Just quickly for now:
- in the 1st xo schematic, the trio of 1041 parts is the tweeter resonance compensation filter
- generally you don't want to add series resistance to the woofer circuit - it's a waste of power and it alters the woofer's box response in the low end
- I would say that between about 1500Hz and 4000Hz, it's playing a little too loud. If anything, this area might sound better just slightly attenuated from the main level.
- matching the phase at the point where it goes vertical looks good but isn't necessarily the best alignment. Look where your xo point is and what the phase matching looks like above and below that point.
- re the very top end tweeter notch: if you look at how Troels has used that tweeter, I don't think he's ever been concerned about the rising top end above 10kHz.
Another thought on your cabinet if 1 is already made and permanently closed up: use a jigsaw to cut out the bottom panel. Use a router with a flush trim bit to clean it up so the sides are all the same thickness as the side and front and back panels. Now glue in a new bottom panel but higher up inside the cabinet, thus decreasing Vb and now creating a space to put your xo underneath the speaker.
- in the 1st xo schematic, the trio of 1041 parts is the tweeter resonance compensation filter
- generally you don't want to add series resistance to the woofer circuit - it's a waste of power and it alters the woofer's box response in the low end
- I would say that between about 1500Hz and 4000Hz, it's playing a little too loud. If anything, this area might sound better just slightly attenuated from the main level.
- matching the phase at the point where it goes vertical looks good but isn't necessarily the best alignment. Look where your xo point is and what the phase matching looks like above and below that point.
- re the very top end tweeter notch: if you look at how Troels has used that tweeter, I don't think he's ever been concerned about the rising top end above 10kHz.
Another thought on your cabinet if 1 is already made and permanently closed up: use a jigsaw to cut out the bottom panel. Use a router with a flush trim bit to clean it up so the sides are all the same thickness as the side and front and back panels. Now glue in a new bottom panel but higher up inside the cabinet, thus decreasing Vb and now creating a space to put your xo underneath the speaker.
Thanks jReave,
That helps explain some things that are still not clearly in focus. I will remove 1 ohm resistor in woofer circuit. I was using it to tame a big bulge in the lower region. If I can't remove that I will look to your expertise for assistance. With regard to phase matching, I probably have concentrated to much with the vertical matching and not enough above and below as you pointed out. As for cabinet, your thoughts are exactly what I had planned as well. I wanted to keep the same overall cabinet height but lower the vb. I briefly considered cutting the bottom of the box to achieve desired vb, but removing approximately 8 inches would have made box way to short.
Best,
Rich
That helps explain some things that are still not clearly in focus. I will remove 1 ohm resistor in woofer circuit. I was using it to tame a big bulge in the lower region. If I can't remove that I will look to your expertise for assistance. With regard to phase matching, I probably have concentrated to much with the vertical matching and not enough above and below as you pointed out. As for cabinet, your thoughts are exactly what I had planned as well. I wanted to keep the same overall cabinet height but lower the vb. I briefly considered cutting the bottom of the box to achieve desired vb, but removing approximately 8 inches would have made box way to short.
Best,
Rich
Hi jReave,
I have attached a few more screenshots of my latest invention. Its a bit bumpy in places but I'm having a tough time smoothing out the same troublesome peaks without throwing a bunch of components into the design. Even then, it has proven very challenging to add in notch filters and still keep impedance from dipping way low. Anyway, your help would be very much appreciated if further tweaking is necessary.
Best Regards,
Rich
I have attached a few more screenshots of my latest invention. Its a bit bumpy in places but I'm having a tough time smoothing out the same troublesome peaks without throwing a bunch of components into the design. Even then, it has proven very challenging to add in notch filters and still keep impedance from dipping way low. Anyway, your help would be very much appreciated if further tweaking is necessary.
Best Regards,
Rich
Attachments
I think I'll just attach what I've come up with when I use Troels' style of xo. I like how it smooths out the woofer response more than our previous attempts. But it does take more parts and the impedance does go down to about 4ohm although I don't think that's problematic for most amplifiers.
So the choice is up to you which way you want to go with it. Just sitting here working the simulations, it's not clear to me which one is going to sound better, if at all, although I may lean a little more toward this latest one. If it comes down to xo costs, for the large capacitors, you can get away with cheaper, non polarized electrolytics which makes a big difference money-wise.
