Grant, I have to clarify on when a crossover is 6db, or something else, and I have written about it before, but you might have missed it
It must be very clear that the slopes that counts is MEASURED ACOUSTICAL SLOPES, and not how many components that are used
This means that a single inductor(6db electrical) can give a 12db acoustical rolloff......if placed closer to drivers natural rolloff
Likewise can a electrical 12db give a acoustical 24db
The reason for this is drivers own rolloff, which always must be taken into calculations when working with low order slopes...
This offcourse is true for both lowpass and highpass
As an excample we can look at you mids low crosspoint
THAT single series C is most likely to give a 12db rolloff
AND that means that your box calculations can play a very important role
AND that is one reason why it might work better with a higher crosspoint than 2.7khz ... to have a 12db slope with a single inductor
FR rolloff is an unpredictable "player" as well as is the impedance
About zobels...
Ofcourse wrong adjusted zobels are worse than no zobel
It must be very clear that the slopes that counts is MEASURED ACOUSTICAL SLOPES, and not how many components that are used
This means that a single inductor(6db electrical) can give a 12db acoustical rolloff......if placed closer to drivers natural rolloff
Likewise can a electrical 12db give a acoustical 24db
The reason for this is drivers own rolloff, which always must be taken into calculations when working with low order slopes...
This offcourse is true for both lowpass and highpass
As an excample we can look at you mids low crosspoint
THAT single series C is most likely to give a 12db rolloff
AND that means that your box calculations can play a very important role
AND that is one reason why it might work better with a higher crosspoint than 2.7khz ... to have a 12db slope with a single inductor
FR rolloff is an unpredictable "player" as well as is the impedance
About zobels...
Ofcourse wrong adjusted zobels are worse than no zobel
Hi Omni, re: midrange enclosure (- a reply to your previous one)
I read Vance again on your suggestion, and it seems that my existing 1.5Litre sealed and fibre-
filled mid-boxes may just be ok for the CA15's at HP 270Hz. As you say the driver will be
(partially?) operating in its piston range, so (from Vance) a 'sealed enclosure is required when
the Fb>=XO/2 to minimize phase disturbance', that is, from 135 to say possibly 150Hz.
I assume this applies to 1st order filters. But what is the Fb for the CA in 1.5L?
For closed box systems, the Fb is evidently dependent on the driver T/S parameters.
(I'll use Zaphs figures for the CA, ie Qts=.411, Fs~56Hz, Vas=8.78 L).
Also because Qts>0.4 it seems ok to use a closed box.
So for the 'Chebychev Qtc=1.1 alignment', Fb=(1.1 * 56) / .411 = 150Hz ...bingo!
To check this, alpha=5.85, so reading from the Qtc=1.1 table for Qts=.41,
Fb=56 * multiplier(2.68) = 150Hz. Does this appear correct to you?
If not, apparently an infinite baffle for tweet/mid sitting on top of woofer box has many
advantages but I have no idea what dimensions are required for response down to 270Hz?
The existing baffle could be easily removed and replaced with one for woofer and port only.
I think Troels G's site has some examples. thanks omni....grant
thanks for latest comments - reading now, will reply soon.
I read Vance again on your suggestion, and it seems that my existing 1.5Litre sealed and fibre-
filled mid-boxes may just be ok for the CA15's at HP 270Hz. As you say the driver will be
(partially?) operating in its piston range, so (from Vance) a 'sealed enclosure is required when
the Fb>=XO/2 to minimize phase disturbance', that is, from 135 to say possibly 150Hz.
I assume this applies to 1st order filters. But what is the Fb for the CA in 1.5L?
For closed box systems, the Fb is evidently dependent on the driver T/S parameters.
(I'll use Zaphs figures for the CA, ie Qts=.411, Fs~56Hz, Vas=8.78 L).
Also because Qts>0.4 it seems ok to use a closed box.
So for the 'Chebychev Qtc=1.1 alignment', Fb=(1.1 * 56) / .411 = 150Hz ...bingo!
To check this, alpha=5.85, so reading from the Qtc=1.1 table for Qts=.41,
Fb=56 * multiplier(2.68) = 150Hz. Does this appear correct to you?
If not, apparently an infinite baffle for tweet/mid sitting on top of woofer box has many
advantages but I have no idea what dimensions are required for response down to 270Hz?
The existing baffle could be easily removed and replaced with one for woofer and port only.
I think Troels G's site has some examples. thanks omni....grant
thanks for latest comments - reading now, will reply soon.
Thanks very much Tinitus!
Yes I see! total rolloff depends on measured acoustical slopes, as well as drivers own rolloff,
(quote) "....if placed closer to drivers natural rolloff". I'm not sure that I understand how close is close!!! I've been trying to 'fathom' summation response for a while by studying Vance's examples.
