Instead of hijacking the OB thread... how does a 2xRS225, Neo8, Neo3 OB sound?

[diy speaker guy]

Re: Instead of hijacking the OB thread... how does a 2xRS225, Neo8, Neo3 OB sound?

Now all this talk about the Q of the speaker at resonance has made me go back and think hard about my own budding OB design. So I thought: man, what if I can get real low end extension with a higher Q driver at FS! Nope, as I see it, that is delusion.

You need to read MJK's OB papers. Alpha 15 gets down to around 50 hz on an 18 inch baffle with good sensitivity.

And you cannot hear q if the q is appropriate for the design.

[JasonP]

Re: Instead of hijacking the OB thread... how does a 2xRS225, Neo8, Neo3 OB sound?

You need to read MJK's OB papers. Alpha 15 gets down to around 50 hz on an 18 inch baffle with good sensitivity.

And you cannot hear q if the q is appropriate for the design.

I have that exact sheet, and yes that is truth. What I'm kind of getting at obliquely is that, the Q of the driver is only one part of the speaker parameters. For instance, if I change the Q parameter and even the FS of my Dayton DA270 to create a fictional driver more like the Eminence Alpha 15, it does /almost nothing/ in my worksheet. Its the sum of all the thiele/small parameters that make the driver what it is, not just one part of the spec. If you can see what I'm saying.

In short, I'm not saying that its not possible to choose a driver that would go much lower and that driver would likely have a higher Q, its just that looking at only that isn't the solution.

[Face]

Re: Instead of hijacking the OB thread... how does a 2xRS225, Neo8, Neo3 OB sound?

Plus, of course, the "hear the Q" sound you mention . . .

Can you elaborate on this?

[diy speaker guy]

Re: Instead of hijacking the OB thread... how does a 2xRS225, Neo8, Neo3 OB sound?

I have that exact sheet, and yes that is truth. What I'm kind of getting at obliquely is that, the Q of the driver is only one part of the speaker parameters. For instance, if I change the Q parameter and even the FS of my Dayton DA270 to create a fictional driver more like the Eminence Alpha 15, it does /almost nothing/ in my worksheet. Its the sum of all the thiele/small parameters that make the driver what it is, not just one part of the spec. If you can see what I'm saying.

In short, I'm not saying that its not possible to choose a driver that would go much lower and that driver would likely have a higher Q, its just that looking at only that isn't the solution.

I'm not sure what worksheet you are using but I have a simulator and I can SHOW you results. Pick a low frequency cutoff (30 hz or higher) and spl goal and I'll show you a real driver (or drivers) that can do it on a narrow baffle (2 inches wider than the driver) and with very simple crossover requirements.

But if you are dead set against high q drivers there's not much point...

(If doubling the driver's qts doesn't do anything, your worksheet is broken.)

[JasonP]

Re: Instead of hijacking the OB thread... how does a 2xRS225, Neo8, Neo3 OB sound?

I'm not sure what worksheet you are using but I have a simulator and I can SHOW you results. Pick a low frequency cutoff (30 hz or higher) and spl goal and I'll show you a real driver (or drivers) that can do it on a narrow baffle (2 inches wider than the driver) and with very simple crossover requirements.

But if you are dead set against high q drivers there's not much point...

(If doubling the driver's qts doesn't do anything, your worksheet is broken.)

I'm using Martin King's MathCAD sheet: OB_2_Drivers_Corner_Passive_4_12_11.mcd

I'm trying to say its more than just the Fs and Qts that matters, as I can just change those values and it has little effect on the resultant output. I made the Qes, Qms, and Fs of the DA270-8 the same as the Eminence and it had little effect. Its the sum of all the parameters: Vas, Sd, Bl, and so on that make the driver. I'm not trying to get into an argument over the merits of high Q and low Q drivers, and I'm not dead set against anything (well expect cheap motors, but I think many of us can agree on that :D) I think we are both in agreement here, but are just having misunderstanding due to miscommunication?

[diy speaker guy]

Re: Instead of hijacking the OB thread... how does a 2xRS225, Neo8, Neo3 OB sound?

I think we are both in agreement here, but are just having misunderstanding due to miscommunication?

