Chuck was #11.
Jack was #9 and i have his schematic.
EQ was #2 and Jack has his boards. Maybe i can get it from EQ though.
Jon, i believe the breakup on mine was lower initially as I'm seeing it. I was 38dB or lower above 3.2kHz. You had 38dB at 3.5kHz, but yours did continue to tank hard. So, I was only ahead for 300Hz.
I am interested in a graphed comparison, should you so choose.
@dcibel said:
Just make sure you include 60us of delay for the woofer when you generate those curves
I'd be happy to send you woofer data with the 60us delay included and nearfield / port response included for the low end, if you want.
That would be great. Just zip up the corrected woofer files and attach them. I'll unzip and replace all the woofer files in my model with your corrected ones. This will save me the time of including the delay and then re-doing my NF/FF/Diff model again.
I did not notice this delay problem when I created my VituixCad model, as I only used the polar data set when working on this model. However, several months ago I put together a simple XSim model using the on-axis offset files and I thought it was strange that zero model delay was required when I did the "get file" matching procedure. The delay was already included in the files, which never happens when I use OmniMic.
I did not notice this delay problem when I created my VituixCad model, as I only used the polar data set when working on this model. However, several months ago I put together a simple XSim model using the on-axis offset files and I thought it was strange that zero model delay was required when I did the "get file" matching procedure. The delay was already included in the files, which never happens when I use OmniMic.
Exactly right. You will find the same result with your new mic and that fancy jig you've built
Attached the corrected and spliced woofer responses. These can be loaded directly to VituixCAD, zero offset of z axis required, just 145mm y axis offset.
I have the VituixCad model set up with dcibel's corrected woofer data files and am beginning to enter crossovers. I have a clarification question for Billet (the other Bill S.). I have your xover entered as variant #6 and I get a very good match to your posted impedance curve and on-axis FR graph when the tweeter polarity is NOT reversed. But your schematic shows the tweeter reversed. If I reverse it, I get a huge suckout near the xover (6498Hz). Would this be a simple typo?
My sim certainly has the tweeter polarity reversed. It is not a typo, it is screen shot from Boxsim. I also checked it in WinPCD, it indicates that the tweeter should be reversed.
I tried adding a 25mm offset to the woofer and I now see what you are seeing. A large dropout from about 4500 to 7000 Hz, no wonder it sounded dull...
It seems that my crossover may have been out of phase? Isn't that funny...
It involves notching the response just to the other side of the xover point in the out of bandwidth areas for the drivers. This steepens the rolloffs albeit with a bit more ripple than conventional methods outside of that range. Basically, it's a type of cauer-elliptic filtering.
So my thought other than playing with off axis averages, was to tank the responses with 8th order slopes. I took a bit more liberty with the woofer break up and pushed down the ripple.
@4thtry said:
I have the VituixCad model set up with dcibel's corrected woofer data files and am beginning to enter crossovers. I have a clarification question for Billet (the other Bill S.). I have your xover entered as variant #6 and I get a very good match to your posted impedance curve and on-axis FR graph when the tweeter polarity is NOT reversed. But your schematic shows the tweeter reversed. If I reverse it, I get a huge suckout near the xover (6498Hz). Would this be a simple typo?
When you publish these sims, can you show mine with the tweeter in the "correct" polarity? I certainly did not intend to have a large on-axis dip at 6.5k.
@4thtry said:
I have the VituixCad model set up with dcibel's corrected woofer data files and am beginning to enter crossovers. I have a clarification question for Billet (the other Bill S.). I have your xover entered as variant #6 and I get a very good match to your posted impedance curve and on-axis FR graph when the tweeter polarity is NOT reversed. But your schematic shows the tweeter reversed. If I reverse it, I get a huge suckout near the xover (6498Hz). Would this be a simple typo?
When you publish these sims, can you show mine with the tweeter in the "correct" polarity? I certainly did not intend to have a large on-axis dip at 6.5k.
@a4eaudio said:
Bill, I know you said you were doing that for your notes but will you post a summary here? Either the VituixCAD 6 pack for each and/or overlays of the SPL curves?
Update: I currently have 8 of the 11 xovers entered into the VituixCAD crossover variant tabs. Still waiting for 3 schematics. In the meantime, here is the VituixCAD project file with all the xovers. It has the extension .vxp when unzipped. All you need to do is open this file in VituixCAD and then 1) re-load the woofer driver files with dcibel's corrected .txt files, and 2) re-load the tweeter frd files with the original spinorama set and 3) re-load the woofer and tweeter zma files from the original set. You will then be able to quickly compare all crossovers by clicking through the variant tabs. Each xover is named for reference.
Updated comparison picture show to include above entries isaeagle and Jack. I've also renamed the files with bag # included. Just missing the winner now..
