I am wondering if there are any problems building a 2.5 way system with different upper and lower midbass drivers. Using the natural baffle step design by running both only at lower frequencies, but one of them being of a different type. This is an issue of necessity as I can't find any more of the Focal 6W 4254 drivers. However, I was thinking that many manufacturers probably design similar designs using an altered lower bass driver better designed for low bass. While I don't have that ability, I could use something like the Dayton Reference 7" midbass. Another option would be one of the 6.5" woofers with extreme amounts of xmax, such as the exodous EX-6.5 woofer. Let me know if you see any problems with this. I really like the sound of my Focal towers at the moment, but I can't deal with the ridiculously low efficiency. It has killed dynamics and max output too much.
This thread may help you.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=117441&highlight=
http://www.diyaudio.com/forums/showthread.php?s=&threadid=117441&highlight=
Thanks, so from what Sreten has stated, its possible. my only problem with his reply is that he doesn't really state what difficulty you will have in mating the two drivers, and has tailored his response for someone who is not designing the crossover. He talks about a fully BSC vs not, well mine is right now, but I will simply redesign the crossover for this. I have models of the setup with the alternative drivers, and it looks pretty good. As I said before, I'm not sure how it will sound until I do it.
My concern is that, while the models look good, the drivers will sound so different that it won't sound good. Another option, which wont be nearly as affordable to do, would be to use the 7" Focal utopia midbass driver as the lower bass driver. This driver has good xmax, would work in my enclosure, and keep the same family sound.
Anyone know why in a 2.5way design Sreten says you gain 6db's of voltage sensitivity. I'm confused by this, this increase in sensitivity is only happening at the range where both drivers are operating, right. I mean, when I model this, what I see is something that looks just like my fully BSC design, but with more sensitivity. However, its not 6db's more sensitive from 500hz on up if it wasn't a BSC design. It's simply more sensitive than the full BSC design, by roughly 6db's, which doesn't happen at 4 ohms, as he mentions, but whatever impedance the system is, accept in that lower bass range, right?
I wish I had even more sensitivity, but I don't think I can afford my own desires here. It kills me that this tweeter is so sensitive, and I'm throwing away so much through resistors because of the midbass drivers. Now I suppose I could take 4 drivers, and make a sort of 2.5 way design, with two paralleled drivers, which gives me the baffle step compensation and an increase in efficiency. In order to keep the tweeters at the right height it would have to be an MMTMM design, which would begin to look a bit monstrous.
My concern is that, while the models look good, the drivers will sound so different that it won't sound good. Another option, which wont be nearly as affordable to do, would be to use the 7" Focal utopia midbass driver as the lower bass driver. This driver has good xmax, would work in my enclosure, and keep the same family sound.
Anyone know why in a 2.5way design Sreten says you gain 6db's of voltage sensitivity. I'm confused by this, this increase in sensitivity is only happening at the range where both drivers are operating, right. I mean, when I model this, what I see is something that looks just like my fully BSC design, but with more sensitivity. However, its not 6db's more sensitive from 500hz on up if it wasn't a BSC design. It's simply more sensitive than the full BSC design, by roughly 6db's, which doesn't happen at 4 ohms, as he mentions, but whatever impedance the system is, accept in that lower bass range, right?
I wish I had even more sensitivity, but I don't think I can afford my own desires here. It kills me that this tweeter is so sensitive, and I'm throwing away so much through resistors because of the midbass drivers. Now I suppose I could take 4 drivers, and make a sort of 2.5 way design, with two paralleled drivers, which gives me the baffle step compensation and an increase in efficiency. In order to keep the tweeters at the right height it would have to be an MMTMM design, which would begin to look a bit monstrous.
Hi,
Implementing BSC with a 2.5way gives you +6dB midrange and treble
sensitivity [and forces full (6dB) BSC unless higher driver has a rising
response] compared to a single driver 2-way with the same BSC.
For the two cases the tweeter is cut by 6dB for the latter.
Have you considered line level BSC ? would be a lot easier.
/sreten.
Implementing BSC with a 2.5way gives you +6dB midrange and treble
sensitivity [and forces full (6dB) BSC unless higher driver has a rising
response] compared to a single driver 2-way with the same BSC.
For the two cases the tweeter is cut by 6dB for the latter.