The CA15's 60deg off-axis rolloff is about 10dB/Oct from 2Khz and about 15dB/Oct from 4Khz, wheres the on-axis goes all the way to 9Khz! I chose 2700 only because I wanted to retain as much off-axis as possible, but I'm certainly very open to advise from people wiser than me!
I'll have to think about and work through your comments on changing C and L till I fully understand the outcomes and get back to you if thats OK?
"Wrongly adjusted Zobels are worse than no Zobels" - understood! Many,many thanks Tinitus
Yes I see! total rolloff depends on measured acoustical slopes, as well as drivers own rolloff,
(quote) "....if placed closer to drivers natural rolloff". I'm not sure that I understand how close is close!!! I've been trying to 'fathom' summation response for a while by studying Vance's examples.
The CA15's 60deg off-axis rolloff is about 10dB/Oct from 2Khz and about 15dB/Oct from 4Khz, wheres the on-axis goes all the way to 9Khz! I chose 2700 only because I wanted to retain as much off-axis as possible, but I'm certainly very open to advise from people wiser than me!
I'll have to think about and work through your comments on changing C and L till I fully understand the outcomes and get back to you if thats OK?
"Wrongly adjusted Zobels are worse than no Zobels" - understood! Many,many thanks Tinitus
Thanks again Omni!
Yes, I have a fear that my amp might be blown up again, and also considering Tinitus comment:
"a wrongly configured Zobel is worse than no Zobel !" just about seals the argument for me!
I will go into Troels site again to look at infinite baffles and also to find what Sreten was referring to.
If I ever do get the drivers, I may experiment with Zobels at low power.
Our posts crossed, I did read Vance again re the midrange cavity size. Maybe 1.5L is not big enough as you suggest and my calculations are flawed, my other post refers (which I tried to squeeze down to the bare minimum - difficult for me not to get overly verbose, lol)
Yes I am enjoying this, especially all the wonderful help you kind folk generously provide.
TINITUS, I forget to mention that I do like to experiment a little. My crossovers are in plastic boxes on the back of cabs, and still only jumper-wired with alligator clips (strong ones so no chance of
losing a connection) I never bothered to solder them up, and if I want to modify something, its so easy.
Yes, I have a fear that my amp might be blown up again, and also considering Tinitus comment:
"a wrongly configured Zobel is worse than no Zobel !" just about seals the argument for me!
I will go into Troels site again to look at infinite baffles and also to find what Sreten was referring to.
If I ever do get the drivers, I may experiment with Zobels at low power.
Our posts crossed, I did read Vance again re the midrange cavity size. Maybe 1.5L is not big enough as you suggest and my calculations are flawed, my other post refers (which I tried to squeeze down to the bare minimum - difficult for me not to get overly verbose, lol)
Yes I am enjoying this, especially all the wonderful help you kind folk generously provide.
TINITUS, I forget to mention that I do like to experiment a little. My crossovers are in plastic boxes on the back of cabs, and still only jumper-wired with alligator clips (strong ones so no chance of
losing a connection) I never bothered to solder them up, and if I want to modify something, its so easy.
hello again, Tinitus,
You and Omni have already helped me a lot, and I'm worried that I might be too bothersome or annoying because I have asked lots of questions (but my appreciation is immense, as you know,
and I think I'm slowly learning - new concepts can be a bit difficult, I'm not as 'sharp' as I used to be! lol)
So with a little trepidation, may I ask these dumb questions:
1) I now know a reversed-tweeter 2nd Butterworth will work with 1st (also Butterworth!) on mid and
woofer, but, WILL a LR2 (2nd order Linkwitz-Riley) also work for tweeter in this case? Untill now I had just assumed that an LR2 network had to be applied to each driver, don't know why, maybe
I'm wrong. If the answer is yes, then it would greatly simplify component values, and eliminate 'fudging' the capacitor value in my first tweeter circuit.
2) If this is indeed possible, and considering your comments re the Mid C and L values, which I'm still trying to understand, I'm looking at ways to implement your advice. I have been playing in Lalena and it seems very possible that I can reduce (as you suggest) the Mid inductor (to 0.39mH) by increasing the XO to 2950Hz and still use standard-sized inductors elsewhere. Which was my initial design goal.
3) the revised XO for 295/2950 (LR2 tweeter) attached may be more suitable? And I would only lose several dB of off-axis. Presumably, at this new XO-point, 'summed response' for the Mid will decay at around 16dB/Oct (10dB for the CA15 and 6dB for the filter).
4) If the mid chamber is appropriate; baffle-step and the previously mentioned woofer response is ok or able to be tamed, then I'm well on my way!
