Experimentation can be fun!!
You definitely have a good mindset on this.... (experimentation) one of the reasons Pano & I released the project. As I have said before, due to the relatively light weight of the components you can work up a simple card board baffle. Some have reported to me via email of some improvement with a small rearward mounted added tweeter. I did play with that a while but kept it out of the design to help with the simplicity issue - goal. Will be interested to hear how your design works out!
You definitely have a good mindset on this.... (experimentation) one of the reasons Pano & I released the project. As I have said before, due to the relatively light weight of the components you can work up a simple card board baffle. Some have reported to me via email of some improvement with a small rearward mounted added tweeter. I did play with that a while but kept it out of the design to help with the simplicity issue - goal. Will be interested to hear how your design works out!
Inductors
John, Pan
You two have explained how to tweak the Ultra and Manzi crossovers to best effect. It's been pretty rewarding adding the woofer trap, changing resistors and adding the tiny bipass caps to the existing caps. Now it's time to do some more tweaking.
In post #1301 Pano mentioned the inductors:
"You really DO need to adjust the crossover, tho. When I first built them I found John's crossover too rolled off at the top for my room. The Vifa response really gets bright at the top, but John had tamed it too much for my room. I had to change the trap inductor and resistor some. Tweak that inductor!"
Long one short I'm hoping to open up the midrange and high end on the Ultra in a way that the resistor changes don't quite manage. Is it possible with an inductor change? Don't know but it seems worth the effort to try.
Any thoughts on the range of values to try with different inductors? I know it's system and room dependent but a decent starting point would help those of us that are less experienced with making crossover changes. There aren't any odd value inductors lying around the parts bin in my house so just ordering random values and hoping for the best will get a bit pricey real fast. And I'm cheap.
The Manzi crossover still calls for a 2.0 mh inductor. Would that be a reasonable place to start on the Ultra?
Many thanks,
Barry
John, Pan
You two have explained how to tweak the Ultra and Manzi crossovers to best effect. It's been pretty rewarding adding the woofer trap, changing resistors and adding the tiny bipass caps to the existing caps. Now it's time to do some more tweaking.
In post #1301 Pano mentioned the inductors:
"You really DO need to adjust the crossover, tho. When I first built them I found John's crossover too rolled off at the top for my room. The Vifa response really gets bright at the top, but John had tamed it too much for my room. I had to change the trap inductor and resistor some. Tweak that inductor!"
Long one short I'm hoping to open up the midrange and high end on the Ultra in a way that the resistor changes don't quite manage. Is it possible with an inductor change? Don't know but it seems worth the effort to try.
Any thoughts on the range of values to try with different inductors? I know it's system and room dependent but a decent starting point would help those of us that are less experienced with making crossover changes. There aren't any odd value inductors lying around the parts bin in my house so just ordering random values and hoping for the best will get a bit pricey real fast. And I'm cheap.
The Manzi crossover still calls for a 2.0 mh inductor. Would that be a reasonable place to start on the Ultra?
Many thanks,
Barry
Tweaking the Vifa mid-tweet trap
Happy Halloween 2016 Barry!
The trap is pretty simple once you establish what you need to have it do.
The trap inductor value sets the frequency the trap roll off starts to take effect. Larger the value, lower the frequency, smaller, the higher the frequency. Left alone it would act as a first order low pass filter.
The parallel cap sets the frequency where the inductor part of the trap stops having its roll off effect. Larger cap values narrow the traps frequency width, smaller values allow the inductor to be in effect higher up in frequency. The cap in effect acts as a frequency dependent by pass for the inductor. The cap by itself would be a first order high pass filter.
So, the inductor rolls off the naturally rising response of the Vifa which primarily occurs in the upper mid range and lower treble. Left alone, the inductor would continue to roll off the high end way beyond what is needed. The cap provides a bypass for the higher frequencies we want to keep. By balancing to two you create a very specific band pass filter in the frequency domain.
One thing that complicates setting the inductor and cap values is the rear radiation (frequency response) of the Vifa is very different than the front frequency response. There are several reasons for this, but that is another subject for another time.
Cutting to the chase, if you feel you need more mid range you would first reduce the pad resistor by an ohm or two and then reduce the inductor value. The question is, how much? One nice thing about inductors is that you can unwind them to obtain a lower value. The cap side of things is a bit more trouble to adjust. You have to have exact values or start with an estimated smaller value and then add small parallel caps until you hit the magic spot.
Remember, by lowering the pad value and the inductor value you have changed the "resistive load" the first order capacitor sees, and it will have to be increased to maintain the original crossover frequency.
