Morbo:
I looked at some of those Audax tweeters, but none of them "jumped out" at me and made me think they would really shine. I was thinking about tweeter criteria on the airplane today and I realized that if I began to roll-off the tweeter at around 12K with a 1st order x-over, that I would be down 18db by 1.5K (assuming that I also include impedance correction in the circuit. One of my priorities is to have excellent off-axis response out of the tweeter as well as excellent overall sound quality to ensure that the tweeter will mate nicely with the new midrange/woofer. I'd like to know which Audax tweeters you were thinking about using. For about $30, some of those 19/20mm Seas tweeters look pretty good. I'd suggest that we'd want the tweeter to have an Fs of no higher than 1500hz, and of course, anything lower would be better. Some of the top mount Morel tweeters also look interesting.
I fully agree with you regarding the cone excursion issue below the tuning frequency with this mid-woofers. Even a first order high pass filter set between 80-100hz, would make a tremendous difference for power handling. I know that everyone complains about the size of the caps required, but I really don't think this is a big deal, NP electrolytics with a poly bypass should be just fine. I've had great success with Black Gate caps in my amplifiers and I'd probably use those. They sound terrific even without the poly bypass cap. I'm considering making this high pass cap "switchable" so that I can decide on the fly whether or not to activate it. The other options are to just use a built in active x-over either in the pre-amp or in a subwoofer amp.
I'm still looking for thoughts on the tweeter selection. I'm going to post on the PE and Madisound boards to see if I can get some recommendations.
Andy
I looked at some of those Audax tweeters, but none of them "jumped out" at me and made me think they would really shine. I was thinking about tweeter criteria on the airplane today and I realized that if I began to roll-off the tweeter at around 12K with a 1st order x-over, that I would be down 18db by 1.5K (assuming that I also include impedance correction in the circuit. One of my priorities is to have excellent off-axis response out of the tweeter as well as excellent overall sound quality to ensure that the tweeter will mate nicely with the new midrange/woofer. I'd like to know which Audax tweeters you were thinking about using. For about $30, some of those 19/20mm Seas tweeters look pretty good. I'd suggest that we'd want the tweeter to have an Fs of no higher than 1500hz, and of course, anything lower would be better. Some of the top mount Morel tweeters also look interesting.
I fully agree with you regarding the cone excursion issue below the tuning frequency with this mid-woofers. Even a first order high pass filter set between 80-100hz, would make a tremendous difference for power handling. I know that everyone complains about the size of the caps required, but I really don't think this is a big deal, NP electrolytics with a poly bypass should be just fine. I've had great success with Black Gate caps in my amplifiers and I'd probably use those. They sound terrific even without the poly bypass cap. I'm considering making this high pass cap "switchable" so that I can decide on the fly whether or not to activate it. The other options are to just use a built in active x-over either in the pre-amp or in a subwoofer amp.
I'm still looking for thoughts on the tweeter selection. I'm going to post on the PE and Madisound boards to see if I can get some recommendations.
Andy
Just an FYI, Dan has done a design for the WR125s here:
http://www.adireaudio.com/Files/WR125S-TM025F1Design.pdf
The driver is full range and the tweeter just fills in the top end.
http://www.adireaudio.com/Files/WR125S-TM025F1Design.pdf
The driver is full range and the tweeter just fills in the top end.
Grahamt:
Thanks for the link, I hadn't seen that configuration. One thing which concerns me is the woofer's off axis response above 2K. On axis and 15 degrees off axis looks okay up to about 10K, however, the 30 and 45 degree off axis curve don't look nearly as good. I'm guessing that this woofer might do really well mated with a tweeter using only 1st order x-overs with a full 3 octaves in the cross over range. I'm guessing that this would keep the parts count low and might also provide a very smooth transition from tweeter to woofer. Since I don't have the woofers in yet, I'll have to test them once I get them. I suspect that Dan was trying to showcase the high frequency capabilities of the driver and that is the reason why he configured the tweeter the a 4th order cross over and no attenuation on the upper octaves of the woofer . . .