Thanks very much Tinitus and Omni for all of your time, grant
(I do realise that 3way's are very difficult to design, especially for a beginner, but it seems feasible,
and I would learn a lot more than building a published design!)
...new XO attached...
You and Omni have already helped me a lot, and I'm worried that I might be too bothersome or annoying because I have asked lots of questions (but my appreciation is immense, as you know,
and I think I'm slowly learning - new concepts can be a bit difficult, I'm not as 'sharp' as I used to be! lol)
So with a little trepidation, may I ask these dumb questions:
1) I now know a reversed-tweeter 2nd Butterworth will work with 1st (also Butterworth!) on mid and
woofer, but, WILL a LR2 (2nd order Linkwitz-Riley) also work for tweeter in this case? Untill now I had just assumed that an LR2 network had to be applied to each driver, don't know why, maybe
I'm wrong. If the answer is yes, then it would greatly simplify component values, and eliminate 'fudging' the capacitor value in my first tweeter circuit.
2) If this is indeed possible, and considering your comments re the Mid C and L values, which I'm still trying to understand, I'm looking at ways to implement your advice. I have been playing in Lalena and it seems very possible that I can reduce (as you suggest) the Mid inductor (to 0.39mH) by increasing the XO to 2950Hz and still use standard-sized inductors elsewhere. Which was my initial design goal.
3) the revised XO for 295/2950 (LR2 tweeter) attached may be more suitable? And I would only lose several dB of off-axis. Presumably, at this new XO-point, 'summed response' for the Mid will decay at around 16dB/Oct (10dB for the CA15 and 6dB for the filter).
4) If the mid chamber is appropriate; baffle-step and the previously mentioned woofer response is ok or able to be tamed, then I'm well on my way!
Thanks very much Tinitus and Omni for all of your time, grant
(I do realise that 3way's are very difficult to design, especially for a beginner, but it seems feasible,
and I would learn a lot more than building a published design!)
...new XO attached...
Grant, The issue of the cabinet size for the CA15RLY was one which I agonized for a long time during the initial design phase, as well as the overall cabinet size for my woofer. During that point I was discussing with Tinitus the virtues of what I thought would be a good volume for my woofer and he and a few others offered that I was thinking too small. I ultimately changed my box size from 2 cubic feet to 4 cubic feet for my woofer, which is sealed. I come from the school of sealed enclosures. The next step brought me to the midbass. Based on the fact that I was going with a sealed enclosure, which was simpler for me, and from what I have read and researched, I designed the midbass enclosure volume to critically damp its performance, ie, a Qtc in the range of .5. Also, since my crossover point is in the 300 Hz range, I wanted to optimize the box to tune to at least 2 octaves below the crossover point of 300 Hz. This is essentially 75 Hz. Using the downloadable programs: WinISD alpha, WinISD beta, I modelled many scenarios which rendered a workable enclosure size of around 15-18 Litres for the CA. The final enclosure size ended up being 15 Litres net after I deducted volume for the braces and magnet structures. Box tuning is in the area of about 65-75 Hz, which was my target. This renders a Qtc of about .517-.53, so the end result will be a midbass which will perform in the frequency range where pistonic motion will be there, however, it will be "tamed", thereby maintaining good lower midbass performance, along with the desired smoothness as it goes up in frequency......................I hope!..........You mention the difficulty not to get overly verbose, and as you can see from my posts, I suffer from that affliction as well...............Oh well, rather be more verbose than clear as mud....................Regards, .................Omni
tinitus, its me again! lol
re: (quote) "As an excample we can look at you mids low crosspoint
THAT single series C is most likely to give a 12db rolloff
AND that means that your box calculations can play a very important role AND that is one reason why it might work better with a higher crosspoint than 2.7khz ... to have a 12db slope with a single inductor "
hmm, I'm still struggling with these concepts, could you please elaborate (for a beginner?) I really need to understand this.
(risking 'wearing out my welcome' here!) ...grant
re: (quote) "As an excample we can look at you mids low crosspoint
THAT single series C is most likely to give a 12db rolloff
AND that means that your box calculations can play a very important role AND that is one reason why it might work better with a higher crosspoint than 2.7khz ... to have a 12db slope with a single inductor "
hmm, I'm still struggling with these concepts, could you please elaborate (for a beginner?) I really need to understand this.
(risking 'wearing out my welcome' here!) ...grant
Omni,
It seems that I am completely 'way-off-target' WRT the CA mid chamber! (EVEN IF you are targetting
a Qtc ~ 0.5!) My calc. for 1.5 Litres @ Qtc=1.1 alignment(much? smaller sealed box) seems by
comparison horribly inaccurate and wrong! Where did I go wrong? - no idea, I thought I reasonably
followed Vance. Could my understanding be that poor?
