Project Update
Folks:
I know there is interest in the design I've been working on. As you know from my prior update, the x-over design is coming along quite nicely however, as I've been listening to the x-over design over the last few weeks I've noticed some odd behavior in the upper mid-range area. Actual response testing has shown me that there is a rising output above 4K which needs to be dealt with as well as a big spike around 15k. I've spent some time testing various L-PAD network attenuators on the tweeter to try and resolve the issue. At this point, being the perfectionist I am, I am going to try and model some new x-over designs for the tweeter.
In addition to the frequency response problem described above, I am also having another issue which has to do with making sure the L-PAD network on the tweeter doesn't overheat. The tweeter is so responsive relative to the woofer that a significant amount of attenuation is required. I've noted that the 10 watt non-inductive resistors I'm using are getting extremely hot at high volume levels. I need to make sure that the resistors are able to effectively dissipate enough energy fast enough that they don't overheat.
I still feel that, despite the problems I'm working through now, that this design will end up being quite good. Furthermore, given the price drop on the WR125S from CSS, it might make a lot of sense for me to consider running the system with 2 woofers instead of one and thereby improve power handling and efficiency overall.
I will post another update as I find the time to continue with this project . . .
Regards,
Andrew
Folks:
I know there is interest in the design I've been working on. As you know from my prior update, the x-over design is coming along quite nicely however, as I've been listening to the x-over design over the last few weeks I've noticed some odd behavior in the upper mid-range area. Actual response testing has shown me that there is a rising output above 4K which needs to be dealt with as well as a big spike around 15k. I've spent some time testing various L-PAD network attenuators on the tweeter to try and resolve the issue. At this point, being the perfectionist I am, I am going to try and model some new x-over designs for the tweeter.
In addition to the frequency response problem described above, I am also having another issue which has to do with making sure the L-PAD network on the tweeter doesn't overheat. The tweeter is so responsive relative to the woofer that a significant amount of attenuation is required. I've noted that the 10 watt non-inductive resistors I'm using are getting extremely hot at high volume levels. I need to make sure that the resistors are able to effectively dissipate enough energy fast enough that they don't overheat.
I still feel that, despite the problems I'm working through now, that this design will end up being quite good. Furthermore, given the price drop on the WR125S from CSS, it might make a lot of sense for me to consider running the system with 2 woofers instead of one and thereby improve power handling and efficiency overall.
I will post another update as I find the time to continue with this project . . .
Regards,
Andrew
Hi ABS, are you still working with the TN25 tweets? I bought a pair and have mounted them on top of my WR125 boxes, C-C distance is ~ 4.5". I threw a 22uf cap on them for protection since I'm crossing them actively, and tried crossing at 2nd order and 4th order, various frequencies from 3.5-6k.
It was certainly an interesting learning experience, playing with these drivers and the KXproject active XO drivers... At the touch of the mouse wheel XO frequencies, relative levels, delay, etc can all be adjusted while the music plays.
Keeping in mind that my setup is nearfield, in that I sit about 1.5 feet away from the speakers at all times, here are my findings: With the tweeter active, I was never able to get the same imaging I had with the single driver, no matter how well integrated the sound was. The position of vocalists and instruments just seemed somehow more 'vague'. Playing with the delay on the tweeter helped here, but it mainly changed where the center image appeared, varying from about 2 feet behind the speakers to about 6 inches in front of them.
What the tweeters did help was the very small sweet spot, ie I could move my head around much more and not have the top end change as much. Once it did start to change though, it was in very ugly ways, with just the WR125, the highs roll off drastically when you get a bit off axis (remember this is nearfield), with the WR125 + tweeter, the highs did not roll off much, but get far enough off axis and the sound turned really 'phasey' for lack of a better word, with noticeable lobing effects as I moved my head up and down a few inches. FWIW, I tried listening to the combo a few feet further back, and the phasiness was much reduced, and coherence increased a good deal as well. They also add that 'sparkle' on some tracks, but not nearly as well as I expected, which I suspect has to do with the tweeter or the HF limitations of my sound card/cheap amps.
In the end I decided that the extra complexity, amp, and the weird 'phasiness' off axis were not worth the extra sparkle and somewhat wider listening window. I will probably end up replacing the WRs with the new FR version when it comes out (provided they will work in the same box/cutout), and maybe use the TN25s in a future project aimed at a more conventional listening distance, with a pair of WR125s in a 2.5way arrangement.
