[Danny Richie]

In Eric Steg's design the single 8" didn't play high enough to reach the tweeter and the 5" didn't play low enough to reach the subs.  That is why he landed at using both 5" and 8" drivers.

I am not sure what his thinking is there. If he made all of those designs then it looks like a trial and error learning curve took place. 500Hz is not a place for a crossover. That is right in the heart of the midrange.

Danny do you think that by using 2x5" and 4x8" drivers he was adding bandwidth and output to get the sensitivy up?  Somewhere in that thread I remember reading he was still padding down the 99db tweeter by ~7-8db.

Well, he's getting the sensitivity up for sure. His impedance will be really low though. 

What would it take in the midrange to run without padding down a tweeter like that?

You could reach that sensitivity level with two high sensitivity 8" woofers and still just play just low enough to reach the 200Hz range. The problem though is that the acoustic centers get so far apart that it creates that problem to overcome.

This is yet another unfounded comment based on your misunderstanding of the design.  I'm not sure why you are continuing to bash the speaker given that you have stated to understand the concepts and goals of it.    

Get over it Lowtech. I am not bashing that speaker. It is a great speaker, but limited SPL levels and dynamics are a known limitation. I have had many owners tell me that as well.

[Nyal Mellor]

Nyal, you will not get enough output from a single 5" driver as low as 200 Hz.  If you're trying to estimate SPL using x-max, that's not realistic when it would take kilowatts to reach x-max at 200 Hz.  I haven't tried to simulate that but it's just a general rule, you can't move much air with a driver that size.   However, it doesn't necessarily change the concept if you're working with a DEQX, when you can figure this out experimentally.  So long as the woofer is free from cavity resonance (no deep cabinet shape) it can just cross higher to the mid, like 300-350 Hz.

If you can't model SPL using xmax what other factors do you have to consider? I found this on the Linkwitz site http://www.linkwitzlab.com/dipole_spl_limit.htm

I still like the idea of attempting to run a mid below its dipole peak. It seems I could do that with a 5" driver and a lowish XO to the tweeter.

[lowtech]

Get over it Lowtech. I am not bashing that speaker. It is a great speaker, but limited SPL levels and dynamics are a known limitation. I have had many owners tell me that as well.

The Orion is SPL limited below 120Hz.  This is clearly stated as a design tradeoff and the Thor subwoofer was designed to address it.

You have "many owners" tell you this?  Do you mean one owner or more?  The one owner that I'm aware of was attempting to drive his Orions with a Pioneer home entertainment receiver.  And to make matters worse, he was running two 4-ohm XLS woofers off of one channel!  It's not too surprising that he wan't pleased with the results and I'm not surprised that the amplifier did not perfoirm properly running LESS THAN A TWO OHM REACTIVE LOAD.

If you do understand the design of the Orion and its tradeoffs (as you have stated), and are aware that your customer was using them with sub-optimal amplification, then you are intentionally bashing the speaker.   

[HT cOz]

This is yet another unfounded comment based on your misunderstanding of the design.  I'm not sure why you are continuing to bash the speaker given that you have stated to understand the concepts and goals of it.   

I dont think it is so crazy to think that SPL could become an issue.  I created a grid for the W22EX001 woofer and keeping in mind that a solo grand piano plays at 106db, orchestra 109, and a rock band 120db.  I think you can see that the woofer is pretty maxed out at reasonable seating distances.  The stated short term power handling is 300W.  I certainly would not say that the design has a lot of headroom.  I'm by no means a pro at this and my grid could be off.  I am still learning here 




Edit for new information




Kind Regards,
Robert

[Nyal Mellor]

I dont think it is so crazy to think that SPL could become an issue.  I created a grid for the W22EX001 woofer and keeping in mind that a solo grand piano plays at 106db, orchestra 109, and a rock band 120db.  I think you can see that the woofer is pretty maxed out at reasonable seating distances.  The stated short term power handling is 300W.  I certainly would not say that the design has a lot of headroom.  I'm by no means a pro at this and my grid could be off.  I am still learning here 




Kind Regards,
Robert

Shouldn't a stereo pair be +6db?

[Danny Richie]

If you can't model SPL using xmax what other factors do you have to consider? I found this on the Linkwitz site http://www.linkwitzlab.com/dipole_spl_limit.htm

I still like the idea of attempting to run a mid below its dipole peak. It seems I could do that with a 5" driver and a lowish XO to the tweeter.

Drivers have mechanical and thermal limits.

You can use a single 5" driver if you want, but you will have the dynamics and SPL limitations of a mini-monitor.