My ported cabinet is supposedly optimised @ 72 Litres, woofer Vas=178 Litres but your 300HzXO
mid chamber (tuned to 75Hz) @ 15 Litres is a huge proprtion of that. I thought that 1.5 Litres may
be too small, but 15 Litres is an order of magnitude greater, even considering Qtc~0.5. Whoa, lol.
I've got some work to do! Are the WINISD programs free?, if so, I will download.
I can understand that your sealed cabinets are largish - but 4 cubic feet seems huge! (Litres?) for the
Peerless 830669: Vas~172 Litre, is less than my P25 (Vas=178 Litre!) I must have been misinformed
in the original design, unless a sealed box makes such a HUGE difference?
Btw, 'old school' sealed seems good (I'm old school in most things!), supposedly better transient
response.
Unless there is a way around this, it seems that I may need a seperate Tweet/Mid enclosure on top
of the woofer boxes OR maybe an infinite baffle similarly for Tweet/Mid also on top, but difficult
(Troels has a section on that) grant....pondering.....
It seems that I am completely 'way-off-target' WRT the CA mid chamber! (EVEN IF you are targetting
a Qtc ~ 0.5!) My calc. for 1.5 Litres @ Qtc=1.1 alignment(much? smaller sealed box) seems by
comparison horribly inaccurate and wrong! Where did I go wrong? - no idea, I thought I reasonably
followed Vance. Could my understanding be that poor?
My ported cabinet is supposedly optimised @ 72 Litres, woofer Vas=178 Litres but your 300HzXO
mid chamber (tuned to 75Hz) @ 15 Litres is a huge proprtion of that. I thought that 1.5 Litres may
be too small, but 15 Litres is an order of magnitude greater, even considering Qtc~0.5. Whoa, lol.
I've got some work to do! Are the WINISD programs free?, if so, I will download.
I can understand that your sealed cabinets are largish - but 4 cubic feet seems huge! (Litres?) for the
Peerless 830669: Vas~172 Litre, is less than my P25 (Vas=178 Litre!) I must have been misinformed
in the original design, unless a sealed box makes such a HUGE difference?
Btw, 'old school' sealed seems good (I'm old school in most things!), supposedly better transient
response.
Unless there is a way around this, it seems that I may need a seperate Tweet/Mid enclosure on top
of the woofer boxes OR maybe an infinite baffle similarly for Tweet/Mid also on top, but difficult
(Troels has a section on that) grant....pondering.....
Grant, your math is not that wrong as you may feel. I ran a model in the WinISD alpha and beta for the CA with the driver database parameters and the parameters I inserted from Zaphs measurements. At 1.5 litres I got Qtc of 1.025 with a Fs of 152 Hz and in another I got Qtc of 1.061 with Fs of 146 Hz. so don't despair, your calculations seem to be right in the ballpark..............However, with that high of a Qtc for the midbass, the small 1.5 cabinet is NOT providing much damping to the driver. This is a problem for smooth midbass/midrange performance. I also checked out your woofer in the WinISD, but not thoroughly, however it seems that the EBP of 80 something suggests the driver is better suited to a ported enclosure. The 4 mm Xmax also suggests that as well. There are differences in optimized cabinet volumes when you compare sealed vs. vented, so my 4 cubic feet sealed enclosure with a 12 inch woofer compared to your vented enclosure with a 10 inch woofer is like comparing apples to oranges................But, as far as the CA15RLY is concerned, you can take a variety of approaches. SEAS, on their website recommends a box size of 6-9 liters vented box, I believe, but not certain, and these are only guesses from memory, so you may want to check their website. I am also of the thought that these alignments are based on a 2 way system, using the CA as the primary woofer. But Vance Dickason suggests that we are tuning the midbass enclosure to optimize it similar to a bass enclosure. It describes the alignments but I am not certain if it goes into port diameters and lengths........... Again, I chose to go sealed to eliminate hassles or problems in dealing with a port, not to mention my lack of working knowledge of vented systems, in addition to my research, but also, I believe the sealed midbass cabinet will provide for a more robust midrange performance..........Your idea of building a seperate cabinet for the midbass and tweeter may likely prove to be a good choice that you wouldn't regret. I was originally going to use my old cabinets which were only 2 cubic ft, but when I started dialoguing with some people on this forum and met Tinitus, They convinced me to rethink my design. I knew from the start that it would be a pain in the *** to do completely new cabinets, but now I believe it was the right choice to make, not to mention the joy that has come from this endeavor, albeit fraught with delays and a few cost overruns............As far as my bass cabinets being huge, here are a few considerations I had to face:....SPL sensitivity of the CA15RLY is around 86dB. My