Dunno, depends on what the drivers actually measure. Regardless, the way I figure stuff, the XO point would normally be ~349 Hz, though considering the under-damped alignments would opt for ~108 Hz (or 'close enough' THX LFE 120 Hz or ~195 Hz if doing passive due to component cost) to isolate the mid-bass and make the divided! box either 15 L each with the HF ~aperiodic and maybe experiment with the LF box taking up most of the space, though either way removes any LF influence on the HF with the XO trimming it even more as we're mainly concerned with getting the > 250 Hz BW 'squeaky' clean.
My guess is the T/L parameters that you entered are different to what I used.
I used the manufacturer's data for both drivers.
Below you'll find 3 images of the system parameters window, one for each driver as per the manufacturer and 1 for a real SB20FR which I measured myself using LIMP.
Also make sure that you have the room gain enabled for both drivers in your simulation.
Attachments
Thanks GM. I read that three times and I’m not sure I follow, but then I often don’t understand half of what you say lol. You’re just too advanced for me I guess. I’m guessing that your just screwing around with those very specific XO points? Anyways, this is a .5-way not a 2-way, so it doesn’t have a XO per se.
Sounds like divided box is the consensus, which kills my design. Maybe I should just use two of the of the full range in a .5 alignment. I’ve done 2.5 ways before and I don’t use separate enclosures with no ill effect. My only real issue with that is I don’t want the full baffle step comp.
Stanislav, I used the manufacture data and I triple checked it. I don’t think I have room gain enabled though. I was waiting to mess with that until I got everything else working. I think I’ll just start over. After first entering the FR data, I believe the sensitivity was correct. Then it all changed for some reason.
Still confused about the phase rotation though. There basically shouldn’t be any. And there’s no way to unwrap it that I see.
Also, the measured Qts is up there. Were your drivers broken in?
Still confused about the phase rotation though. There basically shouldn’t be any. And there’s no way to unwrap it that I see.
Also, the measured Qts is up there. Were your drivers broken in?
You're welcome!
No, not 'screwing around', the pioneer's were 'serious as a heart attack' WRT to optimizing a theoretical design, though for the 'bean counters' they had to maximize a wide range of designs using the least number of components, so obviously either a lot of compromises and/or design components to 'fit' the widest range of apps, so chose a combination of the two, leaning towards getting max speech intelligibility over wide arcs at high SPL as practical.
We OTOH can only choose from what others have thought the most profitable, which for the most part is to 'don't fix what ain't broke'/~mirror the pioneer's choices and me preferring to optimize them the way the designers wanted to make them.
Your two box design match-up is a mechanical XO in concept, ergo XO design points apply.
Yeah, personally would use identical drivers if a single box.
Thanks for that. I did not realize that anyone was ever that exact with a crossover point. Component variances would generally make that goal less than feasible.
Different drivers different boxes, same drivers shared is ok. This actually clears up a lot for me. I figured that might be the case but I have no experience with this and I haven’t looked for another design that does this. Trolls actually uses different drivers in .5 arrangements, I should have looked to see how he handled that. Normally a divider is no issue, in fact it just makes for a simpler brace. Unfortunately, my cabinets are usually more complicated than they need to be, but I’ll take a look at redesigning before I move on from the woofer. Using a FR driver as a .5 woofer seems like a bit of an odd fit when there’s a solid match woofer for the picking.
Different drivers different boxes, same drivers shared is ok. This actually clears up a lot for me. I figured that might be the case but I have no experience with this and I haven’t looked for another design that does this. Trolls actually uses different drivers in .5 arrangements, I should have looked to see how he handled that. Normally a divider is no issue, in fact it just makes for a simpler brace. Unfortunately, my cabinets are usually more complicated than they need to be, but I’ll take a look at redesigning before I move on from the woofer. Using a FR driver as a .5 woofer seems like a bit of an odd fit when there’s a solid match woofer for the picking.
Sure…..you asked. So when moving on a design, it’s best before finalizing the the design ( enclosure size, alignment, shaping components, etc) to just LISTEN to the driver on its own……either on a large open baffle or box that’s large enough to not load the driver at rational levels ( you don’t want to induce fatigue). This is the beginning of the voicing process…….you’re going to get an idea on the timbre of the driver on its own….where it’s lacking and where it excels…..and what, if anything you can do about it. Listen first……build after.
You're welcome!
It's about polar response matching and using 1st order XOs plus they used tapped autoformers and could design/build whatever they pleased in their early development days Thanks to MGM's mega bucks before the 'bean counters' took over, ergo nowadays either DSP or get reasonably close; I didn't mean it 'on the nose', just that they're (radically) different from today's norm in many cases.