Maybe but I'm not really sure because I'm not exactly sure what your point is. I'm just responding to this:

Now all this talk about the Q of the speaker at resonance has made me go back and think hard about my own budding OB design. So I thought: man, what if I can get real low end extension with a higher Q driver at FS! Nope, as I see it, that is delusion. Here is what I mean, lets look at the output of a driver with various Qs at its resonance point:

And I'm saying it's not delusional, it's actually easy. I showed how to do it with 6 inch drivers on an 8 inch baffle, MJK showed how to do it with 15 inch drivers on a 20 inch baffle, and I can show you how to do it with any (reasonable) LF cutoff on a small baffle with existing drivers. As long as you balance fs, qts and baffle size you can get relatively flat extension down as low as you want (within reason, say 30 hz).

You lose a bit of sensitivity doing this (you don't retain the driver's IB sensitivity) but that's common sense, you have to fight a 6 db roll off. Going back to Pete's example, he's already down to 75 db at 100 hz and if he tried to passivly eq it flat to an even lower frequency he would lose even more sensitivity, and fast. If he chose 30 hz as the low end cutoff the sensitivity would be 55 db, as you can see on the graph he posted. Using a high q driver you can get the same 30 hz with much higher sensitivity, and it won't need such an aggressive crossover to whip it into shape.

The point is that you can force low q drivers to do what you want but it's more expensive than choosing a driver that already agrees with your plans, and sensitivity gets worse the lower you try to push it.

I'm not going to comment on "cheap" motors, other than to say that you only need enough magnet to achieve your desired system q. I realize you and Deward believe otherwise (strongly) but passive OB systems with high q drivers win awards so how bad can they really be?

[martin]

Re: Instead of hijacking the OB thread... how does a 2xRS225, Neo8, Neo3 OB sound?

I'm trying to say its more than just the Fs and Qts that matters, as I can just change those values and it has little effect on the resultant output. I made the Qes, Qms, and Fs of the DA270-8 the same as the Eminence and it had little effect. Its the sum of all the parameters: Vas, Sd, Bl, and so on that make the driver.

Please be careful changing T/S parameters on the fly. While it seems simple enough to just adjust the Qts of a driver, what you are really doing is adjusting all of the driver's physical properties that combine to become the T/S parameters. Sometimes these adjustments are not physically possible and you are left with "Garbage In --> Garbage Out". I would recommend using one of my two T/S consistency checking worksheets to make sure what you are attempting to do makes sense, if the mass or stiffness starts to go negative you have entered a new area of physics. I am surprised that you do not see any signficant change in the response which makes me suspicious that something is messed up.

[JasonP]

Re: Instead of hijacking the OB thread... how does a 2xRS225, Neo8, Neo3 OB sound?

Please be careful changing T/S parameters on the fly. While it seems simple enough to just adjust the Qts of a driver, what you are really doing is adjusting all of the driver's physical properties that combine to become the T/S parameters. Sometimes these adjustments are not physically possible and you are left with "Garbage In --> Garbage Out". I would recommend using one of my two T/S consistency checking worksheets to make sure what you are attempting to do makes sense, if the mass or stiffness starts to go negative you have entered a new area of physics. I am surprised that you do not see any signficant change in the response which makes me suspicious that something is messed up.

Ok, right, that thought has crossed my mind. If all the T/S params are interconnected, changed just a couple is likely to create an unrealistic or impossible driver (I had said something about fictional above). This is surely something worth remembering for the future.

[JasonP]

Re: Instead of hijacking the OB thread... how does a 2xRS225, Neo8, Neo3 OB sound?

Maybe but I'm not really sure because I'm not exactly sure what your point is. I'm just responding to this:

...

And I'm saying it's not delusional, it's actually easy. I showed how to do it with 6 inch drivers on an 8 inch baffle, MJK showed how to do it with 15 inch drivers on a 20 inch baffle, and I can show you how to do it with any (reasonable) LF cutoff on a small baffle with existing drivers. As long as you balance fs, qts and baffle size you can get relatively flat extension down as low as you want (within reason, say 30 hz).

