[goskers]

Andy,

I urge you to search SL's page for proper information regarding the orion design.  Nothing was left to chance by SL.    Many 'typical' rules were broken by him in this design but I think he knows what he's doing  

The passband of the 8" driver is not as limited as the low end excursion of the 2 10" woofers so the usage of the 8" mid for low midbass duties is not a problem.

[eweitzman]

Can any of you Orion owners shed any light on why Siegfried put the bass/mid crossover down at 120Hz?

Might I respectfully suggest that you post your question on the Orion forum at http://orion.quicksytes.com/index.php?

- Eric

[JoshK]

Can any of you Orion owners shed any light on why Siegfried put the bass/mid crossover down at 120Hz?  It seems to me the Peerless 10" woofers are not doing enough work and the Seas mid is doing too much?

My understanding is he choose 120 to avoid the resonance that the H woofer baffle creates above this freq.

[awm]

I'd imagine that if the resonance avoidance that joshk mentions was the primary reason, there were still numerous others.  Perhaps he wanted to move the xover point below the regions where human hearing is most sensitive.  Or, on the other side of the coin, maybe it was to utilize just one driver for the entire mid-range, etc.  (And, 3khz is not mid-range, IMHO, though others might disagree...)

But, funny you should mention the Peerless woofers not appearing to do enough.  They are in a dipole config and as such, have to move HUGE amounts of air to create the necessary spl's.  I was thinking that they were doing too much!  In my Orions and others I've listened to and whose owners I've spoken with, the bass has always been the rate limiter.  I used the boost on the active XO to produce what I consider more realistic representations.  It made a very large difference -- more than I'd anticipated.  The mid-range has never been a problem in mine.  

Also, someone below implied that the slopes were 1st order -- 6db/octave.  I'd swear that this is inaccurate and that the active XO achieves 4th orders across the board -- but, i do have to recheck it myself.  

Finally, ewietzman, how did you access the Orion forum in the link you provided?  Didn't work for me, and I'd heard of this mythical forum awhile back, but forgot to continue my quest for it till now...  Thanks

Andy

[andyr]

Can any of you Orion owners shed any light on why Siegfried put the bass/mid crossover down at 120Hz?

Might I respectfully suggest that you post your question on the Orion forum at http://orion.quicksytes.com/index.php?

- Eric

Thanks, Eric,

Good idea - I will as well !!

Regards,

Andy

[andyr]

... My understanding is he choose 120 to avoid the resonance that the H woofer baffle creates above this freq.

Maybe, Josh ... but the Orion has a straight "open" baffle - with one 10" Peerless pointing forwards (at you) and the other pointing backwards (away from you).

The Phoenix I think is the one with the H-woofer-baffle.

Regards,

Andy

[JoshK]

The Orion woofer is considered an H baffle and will have a resonance.  I will try to look up on linkwitzlabs were SL stated this himself.

[JoshK]

From

http://www.linkwitzlab.com/orion_challenge.htm
* Crossovers at 120 Hz and 1440 Hz, both LR4 (24 dB/oct)
* Woofer - two Peerless 10" XLS, 830452
Push-pull mounted in H-frame of 11.5" x 11.5" x 24.5" OD

http://www.linkwitzlab.com/orion-faq.htm
*  A woofer H-frame tends to have an acoustic resonance at the high end of its frequency range.

[andyr]

... But, funny you should mention the Peerless woofers not appearing to do enough. They are in a dipole config and as such, have to move HUGE amounts of air to create the necessary spl's. I was thinking that they were doing too much! In my Orions and others I've listened to and whose owners I've spoken with, the bass has always been the rate limiter.  ...

Thanks, andy.

It's nice to hear that - then the "ultimate" Orion definitely has a AKSA 100 module on each bass driver!  
 
The reason I'm worrying about this is that my current 3-way active speakers use 2 x 55s and a 100 ... so, for the 4th amp for the Orions, I was wondering whether I need a 2nd AKSA 100 (to have one on each bass driver) or a 3rd 55 (so my existing 100 goes on the mids).

So now all I have to do is work out is whether the mids could really do with a 100 too!  

Regards,

Andy

[andyr]

The Orion woofer is considered an H baffle and will have a resonance.  I will try to look up on linkwitzlabs were SL stated this himself.

