Would it be possible for anyone skilled with Vituixcad to eyeball my work for errors. I am a newbie. I started trying to design a crossover to replace drivers in nice old cabinets. As a consequence, my cabinet dimensions, including driver hole sizes are fixed. I have done 7 different designs and entered each into LibrePCB and have completed board layout, but have not had any PCB built so far. Being a beginner, it sure would be nice to have a set of eyes on my work before building the PCB.
I enjoy the work and don't want anyone to do it for me, but I've exhausted what I know to check. If anyone would care to say, something like "are you sure you used the right configuration of ...", or "there appears to be an issue on graph x ..." it would be of enormous help. I attached the most basic graphs and schematic of my most likely to succeed version - there's more for the asking.
The drivers I expect to use: Dayton SIG180-04 (woofer) and Tymphany XT25BG60-04 (tweeter)
Since I feel I owe this sight some payback, I created a library of resistors, capacitors, inductors, and a few miscellaneous items - enough to allow others a pretty fast way to produce a PCB in LibrePCB. Unlike Vituixcad, my background favors this latter undertaking. I'm happy to share this library and could use a pointer of how to do it on diyAudio.
Hopefully I've not offended anyone with this post.
Thanks,
Dave
I enjoy the work and don't want anyone to do it for me, but I've exhausted what I know to check. If anyone would care to say, something like "are you sure you used the right configuration of ...", or "there appears to be an issue on graph x ..." it would be of enormous help. I attached the most basic graphs and schematic of my most likely to succeed version - there's more for the asking.
The drivers I expect to use: Dayton SIG180-04 (woofer) and Tymphany XT25BG60-04 (tweeter)
Since I feel I owe this sight some payback, I created a library of resistors, capacitors, inductors, and a few miscellaneous items - enough to allow others a pretty fast way to produce a PCB in LibrePCB. Unlike Vituixcad, my background favors this latter undertaking. I'm happy to share this library and could use a pointer of how to do it on diyAudio.
Hopefully I've not offended anyone with this post.
Thanks,
Dave
Attachments
😊
If I am not mistaken you did not include the effect of baffle (baffle step and all other diffraction effects), or did you?
I also suggest you build a non-pcb version of your crossover to check whether you like the sound and only then (If you do like IT) make a PCB.
Info on the front baffle and the driver separations centre to centre with distance to the cabinet edges would allow a quick simulation for diffraction or baffle step.
I am assuming the use of the Dayton FRD and ZMA data for the midwoofer. The data for the Peerless tweeter would have to be curve traced and input into VituixCAD.
I am assuming the use of the Dayton FRD and ZMA data for the midwoofer. The data for the Peerless tweeter would have to be curve traced and input into VituixCAD.
Thank you both for you responses!
@stv: To be honest, I don't really understand the picture in the diffraction tool or how to get the results back into the curves I attached. Its a 2D rendering of a 3D problem - making it hard for me to grasp what is the mic location, driver locations, and what to do about the freestanding tweeter.
@raymondj: I have done some modelling of the baffle for the woofer, but the tweeter is mounted outside of an enclosure. The plane of the old tweeter mounting flange is 18mm further from the listener than that of the woofer. However, I am buying new woofers and tweeters and have estimated an acoustic center difference of 15mm (but I am far from sure about this). That said, it is easy for me to move the tweeter forward or back. The center of the woofer lies 187mm below the tweeter center - you'll see those values in the schematic. Cabinet outside dimensons: 9.5" x 17.5" x 11.5" with the baffle being the first two dimensions. The wall are 3/4" or about 19mm and the edges are of a radius 3/4". It is a ported design with a 2" ID and length of 3.3". I have both Dayton and curved traced SPL and impedance for the woofer and curved traced SPL and impedance for the tweeter (a handy tool). Picture below shows the cabinet
@stv: To be honest, I don't really understand the picture in the diffraction tool or how to get the results back into the curves I attached. Its a 2D rendering of a 3D problem - making it hard for me to grasp what is the mic location, driver locations, and what to do about the freestanding tweeter.