Have you considered line level BSC ? would be a lot easier.
/sreten.line level BSC doesn't change the efficiency issue though really. I have all the necassary crossover parts, remaking the front baffle is no big deal, and so its really just a matter of buying the drivers.
What is easier about the line level BSC? While I have some of the parts I would need to build such a circuit, there is a lot more involved. I would have to construct a simple regulated supply, the eq with appropriate opamp buffers on both sizes, etc. Any links to a simple circuit for this?
What is easier about the line level BSC? While I have some of the parts I would need to build such a circuit, there is a lot more involved. I would have to construct a simple regulated supply, the eq with appropriate opamp buffers on both sizes, etc. Any links to a simple circuit for this?
I don't think matching the driver exactly is that important in a 2.5 way. Manufacturers use different drivers all the time in this situation. Also, I think a slight timbre variation between bass range drivers would be less of a problem than a boring undynamic speaker.
OF course it's best if you have drivers with a similar cone material but below 500 Hz drivers of this size should be operating in nearly pistonic motion and tone variations will not be as pronounced this far away from the breakup frequencies.
A more important question is weather your amps can drive a low impedance speaker. If your using an integrated amp it would be preferable to have an easier to drive speaker. This can probably be easier to accomplish with a 3 way if you set the lower crossover region at the BSC frequency and use a higher sensitivity woofer.
Also, you need to think about placement. If the speakers are not going to be well out in the room you may not want more then 2-3db of BSC. If the speakers are going to be well out in the room perhaps a full 6db of BSC will be warranted.
IF you want a speaker that is as sensitive as the tweeter you will likely have to go with a MTMWW type of set up. This is what I am designing at the present time using some leftover drivers. It will be an interesting frankenstein of a speaker system.
OF course it's best if you have drivers with a similar cone material but below 500 Hz drivers of this size should be operating in nearly pistonic motion and tone variations will not be as pronounced this far away from the breakup frequencies.
A more important question is weather your amps can drive a low impedance speaker. If your using an integrated amp it would be preferable to have an easier to drive speaker. This can probably be easier to accomplish with a 3 way if you set the lower crossover region at the BSC frequency and use a higher sensitivity woofer.
Also, you need to think about placement. If the speakers are not going to be well out in the room you may not want more then 2-3db of BSC. If the speakers are going to be well out in the room perhaps a full 6db of BSC will be warranted.
IF you want a speaker that is as sensitive as the tweeter you will likely have to go with a MTMWW type of set up. This is what I am designing at the present time using some leftover drivers. It will be an interesting frankenstein of a speaker system.
Some of your replies, though perfectly true, do make me laugh a bit. You ask if my amplifiers are capable of handeling low impedance loads. Well I made them all myself from Aussie Amps modules, they can handle 4 ohm loads just fine. In fact, the power supplies are so robust over sized, and the heatsinking so large, that I am quite confident they will handle much lower impedances as well. I have measured the with 4 and 8 ohm loads, and I will be recieving the necessary parts soon to create a 2 ohm load for testing. Into an 8 ohm load they can produce 350 watts with less than .01% distortion (The best I can measure right now), and it does over 650 watts into a 4 ohm load. Distortion doesnt change, and I believe it is actually much much lower. The amplifiers will handle it I think.
I am basing my BSC on actual measurements. I have a graph I made of the in room measurements of the response of these speakers with no BSC. They were almost exactly 6db's down starting below 500hz, exactly as predicted. Adding 6 db's of baffle step brought things into +/- 1.5db's from 60hz to 15khz. However, adding 6db's of baffle step compensation means I lost 6db's of sensitivity. Given that the speakers actual sensitivity is around 82-83db's with the 2.83 volts at 1 meter, and the woofers power handling being what it is means the maximum dynamic range will be somewhat limited. Even if for short bursts the tweeters, crossover parts, and midbass drivers could handle peak power burts in excess of 500 watts, I feel like things sound dynamically over dampened. I'm not sure if there is any truth to that, but thats how I feel. I played some very dynamic symphony music, and felt like it just wasn't handling the dynamic range correctly. I did take some measurements to see what kind of peak voltages I was hitting with the music using my peak hold recording feature of my multimeter, and it actually never went over 5 volts. I guess I would need to know the dominate frequencies of the peak to truely say, but given an estimated 10 ohm impedance, and 8 volt peaks, we are actually talking about less than 10 watts. I dont really listen to music very loud, so maybe its something else. Like I said, it sounds dynamicly over dampened to me.
