In an earlier thread I inquired about active crossover options, and I am still going that route, (DBX 260) but I really like the idea of *not* having a rack of amplifiers taking up space in my studio. More than that, I am really enjoying learning about crossover design and have been fiddling around in xsim/reading about crossovers for a couple of weeks now.
I'd love yalls thoughts on this hypothetical xover. 2nd and 3rd order to keep the cost and component count down, but the bandpass on the midrange may need something more as there are dips and bumps that I can't seem to tame while keeping the impedance curve reasonable.
I'd love yalls thoughts on this hypothetical xover. 2nd and 3rd order to keep the cost and component count down, but the bandpass on the midrange may need something more as there are dips and bumps that I can't seem to tame while keeping the impedance curve reasonable.
Attachments
' dips and bumps' - don't worry about them until you hear how they sound
I assume R3 is the resistance of the inductor; consider using a cored conductor (ferrite) for minimum resistance & power loss
I assume R3 is the resistance of the inductor; consider using a cored conductor (ferrite) for minimum resistance & power loss
Are you using the frd & zma files from the manufacturer?
They're decent for practicing with, but note that once you put the drivers in a real box will baffle effects, you'll get different starting responses.
VituixCad is a little more complex than XSim but can simulate more of these effects if you want to try to get to grips with them too.
They're decent for practicing with, but note that once you put the drivers in a real box will baffle effects, you'll get different starting responses.
VituixCad is a little more complex than XSim but can simulate more of these effects if you want to try to get to grips with them too.
If this is just for playing around with networks then you're not doing too badly. You mention impedance, so try activating the individual phase plots instead of the system phase.
The resistor in series with your woofer should hopefully be the DCR of the coil only. To add resistors in series with woofers is not typically a good idea.
Wolf
Wolf
This is indeed me just playing around. The resistors in series with the mid and low driver were an attempt to bring the sensitivity down a touch, but I now realize I should be using an L-pad, which probably explains why the dissipation of R1 is so high? I’ll have to play around with that.
I am using the FRD and ZMA files provided by Dayton, yes. I do realize that once these are put into a baffle the response will change, and I will likely have to compensate for that. I was really just trying to see how nice and flat of a curve I could get. I cut all of my MDF panels for this build today, as well as a test baffle which I will use to see how these drivers respond once they’re in place.
I am using the FRD and ZMA files provided by Dayton, yes. I do realize that once these are put into a baffle the response will change, and I will likely have to compensate for that. I was really just trying to see how nice and flat of a curve I could get. I cut all of my MDF panels for this build today, as well as a test baffle which I will use to see how these drivers respond once they’re in place.
As Wolf suggests, it's a good idea to work the sensitivity to fit in with the woofer. This way there are no resistors in the woofer circuit for that purpose (there are other reasons for using resistors there, usually in connection with resonances).
Your woofer will only be so loud in an infinite baffle. If you don't have that and you consider the baffle step you will not need any resistors for the woofer.
Looking only at on-axis frequency response is only one part of the story. What does this crossover look like off axis, e.g. horizontal plane, 30 and 60 degrees?
answers to rate my xover
Most important is what you think of the sonics your xover produces ...What are you trying to achieve. ? .
Sadly This is a question for all the techies who like playing around with mathematical formulae. Unless it is doing something that effects the rest of your system.. then Does it matter.? The purpose of a crossover to me is to acquire a sound that closely represents accuratly the original signals recorded. Adding numerous parts to a circuit can only add or detract to the signal.Not always an improvement ????? Even if you had the best most accurate sounding speaker units in the world ,along with perfect speaker leads,then connectors then the numerous extra crossover parts designed to produce a so called more accurate set of numerics is ignoring the fact that their are so many of these mathematical calculations ? not just the ones that are favourites to be held in high esteem because they are currently in fashion ? and not all individually are relevant . Some mathematicall improvements would probably and usually sound worse. Proved it many times....Mainly due to the side effects of the extra parts and unconsidered or ignored other effects.... Most of the posts and advice on this site seems to be given by academics not music listeners. And very few listen in high resolution or to original sound , Its just a sound to them (if they listen? ) If the calculations they agree with are present,then the sound must be correct . Some just want to produce their own personal sound..
Recently I heard someone state that they thought a particular cheap resistor had colourations therefore sounded better that non coloured Vishays. ?
I bet they could not identify a good violin from a strad as they all sound the same.
Rant over.
Hope it effects at least one person...ie 0ne small step for mankind towards music listening and semi accurate music reproduction.
I suggest you try your own or a simplified version and listen. Then note how much better or worse ? it sounds with changes.
My phase is better than your phase.My base drum sounds louder etc etc Give me a break. Best Tools. are EARS.
The systems weakest link is usually the most relevent factor.
Most important is what you think of the sonics your xover produces ...What are you trying to achieve. ? .