@raymondj: I have done some modelling of the baffle for the woofer, but the tweeter is mounted outside of an enclosure. The plane of the old tweeter mounting flange is 18mm further from the listener than that of the woofer. However, I am buying new woofers and tweeters and have estimated an acoustic center difference of 15mm (but I am far from sure about this). That said, it is easy for me to move the tweeter forward or back. The center of the woofer lies 187mm below the tweeter center - you'll see those values in the schematic. Cabinet outside dimensons: 9.5" x 17.5" x 11.5" with the baffle being the first two dimensions. The wall are 3/4" or about 19mm and the edges are of a radius 3/4". It is a ported design with a 2" ID and length of 3.3". I have both Dayton and curved traced SPL and impedance for the woofer and curved traced SPL and impedance for the tweeter (a handy tool). Picture below shows the cabinet
Attachments
Last edited:
Hi Dave,
I will have a go at this tomorrow afternoon. I probably would have had a try at this tomorrow a.m. but just out of the blue today I had the offer of new electric and gas smart meters installed tomorrow between 8-12 in the morning. A poor and lame excuse I know. 🙂
I am not sure it will add much worth but until I have a try I wont know.
I will have a go at this tomorrow afternoon. I probably would have had a try at this tomorrow a.m. but just out of the blue today I had the offer of new electric and gas smart meters installed tomorrow between 8-12 in the morning. A poor and lame excuse I know. 🙂
I am not sure it will add much worth but until I have a try I wont know.
Its been a long day.
I think the installer of the new electricity and gas meters was happy to get away. Mixing old and new is always tricky.
I haven't had much time with this so kindly take with a piece of caution. The most worrying things is the bump at 2-3Khz in the first shot.
The second is an attempt to do something about it, not having the drivers here is tricky to know what you can do and cannot do with them especially a new metal cone driver. Also there is an element of guesstimate on the way the tweeter and its baffle interact.
As this is a simulation playing with the tweeter z does change things, and again I am not yet convinced I should have made its mount part of a more complex shape.
Time to start cooking my wife's evening meal.
I think the installer of the new electricity and gas meters was happy to get away. Mixing old and new is always tricky.
I haven't had much time with this so kindly take with a piece of caution. The most worrying things is the bump at 2-3Khz in the first shot.
The second is an attempt to do something about it, not having the drivers here is tricky to know what you can do and cannot do with them especially a new metal cone driver. Also there is an element of guesstimate on the way the tweeter and its baffle interact.
As this is a simulation playing with the tweeter z does change things, and again I am not yet convinced I should have made its mount part of a more complex shape.
Time to start cooking my wife's evening meal.
Attachments
I tried to get diffraction plots for the left and right speakers. The right is further from the wall than the left. Because of the way the tool works I modeled both as a distance from the left wall with the right speaker being farther. Here they are:
Attachments
The images I posted earlier are of course the wrong way round the second one has your original component values and the first my attempt to flatten things a bit.
As its a metal driver and the Dayton info pack for it had lots of information including a distortion plot, I thought I should try to get the woofer rolled off relatively sharply. Additionally trying to decide on dimensions and baffle may need a few iterations. Can you measure this when the drivers are fitted. All that felt is probably going to modify the response as well.
I made the woofer baffle at 241 x 444 mm, with the woofer centred at 290 mm vertical height and centre width at 120mm. with mic on axis at the same location.
The tweeter baffle I assumed at width 164 mm by height 144. with tweeter and woofer aligned in the middle for its measurement, maybe combining the baffles would make it look nicer and improve the vertical directivity.
Impedance for the woofer from the data pack and for the tweeter via VituixCAD SPL trace and export.
I added the woofers parameters to the driver database on the computer here and 18 to 22 Litres seems fine, I could have exported that from the enclosure modeller and it would give you the two impedance peaks of a reflex tuned box, but i decided to skip that as I was more interested in the impedance in the crossover region for now.
I would treat the data as a starting point as so many things are likely to subtly change or become less discernible when you start trying to integrate them for the first time.
I haven't seen much about the sig 180's here so it will be interesting to see how you adjust the crossover to work with them.
Final thought. Dont be afraid to flip the tweeter phase and see what that does for you.
As its a metal driver and the Dayton info pack for it had lots of information including a distortion plot, I thought I should try to get the woofer rolled off relatively sharply. Additionally trying to decide on dimensions and baffle may need a few iterations. Can you measure this when the drivers are fitted. All that felt is probably going to modify the response as well.