I was also going to get you to work on another xo in which you look at trying a little less baffle step compensation, but given your room and speaker placement, I don't think that that's going to be what you need so I kind of think you're ready to move on to the next step unless you have some further questions.
So the choice is up to you which way you want to go with it. Just sitting here working the simulations, it's not clear to me which one is going to sound better, if at all, although I may lean a little more toward this latest one. If it comes down to xo costs, for the large capacitors, you can get away with cheaper, non polarized electrolytics which makes a big difference money-wise.
I was also going to get you to work on another xo in which you look at trying a little less baffle step compensation, but given your room and speaker placement, I don't think that that's going to be what you need so I kind of think you're ready to move on to the next step unless you have some further questions.
Attachments
Thanks jReave,
I really appreciate your assistance on this journey. It's been an eye opening experience and one that been very enjoyable. The only question I have is: trying to get my head around mod delay in Xsim. I noticed it says positive delay moves driver away from the mic. My confusion comes from the idea of the woofer timing being behind the tweeter, it seems in my mind that the woofer needs to be brought closer to the mic (if mic represents the listener). I know there is a reasonable explanation that I have failed to consider. Most everything else throughout the process has registered in my mind but this one area has thrown me a curve ball.
Best Regards,
Rich
I really appreciate your assistance on this journey. It's been an eye opening experience and one that been very enjoyable. The only question I have is: trying to get my head around mod delay in Xsim. I noticed it says positive delay moves driver away from the mic. My confusion comes from the idea of the woofer timing being behind the tweeter, it seems in my mind that the woofer needs to be brought closer to the mic (if mic represents the listener). I know there is a reasonable explanation that I have failed to consider. Most everything else throughout the process has registered in my mind but this one area has thrown me a curve ball.
Best Regards,
Rich
You have the concept down. Woofer acoustic center is behind the tweeter acoustic center. If you wanted to physically align them, you would indeed bring the woofer forward as Troels did for eg with his Discovery 18 by adding another piece of wood to the front baffle (suggesting that the acoustic center difference between drivers in actuality is close to 3/4").
In either case though, all you are doing is telling the xo program what this distance actually is. In Xsim, the author has chosen to conceptualize this as a positive distance. On the other hand in PCD, Jeff Bagby chose to think of it in negative terms and you input the distance as a negative number.
Now in PCD, you also have to tell the program what the height differential is between the 2 drivers as whereas in XSim you do not. Because if you stop to think of it for a second or perhaps draw up a diagram, if you are listening on the tweeter axis then the distance to the woofer is actually a little longer than just the distance to the tweeter plus the extra distance to the woofer acoustic center because you are also dealing with a downward angle, hypotenuse of a triangle and all that. Thus I had you use a value for the woofer delay in XSim that was a little larger than you measured on the woofer diagram because XSim doesn't include driver height information.
Not sure if that clears it up or makes it more confusing.
In either case though, all you are doing is telling the xo program what this distance actually is. In Xsim, the author has chosen to conceptualize this as a positive distance. On the other hand in PCD, Jeff Bagby chose to think of it in negative terms and you input the distance as a negative number.
Now in PCD, you also have to tell the program what the height differential is between the 2 drivers as whereas in XSim you do not. Because if you stop to think of it for a second or perhaps draw up a diagram, if you are listening on the tweeter axis then the distance to the woofer is actually a little longer than just the distance to the tweeter plus the extra distance to the woofer acoustic center because you are also dealing with a downward angle, hypotenuse of a triangle and all that. Thus I had you use a value for the woofer delay in XSim that was a little larger than you measured on the woofer diagram because XSim doesn't include driver height information.
Not sure if that clears it up or makes it more confusing.
Hi Rich,
Just because I can't help myself sometimes.....
Thinking that 2nd order acoustic might be the basis for a better sounding xo, I thought I would try one more time attempting not to use that notch filter on the woofer with such big values this time.
I started by increasing L1 on the woofer but still found the need for the notch filter but this time with a smaller cap but a bigger inductor. I needed to add another notch filter to the tweeter but this time the values are all small so that won't be much additional cost at all.
The summed FR is about the same as the previous one, but this time the phase looks a little better and the impedance doesn't drop as low.
Might be a keeper. I just haven't double checked my values against what's available yet though.
Just because I can't help myself sometimes.....
Thinking that 2nd order acoustic might be the basis for a better sounding xo, I thought I would try one more time attempting not to use that notch filter on the woofer with such big values this time.
