-
Hi-Vi Research DIY project: any advice?
I'm considering a project, (my 3rd), using the Hi-Vi B3N driver which is so-called full or extended range. I know there are smart, experienced people out there, and I love advice.
In fact I thinking of using them in a 3-way, two-box, all-Hi-Vi system something like this:
- Tweeter: one Hi-Vi RT1L round planar: 3500 Hz up
- Mid-range: four Hi-Vi B3N: 300 - 3500 Hz
- Bass: one or two Hi-Vi M8a: 0 - 300 Hz
I thinking of four B3N's because one has pretty limited power handling. I think a 2nd order crossover would work with these drivers at these frequencies.
First box would be tweeter and mids in a TMMMM or MMTMM configuration. Second would be a sealed box for the woofer(s). I'm thinking of two boxes because it would be easier to achieve a ridgid, well-sealed result, and I could use either pair separately in some later project.
The driver specs are here ...
-
I just finished a small, full-range HTIB type system with the B3N. Bass power handling is quite low, but I don't see any need for more than 1 of those if you're looking for 3-way system. They'll handle midrange power , guess it depends how much juice you plan on feeding them.
As a standalone speaker they are surprisingly good. Vocals are clear and the imaging and soundstage are well above what you could get in most HTIB systems. I've owned $200 small bookshelf speakers that couldn't keep up with the midrange. I think they'd be an excellent, low-cost mid-range driver.
If you use more than one, driver spacing will be critical I think, I wouldn't want too much physical separation across the critical midrange.
They have very low efficiency, around 80-81 dB. This can require some creativity to overcome if you're using them in a 3-way with more efficient woofers or tweeters (87 dB or so in the case of your chosen drivers).
Sounds like an interesting project.
-
Kex, based on your comments ...
[QUOTE=kexodusc]I just finished a small, full-range HTIB type system with the B3N. Bass power handling is quite low, but I don't see any need for more than 1 of those if you're looking for 3-way system. They'll handle midrange power , guess it depends how much juice you plan on feeding them.
...
If you use more than one, driver spacing will be critical I think, I wouldn't want too much physical separation across the critical midrange.
They have very low efficiency, around 80-81 dB. This can require some creativity to overcome if you're using them in a 3-way with more efficient woofers or tweeters (87 dB or so in the case of your chosen drivers).
... QUOTE]
I might consider using only two B3N; running two in parallel would presumably increase their output due to lower total impedance by 6 dB, (or only 3dB ??). I would place them as close together as physically possible -- this might argue for a TMM configuration.
-
Yeah, you gain 3 dB from doubling the driver, 3 dB from running them in parallel and halving the impedence....
Your nominal impedence will depend mostly on the woofers...or the B3N if it is acting as the woofer. Just keep that in mind.
-
Quote:
Originally Posted by kexodusc
If you use more than one, driver spacing will be critical I think, I wouldn't want too much physical separation across the critical midrange.
Sounds like an interesting project.
MTM is a very proven design, are you arguing against it?
-
Quote:
Originally Posted by Geoffcin
MTM is a very proven design, are you arguing against it?
Not at all, not sure where you got that idea from. If you read again you'll see Feanor mentioned using multiple B3N's in either a TMMM or MTM alignment. MMTMM's are known for further enhancing the lobing problems along the vertical axis that MTM's are famous for (and why designers hate MTM's for center channels). Just requires a bit more work to compensate for this, that's all. Drivers spacing and speaker height are critical for MTM's.. The most common mistake for newbies (present company included when I made a pair) is to start with an TM, 1st or 2nd order crossovers, take half the values for the inductors, and twice the values for the caps, and fire away. On axis response is fine, but off-axis is usually terrible.
Interstingly enough, Joe D'Appolito who is credited with creating the MTM alignment has rethought his original assesment twice now. The orignal MTM's called for 3rd order acoustic slopes. A decade later, he changed his mind to 4th order. D'Appolito has lately mentioned that he's leaning towards some hybrid higher order slopes to produce even better results. These difficulties, combined with the dispersion characteristics of many drivers, has led him to recommend TMM's as an alternative more and more.
FYI, the Thor T-line MTM by Joe D is one of the best speakers I've heard under $10 K !!!
-
Whooa! Interesting!
I'm not exactly looking for trouble: if I should avoid MTM, (or MMTMM if ever worse), then I will. I was thinking of a 2nd order crossover, but sounds like that's a bad idea.
On the one hand, I would like to hear more, up-to-date MTM theory -- but in layman's language if possible. Any suggestions?
On the other hand, what about the TMMMM, (my other idea)? Doesn't stacking drivers like that limit vertical dispersion? And isn't that generally a good thing, limiting floor and ceiling reflections?