I made the woofer baffle at 241 x 444 mm, with the woofer centred at 290 mm vertical height and centre width at 120mm. with mic on axis at the same location.
The tweeter baffle I assumed at width 164 mm by height 144. with tweeter and woofer aligned in the middle for its measurement, maybe combining the baffles would make it look nicer and improve the vertical directivity.
Impedance for the woofer from the data pack and for the tweeter via VituixCAD SPL trace and export.
I added the woofers parameters to the driver database on the computer here and 18 to 22 Litres seems fine, I could have exported that from the enclosure modeller and it would give you the two impedance peaks of a reflex tuned box, but i decided to skip that as I was more interested in the impedance in the crossover region for now.
I would treat the data as a starting point as so many things are likely to subtly change or become less discernible when you start trying to integrate them for the first time.
I haven't seen much about the sig 180's here so it will be interesting to see how you adjust the crossover to work with them.
Final thought. Dont be afraid to flip the tweeter phase and see what that does for you.
You have one thing most of us would do differently: you measured everything and simulated everything where listening point is not in line with tweeter. From the picture in #4 post I see that tweeter is more or less in ear height, so measurements, simulations and everything else should be with X, Y, Z = 0 for tweeter.
Then... You are Imperial units dude and probably do not understand, that "m" and "mm" kinda differs - 1 "m" = ~1.1 yards, 1 "mm" = 0.001 "m" = ~1/25.4 of inch. In your case distance between tweeter and woofer is ~ 187mm. Picture shows 187m
Forget about the room reflections in VituixCAD. First make the neutral crossover, simulate it and then make room corrections if needed. Diffraction tool should be used without floor and wall bounces for this.
Ok. That is the first part. You are making huge jumps in design, but are making big miscalculations. The best bet would be to follow this:
https://kimmosaunisto.net/Software/VituixCAD/VituixCAD_Measurement_REW.pdf - it is 99% to 100% guaranteed to have a good measurement result. I am following everything what's there, though there are some steps, which are not very clear to me - it still works. Simulation and final result match. I tried to make some simulation using manufacturer data - it does not work. It just does not. You have to measure all of it yourself. Every driver, in 10-15 degrees increments on hor and ver planes 🙁
Then... You are Imperial units dude and probably do not understand, that "m" and "mm" kinda differs - 1 "m" = ~1.1 yards, 1 "mm" = 0.001 "m" = ~1/25.4 of inch. In your case distance between tweeter and woofer is ~ 187mm. Picture shows 187m
Forget about the room reflections in VituixCAD. First make the neutral crossover, simulate it and then make room corrections if needed. Diffraction tool should be used without floor and wall bounces for this.
Ok. That is the first part. You are making huge jumps in design, but are making big miscalculations. The best bet would be to follow this:
https://kimmosaunisto.net/Software/VituixCAD/VituixCAD_Measurement_REW.pdf - it is 99% to 100% guaranteed to have a good measurement result. I am following everything what's there, though there are some steps, which are not very clear to me - it still works. Simulation and final result match. I tried to make some simulation using manufacturer data - it does not work. It just does not. You have to measure all of it yourself. Every driver, in 10-15 degrees increments on hor and ver planes 🙁
OK, reading the above points from SVP it looks like I will have to flip my Y 187 mm measurement to the woofer and as pointed out keep the tweeter at the 0, 0, 0 position. A good call , I missed that subtlety.
Best if tweeter is all 0. It is a reference point.
And maybe better if you put it on the top in drawings, to mimic its physical location in speaker box.
And maybe better if you put it on the top in drawings, to mimic its physical location in speaker box.
Last edited:
@raymondj Can you describe what metodics you use to get separate speaker data?
If woofer is in bassreflex or any other closed box, use enclosure tool to get at least FRD (impedance) data for woofer. That will be somewhat close to the real world in the lower range.
What you are doing is more or less like a practice. Think like practice, because you STILL need to measure final speaker. It is pretty hard to predict and simulate when the only data is manufacturer provided on some wide baffle, also in Dayton case, data tend to be very cherry-picked and somewhat beautified.