I started by increasing L1 on the woofer but still found the need for the notch filter but this time with a smaller cap but a bigger inductor. I needed to add another notch filter to the tweeter but this time the values are all small so that won't be much additional cost at all.
The summed FR is about the same as the previous one, but this time the phase looks a little better and the impedance doesn't drop as low.
Might be a keeper. I just haven't double checked my values against what's available yet though.
Attachments
Thanks jReave,
Wow! That looks Awesome! You are right that really looks like the best version. I'm assuming because this woofer doesn't have much breakup that there's no reason to try achieve steeper slopes. Btw, your explanation about driver delays and differences between the two programs really helped a lot. As I gain a little more experience, I will probably lean more toward PCD. I have plinked around a little as we have went through this process. Also, if you have any recommendations for crossover parts, I'm all ears. I was leaning toward purchasing Sonic cap gen 1's from Sonicraft for those capacitors in series. But, try to cut down cost by purchasing NPE's or for those in the shunts. I will wait to hear back from you before placing an order on any components.
Best Regards,
Rich
Wow! That looks Awesome! You are right that really looks like the best version. I'm assuming because this woofer doesn't have much breakup that there's no reason to try achieve steeper slopes. Btw, your explanation about driver delays and differences between the two programs really helped a lot. As I gain a little more experience, I will probably lean more toward PCD. I have plinked around a little as we have went through this process. Also, if you have any recommendations for crossover parts, I'm all ears.
I was leaning toward purchasing Sonic cap gen 1's from Sonicraft for those capacitors in series. But, try to cut down cost by purchasing NPE's or for those in the shunts. I will wait to hear back from you before placing an order on any components.Best Regards,
Rich
Hey Rich
So looking at my last version, it's actually got a lot of parts that are going to add up especially if you go for higher end caps. That led me to thinking you might want to look at this as a 2 stage process because what we've come up with now are essentially 2 good xo versions:
1) with 4th order slopes and a xo point at about 1700Hz with a FR that's a little uneven here and there but not actually too bad. (1st three charts below)
2) with 2nd order slopes with a xo point at about 2200Hz with the flattest FR. (next 3 charts)
So here's 1 approach you might want to try. Start off putting option 1 together and listening to that. If you are happy with it, then you're done. If however you think there might be room for improvement (albeit that might actually be hard to know) or if you just can't help yourself, then add in the extra parts to create version 2.
Now I guess if you do end up going to version 2 with this approach, it'll be a little bit more expensive than if you just went there directly to start off with. However, if you look closely at both schematics, what you can do is put v1 together with some smart choices of capacitor values so that none of the extra ones from v1 will go to waste in v2. It'll just be the 2 inductors in v1 that might need replacing and I say 'might' because if you have a multi meter, you can actually unwind the bigger values until you are arrive at the smaller ones. But if not, you'll really need to buy new ones because both of those inductor values have to be pretty much bang on in v2 for everything to work properly.
Anyways, it's just a thought. I'm hoping that you already know how capacitors sum in parallel and how resistors sum in parallel and in series so you will definitely want to buy some extra 1ohm and maybe 2ohm resistors. Maybe some extra 1 and 2uF capacitors as well. In particular, you will want to be playing with your tweeter's series resistance to see what level sounds best to your ears.
In terms of capacitors, I've never gone crazy and spent the money on any of the high end ones and I think you are going to get as many opinions on this subject as there are capacitors out there. Just from what I've read though, the Sonicaps might sound brittle and forward. But on the other hand, Humble Home Hifi likes them and gives you his opinion on all the brands he's tested. I've seen Clarity SA caps recommended for higher end value. Shipping might be a little more expensive with those though, I'm not really sure.
My suggestions would be to keep a balance between the price of your drivers and the price of the xo components in mind. But the tweeter series caps is the right place to splurge. Solens or equivalent should be good for the rest with an NPE for the big sucker in the tweeter LCR filter.
Cheers
So looking at my last version, it's actually got a lot of parts that are going to add up especially if you go for higher end caps. That led me to thinking you might want to look at this as a 2 stage process because what we've come up with now are essentially 2 good xo versions:
1) with 4th order slopes and a xo point at about 1700Hz with a FR that's a little uneven here and there but not actually too bad. (1st three charts below)
2) with 2nd order slopes with a xo point at about 2200Hz with the flattest FR. (next 3 charts)
So here's 1 approach you might want to try. Start off putting option 1 together and listening to that. If you are happy with it, then you're done. If however you think there might be room for improvement (albeit that might actually be hard to know) or if you just can't help yourself, then add in the extra parts to create version 2.
