It's funny I've tended to shy away from the chebychev alignments (modelling, as I've only ever "built" one BR enclosure)... tending towards the butterworth for maximal flatness, but the one akabak came up with had only a very small dip in the response.
I have a new appreciation for my Morel MW-144's now after doing some modeling of the MCA15RCY!! It's main drawback for this project (the mw144) would be it's lower sensitivity (which is why I went MTM) but it really really does a better job (IMO) of being the midbass driver! I must admit I looked at a lot of stuff and did a lot of modelling as well before settling on the MW-144 (I'd originally discounted it before I learnt about MTM configurations). As I said early in this thread though, it Is probably out of the price range. It is around $30 US more per driver than the seas.
I tried modelling the seas as BR, but it had some nasty port resonances right where they would probably be problematic, I like the sealed better for simplicities sake as well.
I've attached a unixbox comparison of the MW-144 vs the MCA15RCY both modeled at 5L enclosure size. I think you will see where I am coming from
For my own system I'm planning to cross actively at 200Hz at 12db/octave.
edit: to get the single MW-144 at the same level as the seas, I had to up the power from 80W (which is what I have unibox set at) to 180W! Just shows how important sensitivity is if you want loud, without too many watts! The MW-144 would be able to handle it (down to 200Hz still within linear range) but the woofer certainly wouldn't
Once I've finished my MTM project I might have to do a writeup on it (and get some of you Aussie guys to have a listen to verify if it really does sound any good). I could then pinch the bass unit from this thread as the woofer part since my Vifa's are no longer available...
edit2: the ts params for the MW-144's were the average of my four drivers as measured, not manufacturers specs.
Tony.
I have a new appreciation for my Morel MW-144's now after doing some modeling of the MCA15RCY!! It's main drawback for this project (the mw144) would be it's lower sensitivity (which is why I went MTM) but it really really does a better job (IMO) of being the midbass driver! I must admit I looked at a lot of stuff and did a lot of modelling as well before settling on the MW-144 (I'd originally discounted it before I learnt about MTM configurations). As I said early in this thread though, it Is probably out of the price range. It is around $30 US more per driver than the seas.
I tried modelling the seas as BR, but it had some nasty port resonances right where they would probably be problematic, I like the sealed better for simplicities sake as well.
I've attached a unixbox comparison of the MW-144 vs the MCA15RCY both modeled at 5L enclosure size. I think you will see where I am coming from
For my own system I'm planning to cross actively at 200Hz at 12db/octave. edit: to get the single MW-144 at the same level as the seas, I had to up the power from 80W (which is what I have unibox set at) to 180W! Just shows how important sensitivity is if you want loud, without too many watts! The MW-144 would be able to handle it (down to 200Hz still within linear range) but the woofer certainly wouldn't
Once I've finished my MTM project I might have to do a writeup on it (and get some of you Aussie guys to have a listen to verify if it really does sound any good). I could then pinch the bass unit from this thread as the woofer part since my Vifa's are no longer available...
edit2: the ts params for the MW-144's were the average of my four drivers as measured, not manufacturers specs.
Tony.
Attachments
Last edited:
Tony
Your post confirms my findings in a few matters.
I found the MCA15RCY in a vented enclosure was a bit eratic and would require more work in the crossover. I'm unable to model a TL but I think would be a better option than vented. Your data confirms why I used an oversized sealed as it produces a smooth acoustic roll off which means the crossover has to do less work to achieve the target roll off as part of the work is already done. The other advantage was to provide a larger enclosure for backwave control. You can get similar roll off to the the Morel with the SEAS but means the enclosure becomes too small IMO for backwave control. The larger mid enclosure has those two advantages and only pays a small price in slightly lower power handling but the crossover can help that.
The other thing is you have confirmed is if you want to cross over lower (say <200Hz), then a mid woofer is the better option over a dedicated mid. For me the MCA15RCY needs to be crossed over >300 for 2nd order and maybe >450Hz for 1st order.
Your post confirms my findings in a few matters.
I found the MCA15RCY in a vented enclosure was a bit eratic and would require more work in the crossover. I'm unable to model a TL but I think would be a better option than vented. Your data confirms why I used an oversized sealed as it produces a smooth acoustic roll off which means the crossover has to do less work to achieve the target roll off as part of the work is already done. The other advantage was to provide a larger enclosure for backwave control. You can get similar roll off to the the Morel with the SEAS but means the enclosure becomes too small IMO for backwave control. The larger mid enclosure has those two advantages and only pays a small price in slightly lower power handling but the crossover can help that.