If woofer is in bassreflex or any other closed box, use enclosure tool to get at least FRD (impedance) data for woofer. That will be somewhat close to the real world in the lower range.
What you are doing is more or less like a practice. Think like practice, because you STILL need to measure final speaker. It is pretty hard to predict and simulate when the only data is manufacturer provided on some wide baffle, also in Dayton case, data tend to be very cherry-picked and somewhat beautified.
Oh, this terrific - lots of help! Okay, let's see if I can respond to the points:
@raymondj - I don't have the driver's yet and will measure them when I get them. Your figures for the box are mostly right. Y values for the speaker beginning at the floor where Y is 0:
The existing tweeter baffle is 102mm square. The new one will be 104mm.
The woofer enclosure is 21.6 Liters. It is ported with a 50mm ID and a length of 84mm. It is not flared at either end. The enclosure is about 20% full of fiberglass insulation a little denser than in USA home walls. If it might improve the box, I have 3mm bitumen sound deadening material that might be adhered to the wall. (difficult to do through the woofer's mounting hole.)
I have toyed with your crossover ideas and concerns - thank you. I'm balancing size, price, and volume. Not sure where I'll end up. Your plots showed more variability than mine. This is not a criticism, I think I'm missing something that would reflect more real-world variability. Thank you for clearing up the ported/not-ported issue.
FYI: when I can, I get driver SPL and impedance curves from, say, hificompass.com and trace them into Vituixcad. The difference can be quite a bit.
@svp I need to come back later this evening to address all your good points.
@raymondj - I don't have the driver's yet and will measure them when I get them. Your figures for the box are mostly right. Y values for the speaker beginning at the floor where Y is 0:
- Floor 0
- Bottom of woofer enclosure 140mm
- Woofer Center 454mm
- Top of woofer enclosure 584mm
- Tweeter Center 641mm
The existing tweeter baffle is 102mm square. The new one will be 104mm.
The woofer enclosure is 21.6 Liters. It is ported with a 50mm ID and a length of 84mm. It is not flared at either end. The enclosure is about 20% full of fiberglass insulation a little denser than in USA home walls. If it might improve the box, I have 3mm bitumen sound deadening material that might be adhered to the wall. (difficult to do through the woofer's mounting hole.)
I have toyed with your crossover ideas and concerns - thank you. I'm balancing size, price, and volume. Not sure where I'll end up. Your plots showed more variability than mine. This is not a criticism, I think I'm missing something that would reflect more real-world variability. Thank you for clearing up the ported/not-ported issue.
FYI: when I can, I get driver SPL and impedance curves from, say, hificompass.com and trace them into Vituixcad. The difference can be quite a bit.
@svp I need to come back later this evening to address all your good points.
@svp you said to use x,y,z all equal to 0 assuming the tweeter is ear height. I'll do that. I was using the woofer center as my origin. That's how I got 187mm for the Y of the tweeter. (Oh yes, Vituixcad that I'm using keeps the unit scale and throws out the unit, so mm displays as m. I don't like it but that's how it works. In my career I work in microns and nanometers all day, every day, in spite of my country's obstinence. At home, however, my tape measure is in feet and inches so I have a lot of conversions to do.)
I've noted your remarks about first creating a neutral crossover first then consider room diffraction measurements. Thank you.
I'll make sure to get impedence data from the enclosure tool.
My goal in simulation is to avoid buying lots of parts to find a combination that works. I currently have no parts because I was very unsure of my competence with Vituixcad. I expect to firm up a crossover design, order parts, measure performance, and hopefully make few minor part size adjustments (hopefully just L-pad changes). I will also move the tweeter forward or backward for best phasing and vertical directionality; and I may need to adjust the felt present on the baffles. I do have the tools needed for this though I'll have to learn how to do it correctly.
Thank you for the link VituixCAD_Measurement_REW.pdf. I'll read it.
I'm not sure how to use the Enclosure Tool with the tweeter. It has its own baffle, but in my installation it will have no enclosure. My choices with the tool are Infinite Baffle, Closed, and Bass Reflex. There are other choices but I don't think they apply. I just want to make sure the woofer and tweeter curves will piece together well and mimic reality.