You're absolutely correct, Josh - my apologies!  

But what of my fundamental Q ... is the mid doing "so much work" bcoz of its extended frequency band that it really needs a 100??  

Regards,

Andy

[eweitzman]

My understanding is he choose 120 to avoid the resonance that the H woofer baffle creates above this freq.

SL says H-frames have a quarter-wavelength (IIRC) resonance near the upper end of their frequency range. The shape of the side baffles isn't there solely to make the speakers look cuter but rather to break up this resonance.

Maybe, Josh ... but the Orion has a straight "open" baffle .. the Phoenix I think is the one with the H-woofer-baffle.

The Orion has sides, which makes the baffle part of the H-frame. The Phoenix uses a W-frame configuration.

I'd imagine that if the resonance avoidance that joshk mentions was the primary reason, there were still numerous others. Perhaps he wanted to move the xover point below the regions where human hearing is most sensitive.

 This is just a guess: maybe he operates the drivers in the regions where they have best performance, such as lower non-linear distortion.

But, funny you should mention the Peerless woofers not appearing to do enough. They are in a dipole config and as such, have to move HUGE amounts of air to create the necessary spl's. I was thinking that they were doing too much! In my Orions and others I've listened to and whose owners I've spoken with, the bass has always been the rate limiter.

I'd like to point out how your comment is both correct and somewhat misleading. The Orions have a -3db point of 30Hz and are -6db at 20Hz. The majority of speakers don't have anywhere near this amount of extension at any SPL.

SL's Audio Artistry Beethoven design used eight 15" woofers per channel to get sufficient excursion to blow out anybody's windows. He didn't want the Orion to be so imposing.

I used the boost on the active XO to produce what I consider more realistic representations. It made a very large difference -- more than I'd anticipated. The mid-range has never been a problem in mine.

Turning up the bass in the ASP will alter the tonal balance, but it won't get you past the "rate limiter". The SPL is ultimately limited -- intentionally -- by the amp's power. Most people who visit me find, at first, that the bass is a bit louder than they're used do. I've measured the speaker's response and find the 0 settings on the woofer and tweeter branches of the crossover are just about right.

A friend came over the other night to "test the bass". (My friends like to try and trip up my gear. Some friends!) He played incessant europop, trance, other electronic music, NIN, Tool, and other stuff I never heard of. I measured 105db peaks, C-weighting, on my ratshack meter. A-weighted averages were often around 95db. He didn't miss anything from the music, even when the bass amps -- as designed -- were clipping and so limiting the woofer excursion. He remarked at least five times about detail and other things in the bass that he'd never heard before. The Orions come very close to subwoofer performance for music at loud levels.

Also, someone below implied that the slopes were 1st order -- 6db/octave.

They're 4th order.

Finally, ewietzman, how did you access the Orion forum in the link you provided?

I emailed the admin. The site will be down for much of today.

"Normal" philosophy would say "the woofers need much more power than the other drivers, so they should each have a 100w AKSA module". The others can have AKSA 55s.

No. SL specifically warns against more than 60 watts per bass driver. As awm pointed out, dipole drivers are asked to move a lot of air as frequency goes down. Dipole output drops 6db/octave, so the crossover boosts the signal 6db/octave. That would ordinarily result in tremendous power going into the drivers at very low frequencies, bottoming out the drivers and potentially damaging the voice coil formers. Instead of another filter in the crossover that limits very low frequency output and would add group delay distortion, SL just let's the amps run out of steam. Clipping harmonics are sent to the woofers and can't blow the other drivers.

... the mid-range spans a relatively high octave-range. If there is a 55w AKSA on each woofer ... is it possible that the mid-range should better have an AKSA 100 rather than an AKSA 55?

I've only seen the midrange amps clip once. I don't recall what I was doing at the time, but it must have been pretty crazy, like playing sine waves at high SPL. During the torture session above, the woofer amps were clipping from time to time because of the beats, but the mid and tweeter amps never clipped.

- Eric

[AKSA]

Hi Andy,

You wrote:

1. In a speaker which has 2 x 10" woofers, 1 x 8" (7"??) mid-range and a tweeter, the crossover points (6dB L-R) are 120Hz and about 1,500Hz.