AS for the TN25, it seems to me that it is a pretty mediocre tweeter, certainly it sounds worse than either the Vifa D27 tweets in my ACI Spirit/Opals, and the Linaeum tweeters on my little LX-5s. Maybe I just haven't heard them in the right configuration yet, future testing will tell.
It was certainly an interesting learning experience, playing with these drivers and the KXproject active XO drivers... At the touch of the mouse wheel XO frequencies, relative levels, delay, etc can all be adjusted while the music plays.
Keeping in mind that my setup is nearfield, in that I sit about 1.5 feet away from the speakers at all times, here are my findings: With the tweeter active, I was never able to get the same imaging I had with the single driver, no matter how well integrated the sound was. The position of vocalists and instruments just seemed somehow more 'vague'. Playing with the delay on the tweeter helped here, but it mainly changed where the center image appeared, varying from about 2 feet behind the speakers to about 6 inches in front of them.
What the tweeters did help was the very small sweet spot, ie I could move my head around much more and not have the top end change as much. Once it did start to change though, it was in very ugly ways, with just the WR125, the highs roll off drastically when you get a bit off axis (remember this is nearfield), with the WR125 + tweeter, the highs did not roll off much, but get far enough off axis and the sound turned really 'phasey' for lack of a better word, with noticeable lobing effects as I moved my head up and down a few inches. FWIW, I tried listening to the combo a few feet further back, and the phasiness was much reduced, and coherence increased a good deal as well. They also add that 'sparkle' on some tracks, but not nearly as well as I expected, which I suspect has to do with the tweeter or the HF limitations of my sound card/cheap amps.
In the end I decided that the extra complexity, amp, and the weird 'phasiness' off axis were not worth the extra sparkle and somewhat wider listening window. I will probably end up replacing the WRs with the new FR version when it comes out (provided they will work in the same box/cutout), and maybe use the TN25s in a future project aimed at a more conventional listening distance, with a pair of WR125s in a 2.5way arrangement.
AS for the TN25, it seems to me that it is a pretty mediocre tweeter, certainly it sounds worse than either the Vifa D27 tweets in my ACI Spirit/Opals, and the Linaeum tweeters on my little LX-5s. Maybe I just haven't heard them in the right configuration yet, future testing will tell.
Oh yeah, one last thing that I found very useful was using the parametric EQ in the KXproject DSP to shape the response. The single biggest improvement here came from imposing a steep rolloff on the frequencies below port tuning. This cleaned up the sound a fair bit on really bass-heavy material. I also did some contouring of the hump at ~ 100hz, and a few peaks higher up, but these changes were far less audible.
Update
Morbo:
Active x-overs can be a lot of fun, the Alpine Head Unit in my car provides active x-over capabilites which I fully utilize. Just for my front tweeters and woofer, I need four channels of amplification. In the car (morel tweeters and seas woofers), I've played with all orders of x-overs. What I've discovered is that the sound quality degrades as the x-over order goes up from 1-4 . . . So, I ended up using the lowest order x-over settings I could. I'm not sure if your project suffers from this as well, but it could be something to consider . . . The listening I'm doing with my WR125S is not nearfield - I want them to be used as the primary speakers in my parents tv/listening room. Eventually, I think they will do quite well as part of a surround sound system or as the mid-tweet combo that can be dropped into a 3-way design . . .
To get specific, I have been modeling with the Vifa DX19 tweeter. This tweeter, although relatively inexpensive provides a very nice sound which, as I understand it, is largely related to its low distortion. The driver is a 21mm diameter so it plays on-axis up to a higher frequency than a traditional 28mm driver. Don't get me wrong, I am not saying that this is a sensationally good tweeter. What I am saying is that it is very, very good for the money. The biggest problem I've had with the tweeter is that it is only 4 ohms and it is extremely efficient relative to the woofer and therefore requires substantial attenuation. Having said that, the recent price drop on the WR125S provide the perfect opportunity for a T-MM configuration with +3dB more output!
I spent a good part of the day today working out the kinks in my x-over design and, I believe I've finally got a result I'm happy with. I decided to simplify the x-over by pulling the impedance compensation circuitry. The sound cleaned up quite a bit afterwards and I ended up altering the x-over point for the midrange by moving it down a bit. I'm now crossing at around 5k with a 2nd order filter on the woofer and a 1st order on the tweeter. This change has smoothed out the upper mid-range dramatically. The speaker now has distinctive mid-forward sound, but it is still easy to listen to and not at all excessive.