Peerless woofer is reported to be around 92dB. So, there exists a disparity in the drivers. As suggested by Tinitus, a larger enclosure will make my woofer less boomy........Have you ever heard a car drive by, blasting its' subwoofers, booming down the road, shaking the entire car and the windows in your house ? This is a result of the driver who obviously does NOT seek MUSICALITY, ....... but primarily from a high Qtc woofer system, generally in small boxes. My larger boxes will transform LOUD boominess into DEEP, TIGHT bass which will not drown out the CA15RLY, which has a lower SPL to start with. As Tinitus expressed to me, it is a way to "extract" more midrange sound. I am already risking the possibility af a laidback midrange, which I don't want, so making the bass cabinet was a start to improving the quality of the bass AND bringing out the midbass more, and I am very grateful he was almost somewhat insistant on this matter...........The WinISD alpha, and WinISD beta are free downloads, but I don't have the website address, as I am sure you can Google it, or someone here may be able to provide the link. Download both, as they have different driver databasses,ie, some drivers listed in one but not the other. You can also enter data from your drivers into the applications....................sorry so long......Respectfully.........Omni
wow, thanks Omni
good to know that 1.5 L is in the ballpark, whew! I will try the new CA (when I get them)in the existing cavity and see how it sounds. I did go back and recheck various alignments (including yours which of course is 'spot on')
Re: high Qtc being a problem for smooth midrange performance (with little driver damping), Oh, OK, I didn't realise that. The 1.5L 'Qtc 1.1 alignment' Vance describes as maximum efficieny and power
(which coincidentally I was hoping for), so I'm wondering then if it will probably exceed the (Zaph) Xmax of 5 mm? ....5mm excursion for a mid/woofer? - a bit optimistic?
Re: apples to oranges....good point, understood!
Yes it seems that sealed midrange cavity is better than ported, and that separate cabinet is better still - I may end up doing that.
Your large enclosure as insisted by Tinitus - good idea. I have the same problem P25 ~90db at 300Hz. CA 86db as you said. But my boxes are already built. So I'm wondering if I could Lpad
the woofer? Or alternatively use a woofer LP inductor with a highish DCR, which is why I asked if the 4.7mH with 1.4 Ohms DCR was suitable. I guess this is obvious, but I would like minimal
insertion loss for the midrange filter (86db is pretty low already?).
I'm sure I'll be able to find WinISD A&B, they seem very useful, thanks.
Write as much as you like omni, I'm eager to learn and enjoy your comments!
Re: my previous post about tweeter LR2 working with mid 1st order LP.....
Looks like it is ok in principle - they are both all-pass types. *BUT*, what I forgot was that 1st Butterworth are 3db down at xo, while LR2 are 6db down at xo. So there is no way it will sum flat.
And consequently my revised crossover is wrong- it won't sum flat.
Its not only untill maybe 2 octaves (back from the xo-point) into the passband that attenuation begins.
I worked out for the LR2, that at approximately half an octave back into the passband, its about 3dB down. So, to sum flat, I would have to move the tweeter xo (from 3000) about one half Octave DOWN in frequency , that is around 2300Hz! If I was using an LR2 for each driver, this 'fudge' wouldn't be required. For 2460Hz (which is probably close enough) I get 5.1uF for Cap and 0.82mH coil. This is great as now all my coils are standard sizes which was one of my
original design goals! (and you say YOU have the verbosity affliction! hehe)
cheers Omni, grant
good to know that 1.5 L is in the ballpark, whew! I will try the new CA (when I get them)in the existing cavity and see how it sounds. I did go back and recheck various alignments (including yours which of course is 'spot on')
Re: high Qtc being a problem for smooth midrange performance (with little driver damping), Oh, OK, I didn't realise that. The 1.5L 'Qtc 1.1 alignment' Vance describes as maximum efficieny and power
(which coincidentally I was hoping for), so I'm wondering then if it will probably exceed the (Zaph) Xmax of 5 mm? ....5mm excursion for a mid/woofer? - a bit optimistic?
Re: apples to oranges....good point, understood!
Yes it seems that sealed midrange cavity is better than ported, and that separate cabinet is better still - I may end up doing that.
Your large enclosure as insisted by Tinitus - good idea. I have the same problem P25 ~90db at 300Hz. CA 86db as you said. But my boxes are already built. So I'm wondering if I could Lpad
the woofer? Or alternatively use a woofer LP inductor with a highish DCR, which is why I asked if the 4.7mH with 1.4 Ohms DCR was suitable. I guess this is obvious, but I would like minimal
insertion loss for the midrange filter (86db is pretty low already?).
