price and description are optional)

C $price desc
1uF 2.20 MKT
4.7uF 2.70 MKT
100uF 1.05 Bipolar
...

Then click "Load List" from the context menu.
Or use "Generate typical values", there is no need to crate the file then.

may be they've had already been answered,
Actually it's all in the pdf and YT examples, but no problem to clarify again.
 
I don't see any good solution at the moment. That would require some sort of circuit analyzing procedure that would check every planned modification whether it will result with shunt resistors. That would dramatically slow down XMachina.

Some workaround that i can think of (and you're probably aware of) would be to setup the task to deliver several versions of the circuits and choose the one in which such a resistor is not present. If it appears in every solution maybe adding capacitor in series wit it will help. Adjust the capacity so the resistor will do the job in the midrange but it will be disconnected in the LF region.
 
I was asked in another thread but I'll answer here too.

XMachina is not available for download until the issue with google drive gets clarified.

I've got an automatic notification (from kind of "no-reply" email address) that the file may be not compliant with terms of google drive service. The message said that if I think that's a mistake I can click a button provided with the email and then a "trust and safety team" will analyze the case. I clicked the button but I haven't received any feedback since then (it's been a week now).
 
There are tons of free file sharing platforms that would do the same job Google Drive does. Would you like me to help you setup something. Here are some names.

  • 1. Jump Share
  • 2: AirDroid Web.
  • 3: Dropbox.
  • 4: iCloud.
  • 5: Box.
  • 6: WeTransfer.
  • 7: One Drive.
  • 8: MediaFire.
 
I doubt they will respond. They do not wish to give out any info on their security system. Or what they are looking for in uploaded files. Its a bit like asking the US visa office why somebody's visa was declined.
In such cases I run the program on a machine with mcafee end point security. It monitors all activity and reports everything the app does.
You could also try compiling the program on another machine. And see if it comes out the same.