The Orion is SPL limited below 120Hz.  This is clearly stated as a design tradeoff and the Thor subwoofer was designed to address it.

Okay we agree on that, but I don't see adding a sealed sub to increase SPL up to 120Hz being a real solution. Adding a sub to cover the first octave a little better will make it a full range speaker, but that is not going to increase its SPL limitations.

You have "many owners" tell you this?  Do you mean one owner or more?

Yes, more than one. I have even sold some of them new speakers. The first guy that complained to me that he loved his Orion's but they didn't have the dynamics or SPL capabilities that he was looking for replaced them with an Alpha LS kit that I sold a long time ago. I think that Alpha LS kit has been discontinued for nearly five years. So this is no new revelation. That guy also complained of lack of low bass.

The Orion's are great speakers (we agree). They have limited SPL capabilities, especially below 120Hz (we agree). The only misunderstanding is that you do not understand that they are not the end all be all, or even the best speaker for everybody, or every room. There are thousands of speakers out there and someone is going to like something else better.

[Danny Richie]

Robert,

After compensating for the baffle step loss, that mid is in the 85db range. And a point source speaker will loose 6db per every doubling of distance.

Also, the mid naturally starts rolling off below 200Hz in an open baffle and it is using gain (more output) to keep it playing down to the crossover point. There could easily be 3db of gain added to it to allow it to crossover there at 120Hz. So instead of using 1 watt to hit 85db you are really using 2 watts.

Nyal is also correct in that adding a second speaker is an increase of 6db. However, that is only for a mono signal that is in phase at the listening position. I'd look at one speaker at a time.

[lowtech]

The only misunderstanding is that you do not understand that they are not the end all be all, or even the best speaker for everybody, or every room. There are thousands of speakers out there and someone is going to like something else better.

No.  I think the misunderstanding begins with the fact that you believe I think "that they [the Orion] are the end all be all, or even the best speaker for everybody, or every room.".  They are not.  However, they do succeed on all counts in meeting their design objectives, which is what you don't seem to understand.

Additionally, I think we may have a slightly different set of compromises that we're willing to accept.  Personally, I would never be happy listening to a pro audio coax driver that was designed primarily to play at ear-bleeding levels.

[Nyal Mellor]

Drivers have mechanical and thermal limits.

You can use a single 5" driver if you want, but you will have the dynamics and SPL limitations of a mini-monitor.

Hi Danny, could you elaborate further and / or reference materials that I could review to better understand this? Thanks.

[Danny Richie]

No.  I think the misunderstanding begins with the fact that you believe I think "that they [the Orion] are the end all be all, or even the best speaker for everybody, or every room.".  They are not.  However, they do succeed on all counts in meeting their design objectives, which is what you don't seem to understand.

Okay fair enough.

Additionally, I think we may have a slightly different set of compromises that we're willing to accept.  Personally, I would never be happy listening to a pro audio coax driver that was designed primarily to play at ear-bleeding levels.

I never thought I would either, but I was wrong. The best sounded, cleanest, and most natural sounding mid-range I have ever heard is from a driver that was originally designed for the pro market.

[Danny Richie]

Hi Danny, could you elaborate further and / or reference materials that I could review to better understand this? Thanks.

I am headed out for lunch and will try to come back to this later today.

[thomasjefferson]

You can use a single 5" driver if you want, but you will have the dynamics and SPL limitations of a mini-monitor.

I don't quite agree with this.  Consider that even some huge hi-fi speakers have ordinary dome tweeters covering the highest frequencies, and small midrange drivers.  They can still present a grand scale acoustically, and much of that depends on limiting the bandwidth of each driver.  The large size mostly serves the bass drivers, which also handle most of the power and give the impression of a big sound. 

A mini-monitor is limited by using a small midwoofer all the way down to subwoofer territory.  Place it on top of a bass bin and cross (actively) at 300 Hz, and you've got a dramatic jump in performance capability.

The point being, just because you use a small format driver (a 1" tweeter, or a 5" midrange) doesn't mean you can't achieve a satisfying or dramatic scale of sound.  Mostly that will depend on limiting the bandwidth of those drivers toward the upper end of their usable range, and then transitioning to progressively larger drivers toward the bass. 

For example there could be a 1" tweeter, a 5" midrange, 12" mid-bass and multiple 15" subwoofers.  It's a high performance system, within the usual limits of hi-fi.  Of course there are people who want to take it further, and they end up using compression drivers and large format mids, etc.  But that's a matter of listening habits and taste. 

Is the goal to generate dazzling special effects at high SPL, or is it more casual listening?  This should be a choice made at the beginning of the system design.