I'm sure I'll be able to find WinISD A&B, they seem very useful, thanks.
Write as much as you like omni, I'm eager to learn and enjoy your comments!
Re: my previous post about tweeter LR2 working with mid 1st order LP.....
Looks like it is ok in principle - they are both all-pass types. *BUT*, what I forgot was that 1st Butterworth are 3db down at xo, while LR2 are 6db down at xo. So there is no way it will sum flat.
And consequently my revised crossover is wrong- it won't sum flat.
Its not only untill maybe 2 octaves (back from the xo-point) into the passband that attenuation begins.
I worked out for the LR2, that at approximately half an octave back into the passband, its about 3dB down. So, to sum flat, I would have to move the tweeter xo (from 3000) about one half Octave DOWN in frequency , that is around 2300Hz! If I was using an LR2 for each driver, this 'fudge' wouldn't be required. For 2460Hz (which is probably close enough) I get 5.1uF for Cap and 0.82mH coil. This is great as now all my coils are standard sizes which was one of my
original design goals! (and you say YOU have the verbosity affliction! hehe)
cheers Omni, grant
grantnsw said:
Hi,
This is not a problem, gives you 4dB baffle step by default.
L-padding woofers is a complete no-no as it seriously afffects source
impedance and thus driver Qts anbd thus Qtc. The same is also true
of mid- drivers bass resonance with L-pads. High DR inductors also
affect Qts and thus Qtc.
See troels 3-ways here :
http://www.troelsgravesen.dk/Diy_Loudspeaker_Projects.htm
/sreten.Grant, The 1.1 Qtc alignment you mention is a description Dickason uses pertaining to woofers operating at low frequencies. The CA15RLY will be going into frequencies well above the frequencies that the woofer will be, so I am not so sure that a 1.1 alignment for the CA is your best bet. Maybe Tinitus can chime in on this ?........ At any rate, the 1.1 Qtc. figure that I mentioned that was in the ballpark was to confirm that your calculations seem to be in the ballpark relative to what I modeled in WinISD, given the size of your current midrange cabinet. I hope this didn't misguide you, I was merely confirming your math...........I would still shoot for a larger cabinet for the CA. This will afford more control on cone movement, help eliminate internal sound reflections, which may cause distorted sound..... A smaller cabinet is gonna allow the cone to excurse farther, and depending on what kind of music you listen to and how loud you play it, if you get too much cone motion in the midbass, your midrange performance going into the higher mid frequencies could end up becoming distorted. This is the delicate challenge of the midrange for me: to produce excellent lower midbass, as well as being able to extend smoothly into the upper midrange frequencies, without distorting. I am of the thought that the larger midbass cabinet will facilitate this. But part of the challenge here too, is to make the midbass cabinet not too large, as to make the midrange sound "thin" acoustically......... I agree with Sreten about the 4dB bafflestep by default as not being a problem. I believe it means that the 4dB loss in the woofer as a result of bafflestep, brings it more in line with the 86dB SPL of the CA15RLY, making the drivers a reasonable match, SPL wise....... You want to minimize insertion loss in midrange, which is a good idea; what guage inductors do you intend to use, and are you gonna repost your crossover schematic?...........Man, I'll tell you what..........Its amazing how we can start a speaker project with the idea that we are simply gonna drop in new drivers and crossover components, but end up going through all this research stuff, learning new things, and complicating the **** out of things, in the hope of achieving the sound we have in our minds............The wonderment of DIY.............Regards...........Omni
Hi sreten,
re: 4db baffle step by default....in a word- brilliant! thank you, this had not 'dawned' on me yet. I was stressing about b-step, now I have ~4db 'slack factor' I didn't realise I had! Your other comments re influence on parameters - understood. TYVM!
(My design hurdles are slowly being overcome by such wonderful help as this! and a little study by me, hehe)
Yes Troels has a great site but its huge. I'm still looking to find your reference on why not Zobeling woofers is 'strange'.
I thought (correct me if I'm wrong) that not Zobeling would allow increased amplifier protection due to higher impedance?
many thanks, grant
re: 4db baffle step by default....in a word- brilliant! thank you, this had not 'dawned' on me yet. I was stressing about b-step, now I have ~4db 'slack factor' I didn't realise I had! Your other comments re influence on parameters - understood. TYVM!
(My design hurdles are slowly being overcome by such wonderful help as this! and a little study by me, hehe)
Yes Troels has a great site but its huge. I'm still looking to find your reference on why not Zobeling woofers is 'strange'.
I thought (correct me if I'm wrong) that not Zobeling would allow increased amplifier protection due to higher impedance?
many thanks, grant
Hi,
Also a mid units bass alignment can have quite a high Q,
combined with its high pass filter, it can be near irrelevant.