The other thing is you have confirmed is if you want to cross over lower (say <200Hz), then a mid woofer is the better option over a dedicated mid. For me the MCA15RCY needs to be crossed over >300 for 2nd order and maybe >450Hz for 1st order.
Hi Rabbitz, I revisited paying attention to the q of the enclosures this time! When you said it was oversized at 5L you weren't kidding!! I had to go to a 1000L box to get the q on the morel that low Definitely very different animals!!!
At 5L the morels box q is 0.718 and the seas 0.44 (according to unibox)
for a q of 0.718 on the Seas, box volume is a meagre 1.4L so I see what you mean about too tight for the backwave!
If I weren't concerned about that backwave I'd prefer to go with the 1.4L and try crossing at 300Hz with a second order electrical to try for a 4th order slope (but bear in mind I'm saying this from a theoretical view point as I haven't ever designed my own crossover).
It looks from unibox that the MCA would be down about 0.5db at 400Hz which is not too bad, but some might say the crossover should be at 500... The rollof is a bit shallower for the qtc of 0.44 is the idea to make it easier to match the woofers rollof to that of the mid (in the crossover region at least), since the woofer will not be rolling off at all (at the crossover freq), so to get a symetrical crossover the woofers filter transfer function will be pretty much all electrical? ie if we were to do the 1.4L and cross at 300Hz with 12db electrical on the mid, we'd need 24db electrical on the woofer (which is going to be expensive and more complicated)?
Am I on the right track? I think I completely missed the point originally
Tony.
edit: it looks like image attachment is currently broken so embedding instead
edit2: since that doesn't appear to have worked either, here is the link http://i109.photobucket.com/albums/n46/_wintermute/seas_vs_morel.png
When you said it was oversized at 5L you weren't kidding!! I had to go to a 1000L box to get the q on the morel that low Definitely very different animals!!! At 5L the morels box q is 0.718 and the seas 0.44 (according to unibox)
for a q of 0.718 on the Seas, box volume is a meagre 1.4L so I see what you mean about too tight for the backwave!
If I weren't concerned about that backwave I'd prefer to go with the 1.4L and try crossing at 300Hz with a second order electrical to try for a 4th order slope (but bear in mind I'm saying this from a theoretical view point as I haven't ever designed my own crossover).
It looks from unibox that the MCA would be down about 0.5db at 400Hz which is not too bad, but some might say the crossover should be at 500... The rollof is a bit shallower for the qtc of 0.44 is the idea to make it easier to match the woofers rollof to that of the mid (in the crossover region at least), since the woofer will not be rolling off at all (at the crossover freq), so to get a symetrical crossover the woofers filter transfer function will be pretty much all electrical? ie if we were to do the 1.4L and cross at 300Hz with 12db electrical on the mid, we'd need 24db electrical on the woofer (which is going to be expensive and more complicated)?
Am I on the right track?
I think I completely missed the point originally Tony.
edit: it looks like image attachment is currently broken so embedding instead
edit2: since that doesn't appear to have worked either, here is the link
http://i109.photobucket.com/albums/n46/_wintermute/seas_vs_morel.png
Last edited:
A mid enclosure with the driver crossing over (2nd order) 2 octaves or more above it's Fs is not going to be working significantly in it's piston range so I don't think it needs to be treated like a woofer box design (someone please correct me if I'm wrong). IMO a very low Qtc isn't a problem because of this and I think use of OB mids confirm this. What's the Q when using an OB mid driver? I don't know.
The bigger concern for me is backwave control as I have experienced it with a PL11MH /P11WH when used in a Qtc=0.7 enclosure. I chased the problem as a crossover issue but ended up being the backwave through the cone so when I changed it to OB mids it fell into place. The only other 3-way I have done used a 18W8531G00 (operating in it's piston range) in a sealed enclosure with a Qtc just over 0.8 and didn't have a backwave issue as the enclosure was 13 litres and a very different shape. I'm a bit paranoid with backwave issues I suppose. I got the taste of a lot of Infinity speakers through the late 70's and 80's that used OB mids as well as I think TL type mid enclosures and the absence of the backwave through the driver was a revelation.
With the crossover we are looking to sum at the crossover point about 6dB with 2nd order. If a mid is already down 1dB around this point I don't believe it's an issue as the crossover values would change to suit to give electrical filter crossover point that would be lower but would sum with the acoustic slope to reach the target.
The woofer crossover would still allow significant mids to be heard and summed with the mid driver to over 1kHz and in this style of speaker is essential to carry the weight and presence of the mids as the small mid could not achieve that on it's own.... it would surely sound weedy if the woofer was cut off too early IMO. A different story of course when using larger or multiple mid woofers for the mids.