Ok, so right, what I'm learning here is that I was just getting carried away with analysis. I was taking the Q out of the T/S and treating it as a "thing" when it means nothing outside of the set of parameters that is T/S. I didn't mean to say that any given driver would have X performance on Y baffle, but I was looking at: What if you just changed the Q? That is where my comment came from. However, as you say and Martin just mentioned, since all the T/S parameters are connected changing the Q implies you are in fact getting just what you described. Awesome! I now have a new view of the parameters that describe drivers, thanks

[fjhuerta]

Re: Instead of hijacking the OB thread... how does a 2xRS225, Neo8, Neo3 OB sound?

BG's definatly will go to 500Hz (12dB), H frame at 200 Hz(12dB), figure in the midbass gain from the top end of said H frame and they blend rather well.

The woopdee-doo @ 200Hz is room junk of course, but all in all not to bad, ehh?
If you do go active between your RS's and the neo's you'll be plenty happy!! Doing it passively can be done as well...the Qts is a little low on the 8ohm units for passive H-frame, the 4's are a bit better, but then you have to have an amp that'll work under a 2 ohm load, or put out some killer watts at 8ohm.

I'm impressed! Do you have a link for your build?!?

[fjhuerta]

Re: Instead of hijacking the OB thread... how does a 2xRS225, Neo8, Neo3 OB sound?

OK... I just read the thread in its entirety. Here are my thoughts and restrictions, ideas, etc...


1) I really don't want to do this passively. I want to learn how to do things active - I have a bit of experience in passive x-over design, and I know beating the RS225 into submission would be expensive without the MiniDSP. I have the Balanced MiniDSP. I have two identical amps (Rotel RB-1070s) with plenty of grunt. I have tons of passive stuff from my old speakers, for the Neo3 / Neo8 x-over. I'll be using the EQ function of the MiniDSP extensively - I know for a fact how difficult the Neo drivers are, and I'm going to beat them to submission with the MiniDSP. I think I have everything I need!

2) I have experience with Edge, and played with it for a bit. It seems to me that going for a narrow baffle on the Neos and a wider one on the RS225s brings the best results. Also, the nicest looking speaker.

3) I'm actually sort of worried about using the RS225s because I like my music LOUD. I do have a nice sub - a Tempest (Remember those?) 15" sealed cube with a 300W amp, so as long as the RS225s can hit, say, 60 ~ 70 Hz I think I'll be happy.

4) My room really is a difficult one. I have *plenty* of room behind the speakers (they are placed near a corner, and their backs point directly to it - I have around 6 feet of space behind them), but the room is full of metal / glass / hardwood. I was hoping that dipoles would help me tame reflections.

5) I'd love to take the Neo8 down to 500 Hz but don't iknow if it'll survive...

I'm picking up my old speakers this weekend, and I'll be simulating stuff these days (we have some time off around here). I'll use Edge and will try to learn how to design an H frame or something similar.... I still don't know whether I should go for a flat baffle or an H frame, BTW...

In a nutshell, a wider baffle will be easier to design a passive xo for the Neo's, but the tradeoff there is irregular off axis dispersion.

Narrow baffle gives you a better off axis dipole response (30 deg -1db, 45 -3db, 60 -6db as defined by Linkwitz), but in most cases requires eq, don't know if a passive approach would work. Depends on baffle size.

According to Edge, a narrow baffle helps flatten the response off axis. Anyway, since I'll EQ them to death I don't care that much. I just want thngs to look as nicely as possible before EQing them.

Piece of cake . . . there are no critical dimensions. A pair of RS225 on a baffle 12-14 inches wide should get you down to 80 Hz. easy enough . . . neither of the neos need baffles at all (or just enough to hold them in place), and they'll work better that way (look at John K's "Note" for an example). The passive crossover should work between the neos, and LR4 is fine for the cross to the RS225s. You *will* need to cross them to a subwoofer (use LR4 for that, too) for decent bottom end, but any small box[es] will do, at least for a start, and what works best below 80-100 Hz depends a lot on your room (I like dipole down to 40-50 Hz., but people argue about that).

One more reason to use the MiniDSP, I guess. A bandpass filter on the woofers around 80 Hz (hopefully lower) and maybe 500 ~ 600 Hz to the Neo8. It sort of sounds like a MartinLogan electrostatic. I hope it doesn't *sound* like one of those, though... (had a pair, never liked them).