Nyal, it might be all you need to add some subwoofers for below 40-50 Hz, as Linkwitz ended up doing with his own Orion setup.  This is a far easier thing to do than starting from scratch, and it would buy a lot of headroom.  If nothing else you could start a future project with the bass taken care of. 

[Danny Richie]

Hold on there Mr. Jefferson. A 5" mid in a box crossed over at 300Hz and one in an open baffle crossed over at 200Hz is a completely different ball game. Let's look at the numbers.

Here are two measurements of the same 5.25" woofer.



The Red one is in an 8" wide mini-monitor box, and the Green one is at the top of an 8" wide open baffle. Note the differences in response. In an open baffle you get that peak at 1kHz, but look how much faster it losses output on the bottom end.

In this example the woofer level is at about 90db, but by 200Hz it is down to 83db levels. In a small ported box, and after compensating for the baffle step loss, this thing is at about 87db. If it were in an open baffle then that would be about 84.5db to 85db. And this doesn't include the reduced output that will be caused by the crossover. This is just the natural roll off.

Two of these woofers in parallel (MTM) hit 90.5db sensitivity with a smooth response in an 8" wide by pretty tall open baffle (small narrow floor standing speaker).

Now let's look at one of these woofers: http://www.madisound.com/store/product_info.php?products_id=1595

It is rated at 87.5 to begin with. But look at the curve. It is well below 85db from 600Hz on down, and that output doesn't go up in an open baffle. You are really more around 84db and that is still before the loss from the baffle. That's 6db down from the woofer above. Let's say you use a 8" wide baffle and get the same results as every other woofer on a baffle that wide. That will put you level at about 78db of output. You could get nearly 79 to 80db out of it if you didn't compensate for all of the loss. You'll be lucky if it is still as high as 74db at 200Hz, and you still have to put a crossover on it.

Now having said that, I also think that the measurements of the Excel woofers, as posted, are off. It really looks to me like they are measuring them in or on a small baffle. If so then those output level in the bottom end are being shown too low. I'd guess as to give them back 3db in that lower range. Even if it was up to 82 or 83db, that is still pretty low. Most small woofers like that are still going to have limited SPL and dynamics.

Still think that you'll have more dynamics and output than a mini-monitor?

The catch 22 is that if you make the baffle wider (and/or taller) then that peak at 1kHz comes down into a lower range. If it is wide enough then the peak can start to fill some of what is rolling off. But the bottom is still going to drop out of it real quick. A nearly 12" to 13" wide baffle will allow the peak to drop back in there and fill in some of the loss making it flat to mid 200's and easy to cross at or below 200Hz. The drawback to the wider baffle is the penalty of surface reflections and reduced imaging and sound stage depth.

[thomasjefferson]

Hold on there Mr. Jefferson.

Oh, call me TJ.

Nyal, I hope you appreciate this analysis, because it's pretty important stuff for your pursuit.

I just want to add that I already suggested the 200 Hz high pass was too low for a 5" midrange in an OB.  Looking at that graphic, the green line beats the red line above 400 Hz.  To make this a high performance design, the mid needs to be limited to a high range, like not below 400 Hz.  Then the question becomes, how do you get high enough (in frequency) with the woofer, when the woofer range needs to also reach down to ~50 Hz?  We've already covered the possibility of a single 12" or 15" AE woofer.  They would probably work pretty well.  The mid/woofer crossover could end up closer to 500 Hz.  It would take some experimentation to determine this, based on directivity of the large cone (increasing as you go higher), rapid rolloff of the small cone (as you go lower) and any response issues of either driver.

[scorpion]

Danny is far too categorical regarding this mid speaker discussion. Here is from my simulation of a baffle-less 18Sound 6ND430 - see 'Simulating a NO Baffle Speaker' below, a 6" unit with 5 mm X-max and 500 W 10 ms power handling. To caluculate simulated Max SPL I used 300 W powerhandling to be on a very safe side. The picture is like this:



This is thus for use without any baffle at all. On top this speaker also sounds very good. 

/Erling

[Danny Richie]

Okay TJ, here is you next catch 22.

If you cross higher then you get up into the mid range. From my experience, putting a crossover right in the heart of the mid-range is a no no. That is a tough place to get right. Splitting a female vocal, for instance, up so that half of her voice is coming from large woofers and half from a smaller one, with difference voice coil offsets, cone material, etc, just doesn't work out very well.

I found it best to cross above the midrange (300Hz to 500Hz) in the 850Hz or higher range, or below it in the 200Hz range or less. And from my experience the 200Hz and under is a magic number.