I'm not going to get into the crossover but I think a 2nd order would be adequate for summing with the mid to give the required targets. More will be revealed I'm sure when Andre posts his crossover simulations.
The bigger concern for me is backwave control as I have experienced it with a PL11MH /P11WH when used in a Qtc=0.7 enclosure. I chased the problem as a crossover issue but ended up being the backwave through the cone so when I changed it to OB mids it fell into place. The only other 3-way I have done used a 18W8531G00 (operating in it's piston range) in a sealed enclosure with a Qtc just over 0.8 and didn't have a backwave issue as the enclosure was 13 litres and a very different shape. I'm a bit paranoid with backwave issues I suppose. I got the taste of a lot of Infinity speakers through the late 70's and 80's that used OB mids as well as I think TL type mid enclosures and the absence of the backwave through the driver was a revelation.
With the crossover we are looking to sum at the crossover point about 6dB with 2nd order. If a mid is already down 1dB around this point I don't believe it's an issue as the crossover values would change to suit to give electrical filter crossover point that would be lower but would sum with the acoustic slope to reach the target.
The woofer crossover would still allow significant mids to be heard and summed with the mid driver to over 1kHz and in this style of speaker is essential to carry the weight and presence of the mids as the small mid could not achieve that on it's own.... it would surely sound weedy if the woofer was cut off too early IMO. A different story of course when using larger or multiple mid woofers for the mids.
I'm not going to get into the crossover but I think a 2nd order would be adequate for summing with the mid to give the required targets. More will be revealed I'm sure when Andre posts his crossover simulations.
I agree, the main problem with the low Q would be if it were running all the way down (which it won't be), in which case it would probably unload at the lower frequencies. On the Q of the open baffle, I'd guess that it would depend on the particular woofer, and would be largely dependent on the compliance of said woofer. When I was trying to match the Q of the seas in unibox in the above exercise it made very little difference between 200L and 1000L at some point it would have been equivalent to running OB I guess
The most impressive speakers I've ever heard were infinities, about $35K and that was in the mid nineties I think. I was depressed when I listened to my lowly 3 ways after that, I haven't really listened to any high end stuff since
I guess I need to get my head out of text book mode
As I said I've never designed my own and only going off theory I think this is the single most important reason for having a good quality woofer! I also think it is why when I changed the 12" plessy drivers in my 3 ways to 10" vifas the improvement was so dramatic. I'm glad you said multiple drivers there, My MTM prototype sounded very good even without the 10" drivers, just a bit lacking in the lower bass
cool I guess I'm going to be learning all about this myself before too long (at least I hope it isn't too long) so may feel better qualified to actually post something useful once I feel I have some experience under the belt!!
Tony.
Last edited:
I was lucky to hear a well set up Infinity RS1 in the 80's and still remains my favourite loudspeaker. They had some great loudspeakers and used some innovative ideas. It's a shame they sold the company to Harman International but they still live on sort of as Genesis (if you have the coin).
Genesis Advanced Technologies
Ah.... the M26WR.... a wonderful driver and a shame it's gone.
Genesis Advanced Technologies
Ah.... the M26WR.... a wonderful driver and a shame it's gone.
You guys actually making me sweat, scared that I will hit my name with a coathanger.
This is my first attempt at a 3-way speaker and was only using WINISD and xoversim so please check and tell where I'm wrong.Since 100L boxes were mentioned, I've stretched it a bit to 110L for the two woofers, hope it is still OK. Seems like either sealed or BR will work, I've tuned the ports for resonance at 25Hz to give better group delay and lower cone excursion at low freq's and to get a similar slope than a sealed box, just lower in FR.
Woofers:
2 x Seas CA26RFX
Box - 110L tuned to 25Hz (2 x 6.8cm dia x 26.6cm rear? ports)
XO - 2nd order, 5mH & 100uF, woofers in parallel
Midrange:
Seas MCA15RCY
Box - 2.57L sealed
XO - Low pass: 0.68mH and (3 Ohm in series with 12uF) to gnd. High pass: 68uF and (3 Ohm in series with 6mH) to gnd.
Tweeter:
Seas 27TFFC
XO - 2nd order, 6.8uF and 0.5mH to gnd.
XO at 310Hz and 2260Hz
Sensitivity 91dB with about 4.5dB BSC.
FR: -6dB at 38Hz and 31kHz
Hope it is of use.