-
Just playing devils advocate ;)
Quote:
Originally Posted by kexodusc
Not at all, not sure where you got that idea from. If you read again you'll see Feanor mentioned using multiple B3N's in either a TMMM or MTM alignment. MMTMM's are known for further enhancing the lobing problems along the vertical axis that MTM's are famous for (and why designers hate MTM's for center channels). Just requires a bit more work to compensate for this, that's all. Drivers spacing and speaker height are critical for MTM's.. The most common mistake for newbies (present company included when I made a pair) is to start with an TM, 1st or 2nd order crossovers, take half the values for the inductors, and twice the values for the caps, and fire away. On axis response is fine, but off-axis is usually terrible.
Interstingly enough, Joe D'Appolito who is credited with creating the MTM alignment has rethought his original assesment twice now. The orignal MTM's called for 3rd order acoustic slopes. A decade later, he changed his mind to 4th order. D'Appolito has lately mentioned that he's leaning towards some hybrid higher order slopes to produce even better results. These difficulties, combined with the dispersion characteristics of many drivers, has led him to recommend TMM's as an alternative more and more.
FYI, the Thor T-line MTM by Joe D is one of the best speakers I've heard under $10 K !!!
As I was wondering myself how speaker spacing had an effect on performance. What's your take on multiple driver systems like the Axiom TTMMWW, and the Paradigm Signatures TMWWWW?
-
Quote:
Originally Posted by Geoffcin
As I was wondering myself how speaker spacing had an effect on performance. What's your take on multiple driver systems like the Axiom TTMMWW, and the Paradigm Signatures TMWWWW?
Nothing wrong with multi driver speakers at all. I don't see a need for 3 or 4 woofers myself, that introduces way too much complication IMO, but I'm sure Paradigm has their reasons. Why not just invest that money into a better, common woofer? Do you really need 4? The difficulty of pulling that off, and added cost in the crossover had better be justified. I don't like the Signatures for the money, and I haven't heard Axiom's TTMMWW (wow).
But keep in mind my perspective is coming from that of a modest DIY-er with limited measurement and CAD resources. Keep in mind most speaker companies have access to sophisticatd measurement setups, abilities to measure on axis and off-axis response, phase tracking etc. Axiom and Paradigm are a stone toss away from the NRC in Canada. Most DIY-ers don't, or have limited measurement setups and rely on good techniques.
Driver spacing depends on the speaker design. The general rule of thumb (there's always exceptions, but expect more difficulty) is to have the woofer and tweeter as compact as possible, with center to center spacing less than 1 wavelength at the crossover frequency. MTM's dictate equal spacing for each woofer from the tweeter. Consider that in the crossover region, two or more speakers are emitting the same frequency. Ideally you want this common sound to reach you at the same time. In the case of an MMTMM where all 4 midwoofers are crossed over at the same frequency, you'd will always have 2 mid-woofers that are closer to your ear than than the other two. At some point it's foreseeable that there will be a significant delay between the arrival of sound from the closer woofers and more distant woofers. Kind of like a hangover effect. If it gets really bad, there can be cancellations. Guess it' s no different than speaker placement where even an inch or two offset can produce undesireable results, except it's even more so as there's a uneveness at some frequencies. If you're using 4 midwoofers, you'll have to watch this. At an XO frequency of 3 KHz, your c-to-c spacing for the woofers and tweeters should be no greater than 4.5 inches..kind of hard to do with 4 of those drivers. You could do it with 2.
With MTM's you want both woofers the same distant from your ear ideally, as the midrange frequencies are split between the two. Getting closer to one can have the effect of the tweeter appear louder or more prominent, or more worse, cancellations of some frequencies.
TMM's are often a bit safer, the crossover for the low woofer is much lower so the the spacing rule is a bit more generous. The size of the wavelength at the lower xo point is much larger. I think generally though, you lose the higher efficiency and dynamics that MTM's are known for. TMM's as 2.5 ways offer an easy way to counter baffle-step loss.
Six of one, half dozen of the other. The subject of TMM vs MTM comes up a lot on some other forums I visit, and people there with waaay more knowledge than me are split as to which is "better". My take is it depends on the design goal and execution.
Feanor - I don't see anything wrong with an MTM or a TMM if that's what you want to do. But in an MTM my understanding is it's critical to have the tweeter as close to ear-level as possible for best results. Keep that in mind if you're going to add another woofer underneath the MTM section of the speaker.
-
Thanks Kexo!
A more coherent discription of the trade-offs vs. good points of mulitple speaker design I've yet to read!
-
Good stuff!
Quote:
Originally Posted by kexodusc
...
Feanor - I don't see anything wrong with an MTM or a TMM if that's what you want to do. But in an MTM my understanding is it's critical to have the tweeter as close to ear-level as possible for best results. Keep that in mind if you're going to add another woofer underneath the MTM section of the speaker.