Compared to my 2-way Scan-Speak monitors (7" Carbon Fiber woofer and 9300 tweeter), these speakers have a much more cohesive sound and much crisper through the mid--band. Obviously the WR125S doesn't have nearly the same low end as the SS 7", but we all know that for a 4.5" driver the WR125S provides exceptionally good bass output.
I need to do some more listening tests before I finalize the x-over design, but it is nearly done. Once I close the book on this one, I'll be sure to post the design, or better yet, perhaps publish it to my web site.
Andrew
Morbo:
Active x-overs can be a lot of fun, the Alpine Head Unit in my car provides active x-over capabilites which I fully utilize. Just for my front tweeters and woofer, I need four channels of amplification. In the car (morel tweeters and seas woofers), I've played with all orders of x-overs. What I've discovered is that the sound quality degrades as the x-over order goes up from 1-4 . . . So, I ended up using the lowest order x-over settings I could. I'm not sure if your project suffers from this as well, but it could be something to consider . . . The listening I'm doing with my WR125S is not nearfield - I want them to be used as the primary speakers in my parents tv/listening room. Eventually, I think they will do quite well as part of a surround sound system or as the mid-tweet combo that can be dropped into a 3-way design . . .
To get specific, I have been modeling with the Vifa DX19 tweeter. This tweeter, although relatively inexpensive provides a very nice sound which, as I understand it, is largely related to its low distortion. The driver is a 21mm diameter so it plays on-axis up to a higher frequency than a traditional 28mm driver. Don't get me wrong, I am not saying that this is a sensationally good tweeter. What I am saying is that it is very, very good for the money. The biggest problem I've had with the tweeter is that it is only 4 ohms and it is extremely efficient relative to the woofer and therefore requires substantial attenuation. Having said that, the recent price drop on the WR125S provide the perfect opportunity for a T-MM configuration with +3dB more output!
I spent a good part of the day today working out the kinks in my x-over design and, I believe I've finally got a result I'm happy with. I decided to simplify the x-over by pulling the impedance compensation circuitry. The sound cleaned up quite a bit afterwards and I ended up altering the x-over point for the midrange by moving it down a bit. I'm now crossing at around 5k with a 2nd order filter on the woofer and a 1st order on the tweeter. This change has smoothed out the upper mid-range dramatically. The speaker now has distinctive mid-forward sound, but it is still easy to listen to and not at all excessive.
Compared to my 2-way Scan-Speak monitors (7" Carbon Fiber woofer and 9300 tweeter), these speakers have a much more cohesive sound and much crisper through the mid--band. Obviously the WR125S doesn't have nearly the same low end as the SS 7", but we all know that for a 4.5" driver the WR125S provides exceptionally good bass output.
I need to do some more listening tests before I finalize the x-over design, but it is nearly done. Once I close the book on this one, I'll be sure to post the design, or better yet, perhaps publish it to my web site.
Andrew
I've got some really terrific news - my design is finally finished. Last week I realized that I had the L-PAD on the tweeter position BEFORE the 1st order x-over (capacitor) instead of AFTER the x-over. Changing the physical layout of the components made all the difference. The speakers sound extremely good. They are very detailed, the tweeter blends nicely. The mid-range and high end is slightly forward, but not too much so. I will get some documentation together and post it to this site with instructions on how to go about building the same setup. I really couldn't be happier with the sound of these speakers! Even my brother, who is an incredibly critical listener, is impressed with them. Due to the frequency at which I am rolling off the WR125 and the use of the 3/4 tweeter, the off-axis response is very good. The only thing I don't like is that they are not very efficient. For my next project, I may try the same drivers in a TMM configuration to try and get a big improvement in efficiency . . .
Regards,
Andrew
Regards,
Andrew
ABS said:
Our 1st designs with the FR125S are in mine & Chris' living rooms. Even with very low hours, i am very impressed.
We also listened to the WR in the same boxes. They are also very good, but with the extra extension the FR looks to be a real winner. It is easy to see how WR with a well integrated tweeter could be a real killer.