André
Hi Andre, don't sweat!! A lot (most?) of the participants here are learning I have a lot of experience "experimenting" with three ways, over the last 20 odd years, but most of that was not based on any science.. It did teach me a lot about what works and what doesn't (and ultimately why when I tried to understand what was going on). If any of the people here on diyAudio who have built a decent set of speakers were to listen to my three ways, I'm sure they would tear them to shreds (or just run out of the room holding their ears)! But for what I have spent on them they give reasonably good bang for the buck
Rabbitz posted earlier in the thread that he wasn't going to venture into the design because he sucked at three ways and knew his limitations (that could be slightly paraphrased), But he has none the less stood up and put up a design for us to start working on which is to be commended! I'm hesitant to do anything but ask dumb questions, because I don't have any experience with crossovers other than two different pre-built ones (which were like chalk and cheese) that I have used on my three-ways, the lpads that I added, and multiple different mids and tweeters used over the years, some vastly better than others.
In the end, until someone buys the drivers, does some measurements, and starts to do some serious design work on the crossover, anything that we post here in the mean time will be just a suggested starting point IMO
Tony.
PS. attached is a graph from my three ways, taken quite a few years ago with a computer mic (so who knows what it's freq response was like) at 0.5M in room. I originally thought it was horrid, but with what I know now (especially about how much of an effect room interaction has) It really wasn't that bad at all (which janemann told me at the time). The drivers are being used outside of their ranges though (as they don't match well) so that is the cause of some of the problems with them.
Tony.
I have a lot of experience "experimenting" with three ways, over the last 20 odd years, but most of that was not based on any science.. It did teach me a lot about what works and what doesn't (and ultimately why when I tried to understand what was going on). If any of the people here on diyAudio who have built a decent set of speakers were to listen to my three ways, I'm sure they would tear them to shreds (or just run out of the room holding their ears)! But for what I have spent on them they give reasonably good bang for the buck Rabbitz posted earlier in the thread that he wasn't going to venture into the design because he sucked at three ways and knew his limitations
(that could be slightly paraphrased), But he has none the less stood up and put up a design for us to start working on which is to be commended! I'm hesitant to do anything but ask dumb questions, because I don't have any experience with crossovers other than two different pre-built ones (which were like chalk and cheese) that I have used on my three-ways, the lpads that I added, and multiple different mids and tweeters used over the years, some vastly better than others. In the end, until someone buys the drivers, does some measurements, and starts to do some serious design work on the crossover, anything that we post here in the mean time will be just a suggested starting point IMO
Tony.
PS. attached is a graph from my three ways, taken quite a few years ago with a computer mic (so who knows what it's freq response was like) at 0.5M in room. I originally thought it was horrid, but with what I know now (especially about how much of an effect room interaction has) It really wasn't that bad at all (which janemann told me at the time). The drivers are being used outside of their ranges though (as they don't match well) so that is the cause of some of the problems with them.
Tony.
Attachments
I've used the specs on a Solen 14AWG air core, although I would prefer a 12AWG, the resistance is 0.59 Ohm.
Yes, both front firing, the filter start quite early to set the amount of BSC since the sensitivity of the two drivers together is a bit high. This allow for a bit more SPL at low freq.
A TL mid enclosure should be ideal for this mid. Using a larger mid or a midwoofer will sacrifice Mms for Fs.
Will the sealed box also be 110 liters?
Any idea what the big 5mh and 6mh inductors and 68uf and 100uf caps will cost? I assume the big inductors will not be air core or the caps be polyester/polystyrene/teflon. Actuallly the 6mh inductor can be made of pretty thin wire and have a DCR of 3ohms. Using the formulae in lalena.com
Inductance 6.03 mH
DC Resistance 2.95 Ohms
Wire Gauge 21 AWG
Coil Length 2 in
Coil Inner Diameter 2.25 in
Coil Outer Diameter 2.59 in
BTW this link might be useful.
Loudspeaker enclosure calculating with Thiele Small parameter
baffle step for a baffle of about 340mm will start as early as 60Hz and extend to about 1000Hz. THis covers the woofer and mid sections.
WOW!
I think I have digested all of that??
Just crammed this whole thread.
guys i know i havn't contributed so far but....
Is there any reason not to use a side firing bass driver for the lower octaves?
Thereby allowing a slimmer front?
That way as long as the volume works, you could use any diameter driver.
Anyhow keep it up,I'm wondering where it will come to fruition,should be interesting.
I think I have digested all of that??
Just crammed this whole thread.
guys i know i havn't contributed so far but....
Is there any reason not to use a side firing bass driver for the lower octaves?
Thereby allowing a slimmer front?