Sadly This is a question for all the techies who like playing around with mathematical formulae. Unless it is doing something that effects the rest of your system.. then Does it matter.? The purpose of a crossover to me is to acquire a sound that closely represents accuratly the original signals recorded. Adding numerous parts to a circuit can only add or detract to the signal.Not always an improvement ????? Even if you had the best most accurate sounding speaker units in the world ,along with perfect speaker leads,then connectors then the numerous extra crossover parts designed to produce a so called more accurate set of numerics is ignoring the fact that their are so many of these mathematical calculations ? not just the ones that are favourites to be held in high esteem because they are currently in fashion ? and not all individually are relevant . Some mathematicall improvements would probably and usually sound worse. Proved it many times....Mainly due to the side effects of the extra parts and unconsidered or ignored other effects.... Most of the posts and advice on this site seems to be given by academics not music listeners. And very few listen in high resolution or to original sound , Its just a sound to them (if they listen? ) If the calculations they agree with are present,then the sound must be correct . Some just want to produce their own personal sound..
Recently I heard someone state that they thought a particular cheap resistor had colourations therefore sounded better that non coloured Vishays. ?
I bet they could not identify a good violin from a strad as they all sound the same.
Rant over.
Hope it effects at least one person...ie 0ne small step for mankind towards music listening and semi accurate music reproduction.
I suggest you try your own or a simplified version and listen. Then note how much better or worse ? it sounds with changes.
My phase is better than your phase.My base drum sounds louder etc etc Give me a break. Best Tools. are EARS.
The systems weakest link is usually the most relevent factor.
Raffa, you’re speaking my language! I have been mixing live shows, recording studio albums and playing in bands for 15 years. I am not an audiophile by any stretch, I just love music and being able to reproduce sound accurately in the control room just happens to be something I need to do my job. I’ve been using the Yamaha HS series for years, and they are a far from perfect speaker, but I know them well, as I will get to know these once they’re completed.
The perspective of the hyper analytical and academic types is very helpful to me. It gives me many new things to consider. At the end of the day however, it’s a fun project and a tool I will hopefully use for a very long time. What sounds great to me might not sound great to someone else, and I think that subjectivity is very interesting. The marriage of art and science is a cool thing!
As far as the “sound” of caps/resistors goes, I have yet to hear it, but the placebo effect is very real and I don’t discount people’s experiences in that regard. Cheaper components do generally have lower tolerances, though, and that can be audible, certainly.
The perspective of the hyper analytical and academic types is very helpful to me. It gives me many new things to consider. At the end of the day however, it’s a fun project and a tool I will hopefully use for a very long time. What sounds great to me might not sound great to someone else, and I think that subjectivity is very interesting. The marriage of art and science is a cool thing!
As far as the “sound” of caps/resistors goes, I have yet to hear it, but the placebo effect is very real and I don’t discount people’s experiences in that regard. Cheaper components do generally have lower tolerances, though, and that can be audible, certainly.
Also, I will adjust the drivers around the woofer, thank you all for the input. I am curious how a crossover could affect the on/off axis response? I was under the impression that off axis response was fixed by design of the transducer/baffle. I had no idea passive components could affect response off axis. Any good things to read re: that?
You are correct about that, but there are a couple of small but important exceptions. One is that you 'blend' the directivities when you decide the share between the different ways, and it is possible to make good or not so good decisions there.
The other is that they interact due to their physical separation and you get lobing. A crossover has a hand in how this plays out.
The other is that they interact due to their physical separation and you get lobing. A crossover has a hand in how this plays out.
If you are using the manufacturer's curves, they are most likely to be 2-pi responses. Also known as infinite baffle responses. This is what STV was getting at.
You are going to put these drivers in a box. At low frequencies the box is a free-space radiator, also called 4-pi. At high frequencies it is a 2-pi radiator. There is a 6 dB difference between a 2-pi and a 4-pi radiator, even with the same energy. The transition region from 4-pi to 2-pi is very broad, usually spanning about 3 octaves.
In your case, using published data to make a hypothetical crossover, you should try to get a smooth 6 dB drop from 100 Hz to 1000 Hz, and then a flat response from 1000 Hz on up. This will approximate the baffle step. The result (in theory) would be a flat response for the box.
To achieve this, get rid of the woofer series resistor, and increase the inductance until you get a 6 dB down slope from 100 to 1000 Hz. Then add resistance to the tweeter and midrange to bring them down in level.
Your baffle shape causes peaks and dips in the on-axis response which may not be present in the off-axis. If you design a crossover to be flat on axis, you have inadvertently compensated for those peaks and dips. The off axis response will suffer. In the vertical direction, there is additional driver offset and cancellation effects.
Nice work so far... keep going !
j.
You are going to put these drivers in a box. At low frequencies the box is a free-space radiator, also called 4-pi. At high frequencies it is a 2-pi radiator. There is a 6 dB difference between a 2-pi and a 4-pi radiator, even with the same energy. The transition region from 4-pi to 2-pi is very broad, usually spanning about 3 octaves.
In your case, using published data to make a hypothetical crossover, you should try to get a smooth 6 dB drop from 100 Hz to 1000 Hz, and then a flat response from 1000 Hz on up. This will approximate the baffle step. The result (in theory) would be a flat response for the box.
To achieve this, get rid of the woofer series resistor, and increase the inductance until you get a 6 dB down slope from 100 to 1000 Hz. Then add resistance to the tweeter and midrange to bring them down in level.
Your baffle shape causes peaks and dips in the on-axis response which may not be present in the off-axis. If you design a crossover to be flat on axis, you have inadvertently compensated for those peaks and dips. The off axis response will suffer. In the vertical direction, there is additional driver offset and cancellation effects.
Nice work so far... keep going !
j.