Normally I would too, same as Trolls, though due in my case to how the pioneers chose match ups, but your choice doesn't spec different enough, i.e. how close do they sound when BW limited to at least their ~ -12 dB HF overlap, so with no experience with either drivers chose to err on the side of caution, but in reading earlier in the thread just now I see Stanislav has apparently either already done it or puts a lot of faith in Basta.
No, I have not done anything like this with my SB20FRs, I only measured them with LIMP for their T/L parameters.
My intention is to run them in my 2way OB with a 15" Kawai woofer cut at around 300Hz. Have not done it yet due to other commitments.
Yes, I put a lot of faith in Basta. I've been using it for a good 10 years now for various designs, but what I found it excels in is OB simulation that no other program can provide, not even Hornsresp. Basta can simulate the whole lot, not just the LF part of the design, plus baffle arrangement, plus room involvement, plus XO, baffle step, and so on.
What I find more fascinating is that Basta's predictions turn up very close to the actual result once the system is built and tested.
For a piece of software that costs nothing, takes the minimum of inputs from the user and still generates a very reliable outcome it's simply amazing.
And I've tried many, many different programs with variable success but always come back to Basta for its completeness, flexibility and accuracy.
About aperiodic designs, I like aperiodic loading as much as OB. Mostly because of their outstanding midrange performance where most of the music is.
SB20FR looks and sounds a lot like Seas FA22RCZ at 1/3 the price and will fit a variety of designs including a single driver TQWT.
But will excel when partnered with a helper woofer, being that in a 1.5 or a proper 2way.
Different drivers in 0.5 is fine as long as they are either in OB or in separate volumes. Don't see a point why sharing the same box would be beneficial.
Of course everything is personal and we all judge stuff from our own perspective. I'm not a professional speaker designer and I can not say my way is the right way. I'm only expressing my opinion based on 30 odd years experience as a loudspeaker hobbyist.
So best is to build and judge stuff for one self and mostly have fun.
Cheers
Stan
My intention is to run them in my 2way OB with a 15" Kawai woofer cut at around 300Hz. Have not done it yet due to other commitments.
Yes, I put a lot of faith in Basta. I've been using it for a good 10 years now for various designs, but what I found it excels in is OB simulation that no other program can provide, not even Hornsresp. Basta can simulate the whole lot, not just the LF part of the design, plus baffle arrangement, plus room involvement, plus XO, baffle step, and so on.
What I find more fascinating is that Basta's predictions turn up very close to the actual result once the system is built and tested.
For a piece of software that costs nothing, takes the minimum of inputs from the user and still generates a very reliable outcome it's simply amazing.
And I've tried many, many different programs with variable success but always come back to Basta for its completeness, flexibility and accuracy.
About aperiodic designs, I like aperiodic loading as much as OB. Mostly because of their outstanding midrange performance where most of the music is.
SB20FR looks and sounds a lot like Seas FA22RCZ at 1/3 the price and will fit a variety of designs including a single driver TQWT.
But will excel when partnered with a helper woofer, being that in a 1.5 or a proper 2way.
Different drivers in 0.5 is fine as long as they are either in OB or in separate volumes. Don't see a point why sharing the same box would be beneficial.
Of course everything is personal and we all judge stuff from our own perspective. I'm not a professional speaker designer and I can not say my way is the right way. I'm only expressing my opinion based on 30 odd years experience as a loudspeaker hobbyist.
So best is to build and judge stuff for one self and mostly have fun.
Cheers
Stan
You know, I never do this! I probably rely too heavily on measurements. I listen of course, but not up front, I will give it a try though. Thanks for your input mayhem.
I was just thinking this when I was using it yesterday. To have baffle, room, driver and XO all handled in one place really simplifies things.
My next project was actually supposed to be a three way quasi open baffle with the GRS planner 10” mid, GRS tweeter and a 12” cardioid loaded Faital woofer for midbass. I already bought everything, made foam core mock-ups took measurements and finalized the baffle shape, but the idea of building such a thing, designing the XO and all the endless listening and tweaking has me dragging my feet a bit. I have a 2 year old that’s crazy for one! So, the SB20FRPC peaked my interest as a simple interim project.
Is there any risk of cancellation do to the leakage from aperiodic vents? For instance, If placed near the driver will low frequency response be reduced? Or if placed on the sides of the box is there a risk of inadvertently creating an aperiodic response?
I’m looking at redesigning my 1.5 way idea so that I can divide the interior volume and I’m not sure if I’m constrained on the location of the vent area.
I’m looking at redesigning my 1.5 way idea so that I can divide the interior volume and I’m not sure if I’m constrained on the location of the vent area.
I never found a way to import frd or zma files in Basta!
One needs to remember that Basta is a rather old unsupported piece of software and has its limitations.