That way as long as the volume works, you could use any diameter driver.
Anyhow keep it up,I'm wondering where it will come to fruition,should be interesting.
My 3-way designs suck. They sounded very good but not great. I've given up on them.... I know my limitations. I'll watch from the sidelines.
My mains are not 3-ways but use the 810921 + 18W8531G00 + M22WR-09-08 which are a very nice combination.
Thought I better properly quote what rabbitz said rather than my paraphrasing without the full context!!
I've highlighted the important bit, I think he was underselling himself Joz, the main problem with side firing would be whether or not the mid can play low enough (or be crossed steep enough) to not have problems with directionality. Also will limit the number of places the speakers could be used (ie not close to any side walls), which I think would defeat one of the "ideals" of this concept which is to have it be suitable for as wide an audience as possible.
I can't remember off the top of my head how low the crossover freq needs to be in order for side firing to be an option, but I'm pretty sure it is quite a bit lower than what seems like a realistic minimum (of between 300 and 400 Hz) for the MCA15RCY in a sealed enclosure. That might change if someone can come up with a suitable TL or perhaps back loaded horn enclosure for it
Tony.
Hi Navin, increasing the DCR seems to have quite an effect on the reflex enclosures, less so on the sealed, but on both it cuts the level by 2db..
I've also put in a single unit at 72L and tuned to 29.5Hz (the 0.59 DCR actually helps with the low freq extention according to the model) for comparison...
Personally I don't see the need for two 10" drivers. The sensitivity of the 10" and the mid are actually very well matched with a single 10". Going to two in parallel doesn't appear to offer any great advantage in extention (though as Andre said it will give better tansient response) for a vented box (unless you go to a much bigger box of say 200L, but does require quite a 30% bigger box and it will probably need to be padded down to match the mids level... well that's my opinion anyway
Tony.
I've also put in a single unit at 72L and tuned to 29.5Hz (the 0.59 DCR actually helps with the low freq extention according to the model) for comparison...
Personally I don't see the need for two 10" drivers. The sensitivity of the 10" and the mid are actually very well matched with a single 10". Going to two in parallel doesn't appear to offer any great advantage in extention (though as Andre said it will give better tansient response) for a vented box (unless you go to a much bigger box of say 200L, but does require quite a 30% bigger box and it will probably need to be padded down to match the mids level... well that's my opinion anyway
Tony.
Attachments
You guys actually making me sweat, scared that I will hit my name with a coathanger.
Since 100L boxes were mentioned, I've stretched it a bit to 110L for the two woofers, hope it is still OK. Seems like either sealed or BR will work, I've tuned the ports for resonance at 25Hz to give better group delay and lower cone excursion at low freq's and to get a similar slope than a sealed box, just lower in FR.
Woofers:
2 x Seas CA26RFX
Box - 110L tuned to 25Hz (2 x 6.8cm dia x 26.6cm rear? ports)
XO - 2nd order, 5mH & 100uF, woofers in parallel
Midrange:
Seas MCA15RCY
Box - 2.57L sealed
XO - Low pass: 0.68mH and (3 Ohm in series with 12uF) to gnd. High pass: 68uF and (3 Ohm in series with 6mH) to gnd.
Tweeter:
Seas 27TFFC
XO - 2nd order, 6.8uF and 0.5mH to gnd.
XO at 310Hz and 2260Hz
Sensitivity 91dB with about 4.5dB BSC.
FR: -6dB at 38Hz and 31kHz
Hope it is of use.
André
Andre
For a first attempt at a 3-way crossover it looks very good indeed and much better than mine. The only thing I would add is the polarity of the drivers and from my simulation using your values the mid is reversed polarity and others are normal. Is that right? The tweeter may need a slight bit of padding (maybe) with a 1R resistor before the filter. Nice job and well done!
Don't sweat and it's great to have your contribution. Not posting it due to fear would be a crime. Every piece of the puzzle helps.
Navin
I've had success using Janzten P-Core inductors for large values and for 5mH to 6mH would be around A$15 each and they have low DCR (<0R5) depending on the wire size. For large parallel caps I've used Visaton SAF bipolar electrolytic and was about A$10 for 100uF.
Jantzen Audio Denmark - Coils
Visaton SAF - Soundlabs Group
The woofer is crossing over too high in this design for side firing woofers as it contributes to the mids. You really need a crossover <100Hz for side firing to work and a mid that's going to be working that low.
You can do that with something like a 2.1 type speaker where the mid woofer carries some bass and all of the mids and the woofer provides the bass extension. Actually my mains work something like that.