Hope the above helps you with your experimentation. Good luck!
Happy Halloween 2016 Barry!
The trap is pretty simple once you establish what you need to have it do.
The trap inductor value sets the frequency the trap roll off starts to take effect. Larger the value, lower the frequency, smaller, the higher the frequency. Left alone it would act as a first order low pass filter.
The parallel cap sets the frequency where the inductor part of the trap stops having its roll off effect. Larger cap values narrow the traps frequency width, smaller values allow the inductor to be in effect higher up in frequency. The cap in effect acts as a frequency dependent by pass for the inductor. The cap by itself would be a first order high pass filter.
So, the inductor rolls off the naturally rising response of the Vifa which primarily occurs in the upper mid range and lower treble. Left alone, the inductor would continue to roll off the high end way beyond what is needed. The cap provides a bypass for the higher frequencies we want to keep. By balancing to two you create a very specific band pass filter in the frequency domain.
One thing that complicates setting the inductor and cap values is the rear radiation (frequency response) of the Vifa is very different than the front frequency response. There are several reasons for this, but that is another subject for another time.
Cutting to the chase, if you feel you need more mid range you would first reduce the pad resistor by an ohm or two and then reduce the inductor value. The question is, how much? One nice thing about inductors is that you can unwind them to obtain a lower value. The cap side of things is a bit more trouble to adjust. You have to have exact values or start with an estimated smaller value and then add small parallel caps until you hit the magic spot.
Remember, by lowering the pad value and the inductor value you have changed the "resistive load" the first order capacitor sees, and it will have to be increased to maintain the original crossover frequency.
Hope the above helps you with your experimentation. Good luck!
John gives a good run down of how that parts work and what they do above. If you want to use a crossover simulator software, like XSim or PCD, you can virtual change the values and see the resulting changes in the plots. It's a good teaching tool.
Inductors are the expensive parts to play with, for sure. As John mentions, you can unwind from a higher value if you need to.
Resistors and caps are cheap enough that you can afford to play. If you're on a tight budget, buy non-polarized electrolytic caps from Parts Express or somewhere. They are just fine for research. Once you find the values you need, you can replace them with film caps if that's what you want.
Inductors are the expensive parts to play with, for sure. As John mentions, you can unwind from a higher value if you need to.
Resistors and caps are cheap enough that you can afford to play. If you're on a tight budget, buy non-polarized electrolytic caps from Parts Express or somewhere. They are just fine for research. Once you find the values you need, you can replace them with film caps if that's what you want.
John and Pano, those are two great posts. Thanks for the information and guidance.
The possibilities and configurations are far more open ended than expected and a bit overwhelming. It's definitely above my current pay grade but it sure will keep me busy during the winter listening season.
"The parallel cap sets the frequency where the inductor part of the trap stops having its roll off effect. Larger cap values narrow the traps frequency width, smaller values allow the inductor to be in effect higher up in frequency. The cap in effect acts as a frequency dependent by pass for the inductor. The cap by itself would be a first order high pass filter."
I'd never considered the cap and inductor/resistor working off each other. It seemed like the cap would do it's thing as a low end roll off controlling the frequency of the crossover and the resistor would work as a gate to control the overall effect of the inductor. This is far more interesting and complicated.
It's a lot to chew on. It may take simulator software to play with some values and grasp what the changes will do.
Many thanks.
The possibilities and configurations are far more open ended than expected and a bit overwhelming. It's definitely above my current pay grade but it sure will keep me busy during the winter listening season.
"The parallel cap sets the frequency where the inductor part of the trap stops having its roll off effect. Larger cap values narrow the traps frequency width, smaller values allow the inductor to be in effect higher up in frequency. The cap in effect acts as a frequency dependent by pass for the inductor. The cap by itself would be a first order high pass filter."
I'd never considered the cap and inductor/resistor working off each other. It seemed like the cap would do it's thing as a low end roll off controlling the frequency of the crossover and the resistor would work as a gate to control the overall effect of the inductor. This is far more interesting and complicated.
It's a lot to chew on. It may take simulator software to play with some values and grasp what the changes will do.
Many thanks.
That's they way I learned it. Instead of wasting time watching TV, I wasted time playing the crossover simulator game. 
Build a simple crossover or trap, then start moving values around and look at the resulting curves. It will take time, but it's a good way to quickly see and understand what each part does, and how it interacts with the other parts.
You can start on simple flat response curves and flat impedance to get a feel. When you add in real speaker impedance and response curves, you'll start to see a whole new level. It's fun.