I just want to ask, what size room do you have? and how far from the wall will the rear of your dipole sit, and how far to the wall behind you? I would look at this if I was you. You *might* find problems with the Neo8. I would look into foam and stuffing behind the rear of the Neo8 in a small chamber. I would do a damped Uframe type build for the RS225. I did this once with 2x DCX crossovers. I never nailed it down, and sold all the drivers. I never made it past raw mdf. It was fun though.

The room is pretty big. Around 150 square meters. It may be too big... I don't know.

I experimented with this very combination of drivers with a passive crossover. Never personally liked the NEO8 running open baffle (at least as a single). The rear chamber sounded better to me. Doesn't mean it can't be done, but I didn't think it was worth it.

Yeah, I actually followed your build thread back when I was doing my own Neo8 / Neo3 speaker. I know you were disappointed in the Neo8. I was, for a long time. It took a lot (and I mean *years*) of tweaking until I got it finally right. The only reason why I'm doing it all over again is because the Neos are dipoles (and I'd have a true full range dipole) and now I have an EQ...

I've been playing with the BG parts and have switched to the Neo10 in place of the Neo8. It just works better all around, by far.

As for the RS225... not even on my radar Too small, not enough sensitivity. I'll run dipole to ~40Hz if I ever get this done, but I don't have a space in which to use the results so am in no rush.

I don't know. The Neo10 sounds very difficult to work with without EQ, even more so than the 8! But I agree, if I were to buy the drivers right now, I'd skip the 8 and go for the 10. It just doesn't seem to be much beefier than the Neo3.

B-G Radia already has a similar design--the FS-420. A dipole with two Neo 10 and a Neo 3 with two 6.5" woofers in a sealed box at the base. All in a slim floor standing speaker. The crossover points at 250 Hz and 1.5 kHz.

Thanks for the tip, I forgot about it. I'll check it out very closely...

And because this setup would be so limited by the woofer output, 75dB is not an issue, especially in a smaller room. Using Dual RS225 on a narrow baffle is just fine in a passive implementation.

Foolish? Using Dual RS225 with a separate amp as though it will produce more bass, yeah, pretty foolish.

Will it be as limited if I bandpass filter it at, say, 80 Hz? Why would it be foolish to use a separate amp? I'd think avoiding all passive elements and filtering the signal actively would make far more sense than going all passive....

[diy speaker guy]

Re: Instead of hijacking the OB thread... how does a 2xRS225, Neo8, Neo3 OB sound?

1) I really don't want to do this passively.

That makes things much easier then. I'm not familiar with the mini dsp but if it has built in dsp with flexible filters and and it's capable of providing a healty boost, the sky is the limit. (Actually the limit is driver displacement but you know what I mean.)

I'll use Edge and will try to learn how to design an H frame or something similar.... I still don't know whether I should go for a flat baffle or an H frame, BTW...

The Edge program only simulates baffle response (and only for a flat baffle). It won't show driver response. You can use MJK's or John K's software (both free for U and H baffles, MJK's is $25 for flat baffles.)

H and U frames have a big resonance caused by their depth that has to be dealt with. In an H frame, it's usually around 300 hz or so, in a U it's usually lower. If your mini dsp has flexible dsp filters that's easy enough to notch, otherwise that will probably have to be your crossover point.

It looks like you are in good hands with Deward, your goals and budget match his design style well. Good luck and post it up when it's done.

[fjhuerta]

Re: Instead of hijacking the OB thread... how does a 2xRS225, Neo8, Neo3 OB sound?

Thanks a lot!

I only hope the Neo8 can be crossed over low... I have other mids (ZA14, RS150), but I'd rather have an all Neo array up there. For looks.

Looks do matter, you know ;)

[cjd]

Re: Instead of hijacking the OB thread... how does a 2xRS225, Neo8, Neo3 OB sound?

Go 4 way. Put something below the 225. They can cover the range to the Neo8 and get you something bigger to help push SPL capacity up where you want it.

[MSaturn]

Re: Instead of hijacking the OB thread... how does a 2xRS225, Neo8, Neo3 OB sound?

I already noted the Neo8 can go 'low' just fine. 500Hz LR4 would be easy.