I'm in the middle of building some FAST speakers (i.e. sub and mid - no tweeter) in a transmission line enclosure.
My layman's logic tells me that I test each driver mounted in the enclosure to obtain the correct frd and zma files to use for the crossover design (using Xsim).
But research show that most advice is to just use the frd and zma files from the drivers manufacturer's web site.
Surely the only useable frd and zma files would be from testing in their designated enclosure.
Or am I wrong here?
My layman's logic tells me that I test each driver mounted in the enclosure to obtain the correct frd and zma files to use for the crossover design (using Xsim).
But research show that most advice is to just use the frd and zma files from the drivers manufacturer's web site.
Surely the only useable frd and zma files would be from testing in their designated enclosure.
Or am I wrong here?
You're absolutely right that using measured files is best, but if you want to try out different candidate drivers, that gets expensive.
VituixCad has enclosure and diffraction simulators built in and can merge and combine their outputs with the manufacturer's standard files, so that gets you a lot closer to the correct info if you want to get a rough idea of what will work before buying drivers.
VituixCad has enclosure and diffraction simulators built in and can merge and combine their outputs with the manufacturer's standard files, so that gets you a lot closer to the correct info if you want to get a rough idea of what will work before buying drivers.
I've got the diy bug again after a 20 year layoff.
Doing some casual online reading, I became interested in transmission line and some of the current free software that was available.
I also liked what I read about the Dayton RS225-8 in a transmission line, and that it teamed well with the Vifa TC9FD-08.
So instead of researching suitable drivers (for enclosure and pairing), I jumped straight in and bought the drivers.
So I've probably haven't approached the design like an experienced enthusiast would.
Using Hornresp over several weeks, I believe I have a good TL design and have started to build the enclosure.
The Vifa is in its own sealed enclosure at the top of the box.
The next step will be getting the correct frd and zma files (hence this posted topic).
As this is a one-off project, I don't want to buy DATs. So I'm going to use REW, which I'm very familiar with, and have a calibrated testing mic, but haven't used it yet for impedance testing (another thing to learn).
Then of course I'll have to learn Xsim from scratch. Unless someone can suggest a better program (my first designs, 30+ years ago, were done using a DOS version of Calsod).
Doing some casual online reading, I became interested in transmission line and some of the current free software that was available.
I also liked what I read about the Dayton RS225-8 in a transmission line, and that it teamed well with the Vifa TC9FD-08.
So instead of researching suitable drivers (for enclosure and pairing), I jumped straight in and bought the drivers.
So I've probably haven't approached the design like an experienced enthusiast would.
Using Hornresp over several weeks, I believe I have a good TL design and have started to build the enclosure.
The Vifa is in its own sealed enclosure at the top of the box.
The next step will be getting the correct frd and zma files (hence this posted topic).
As this is a one-off project, I don't want to buy DATs. So I'm going to use REW, which I'm very familiar with, and have a calibrated testing mic, but haven't used it yet for impedance testing (another thing to learn).
Then of course I'll have to learn Xsim from scratch. Unless someone can suggest a better program (my first designs, 30+ years ago, were done using a DOS version of Calsod).
Thanks Allen.
I've already checked out a few videos on using Xsim.
Haven't found any info yet on how to initially select the size of each component.
But I do have an old copy of Vance Dickason to fall back on. 😉
There's an extensive thread on this Forum about an RS225/Scan Speak FAST, in both conventional and TL enclosures, the thread also covers a design with the RS225 and Vifa TC-9, so you might wish to check that out and save yourself some work:
https://www.diyaudio.com/community/...-waw-ref-monitor.273524/page-176#post-6937759
Geoff
https://www.diyaudio.com/community/...-waw-ref-monitor.273524/page-176#post-6937759
Geoff
Thanks Geoff, I've trolled through almost every post in that thread.
Actually that's where I originally got the idea of using the RS225 in a transmission line, and pair it with the Vifa.
Thanks Allen.
I really should get into Xsim and learn how to use it, even with some default or imaginary drivers.
That's how I started with Hornresp.
Sorry guys - another quick question.
What smoothing is best on the SPL graph before tracing, that will give the optimal information in Xsim?
What smoothing is best on the SPL graph before tracing, that will give the optimal information in Xsim?
Xsim is a good tool, easy to learn and use and great for looking at the effects of different crossover parts on drivers; however, it's only as good as what you put in (the G.I.-G.O. principle). You can use manufacturers' graphs and files - which may be accurate, or not - and put them into Xsim using fptrace or whatever, but the results won't be accurate as the driver(s) is not measured in your cabinet. Better than using a 'crossover calculator' but.
As an example, here's a comparison of the Xsim simulation for Paul Carmody's "Classix II" speaker. The top graph is the sim, using maker-supplied FRD and ZMA data for the DC160 and fptrace used to obtain those files from Vifa BC25 graphs. The bottom is Mr Carmody's, properly measured in the specified cabinet. Both use the correct crossover parts.
You can see the difference!
Geoff
While I can see I'd have a few questions about those plots, I see the various choices people make when doing things in those ways so it might just be more constructive to say you can verify results by measuring again after crossing. At this point, the making of relative changes can be more accurately seen on the simulator.
Agreed, hence my remark on GI-GO: my fptrace effort might not have been ideal and the official Vifa FR and ZMA graphs may not have been accurate in the first place. Dayton's supplied files are generally regarded as accurate - but of course, measured on their equipment.
Xsim is great for seeing the effect of various components and filters on a design; sometimes small parts value changes make a real difference, other times not. If I'm making a speaker and don't have the exact part value, Xsim's a good way of seeing whether I might get away with using what I have.
Geoff
Oops I thought no own measurements were meant. In time windowed measurement (correctly applied) smoothing at 1/24 or 1/12 octave is OK. but in fact the time window itself acts as a kind of smoothing, because the frequency resolution is dependent on the time window.
Sorry, I should have explained.
After I complete the cabinet build and mount the speakers, I intend to measure the Freq Response with the speakers outside using REW.
Hence the question of smoothing.
The ground in front of the speaker slopes away, and the only obstacle of concern is the neighbours fence and house approx 30m (100') away (which I could gate if needed).
I intend to measure each driver online, at 1m and 2.83v. I'll also be using REW to measure the impedance of each driver.
This is how I intend to obtain my measurements, but I'm happy to listen to any other ideas.
The enclosure has a cavity under the base where I will locate the Xover.