They don't seem to like SE tube amps (all closer to current sources than voltage sources) thou... means you need to use a PP amp (we used miniA & PP EL34 today, and i'm running them at home on a BrianGT GC) which counters the low efficiency somewhat. I am quite amazed at the bass these do.
dave
Great news from everyone involved! ABS, looking forward to seeing your finished schematic! Dave, glad to hear you have good results with the FRs... I wish I could put mine in, but when I tried to remove my WRs from their boxes, I found they are stuck in there something fierce. Looks like its sawdust time again 
Yup. When I was building them, there was no hint of an upgraded version, so I never planned on taking them out. The driver cutouts were so tight after painting, that I had to go over them with sandpaper to get the drivers in at all. Plus that black 'speaker mounting putty' really grips if compressed enough, I've seen it tear off MDF before letting go. It seems a real shame to take apart already finished boxes, especailly ones that I spent so much time on, so I'm trying to figure out what to do with them...
On that note, if anyone is looking for a pair of very well built, WR125 nearfield monitors and/or center channel at a reasonable price, email me.
tolerances
yeppers, they can be tight as a ..... ; As several have mentioned in the other thread, we chased our tails for at least an hour and a half to isolate the problem to ALL SE tube amps on hand that day, during which time all 6 drivers were re and re-d at least twice in each box.
If your rebated mounting cutout is deadly accurate, it's a be-atch to get the drivers out. Fortunately the flange is thick enough to hook a small pair of curved needle-nose pliers in the screw hole to pry them out.
next time, I think I'll machine the cutouts 1mm oversize, and paint the inside of the rebate with black felt marker.
but boy to these little suckers sound good, particularly the bipoles!
morbo said:Great news from everyone involved! ABS, looking forward to seeing your finished schematic! Dave, glad to hear you have good results with the FRs... I wish I could put mine in, but when I tried to remove my WRs from their boxes, I found they are stuck in there something fierce. Looks like its sawdust time again
yeppers, they can be tight as a ..... ; As several have mentioned in the other thread, we chased our tails for at least an hour and a half to isolate the problem to ALL SE tube amps on hand that day, during which time all 6 drivers were re and re-d at least twice in each box.
If your rebated mounting cutout is deadly accurate, it's a be-atch to get the drivers out. Fortunately the flange is thick enough to hook a small pair of curved needle-nose pliers in the screw hole to pry them out.
next time, I think I'll machine the cutouts 1mm oversize, and paint the inside of the rebate with black felt marker.
but boy to these little suckers sound good, particularly the bipoles!
Now why didnt I think of that??? I better get my hands on some of those ASAP.Folks:
I am sorry to report that recent frequency response testing showed some strange results which I believe are being induced by the phase shifting and correlated cancellation modes between the 1st order HF and 2nd order LF x-overs. I am currently transitioning the HF x-over to a 2nd order filter to resolve the problems. New parts are on the way. I should have an update by the end of the week.
Regards,
Andrew
I am sorry to report that recent frequency response testing showed some strange results which I believe are being induced by the phase shifting and correlated cancellation modes between the 1st order HF and 2nd order LF x-overs. I am currently transitioning the HF x-over to a 2nd order filter to resolve the problems. New parts are on the way. I should have an update by the end of the week.
Regards,
Andrew
Update
Sorry for the delay folks. I now have implemented the 2nd order x-over. Initial impressions are that they sound much better than the 1st order x-over. Much more balanced in both the HF and the MID bands. I hope to have some frequency response testing completed this weekend and will post a follow up update as soon as possible. Once I've confirmed that this new solution is working as well as it sounds.
Andrew
Sorry for the delay folks. I now have implemented the 2nd order x-over. Initial impressions are that they sound much better than the 1st order x-over. Much more balanced in both the HF and the MID bands. I hope to have some frequency response testing completed this weekend and will post a follow up update as soon as possible. Once I've confirmed that this new solution is working as well as it sounds.
Andrew
Latest update - I spent significant time testing/tweaking this past weekend. I have successfully resolved the phase issue I was experiencing, tweaked both the HF and LF x-overs to remove a slight dip in output around 6K and also to provide additional baffle step compensation in the mid-band. I am currently performing listening tests and the speaker sounds much more balanced and natural than before although the excellent accuracy, detail and imaging are all still there . . . I hope to have at long last the design finalized and published very soon.
Andrew
Andrew
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.