Arguing about it when the boxes are already made is pointless.
Generally speaking making Q lower and thus Fb lower increases
excursion in the lower frequencies, it does not reduce it. The c/o
can account for the highish Q if it is a problem.
Also note : for the BSC to work careful alignment of the bass unit
response it needed, and diffraction effects will likely cause a bump
(and dip) in the mid response which also likely will need correction.
/sreten.
Also a mid units bass alignment can have quite a high Q,
combined with its high pass filter, it can be near irrelevant.
Arguing about it when the boxes are already made is pointless.
Generally speaking making Q lower and thus Fb lower increases
excursion in the lower frequencies, it does not reduce it. The c/o
can account for the highish Q if it is a problem.
Also note : for the BSC to work careful alignment of the bass unit
response it needed, and diffraction effects will likely cause a bump
(and dip) in the mid response which also likely will need correction.
/sreten.Hi omni, ( your continued support/comments - so much appreciated! - the cheque is in the mail!, lol)
re: Qtc of mid box. OIC! It NEEDS to be larger. Well my plan is now then to see how the existing sounds, and if no good to my ears, then move mid & tweet into a more appropriately sized box sitting on top of existing woofer box. Still keeping drivers
as closely spaced vertically as possible (woofer near top) and allowing twt/mid to be moved horizontally backwards to
reduce 'ZDP' issues.
Yes, I too would to minimize distortion while possibly? maximising efficiency, a trade-off I guess, but emphasizing "clean-ness'.
I did read Vance and Rod re baffle step a few days ago, but all I can remember now (sheesh, I believe Gingko is good for ones memory?) I think, was that it was about 4dB at around 800Hz, which would make it a mid driver issue in a 3way. Please
excuse my ignorance on this as my head is full with so many other things, eg XO's. But its such addictive fun!
Just when I think I'm getting a little understanding , then I realise I made an error, and Oh, I really am ONLY scratching the surface!
Case in point, my last post re moving XO point down for LR2 tweeter to match 1st order mid LP. Well it could work but it defeats the purpose (unless other issues) because I would have to 'pad' down 6db instead of 2dB (@3000!) Because LR2 is already 6dB down at XO. Took me ~30mins on graph paper to work that one out, lol. There may be multiple ways to do things, the best way is only judged by listening, I guess.
I am thinking now that maybe 'full LR2 - 300/3000' IS the way to go (the twt/mid XO would not be asymetrical, therefore no unintended 'bumps'). Just have to 'bite the bullet' and pay for the special size inductors after all.
Another mistake: Zaphs tweeter attenuation (which I copied) is inapropriate for my design - it offers 8.5dB reduction I think because he crosses the L15 at 750Hz to avoid cone breakup, thus significantly reducing the mid output at tweeter crossover. I reworked mine (by hand then verified Lalena) for 3dB, ie 2 and 15 Ohm for Ztot(tweeter and divider)=6.3.
<rant begins> hehe
I guess what I am trying to achieve is a modelled response (as much as possible with no test equipment) 'on paper' as it were; to identify all solvable obstacles and hopefully 'flatten' them with a view to creating a network/enclosure with the best possible response for my chosen drivers. "Best possible" is unlikely to be anywhere near flat! (I'm realistic - it may have huge +- dB variation). I probably won't notice a small to moderate variation, but I hope I will notice if a driver is 'polarity nulled out'! A 'paper' design at first (before I buy the mids) to ensure I have a good grasp of theory and don't go in 'blind' and at least have some idea of how to fix a major mistake. I'm fairly confident now, that it WILL sound better than my first old attempt with the D75 domes. Besides, its an iterative process, all about the tweaking as Tinitus might say and there is the satisfaction of 'I built it myself' Also having the opportunity to upgrade to 27TDFC if necessary. <rant ends>
Midrange inductor gauge. Not sure, but gauge is 'ga'? in US, here I think its SWG or AWG? Anything thick enough to offer
lowest DCR @ best price (for critical midrange), the 0.39mH (not Solen too expensive for me, except maybe mid DCR~ 0.15 Ohms $25AU? but am I really going to hear 0.15 Ohm difference!?) on 'Rabbitz' (local supplier) site (www.rzaudio.com) I think is around $7AU (0.3 Ohm from memory).
Sure, I'll put a new scematic up, but I'll need to be fairly sure, its an improvement and less laughable than the last one!