Thanks for the link, I hadn't seen that configuration. One thing which concerns me is the woofer's off axis response above 2K. On axis and 15 degrees off axis looks okay up to about 10K, however, the 30 and 45 degree off axis curve don't look nearly as good. I'm guessing that this woofer might do really well mated with a tweeter using only 1st order x-overs with a full 3 octaves in the cross over range. I'm guessing that this would keep the parts count low and might also provide a very smooth transition from tweeter to woofer. Since I don't have the woofers in yet, I'll have to test them once I get them. I suspect that Dan was trying to showcase the high frequency capabilities of the driver and that is the reason why he configured the tweeter the a 4th order cross over and no attenuation on the upper octaves of the woofer . . .
Dan is an ardent believer in high order crossovers. He is especially proud of designs that use 2nd order electrical but yield 6th order acoustical. It would be interesting to see the MLS measurements of this because I'm guessing the WR125S has run out of steam when the tweeter takes over. The XO point bandwidth is probably very small if Dan has stuck to his philosophy.
Although I haven't received my woofers yet, I'm still struggling with the theory behind Dan's design . . . From the curves posted on the CSS web site, we know that the woofer doesn't do so well 30-45 degrees off-axis above 2K. That would make sense since this is typically tweeter territory and the woofer is 4.5" after all. It is quite amazing that the woofer has as much on-axis extension as it does. Despite this fact, if we were going to use this driver in a 2-way design, I just don't understand why we wouldn't opt to run the tweeter lower. I'm not suggesting that the tweeter should be run all the way into the vocal range, but maybe (depending on x-over slope) we'd choose to set our x-over point somewhere between 6-12K? It appears that Dan's x-over point is set around 12K (judging by the FRQ curves only) but with an extremely steep x-over since it appears that he is down 24db by 6K. This steep slope really mitigates the effects of the tweeter in the 2-6K region. Perhaps the woofer's off-axis performance in that 2-6K region is actually better than it appears at first glance?
Andy
Andy
Just noticed something else. I don't think that Audax TM025F1 tweeter is all that good. Check out this link:
http://home.new.rr.com/zaph/audio/audio-speaker14.html
http://home.new.rr.com/zaph/audio/audio-speaker14.html
Hi all,
I designed that crossover to do a few things:
1. Allow easy show-off of the 4" full range (which, by the way, works REALLY well for home theater applications, when high passed at 100 Hz and run full range). Simply disconnect the input to the tweeter crossover network and its the woofer full range.
2. Sonics. We experimented with first order, 2nd order, 4th order, and 6th order networks at a variety of crossover points. After listening, the low orders were simply tossed out - too much sonic "mush" from the comb filtering. Wide bandwidth drivers with lots of overlap really call for higher order crossovers.
So it was the 4th order 6th order solutions, or the solution we ended up with. On axis, and even near on-axis (say, 20-30 degrees off) using just the tweeter as a fill at the top end really worked well. So that was ultimately chosen (the second favorite was a 6th order solution at 7 kHz).
3. The Audax tweeter really isn't that bad. It's one of my three favorites under $20. The slight rise in the top end isn't bad at all, easy to work with. And it does have quite a bit of extension on the low end. Plenty of output, fairly low THD/IMD when used with a higher order crossover (and there's lots of good reasons to use a high order crossover for just about any XO situation). There really aren't any other small faceplate, low cost, video shielded good performing tweeters out there.
Anyway, we've got that design solution running in one of our demo rooms, if anyone wants to stop by and take a listen...
Dan Wiggins
Adire Audio
I designed that crossover to do a few things:
1. Allow easy show-off of the 4" full range (which, by the way, works REALLY well for home theater applications, when high passed at 100 Hz and run full range). Simply disconnect the input to the tweeter crossover network and its the woofer full range.
2. Sonics. We experimented with first order, 2nd order, 4th order, and 6th order networks at a variety of crossover points. After listening, the low orders were simply tossed out - too much sonic "mush" from the comb filtering. Wide bandwidth drivers with lots of overlap really call for higher order crossovers.
So it was the 4th order 6th order solutions, or the solution we ended up with. On axis, and even near on-axis (say, 20-30 degrees off) using just the tweeter as a fill at the top end really worked well. So that was ultimately chosen (the second favorite was a 6th order solution at 7 kHz).