2. So ... what is the "ultimate" AKSA setup for this speaker?

If the 10" woofers were 8R, then a single 100W AKSA on both in parallel would be fine, since the 100 will drop to 2.5R with ease.  However, if the 10" woofers were 4R, you would need two 55W AKSAs because a 100 will not pull a 2R load comfortably.  You can reduce the rail voltage on the 100 to 42V and then do a 2R load, but my preferred solution with 4R woofers would be to use two 55s.

The midrange I would drive with a 55W AKSA, and the tweeter I would drive with a 40W AKSA.  I would choose the 40W variant because it can be refined specifically for the job;  this gives it the sonics of a 2A3 SET amplifier with gonads.......    

3. "Normal" philosophy would say "the woofers need much more power than the other drivers, so they should each have a 100w AKSA module". The others can have AKSA 55s.

4. However, the mid-range spans a relatively high octave-range. If there is a 55w AKSA on each woofer - which seems to be entirely adequate (as we have heard!  ) - is it possible that the mid-range should better have an AKSA 100 rather than an AKSA 55?  IE. against "normal" theory??

I don't feel a 100 is necessary for the midrange, even with extension down to 120Hz.  I may be wrong, but this is a domestic system and the Spls are not that great.  But this is my opinion, Andy, and people with very large listening rooms/high volume preferences/low efficiency speaks might feel differently.  It's very difficult to argue hard and fast on these issues of power - everyone is a bit different.  To someone who likes his 8W 300B SET, all this power looks quite laughable, yet I have a friend who worked in the music industry who argues vehemently that 1200W is the minimum requirement for a domestic room......

Eric,

Many thanks for your comprehensive and detailed explanation of Seigfried's masterpiece.  I really appreciate it;  but I didn't read it until I had completed the above!!  So, if Seigfried suggests two 60W amps for each of the two drivers of the Orion, so be it;  I would not disagree with him on this matter since he designed the speakers, an area where I do not have much knowledge.

Cheers,

Hugh

[andyr]

Hi Eric,

Thanks for your post.  Now, you said:

... No. SL specifically warns against more than 60 watts per bass driver. As awm pointed out, dipole drivers are asked to move a lot of air as frequency goes down. Dipole output drops 6db/octave, so the crossover boosts the signal 6db/octave. That would ordinarily result in tremendous power going into the drivers at very low frequencies, bottoming out the drivers and potentially damaging the voice coil formers. Instead of another filter in the crossover that limits very low frequency output and would add group delay distortion, SL just let's the amps run out of steam. ...

This makes good sense.  However, the AKSA 100 is known to have more "slam" in the bass than the 55 (a separate issue from the fact that it can output much power ... although of course, the two are probably linked!).

It seems to me that the bass drivers would therefore be better driven by the "punchier" amp?  

... If the 10" woofers were 8R, then a single 100W AKSA on both in parallel would be fine, since the 100 will drop to 2.5R with ease. However, if the 10" woofers were 4R, you would need two 55W AKSAs because a 100 will not pull a 2R load comfortably. ...

This makes sense, Hugh, but you left out the final option ... what about an AKSA 100 module for EACH Peerless driver.  So it is driving 4 ohms.

There seems to be a conundrum here!  Eric has made the point that using an AKSA 100 defeats one of SL's design parameters - which was to deliberately use an amp which would "run out of steam", to stop excessive LF voice-coil extension.

Yet the 100 is "punchier" on the bass than the 55.  

Perhaps the best compromise is to get "punch" by having an AKSA 100 on each Peerless driver but downgrade its voltage rails.  Then you get the "punch" from the dual output pairs but only, say, 80w out of them (into 8 ohms).

... I have a friend who worked in the music industry who argues vehemently that 1200W is the minimum requirement for a domestic room. ...

Whew!!     Great in winter (don't need no heating!) but not so good at this time of the year!  

Regards,

Andy

[andyr]

Hi Hugh,

Further to my Q above about maybe using a "downgraded" AKSA 100 module for each Peerless driver, I notice SL makes reference to alternative recommended amps (as the AT amps are not available outside North America) ... including the Alesis RA300 which is rated at 150wpc into 4 ohms ... which could be achieved by an AKSA 100 having lowered power rails??  