"The wonderment of DIY".....chuckle..... (hope to hear from you again Omni, not sure though after this post, LOL)
grant (the verbally afflicted one)
re: Qtc of mid box. OIC! It NEEDS to be larger. Well my plan is now then to see how the existing sounds, and if no good to my ears, then move mid & tweet into a more appropriately sized box sitting on top of existing woofer box. Still keeping drivers
as closely spaced vertically as possible (woofer near top) and allowing twt/mid to be moved horizontally backwards to
reduce 'ZDP' issues.
Yes, I too would to minimize distortion while possibly? maximising efficiency, a trade-off I guess, but emphasizing "clean-ness'.
I did read Vance and Rod re baffle step a few days ago, but all I can remember now (sheesh, I believe Gingko is good for ones memory?) I think, was that it was about 4dB at around 800Hz, which would make it a mid driver issue in a 3way. Please
excuse my ignorance on this as my head is full with so many other things, eg XO's. But its such addictive fun!
Just when I think I'm getting a little understanding , then I realise I made an error, and Oh, I really am ONLY scratching the surface!
Case in point, my last post re moving XO point down for LR2 tweeter to match 1st order mid LP. Well it could work but it defeats the purpose (unless other issues) because I would have to 'pad' down 6db instead of 2dB (@3000!) Because LR2 is already 6dB down at XO. Took me ~30mins on graph paper to work that one out, lol. There may be multiple ways to do things, the best way is only judged by listening, I guess.
I am thinking now that maybe 'full LR2 - 300/3000' IS the way to go (the twt/mid XO would not be asymetrical, therefore no unintended 'bumps'). Just have to 'bite the bullet' and pay for the special size inductors after all.
Another mistake: Zaphs tweeter attenuation (which I copied) is inapropriate for my design - it offers 8.5dB reduction I think because he crosses the L15 at 750Hz to avoid cone breakup, thus significantly reducing the mid output at tweeter crossover. I reworked mine (by hand then verified Lalena) for 3dB, ie 2 and 15 Ohm for Ztot(tweeter and divider)=6.3.
<rant begins> hehe
I guess what I am trying to achieve is a modelled response (as much as possible with no test equipment) 'on paper' as it were; to identify all solvable obstacles and hopefully 'flatten' them with a view to creating a network/enclosure with the best possible response for my chosen drivers. "Best possible" is unlikely to be anywhere near flat! (I'm realistic - it may have huge +- dB variation). I probably won't notice a small to moderate variation, but I hope I will notice if a driver is 'polarity nulled out'! A 'paper' design at first (before I buy the mids) to ensure I have a good grasp of theory and don't go in 'blind' and at least have some idea of how to fix a major mistake. I'm fairly confident now, that it WILL sound better than my first old attempt with the D75 domes. Besides, its an iterative process, all about the tweaking as Tinitus might say and there is the satisfaction of 'I built it myself' Also having the opportunity to upgrade to 27TDFC if necessary. <rant ends>
Midrange inductor gauge. Not sure, but gauge is 'ga'? in US, here I think its SWG or AWG? Anything thick enough to offer
lowest DCR @ best price (for critical midrange), the 0.39mH (not Solen too expensive for me, except maybe mid DCR~ 0.15 Ohms $25AU? but am I really going to hear 0.15 Ohm difference!?) on 'Rabbitz' (local supplier) site (www.rzaudio.com) I think is around $7AU (0.3 Ohm from memory).
Sure, I'll put a new scematic up, but I'll need to be fairly sure, its an improvement and less laughable than the last one!
"The wonderment of DIY".....chuckle..... (hope to hear from you again Omni, not sure though after this post, LOL)
grant (the verbally afflicted one)
Sreten, as to your point about lower Q increasing excursion in lower frequencies, I stand corrected, as I misworded the point I was trying to make about a lower Q box. The point I am trying to emphasize about a lower Q box is that it controls those excursions better than a higher Q box, making it more transient, providing a superior scenario for cleaner, smoother response in the midrange , as it goes up in frequency.......My bad, Grant, I hope I did not mislead you.......I am , however, still of the thought that a larger midbass cabinet will perform better, and since boxes are already made, I understand your comment, Sreten, about the pointless nature of arguing about it, UNLESS, Grant is looking to build new ones. I think it would be remiss, not to offer information, which was so freely given to me, especially by Tinitus........At any rate, Sreten, you mentioned also the concept of partially Zobeling the woofer............can you expand on this? As you may be aware of my fear of amplifier strain, I am still open minded to hear your input on this is. As I pointed out earlier, I am no expert and am always willing to learn from those more experienced than me......Maybe this does offer a reasonable compromise that I haven't thought of yet?.........Grant, I got a kick out of your rant, and totally agree with you as to the frustrations you are experiencing,as I am experiencing similar frustrations. That's cool though, that's what this forum is all about...........Respectfully.............Omni
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.