Comments
Just make sure you include 60us of delay for the woofer when you generate those curves
I'd be happy to send you woofer data with the 60us delay included and nearfield / port response included for the low end, if you want.
I was number 10:
No delay in my Xsim curves, splice, min phase, derive offset
Chuck was #11.
Jack was #9 and i have his schematic.
EQ was #2 and Jack has his boards. Maybe i can get it from EQ though.
Jon, i believe the breakup on mine was lower initially as I'm seeing it. I was 38dB or lower above 3.2kHz. You had 38dB at 3.5kHz, but yours did continue to tank hard. So, I was only ahead for 300Hz.
I am interested in a graphed comparison, should you so choose.
InDIYana Event Website
Wolfy's Photobucket
That would be great. Just zip up the corrected woofer files and attach them. I'll unzip and replace all the woofer files in my model with your corrected ones. This will save me the time of including the delay and then re-doing my NF/FF/Diff model again.
I did not notice this delay problem when I created my VituixCad model, as I only used the polar data set when working on this model. However, several months ago I put together a simple XSim model using the on-axis offset files and I thought it was strange that zero model delay was required when I did the "get file" matching procedure. The delay was already included in the files, which never happens when I use OmniMic.
Exactly right. You will find the same result with your new mic and that fancy jig you've built
Attached the corrected and spliced woofer responses. These can be loaded directly to VituixCAD, zero offset of z axis required, just 145mm y axis offset.
Ben I'll email you my FRDS
I have the VituixCad model set up with dcibel's corrected woofer data files and am beginning to enter crossovers. I have a clarification question for Billet (the other Bill S.). I have your xover entered as variant #6 and I get a very good match to your posted impedance curve and on-axis FR graph when the tweeter polarity is NOT reversed. But your schematic shows the tweeter reversed. If I reverse it, I get a huge suckout near the xover (6498Hz). Would this be a simple typo?
My sim certainly has the tweeter polarity reversed. It is not a typo, it is screen shot from Boxsim. I also checked it in WinPCD, it indicates that the tweeter should be reversed.
I tried adding a 25mm offset to the woofer and I now see what you are seeing. A large dropout from about 4500 to 7000 Hz, no wonder it sounded dull...
It seems that my crossover may have been out of phase? Isn't that funny...
I had to look up NTM style crossover, I hadn't heard of that one. Thanks!
Yeah, what exactly is an NTM style crossover?
Neville Thiele Method.
It involves notching the response just to the other side of the xover point in the out of bandwidth areas for the drivers. This steepens the rolloffs albeit with a bit more ripple than conventional methods outside of that range. Basically, it's a type of cauer-elliptic filtering.
InDIYana Event Website
Wolfy's Photobucket
So my thought other than playing with off axis averages, was to tank the responses with 8th order slopes. I took a bit more liberty with the woofer break up and pushed down the ripple.
https://bssaudio.com/en/site_elements/whiseworks-white-paper#:~:text=A Neville Thiele MethodTM Crossover,(pronounced%20%E2%80%9CTeel%E2%80%9D).
When you publish these sims, can you show mine with the tweeter in the "correct" polarity? I certainly did not intend to have a large on-axis dip at 6.5k.
Thanks!
No problem; will do.
Who let these guys in?
InDIYana Event Website
Wolfy's Photobucket
Is that a GQ fashion shoot?
You guys are working hard not to smile.
Hey, Nick- looks like we missed a photo of your xover for the theme. Neither Bill nor I have it.
InDIYana Event Website
Wolfy's Photobucket
Sorry guys, I already tore it apart and put the parts away. It was nothing to look at any way.
I mean...it was bad enough ya had to put a bag over it😝
Yep, it definitely needed a dark room.
No one is looking at the mantle while their poking the fire.
Update: I currently have 8 of the 11 xovers entered into the VituixCAD crossover variant tabs. Still waiting for 3 schematics. In the meantime, here is the VituixCAD project file with all the xovers. It has the extension .vxp when unzipped. All you need to do is open this file in VituixCAD and then 1) re-load the woofer driver files with dcibel's corrected .txt files, and 2) re-load the tweeter frd files with the original spinorama set and 3) re-load the woofer and tweeter zma files from the original set. You will then be able to quickly compare all crossovers by clicking through the variant tabs. Each xover is named for reference.
I think I'll wait for the movie version to come out...
Picture show attached.
Thank you
This is Jack's entry^^^
He apparently used Xover Pro for the design. I just drew it up in Xsim for clarity.
InDIYana Event Website
Wolfy's Photobucket
Here is my schematic, Entry 11
Updated comparison picture show to include above entries isaeagle and Jack. I've also renamed the files with bag # included. Just missing the winner now..