[sfdoddsy]

I currently have what might be called steroided Orions. Instead of the 10" Peerless, I use a pair of Adire DPL12s a side in an H-frame. Whilst my speakers play flat to 20Hz, and will do so with more volume than the Orions, they are still excursion limited and I cross to monopole subs at 40Hz.

I use a DCX2496 for xover and EQ.

The DPL12s are very good dipole drivers, but previous over-vigorous use of the volume control is perhaps taking its toll and I am looking at moving them to more relaxing duty in a sealed sub.

I was all set on the AE IB12 when I stumbled across the threads about the GR OB servo. The servo theory makes a lot of sense for OBs, IMHO.

From reading the threads, it seems people are extremely happy with them, but I have a couple of questions.

According to SL's MaxSPL spreadsheet at http://www.linkwitzlab.com/dipole_spl_limit.htm my DPL12s are limited to about 85db at 20Hz, and with a bit of room gain this seems to be true in practice. The GR Servos have pretty much identical Vd, and my current H-frame seems to be very close to the one Danny recommends so they should be about the same. Is this what you guys are getting too?

If so are you using them with extra subs for the really deep and/or loud stuff? I've found that one sealed sub driver equals about four of the same driver when used as a dipole down low.

The second question is about the EQ. Does the shelving filter in the amp provide enough, or does the servo add its own?

SL adds over 40dB of boost in the Orions. I don't need as much because of a higher Q driver and bigger H-frame, but I still use over 20dB to get them flat to 20.

Thanks for any info.

 

[sl_1800]

I have Orion's and I have to engage the 50Hz rumble filter so my amp will handle the parallel load.  I have built a single GR OB Sub with two of the 12" servo drivers in it.  I have not measured it to say how low or how loud it is but I can say it has helped my system tremendously on the bottom end.  I'm not sure how much is added in the way of low end boost with the plate amp vs the  servo circuit, maybe Danny can chime in with that info.  I just know what I hear and the servo sub in the OB configuration is fantastic.  If I were a rap music fan or into movies with explosions then I'm sure the sealed servo subs would be best but for music the OB is great.

[sfdoddsy]

Thanks. Where do you crossover to the Servos?

Just as a matter of interest, have you tried replacing the Orion woofers completely? In other words crossing directly from the servo sub to the W22 at 120Hz?

[sl_1800]

I have not tried to replace the Orion woofers with the GR sub, could give that a try this weekend.

I'm running the Orion rumble filter on which is a 6db/oct slope at 50Hz, I have the GR sub amp set at 24/oct at 40Hz seems to give me the best match.

[rythmik]

I have Orion's and I have to engage the 50Hz rumble filter so my amp will handle the parallel load.  I have built a single GR OB Sub with two of the 12" servo drivers in it.  I have not measured it to say how low or how loud it is but I can say it has helped my system tremendously on the bottom end.  I'm not sure how much is added in the way of low end boost with the plate amp vs the  servo circuit, maybe Danny can chime in with that info.  I just know what I hear and the servo sub in the OB configuration is fantastic.  If I were a rap music fan or into movies with explosions then I'm sure the sealed servo subs would be best but for music the OB is great.

The main difference between GR OB driver and Peerless is the Q value. GR OB driver has a Q value specific for OB/IB application. The peerless drive on the other hand, has a Q value that is too low and it will limit the output from 35hz down.   Also GR OB driver is very efficient and one can actually drive up to 4 units with a A370 or A300 amps.  In the end in terms of actual damping of the response, we got the same damping as the peerless, if not better, using servo feedback. And it simply extends lower.

[sfdoddsy]

Thanks Brian. The Peerless driver does need a bunch of EQ, and even then you'll get around 80dB at 20Hz at full excursion.

However, using Linkwitz's spl_max spreadsheet, the maximum output of a pair of the OB servo drivers is only about 6dB more.

So if you like deep low music played at volume, one would still need to cross over to sealed or IB subs at 40Hz or so. Or am I missing something here?

[Danny Richie]

These will play flat to 20Hz and -3db down in the teens in open baffle and hit hard doing it.

While at RMAF, Gary Dodd informed me that measured SPL meter numbers were in the 105db range from the front row in our room when I was in control of the volume. This was with peaks and not continuous. It was also made with only the use of a 15 watt amp driving the upper section. So the subs weren't being pushed hard at all to keep up with them.

There was also a track that we played with an 18Hz fundamental note and we hit it hard with room to spare, and at in room levels that were likely in the 105db range or more.

In my room (at home) the lower end (the servo subs) has to keep up with having the upper end being driven big BIG power amps and they have never failed to keep up. And that is in a good sized room.

[zybar]

These will play flat to 20Hz and -3db down in the teens in open baffle and hit hard doing it.

While at RMAF, Gary Dodd informed me that measured SPL meter numbers were in the 105db range from the front row in our room when I was in control of the volume. This was with peaks and not continuous. It was also made with only the use of a 15 watt amp driving the upper section. So the subs weren't being pushed hard at all to keep up with them.

There was also a track that we played with an 18Hz fundamental note and we hit it hard with room to spare, and at in room levels that were likely in the 105db range or more.

In my room (at home) the lower end (the servo subs) has to keep up with having the upper end being driven big BIG power amps and they have never failed to keep up. And that is in a good sized room.