Now I guess if you do end up going to version 2 with this approach, it'll be a little bit more expensive than if you just went there directly to start off with. However, if you look closely at both schematics, what you can do is put v1 together with some smart choices of capacitor values so that none of the extra ones from v1 will go to waste in v2. It'll just be the 2 inductors in v1 that might need replacing and I say 'might' because if you have a multi meter, you can actually unwind the bigger values until you are arrive at the smaller ones. But if not, you'll really need to buy new ones because both of those inductor values have to be pretty much bang on in v2 for everything to work properly.
Anyways, it's just a thought. I'm hoping that you already know how capacitors sum in parallel and how resistors sum in parallel and in series so you will definitely want to buy some extra 1ohm and maybe 2ohm resistors. Maybe some extra 1 and 2uF capacitors as well. In particular, you will want to be playing with your tweeter's series resistance to see what level sounds best to your ears.
In terms of capacitors, I've never gone crazy and spent the money on any of the high end ones and I think you are going to get as many opinions on this subject as there are capacitors out there. Just from what I've read though, the Sonicaps might sound brittle and forward. But on the other hand, Humble Home Hifi likes them and gives you his opinion on all the brands he's tested. I've seen Clarity SA caps recommended for higher end value. Shipping might be a little more expensive with those though, I'm not really sure.
My suggestions would be to keep a balance between the price of your drivers and the price of the xo components in mind. But the tweeter series caps is the right place to splurge. Solens or equivalent should be good for the rest with an NPE for the big sucker in the tweeter LCR filter.
Cheers
Attachments
Also if you want to see how I've been trying to balance the on and off-axis responses, here's how version 1 looks in PCD. Green is the power response. Notice in particular how the top end of the tweeter falls off as you move off-axis.
Version 2 follows the same pattern but I can't actually construct that exact same xo filter in PCD as I can in XSim. That flexibility is a nice advantage for XSim.
Version 2 follows the same pattern but I can't actually construct that exact same xo filter in PCD as I can in XSim. That flexibility is a nice advantage for XSim.
Attachments
Thanks a Million jReave!
It appears I have choices to go on. I'm leaning towards towards options 2 even though it is more expensive to pull together. I have begun pricing it out through Parts Express and the cost should not be to staggering. The Sonicap idea came from Humble Homemade Hifi. It seems like a very good cost to performance cap. I'm willing to splurge and give them a try. Today I'm working on assembly of cabinet #2. Ordered some veneer from Ebay a couple days ago and once it arrives, I will try applying that for the first time. Hopefully, I can see project completed over the next month or so before things get crazy at work.
Best Regards,
Rich
It appears I have choices to go on. I'm leaning towards towards options 2 even though it is more expensive to pull together. I have begun pricing it out through Parts Express and the cost should not be to staggering. The Sonicap idea came from Humble Homemade Hifi. It seems like a very good cost to performance cap. I'm willing to splurge and give them a try. Today I'm working on assembly of cabinet #2. Ordered some veneer from Ebay a couple days ago and once it arrives, I will try applying that for the first time.
Hopefully, I can see project completed over the next month or so before things get crazy at work.Best Regards,
Rich
Best of luck. Keep us informed of your progress.
If this is going to be the 1st time you put together a xo, these links might prove helpful:
Placement of coils in crossover networks
nodal_analysis
If this is going to be the 1st time you put together a xo, these links might prove helpful:
Placement of coils in crossover networks
nodal_analysis
Well, I finally completed the 2 way tower build that I started earlier this year. Thanks to all who helped make this build possible! A special thanks to diyaudio member jReave for his expert guidance and extreme patience leading me along this journey. This has been quite the learning experience. As the great Freddie Mercury wrote, " Mistakes I've Made A Few". Please don't be to harsh, since this was my first build. The next one is sure to be better.