Regards,

Andy

[eweitzman]

... if Seigfried suggests two 60W amps for each of the two drivers of the Orion, so be it; I would not disagree with him on this matter since he designed the speakers, an area where I do not have much knowledge.

I suppose it best to defer to SL's words about this. You can use bigger amps with the woofers in parallel, but he recommends against it. In the following quote, he's discussing the AT6012 60wpc 12-channel amp and the AT1806 180wpc 6-channel amp. This info is from the middle of a very long page without internal html anchors, so I'll quote SL here.

(6) - The AT6012 is rated at 60 W/channel into 8 ohm ... The actual power output of the individual amplifiers is more than sufficient for midrange and tweeter. A separate amplifier channel is allocated to each of the 10" woofer drivers to obtain output capability that is commensurate with the midrange, yet minimizes bottoming and the risk of mechanical damage to the woofers, or having to reduce the speaker's low frequency extension. At very high sound levels the amplifiers clip first, which gives a clearly audible warning to turn down the volume ...

If the two 10" drivers of the woofer are connected in parallel, then they present a low 2 ohm impedance to the power amplifier around 100 Hz. The power amplifier driving this load must be capable of at least 200 W into 4 ohm. I prefer separate, lower power amplifiers for each 10" driver and consider the AT6012 an optimum match for the Orion, based on extensive experience with it and with the higher power AT1506 amplifier. [replaced by AT1806]

(7) - The AT1806 is a six channel power amplifier rated at 180 W per channel. It is capable of producing slightly higher output volume from the ORION between 20 Hz and 60 Hz with a maximum of 5 dB at 40 Hz. Above 60 Hz it will deliver 2 dB less output than two AT6012 channels driving the two 10" units individually. When using the AT1806, the two 10" woofers of each speaker are connected in parallel to a single amplifier channel. The crossover/equalizer requires no modification.

The higher power amplifier has the potential to bottom out and damage the woofers below 40 Hz and to overheat midrange and tweeter at excessive volume levels. The extra volume capability around 40 Hz can only be used with caution. I find that it exceeds what I would use for normal listening to various types of music in my large size living space. For these reasons I consider the AT6012 of (6) a better match for the ORION.

Another alternative is to drive the woofers in series. The amp would have to deliver enough power into 16 ohms (twice 60w?) and the gain of the amp would have to be 3db (or 6db?) higher than the mid and tweeter amps. You might be able to change the gain in the crossover -- the bass gain is adjustable by +/-2.5db.

- Eric

[SamL]

Don't mind if I ask a side question since so many of you here have experience with Orion. Have anyone here listen to John Kreskovsky (aka John K) NaO and how would you compare it with Orion?
http://www.musicanddesign.com/naomain.html
As NaO only need 4 channel of amplification, it is a better dipole option for me.

TIA,
Sam

[andyr]

Don't mind if I ask a side question since so many of you here have experience with Orion. Have anyone here listen to John Kreskovsky (aka John K) NaO and how would you compare it with Orion?
http://www.musicanddesign.com/naomain.html
As NaO only need 4 channel of amplification, it is a better dipole option for me.

TIA,
Sam

Great link, Sam,

John K has some interesting papers on "Transient perfect crossover design" ... so he's another one who knows his onions!  The NaO might well be equally as good as the Orion - it uses a pair of Peerless 10" drivers for the bass, anyway, and rolled off at 120Hz ... the same!  

Be nice if someone has actually heard both and can comment!

Regards,

Andy

[AKSA]

Hi Andy, you wrote:

Perhaps the best compromise is to get "punch" by having an AKSA 100 on each Peerless driver but downgrade its voltage rails. Then you get the "punch" from the dual output pairs but only, say, 80w out of them (into 8 ohms).

This is, in fact, an option I decided on for Peter, aka PSP, and we went to some detail to spec a 100W AKSA operating from just 42V rails.

This amp would have great current capacity for low impedance loading, with greater SOAR safety margin owing to lower rail voltage.  This certainly would confer the additional punch you need, Andy, but, notwithstanding, I'm still convinced that what we heard at Mark's was more than adequate.