3. The Audax tweeter really isn't that bad. It's one of my three favorites under $20. The slight rise in the top end isn't bad at all, easy to work with. And it does have quite a bit of extension on the low end. Plenty of output, fairly low THD/IMD when used with a higher order crossover (and there's lots of good reasons to use a high order crossover for just about any XO situation). There really aren't any other small faceplate, low cost, video shielded good performing tweeters out there.
Anyway, we've got that design solution running in one of our demo rooms, if anyone wants to stop by and take a listen...
Dan Wiggins
Adire Audio
Dan,
Thank you for sharing your experiences in developing a "good fit" for the high end with this woofer. Understanding your design goals really helps here. It is also good to know that just using a tweeter for the top frequencies really helps with the off-axis HF dispersion. I do have one question though. When you were testing filters at lower frequencies, such as the one you mentioned at 7K, were you still running the woofer full range or were you filtering its high end as well? I'd like to understand if I should anticipate problems using a low pass filter on the woofer . . . Thanks again.
Andy
Thank you for sharing your experiences in developing a "good fit" for the high end with this woofer. Understanding your design goals really helps here. It is also good to know that just using a tweeter for the top frequencies really helps with the off-axis HF dispersion. I do have one question though. When you were testing filters at lower frequencies, such as the one you mentioned at 7K, were you still running the woofer full range or were you filtering its high end as well? I'd like to understand if I should anticipate problems using a low pass filter on the woofer . . . Thanks again.
Andy
Dan - thanks for chiming in with your thoughts on this. Very informative. One question - how baffle dependent is this crossover? I have my WR125 in a box already, the drivers are mounted at the top. How best to implement the tweeter then? If I just mount it in an equal width (6.5") 'baffle extension' piece on top of my existing box, as close to the woofer as possible? Or should I start fresh since my baffle size may be different than yours?
DanWiggins said:
Interesting. A lot of people have the opposite opinion, including me. What about power response, and passing a square wave that looks like a sabertooth saw? Actually, I usually end up going LR4 anyway because it's easier to do with most drivers. But if I had a driver with wide bandwidth, I'd probably go with shallow slopes.
DanWiggins said:
Someone posted a link to a project of mine using the Audax tweeter because I seem to suggest that it's not very good. But the fact is I like it a lot or I wouldn't have posted the project. The only thing about it is that it takes so many components to get it under control, and lately I've come to the conclusion that cheaper tweeters cost more than expensive tweeters because they need more crossover components thrown at them. A $26 Seas can reach a perfect acoustic 4th order with only 2 components, and do it with lower distortion too.
For high quality, wide bandwidth drivers, I highly recommend the Vifa XT19. And if money is tight, the DX19 is nice also. The DX19 has surprising low frequency ability, and 2nd order acoustic slopes at 2500 are no problem. The XT19 can't go quite as low but otherwise it's performance is even better, probably better than the Hiq's.
Bob Reimer was nice enough to send me one of these for evaluation, so I'll be comparing it to other similar size drivers I have in the house, Seas L12 among them. When I was at DIY Chicago, I won a XBL^2 subwoofer as a door prize, so it looks like I'll be giving this technology a good run for it's money. High Xmax never tripped my trigger, but low distortion and smooth response do so I'm curious to see if these live up to it.
John
sbolin said:
The XT19 is a better tweeter to use in a 3-way or for supplementing the high end of a wide range driver. Above 4Khz, I don't think there's a better tweeter. It's built with a copper faraday ring and the rear chamber is well designed, so distortion is very low in the range that it is intended to be used. Right at about 2Khz, 2nd order harmonic distortion starts to rise sharply, so those using this tweeter should really cross over much higher. 2nd order at 3000 or 4th order at 2500 are my recommendations. I'm currently using the XT19 with a GR M130, a driver that has extended response out to 10kHz. I affectionately call this system the MBOW-None. I would guess that the CSS WR125 would be good in a similar system.