Build a simple crossover or trap, then start moving values around and look at the resulting curves. It will take time, but it's a good way to quickly see and understand what each part does, and how it interacts with the other parts.You can start on simple flat response curves and flat impedance to get a feel. When you add in real speaker impedance and response curves, you'll start to see a whole new level. It's fun.
Baffle sides
Can someone confirm the dimensions of the ULTRA. I have 25.4" by 48" The wings should be 6". I have the drawing of the baffle but no mention of the wings. I also read in a post that the on side should be tapered. Which side and how should it be tappered example 2" at top and 6" at bottom?
Thanks
Bonsaimaster
Can someone confirm the dimensions of the ULTRA. I have 25.4" by 48" The wings should be 6". I have the drawing of the baffle but no mention of the wings. I also read in a post that the on side should be tapered. Which side and how should it be tappered example 2" at top and 6" at bottom?
Thanks
Bonsaimaster
Ultra Dimensions
The Ultra dimensions are not critical... close is good enough. It is based on a standard 96" X 48" sheet of plywood. Some lumber yards sell pre-cut 48" X 24" quarter sheets. Anyway, the simplest set up is a 24" wide X 48" high with 6" straight sided front to back total wing depth.
Keeping it a simple as possible to cut and build. You can and are encouraged to experiment with temporary card board wings of various depths and shapes. The main thing is to insure you have a stable, rattle free structure when your done. The 1/4 lb. moving mass 15" cone assembly delivers quite a bit of energy into the structure. Both glue and screws along with clamping during construction are required. Would love to see your finished pair!
The Ultra dimensions are not critical... close is good enough. It is based on a standard 96" X 48" sheet of plywood. Some lumber yards sell pre-cut 48" X 24" quarter sheets. Anyway, the simplest set up is a 24" wide X 48" high with 6" straight sided front to back total wing depth.
Keeping it a simple as possible to cut and build. You can and are encouraged to experiment with temporary card board wings of various depths and shapes. The main thing is to insure you have a stable, rattle free structure when your done. The 1/4 lb. moving mass 15" cone assembly delivers quite a bit of energy into the structure. Both glue and screws along with clamping during construction are required. Would love to see your finished pair!
I've spent many hours poring over this thread in detail... very interesting project and I've got a saved partsexpress cart with all the parts I'll need for a build of the Manzanita Ultra and I'm about to pull the trigger. I have a few questions.
Do you believe there is anything to be gained by constrained layer damping construction (plywood + asphalt/felt + MDF sandwich from front to back) for the baffle? This might make it a bit easier as well for me to get the open area in the back around the TC9FD.
Did you make any cardboard models of tapered asymmetric baffle wings in the Ultra (like you used for the original Manzanita)? Do you think this would this be worthwhile for the Ultra?
Also, there is lots of talk in this and other forums about chamfering front baffle edges to improve diffraction. Did you by chance model this or do you think it might give more even off axis response?
Thanks in advance!
Do you believe there is anything to be gained by constrained layer damping construction (plywood + asphalt/felt + MDF sandwich from front to back) for the baffle? This might make it a bit easier as well for me to get the open area in the back around the TC9FD.
Did you make any cardboard models of tapered asymmetric baffle wings in the Ultra (like you used for the original Manzanita)? Do you think this would this be worthwhile for the Ultra?
Also, there is lots of talk in this and other forums about chamfering front baffle edges to improve diffraction. Did you by chance model this or do you think it might give more even off axis response?
Thanks in advance!
As we wait for John's answers, I'll provide a few from my experience.
A thick, stiff baffle certainly does help. 1" thick can do the trick. I've done some experimenting with layers of tar or roofing felt. Results were inconclusive. Sand filled walls are remarkable good, but remarkable troublesome to build.
Asymmetric wings do work well. They have been dropped in later Manzanita designs for ease of construction.
If you can chamfer the edges, it's a good idea. I've used felt and other things on the baffle to lessen diffraction. It helps in the higher frequencies. As the wavelengths get longer, they don't care much what you are doing.
A thick, stiff baffle certainly does help. 1" thick can do the trick. I've done some experimenting with layers of tar or roofing felt. Results were inconclusive. Sand filled walls are remarkable good, but remarkable troublesome to build.
Asymmetric wings do work well. They have been dropped in later Manzanita designs for ease of construction.
If you can chamfer the edges, it's a good idea. I've used felt and other things on the baffle to lessen diffraction. It helps in the higher frequencies. As the wavelengths get longer, they don't care much what you are doing.