I guess I'm still not certain whether I'll go MMTMM or TMMMM. But I'm hanging in with the four driver mid-range with the Hi-Vi B2N drivers. They only handle 15 watts per driver which doesn't seem like much given I want to cover 300-3500.
Likely I'll go for the B3S driver which is identical to the 'N' model except it's square. Thus two of them can be mounted about .5" closer together than the 'N's.
-
Quote:
Originally Posted by Feanor
I guess I'm still not certain whether I'll go MMTMM or TMMMM. But I'm hanging in with the four driver mid-range with the Hi-Vi B2N drivers. They only handle 15 watts per driver which doesn't seem like much given I want to cover 300-3500.
Likely I'll go for the B3S driver which is identical to the 'N' model except it's square. Thus two of them can be mounted about .5" closer together than the 'N's.
How is that going to affect the final load is what I would be worried about. Even is they are an 8ohm nominal driver, connecting them in parallel would produce 2 ohms, correct? You may have to add resistors to the crossover between the drivers, and that would take away from the drivers already low efficiency. I'm not really up on crossover design, so I may be wrong here. I'ld just hate to see you make this speaker and find out you need a Krell amp to power it!
-
Yeah, paralleling 4 of those is gonna cause some nasty impedence problems. Series crossovers with parallel ones could probably maintain the impedence, but that's adding more complexity.
Feanor, how loud are you planning on playing these? How much power? Above 250 Hz or so these things require very little excursion and can handle a lot of power. Thermal dissipation by the voice coil becomes the concern then. The power ratings on the speakers are refering to the voice coil's ability to dissipate heat, and usually assume running full range. I have no problem pushing mine above 90 dB in room, that's only 8-16 watts or so running full range, these will handle more. You aren't constantly feeding the voice coil with current.
Running 2 in parallel shouldn't cause too much impendance grief. Maybe you can try a higher crossover, 400 or 600 Hz or so?
-
Kex, Geoff,
I was thinking of a series-parallel arrangement for the four B3N's.
Here's my network I modelled using X-over 3 Pro. Note that it's 3-way including the M3a 8" bass even though I intent to put that in a separate box. Also note the crossover points here are 250 and 4000 ...
http://gallery.audioasylum.com/cgi/u...e=&w=807&h=635
Here are the SPL and impedance graphs ...
http://gallery.audioasylum.com/cgi/u...e=&w=800&h=600
To my rank novice eye the relative levels and impedance don't look problematic, but ... any comments you might have would be great.
-
Feanor, I can't access your pics...I even registered and donated to the Asylum, still no pics...
any othe options? Can you post them here, or e-mail them?
-
Yep, I've created an Audio Review Member Gallery
Quote:
Originally Posted by kexodusc
Feanor, I can't access your pics...I even registered and donated to the Asylum, still no pics...
any othe options? Can you post them here, or e-mail them?
Kex,
I'm please you're willing to have a look at my stuff. I think you can now just go to "Gallery" and search for Feanor.
I don't know why you can't see them at AA. I set "Anyone can view", so registered of not, anyone should see them: sorry to put you to that trouble.
-
Here are those pics from the AR Gallery ...
-
Wow, that's neat.
Quote:
Originally Posted by Feanor
Here are those pics from the AR Gallery ...
I like that software! Can you just plug in the parameters from any driver and have it spit out crossover schematics?
-
Yeah, kinda neat
Quote:
Originally Posted by Geoffcin
I like that software! Can you just plug in the parameters from any driver and have it spit out crossover schematics?
Yes, you can do just that. It comes with a quite large database of driver specs too. I had to load the B3N driver info, though.
X-over Pro is extremely easy to use. It will quite quite a few crossover alignments which you just need to select. However it does have some drawbacks.
For example, you can't use different alignments for the top and bottom of the mid-range band pass. Also, it doesn't encorporate notch filters nor baffle shelve compensation. My understanding is a bit weak in these areas, but I believe it doesn't handle experimental impedance or phase measurements. Also it doesn't do stuff like model placement of the speakers on the baffle or that sort of thing.
I have a copy of CALSOD 1.4 but that's such a pain to learn and to use that I haven't gotten around to mastering it. I'm going to work on that though, because CALSOD is apparently much more powerful.
-
CALSOD is more powerful as far as I know. But you can get great results from any number of software. I use mostly free software, so I end up having to learn alot of basics in order to make the limited software useful. For example, baffle-step compensation might not be incorporated in X-over Pro, but if you're aware of it, know where the -3 dB loss is at roughly, you can build it into your speakers slope.
CALSOD is just too much money for me to fork out for now.
|