I've noted your remarks about first creating a neutral crossover first then consider room diffraction measurements. Thank you.
I'll make sure to get impedence data from the enclosure tool.
My goal in simulation is to avoid buying lots of parts to find a combination that works. I currently have no parts because I was very unsure of my competence with Vituixcad. I expect to firm up a crossover design, order parts, measure performance, and hopefully make few minor part size adjustments (hopefully just L-pad changes). I will also move the tweeter forward or backward for best phasing and vertical directionality; and I may need to adjust the felt present on the baffles. I do have the tools needed for this though I'll have to learn how to do it correctly.
Thank you for the link VituixCAD_Measurement_REW.pdf. I'll read it.
I'm not sure how to use the Enclosure Tool with the tweeter. It has its own baffle, but in my installation it will have no enclosure. My choices with the tool are Infinite Baffle, Closed, and Bass Reflex. There are other choices but I don't think they apply. I just want to make sure the woofer and tweeter curves will piece together well and mimic reality.
@Inman, do you have a measurement microphone? If so, you would be better off measuring the frequency responses, as well as the impedances, of the drivers with them mounted into the cabinet. Your model will be more accurate with those measurements. You can use REW, which is free, to make the measurements.
I would forego the PCB and save the cost. Instead, hardwire the crossover components. Indeed, some high end speaker manufacturers tought the use of hardwired crossovers as a feature for the best possible performance. Essentially, it results in lower resistance in the filter networks.
Before completing the assembly, test your crossover to ensure the speaker's frequency response is adequate. If not, fine tune the crossover (this may not be needed if the measurments used in your model are accurate enough).
After you are satisfied with the crossover design, glue the comopnents down to a piece of fiberboard with hot glue or silicon. You can screw that inside your cabinet when doing final assembly, perhaps with a layer of rubber underneath to keep it from vibrating, or you can glue it down. Mount the inductors perpendicular to one another to minimize inductive coupling.
In the crossovers, use air core inductors to avoid core losses. I recommend using a heavy gauge air core inductor for the woofer to minimize the resistance, which adversely affects the damping factor.
I would forego the PCB and save the cost. Instead, hardwire the crossover components. Indeed, some high end speaker manufacturers tought the use of hardwired crossovers as a feature for the best possible performance. Essentially, it results in lower resistance in the filter networks.
Before completing the assembly, test your crossover to ensure the speaker's frequency response is adequate. If not, fine tune the crossover (this may not be needed if the measurments used in your model are accurate enough).
After you are satisfied with the crossover design, glue the comopnents down to a piece of fiberboard with hot glue or silicon. You can screw that inside your cabinet when doing final assembly, perhaps with a layer of rubber underneath to keep it from vibrating, or you can glue it down. Mount the inductors perpendicular to one another to minimize inductive coupling.
In the crossovers, use air core inductors to avoid core losses. I recommend using a heavy gauge air core inductor for the woofer to minimize the resistance, which adversely affects the damping factor.
@TerryForsythe I do have a measurement microphone and noted your recommendation. Given my background, I'm used to simulation as a means to avoid unnecessary product purchases, but I'm getting the feeling that SPL, impedance, and related data are a bit of a wish and a hope. So far, I wouldn't trust my simulations at all. I've been trying to use these specs as a means to find fairly compatible driver/tweeter/box combinations. It's very frustrating. It looks like I'll have to guess on the drivers then measure.
The cost of the PCB is remarkably small $55 for my smaller designs, I can perfectly control distances to fields of the inductors, and the total wiring is actually shorter than my hand wired estimates. I'm very comfortable doing these layouts. If I go this route, I'll take your idea of gluing items down, and I do plan on air core inductors.
I did not know REW was free. I'll look into it. I saw a VituixCad website using it but quit looking figuring it was another expense.
It's very kind of you to help. Thank you!
The cost of the PCB is remarkably small $55 for my smaller designs, I can perfectly control distances to fields of the inductors, and the total wiring is actually shorter than my hand wired estimates. I'm very comfortable doing these layouts. If I go this route, I'll take your idea of gluing items down, and I do plan on air core inductors.
I did not know REW was free. I'll look into it. I saw a VituixCad website using it but quit looking figuring it was another expense.
It's very kind of you to help. Thank you!