However, your listening room is large, so I would never argue that this is a bad option.  After all, there's no substitute for cubic inches, and nothing succeeds like excess, and all that stuff.........

Hi Eric,

Thanks for your post, full of very useful detail.  I don't like running woofers in series as damping factor is greatly affected by mutual damping paths back to the amp.  In any event, you are right about more sensitivity;  a 16R load would require root2 more output voltage to achieve the same power as a 8R load.  This raises the 100W requirement from 80Vpp to 113Vpp, necessitating a redesign.  You'd need a minimum of 60V rails for this, and this would preclude use of 63VW electros in the power supply, greatly increasing cost as 75VW and higher are MUCH more expensive.

Cheers,

Hugh

[andyr]

Hi Andy, you wrote:

This is, in fact, an option I decided on for Peter, aka PSP, and we went to some detail to spec a 100W AKSA operating from just 42V rails.

This amp would have great current capacity for low impedance loading, with greater SOAR safety margin owing to lower rail voltage.  This certainly would confer the additional punch you need, Andy, but, notwithstanding, I'm still convinced that what we heard at Mark's was more than adequate. ...

Thanks Hugh, but why do you need "great current capacity for low impedance loading" if you have 4 AKSA modules for the 4 Peerless drivers ... since each one will surely see 4 ohms?

Yes, what we heard at Marks CERTAINLY WAS more than adequate but I don't currently have 8 modules of AKSA 55 ... I only have 6 modules in all.  So I need to get 2 more modules and my dilemma is...  do I buy another two 55 modules (and use my two 100 modules on the mid-range) or do I buy another two 100 modules, so each bass driver can have a downgraded 100?

This is what I thought would be best, given the "slam" assoviated with the 100 ... so what power does reducing the rails from 49v to 42v result in?

Regards,

Andy

[PSP]

Andy,
I struggled with the same question some months ago... bass driven by one souped-up 100 driving woofers in parallel or one 55w per driver?  SL certainly recommends one 60w amp per driver, and if I was in your position with six 55w modules, I'd buy two more and be done with it.  In my case, I only had four 55w modules and so it was a matter of buying four more 55w modules or two 100w modules.

With no passive xo components in the chain, 55w driving a 90.5db midrange driver should have plenty of punch.

After a lot of e-mail discussion with Hugh, I decided to go with the 100 for the paralleled woofers.  He was confident that it was up to the task, and it was the lower cost option.

The question of how much power is needed to drive a dipole woofer is complex... depending on driver T/S params and the characteristics of the amp, you get frequency ranges in which SPL is limited by driver Xmax, amp voltage output, and the amp's current output.  For details, see:  http://www.linkwitzlab.com/dipole_spl_limit.htm

Using closedbox-1.xls (which you can download from the above web page) you can evaluate dipole woofer performance for various combinations of driver and amp parameters.  For analysis of open baffle systems, SL recommends setting the Internal Volume to something much larger than Vas; I chose 10,000 liters.  The Dipole Path Length = 380mm for the Orion.

When I plugged in the T/S params for the 10" Peerless XLS and estimated amp characteristics (based on discussion with Hugh) I found that SPL would be current limited above 80Hz unless I could get 10A/channel out of the amp.  The amp itself can handle that, but I needed to use 625VA transformers (rated to 10.5amps), 20amp diodes, high current power switches (I used NKK JWL Series, 16 amp, 163amp peak inrush current) and an LC Audio soft start circuit (10 amps) for each channel to handle the inrush.

To summarize:

100w AKSA N+ driving 2x10" XLS in parallel, decision based on cost
43v rails for stable amp operation into 2.45R (XLS in parallel).  Hugh estimates that this will deliver 190w into 2.45R  
Enable 10amp/channel to avoid current limiting the amp.[/list:u]
Now I know that I'm not going to drive this thing to 10A/channel very often, but I do not want to build something that I know is hobbled before I even apply power to it.

My purpose here is not to advocate for this approach, but simply to document what I've done and how I thought about the problem.  I'll let you know how it works.  Since I'm still doing the metal work on the amp, then I need to build the XO, then I need to build the cabinet, it will be a while before we know the answer.

Peter