I guess VituixCAD will be more usefull in that respect, but I didn't find it as easy to use as Basta.
Re. cancellation, don't think so, but I'll be the first one to admit that my experience at what you are asking is limited.
I've placed aperiodic vents (dynaudio variovents and home made vents) at the back of the box or at the front, right next to the driver with no adverse effect.
Check Dutch & Dutch 8c which employs side aperiodic vents for the mid driver, an interesting concept to help achieve constant directivity over a wide range.
Have a look at these files, may give you some ideas.
One needs to remember that Basta is a rather old unsupported piece of software and has its limitations.
I guess VituixCAD will be more usefull in that respect, but I didn't find it as easy to use as Basta.
Re. cancellation, don't think so, but I'll be the first one to admit that my experience at what you are asking is limited.
I've placed aperiodic vents (dynaudio variovents and home made vents) at the back of the box or at the front, right next to the driver with no adverse effect.
Check Dutch & Dutch 8c which employs side aperiodic vents for the mid driver, an interesting concept to help achieve constant directivity over a wide range.
Have a look at these files, may give you some ideas.
Attachments
FYI/FWIW, the pioneers concluded that the ideal driver/vent location was the driver in the vent or as close as practical, so no.
Thanks. So Basta! is good for initial, rough modeling, but if it doesn’t account for the drivers true, non linear response then how are you getting very accurate results that match well with your final measured response? I’m genuinely interested because the method I use is kind of arduous.
Yep, I’m aware that Dynaco placed the vents on the front for a number of models. Thanks you taking he time to post those links. That’s what I was thinking of doing for the next design attempt, but I can’t see how it wouldn’t cause some bass cancellation. Good to hear that you’ve had no problem with this though and as I’ll be crossing to a sub then it’s probably fine. It seems like the wavelength is too long to be absorbed by any stuffing though, but maybe there’s something going on that I don’t understand. And yes I could place them on the sides to possibly create a cardiod type response, but that wasn’t my intention with this speaker. That is exactly what I was planning for the bass section of my OB top section three way, but I understand that that takes a bit of experimentation and measuring to get right. I just want this to be simple! Wishful thinking maybe.
Yep, I’m aware that Dynaco placed the vents on the front for a number of models. Thanks you taking he time to post those links. That’s what I was thinking of doing for the next design attempt, but I can’t see how it wouldn’t cause some bass cancellation. Good to hear that you’ve had no problem with this though and as I’ll be crossing to a sub then it’s probably fine. It seems like the wavelength is too long to be absorbed by any stuffing though, but maybe there’s something going on that I don’t understand. And yes I could place them on the sides to possibly create a cardiod type response, but that wasn’t my intention with this speaker. That is exactly what I was planning for the bass section of my OB top section three way, but I understand that that takes a bit of experimentation and measuring to get right. I just want this to be simple! Wishful thinking maybe.
In the vent? Is the sound that leaks from the vent in phase with the sound from the front of the cone?
Side vents, a la passive cardioids certainly get cancellation. Off axis at least. But that requires delay (distance in this case), so maybe that’s the difference. And as I understand it, they are only effective down to around 100Hz, so apparently lower bass isn’t affected (cancellation) by aperiodic loading. For it to cancel, I have to assume that the sound escaping those vents are out of phase.
Actually GM, I wasn’t able to open the link you posted for some reason before I posted, but now it works. I’d bet that what they are measuring there is dipole cancellation. Or maybe rolloff below a very high vent tuning. I doesn’t look steep enough for that, but combined with the underdamped raw response maybe it’d look something like that?
Anyways, that doesn’t appear to be aperiodic, the enclosure is lined but there’s no stuffing. It looks like whole bunch of short ports that would result in a very high fb or a quasi baffleless, dipole response. Please correct me if I’m wrong.
Anyways, that doesn’t appear to be aperiodic, the enclosure is lined but there’s no stuffing. It looks like whole bunch of short ports that would result in a very high fb or a quasi baffleless, dipole response. Please correct me if I’m wrong.
I don't know, I wish I could answer your questions but my experience tells me exactly this. I'm not overly concerned about measurements, mostly do a few impedance sweeps with LIMP and a few FR sweeps with REW and oddly enough a little Android app called Spectroid which is sort of real time spectrum analyser. Funnily enough what I see with Spectroid closely resembles Basta's predictions.
I would say if you contact Svante he might be willing to elaborate how everything works.
If he's still here at Diyaudio. Otherwise try Tolvan Data contact.
No problem. I’m actually enjoying it quite a bit. It’s a much more useful modeling tool than a simple box program or baffle diffraction tool. Especially with the ability to add crossover components. I’ll probably stick to PCD for full crossover modeling but that’s a more tedious multistep process that I tend to avoid getting into until I’m pretty sure of the design and drivers