Based on recommendation by jReave, I decided to use Scanspeak 18/8434G-00 Discovery 7" midwoofers and Scanspeak D2604/8330 tweeters. This was an excellent recommendation (especially for the cost to performance ratio). The sound quality is outstanding (at least to my ears). The frequency respose seems well balanced. If anything, bass may be a little thin giving their current room setup. But, my wife kinda claimed them for the living room when they were originally intended for my office which is a bit smaller. I don't mind because all that means is there is another project in the hopper. Still yet, they sound very nice and the wife is impressed. Since she will let me build about anything I desire; her opinion matters. As a person new to this hobby, it is difficult to describe aspects of sound. Maybe best to try and answer questions from the group. I attached some photos to take a look at. My deepest appreciation to this great forum!!!
Best Regards,
Rich
The next one is sure to be better. Based on recommendation by jReave, I decided to use Scanspeak 18/8434G-00 Discovery 7" midwoofers and Scanspeak D2604/8330 tweeters. This was an excellent recommendation (especially for the cost to performance ratio). The sound quality is outstanding (at least to my ears). The frequency respose seems well balanced. If anything, bass may be a little thin giving their current room setup. But, my wife kinda claimed them for the living room when they were originally intended for my office which is a bit smaller. I don't mind because all that means is there is another project in the hopper. Still yet, they sound very nice and the wife is impressed. Since she will let me build about anything I desire; her opinion matters. As a person new to this hobby, it is difficult to describe aspects of sound. Maybe best to try and answer questions from the group. I attached some photos to take a look at. My deepest appreciation to this great forum!!!
Best Regards,
Rich
Attachments
Hey Rich,
Those look fantastic. That is some very fine woodworking there my friend. Very glad that your wife likes them too. Nice simple xo in the end and a decent looking FR too although what really matters is how they sound to you (well and to your significant other too.)
Concerning the slight lack of bass, I'd first probably suggest listening to them for a while to become accustomed to them and let them get well worked in if you haven't already done so. Perhaps you might be used to speakers that actually have a little more bass than perfectly anechoically flat and you might come to appreciate these as they are.
If on the other hand your personal tastes would like a little more on the bottom end, I guess here are a few things you might try;
1) Move them closer to the back and/or side wall for a little more boundary reenforcement. (Depends on your seating distance, but if I was listening to music, I would probably have them further apart as well.)
2) Dial up the bass on your receiver or run its equalization program if it's got one.
3) Not sure if you have access still to the interior but you could try shortening the port in small increments. A little more internal volume would be better but somehow I think that's not going to work out at this point. Maybe try it first in a modeling program to see the effects on the FR.
4) And the most difficult probably at this point, start playing with the xo again.
Once again, very nicely done.
Those look fantastic. That is some very fine woodworking there my friend. Very glad that your wife likes them too. Nice simple xo in the end and a decent looking FR too although what really matters is how they sound to you (well and to your significant other too.
)Concerning the slight lack of bass, I'd first probably suggest listening to them for a while to become accustomed to them and let them get well worked in if you haven't already done so. Perhaps you might be used to speakers that actually have a little more bass than perfectly anechoically flat and you might come to appreciate these as they are.
If on the other hand your personal tastes would like a little more on the bottom end, I guess here are a few things you might try;
1) Move them closer to the back and/or side wall for a little more boundary reenforcement. (Depends on your seating distance, but if I was listening to music, I would probably have them further apart as well.)
2) Dial up the bass on your receiver or run its equalization program if it's got one.
3) Not sure if you have access still to the interior but you could try shortening the port in small increments. A little more internal volume would be better but somehow I think that's not going to work out at this point. Maybe try it first in a modeling program to see the effects on the FR.
4) And the most difficult probably at this point, start playing with the xo again.
Once again, very nicely done.
Thanks jReave,
There is some errors in regard to craftsmanship, but they sound much better then they look (IMHO). I appreciate the input you provided on boosting bass response from these speakers. I'm going to take your advice and live with them for a while before making any changes. If at some point bass or lack thereof becomes an issue: I will start implementing the suggestions you offered. I suppose that is somewhat of the addictive part of our hobby. The thought, it can always be a little better. It is certainly great hearing from you and hope you are doing well!
Best Regards,
Rich
There is some errors in regard to craftsmanship, but they sound much better then they look (IMHO). I appreciate the input you provided on boosting bass response from these speakers. I'm going to take your advice and live with them for a while before making any changes. If at some point bass or lack thereof becomes an issue: I will start implementing the suggestions you offered. I suppose that is somewhat of the addictive part of our hobby. The thought, it can always be a little better.
It is certainly great hearing from you and hope you are doing well!Best Regards,
Rich
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.