This could be all in my head, I should probably spend some time listening more. I have actually never implemented a full 6db's of baffle step because of the efficiency loss. I had the parts and wanted to try it, and I am very impressed with the sound over all. It's really just the dynamics now I don't like. You know I was thinking, another way to do "line level" BSC would be to add the extra drivers with an active crossover and their own amplifier. I'm still toying with the WWMTM or WMTMW idea though, it would be kind of cool.
I am basing my BSC on actual measurements. I have a graph I made of the in room measurements of the response of these speakers with no BSC. They were almost exactly 6db's down starting below 500hz, exactly as predicted. Adding 6 db's of baffle step brought things into +/- 1.5db's from 60hz to 15khz. However, adding 6db's of baffle step compensation means I lost 6db's of sensitivity. Given that the speakers actual sensitivity is around 82-83db's with the 2.83 volts at 1 meter, and the woofers power handling being what it is means the maximum dynamic range will be somewhat limited. Even if for short bursts the tweeters, crossover parts, and midbass drivers could handle peak power burts in excess of 500 watts, I feel like things sound dynamically over dampened. I'm not sure if there is any truth to that, but thats how I feel. I played some very dynamic symphony music, and felt like it just wasn't handling the dynamic range correctly. I did take some measurements to see what kind of peak voltages I was hitting with the music using my peak hold recording feature of my multimeter, and it actually never went over 5 volts. I guess I would need to know the dominate frequencies of the peak to truely say, but given an estimated 10 ohm impedance, and 8 volt peaks, we are actually talking about less than 10 watts. I dont really listen to music very loud, so maybe its something else. Like I said, it sounds dynamicly over dampened to me.
This could be all in my head, I should probably spend some time listening more. I have actually never implemented a full 6db's of baffle step because of the efficiency loss. I had the parts and wanted to try it, and I am very impressed with the sound over all. It's really just the dynamics now I don't like. You know I was thinking, another way to do "line level" BSC would be to add the extra drivers with an active crossover and their own amplifier. I'm still toying with the WWMTM or WMTMW idea though, it would be kind of cool.
pjpoes said:
Sounds like you have some nice amps. I don't think the hampered dynamics is in your head for several reasons. And it sounds like from your measurements you might indeed need the full 6 db of BSC.
First, even a few db is audible when it comes to fast transient passages in the music. Second, accomplishing the BSC with the two way may have caused you to use too many components in the crossover. This can often rob the system of life and dynamics.
At low background levels the speakers may sound good. But indeed they are probably dynamically flat since you have had to pad down part of the frequency range by 6db!! It's one thing to pad down a tweeter but doing that to a mid bass driver often leaves it lifeless sounding.
Try to get a similar cone material Focal driver of 7 inches or so and built it as a 2.5 way. I think you will be much happier.
I think so too. I think I will put in an order for the 7" drivers next week and cut new test baffles to see how things go. I know this is a strange thing to have just sitting around, but I happen to have a pair of CFAC 14 guage inductors at 3.8 mh's and some 18uf clarity caps. these values seem to work quite well with the .5 woofer.
pjpoes said:
Hmmm........
Line level BSC does change the efficiency issues. You get an extra
6dB of headroom mid and treble. Does not change the fact an 89dB
bass driver into half space is 83dB into full space, that is physics.
Line level can be done with two or three passive components.
e.g. http://sound.westhost.com/bafflestep.htm
/sreten.Hi pjpoes,
Do you have a target SPL in mind? (say allowing 10dB headroom for crescendos). Also - what is your current system sensitvity?
Reason being - a TMM 2.5 way may still not cut it. With your sort of amplifier power - you are always going to be excursion limited - probably around the 50 - 75 watt mark with your nominal 6.5" driver based TMM.
David.
Do you have a target SPL in mind? (say allowing 10dB headroom for crescendos). Also - what is your current system sensitvity?
Reason being - a TMM 2.5 way may still not cut it. With your sort of amplifier power - you are always going to be excursion limited - probably around the 50 - 75 watt mark with your nominal 6.5" driver based TMM.
David.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.