I wasn't going to test the WR125 with the other 3" drivers because it's really in a different class, both price and size-wise. If things go well, I'll be testing the 3" this weekend. The guy who was going to send me the Fostex didn't come through, so now the list is:
TB W3-871
TB W2-880S
Aurasound NSW1-205-8A
Aurasound NSW2-326-8A
Aurasound NS3-193-8A
Aurasound NS3-194-8E (New model)
Hi-Vi B3S
Hi-Vi A2S
MCM 54-606
MCM 54-609
Zaph wrote:
Passing a square wave really only happens easily with a first order filter, and then only at a prescribed small sweet spot. You can do higher order linear phase networks, but they do get quite tricky above 2nd order. The goal of linear phase is admirable, but the scientific community is still undecided about the audibility of linear phase (note that this is whether you can even hear it, let alone if it's required for good perceived sound).
HOWEVER, with a first (or even a second) order filter, you automatically accept:
- higher THD
- higher IMD
- higher power compression
- worse off axis response
- smaller sweet spot
- higher component count
relative to a higher order crossover. Many of these artifacts ARE known to be audible and detrimental to the perceived results of the speaker.
So, in my mind, it's either shoot for a potentially good thing at the expense of many known pitfalls, or eliminate one potentially good thing and avoid many known pitfalls. Since I see engineering as the art of making the perfect set of compromises, I usually go for higher order. Rarely do I do a speaker with lower than 4th order (usually 6th or 8th).
Just my design philosophy! I've done first and second order networks when the client requests it, but if they don't state a specific requirement for crossover order, I go high...:
I know Bob has some other kit designs for the WR125S coming out, with different tweeters, but I put one out for general info that used tweeters I had.
Dan Wiggins
Adire Audio
Well, using a critically damped filter (Q=0.5) you get flat power and frequency response at the crossover point.
Passing a square wave really only happens easily with a first order filter, and then only at a prescribed small sweet spot. You can do higher order linear phase networks, but they do get quite tricky above 2nd order. The goal of linear phase is admirable, but the scientific community is still undecided about the audibility of linear phase (note that this is whether you can even hear it, let alone if it's required for good perceived sound).
HOWEVER, with a first (or even a second) order filter, you automatically accept:
- higher THD
- higher IMD
- higher power compression
- worse off axis response
- smaller sweet spot
- higher component count
relative to a higher order crossover. Many of these artifacts ARE known to be audible and detrimental to the perceived results of the speaker.
So, in my mind, it's either shoot for a potentially good thing at the expense of many known pitfalls, or eliminate one potentially good thing and avoid many known pitfalls. Since I see engineering as the art of making the perfect set of compromises, I usually go for higher order. Rarely do I do a speaker with lower than 4th order (usually 6th or 8th).
Just my design philosophy! I've done first and second order networks when the client requests it, but if they don't state a specific requirement for crossover order, I go high...:
Good information! I'll have to consider those... Another reason I used the Audax was I have a supply of them here...
I know Bob has some other kit designs for the WR125S coming out, with different tweeters, but I put one out for general info that used tweeters I had.I look forward to your feedback! We're picking up new OEMs weekly after playing with XBL^2 drivers. One new convert (a very skeptical engineer) wrote us that "The quality of bass and even lower midrange was different, less muffled and less congested. Quite noticeable difference" for an 8" subwoofer we designed for him.
Dan Wiggins
Adire Audio
Interesting posts about the various tweeters Zaph.. I got the impression from your comments after testing that the DX19 is a pretty mediocre tweeter... it is fairly cheap here, so if it is that good, I may try that out as well. The XT19 is a bit much for me $$$ wise I think. Would crossing above 5 or 6k ameliorate the stored energy @ 4k issue much?
The other tweeter I've had my eye on (mainly because of cost) is the little Audax 10mm gold dome:
It would also allow for very tight driver-driver spacing, although it would need a relatively high XO, but thats something the WR125 can help with I think
A question - if I'm listening very close, ie 1-2 feet away from the speakers, how much do I need to worry about what they do @ 30-45 deg. off axis?
The other tweeter I've had my eye on (mainly because of cost) is the little Audax 10mm gold dome:
An externally hosted image should be here but it was not working when we last tested it.
It would also allow for very tight driver-driver spacing, although it would need a relatively high XO, but thats something the WR125 can help with I think
A question - if I'm listening very close, ie 1-2 feet away from the speakers, how much do I need to worry about what they do @ 30-45 deg. off axis?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.