Scorpion, I think you are finding an exception. That is a really high power handling driver.

By the numbers though, my 5.25" mid is 89db sensitivity. Two of them in parallel (they are 16 ohm woofers) gives me a 90.5db level in an open baffle crossed at 200Hz. By those numbers, one woofer would then be 84.5 in the same application. So if the mid in consideration is rated at, and is a solid 87db rating, then it is really going to be level at about 82.5db in the same application. All of these numbers are 1 watt of course. 

[Nyal Mellor]

Danny is far too categorical regarding this mid speaker discussion. Here is from my simulation of a baffle-less 18Sound 6ND430 - see 'Simulating a NO Baffle Speaker' below, a 6" unit with 5 mm X-max and 500 W 10 ms power handling. To caluculate simulated Max SPL I used 300 W powerhandling to be on a very safe side. The picture is like this:



This is thus for use without any baffle at all. On top this speaker also sounds very good. 

/Erling

So basically the conclusion is that you can use a smaller driver, you just need a very large amplifier / high power handling to produce the SPL? Maybe some form of pro driver, if one could find it in a 5/6" form factor would be suitable.

If you equalize the driver to flat at line level using an active circuit, what happens to the power required to achieve xmax? Haven't you reduced it?

[scorpion]

EQ won't change these Max capabilities but as Danny says you have to choose the right driver for OB use.
Getting a 90 dB/watt sensitive 3 mm X-max 5-6" driver and highpassing Linkwitz-Riley 12 dB/octave at 300 Hz,
then this driver will not need more than 8 watts to reach 100 dB. And that will a lot of drivers do also on so small baffles
as to allow operation below the first dipole baffle peak.

Crossing at 200 Hz, which I do, agreeing with Danny about this frequency - I cross at 300 Hz also but then 24 or 48 db/octave steep, needs more careful computations and is very much
easier with active drive.

But going much higher most small drivers will be power constrained rather than having too small X-max, like my example shows.

/Erling

[thomasjefferson]

I'm not so concerned about having a crossover point in the 500 Hz region, if it's done well.  Using an active crossover helps get a good splice, and you also need to choose drivers that have similar behavior in this area--low distortion and similar dispersion.  Having the drivers close together also helps. 

However--you can still use a small upper midrange while following Danny's recommended crossover locations, if you're willing to go 4-way (above a subwoofer, making a 5-way system).  That means the upper mid crosses around 1 kHz, and the lower mid goes from there down to around 200 Hz.  Then the woofers take it from there down to 50 Hz. 

John Kreskovsky's NaO Note is an example of this, more or less.  He's got a 10 cm upper midrange covering from 1 kHz to 6 kHz.  His woofer design doesn't go up to 200 Hz but I'd suggest using an OB woofer that does go that high.  Here's one possible 5-way system lineup:
 
>5 kHz: tweeter.  Small format, 19-20 mm.
1 kHz - 5 kHz: upper midrange (OB).  9-12 cm.
200 Hz - 1 kHz: lower midrange (OB).  18-22 cm.
50 Hz - 200 Hz: woofer (OB).  Single 15", or multiple smaller drivers.
<50 Hz: subwoofer (closed box).

If the upper mid goes as high as 5 kHz in an open baffle, there may not be much point to adding a rear tweeter.  You're not likely to hear much reflected energy from a rear tweeter above 5 kHz anyway, at least when listening in the front hemisphere.  Obviously this complex of a design wouldn't be easy to pull off.  But that's what makes DIY interesting, right?

[lowtech]

John Kreskovsky's NaO Note is an example of this, more or less.  He's got a 10 cm upper midrange covering from 1 kHz to 6 kHz.  His woofer design doesn't go up to 200 Hz but I'd suggest using an OB woofer that does go that high.  Here's one possible 5-way system lineup:
 
>5 kHz: tweeter.  Small format, 19-20 mm.
1 kHz - 5 kHz: upper midrange (OB).  9-12 cm.
200 Hz - 1 kHz: lower midrange (OB).  18-22 cm.
50 Hz - 200 Hz: woofer (OB).  Single 15", or multiple smaller drivers.
<50 Hz: subwoofer (closed box).

If the upper mid goes as high as 5 kHz in an open baffle, there may not be much point to adding a rear tweeter.  You're not likely to hear much reflected energy from a rear tweeter above 5 kHz anyway, at least when listening in the front hemisphere [wrong!].  Obviously this complex of a design wouldn't be easy to pull off.  But that's what makes DIY interesting, right?

Right.