I own a pair of the OB subs with two drivers in each "cabinet" and they do indeed play into upper teens.

If I have time this weekend, I will see how loud they can get at a few frequencies from listening position (about 20' away from the subs).

George

[Davey]

Steve,

If the volume displacement of the GR OB servo drivers is "pretty much identical" to your Adire DPL12 drivers how did you calculate you'd get "only about 6db more?"  Volume displacement is directly related to SPL capability so your results should be pretty much identical.  To yield 6db more capability one driver would need to have double the Xmax or a larger Sd or a combination of those two.  I calculate a 0.697db difference between the two drivers.  (Somebody can check my math.)

Equalization is a technique to extend response, alter Q, etc, but it can not change the basic SPL capabilities of the driver.

Switching to servo drivers may have other advantages like reduction of distortion, but it won't allow you to play your system any louder......all other things being equalized. 

Cheers,

Dave.

[rythmik]

Steve,

Switching to servo drivers may have other advantages like reduction of distortion, but it won't allow you to play your system any louder......all other things being equalized. 

Cheers,

Dave.

The key to the answer is the max excursion is not a fixed value at all frequencies.  For instance, if the 300W power can push a driver to 30mm excursion at 10hz and only 15mm excursion at 20hz in simulation while the max physical excursion is only 15mm, would one want to still use 300W amp, or would one want  to use 300W/4=75W amp to keep the max excursion at 15mm so that it will stay within the physical limitation?  If we do this, the excursion at 20hz will drop to 7.5mm.  That is the limitation imposed by the physical world. Now if you use Sigfried's speadsheet, the conclusion is wrong because he hasn't considered the limitation from real world due to various Q value. If one insists and still use 300W, the consequence is he/she can really distory the driver if not careful. 

On the other hand, with servo, we can design servo drivers to have the physical parameters so that it gives us 30mm at both 10hz and 20hz with 300W power. Even though we still need to bring down the power to 75W, but both 10hz and 20h gets 15mm and that is a big improvement over the first case which is a low Q driver.  Our servo driver is high Q driver.  For nonservo, a high Q driver means there is a ringing in time domain response. Some may say it can be equalized by an EQ. Unfortunately, if that works, the ringing in metal cone midrange driver can be easily resolved. So how does servo resolve this issue?  Servo feedback effectively reduce the "equivalent" DCR to 1/3  of its physical value or even lower and that reduces the Q value by 3x or more, from say 0.9 to 0.3, or even lower. This is not done by EQ. It is done by servo feedback. As we all know the ringing won't start until 0.5 and we have plenty of margin. The ringing from the servo driver is very low if there is any.  Now if we do want ringing, we have low damping setting on our plate amp that can introduce ringing on purpose. That is still an improvement over nonservo which gets ringing from mechanical parameters. Ringing generated by electrical circuit is always more stable and consistent than the one from mechanical parameters.   So we really have a win-win solution. I know DPL also claim to be OB friendly. However, SW12-16 has only 90g of moving mass. how much lighter is that? Lower moving mass means less energy is required to push the cone and less force means less reaction force acted on the H frame, which leads to less vibration... 

   

[dvenardos]

I can't wait to get these. 
I have a flat pack for the dual OB and will be working on assembling and finishing that while saving up for the amp/speakers.

These will play flat to 20Hz and -3db down in the teens in open baffle and hit hard doing it.

While at RMAF, Gary Dodd informed me that measured SPL meter numbers were in the 105db range from the front row in our room when I was in control of the volume. This was with peaks and not continuous. It was also made with only the use of a 15 watt amp driving the upper section. So the subs weren't being pushed hard at all to keep up with them.

There was also a track that we played with an 18Hz fundamental note and we hit it hard with room to spare, and at in room levels that were likely in the 105db range or more.

In my room (at home) the lower end (the servo subs) has to keep up with having the upper end being driven big BIG power amps and they have never failed to keep up. And that is in a good sized room.

Thanks for the explanation Brian.

[TRADERXFAN]

I am working on my second one just today...  can't wait to see what 2 can do!


So you don't have the driver currently, correct?
If this is the W version, you should drill out the screw holes out before you glue your flatpack up (if its not predrilled, that is...)


-Tony

[sfdoddsy]

Steve,

If the volume displacement of the GR OB servo drivers is "pretty much identical" to your Adire DPL12 drivers how did you calculate you'd get "only about 6db more?"  Volume displacement is directly related to SPL capability so your results should be pretty much identical.  To yield 6db more capability one driver would need to have double the Xmax or a larger Sd or a combination of those two.  I calculate a 0.697db difference between the two drivers.  (Somebody can check my math.)

Equalization is a technique to extend response, alter Q, etc, but it can not change the basic SPL capabilities of the driver.

Switching to servo drivers may have other advantages like reduction of distortion, but it won't allow you to play your system any louder......all other things being equalized. 

Cheers,

Dave.

Davey,

I meant the Servo OB12s have 6dB over the 10" Peerless drivers in the original Orion. I get the same result as yourself when comparing the DPL12 and the OB12, ie virtually identical.

And I agree that EQ can't change the basic SPL capability of the driver. Nor, one would think, can servo.

Which brings me back to Danny's comment about hitting 105dB at 18Hz. According to the Linkwitz Spl_max formula (www.linkwitzlab.com/spl_max1.xls) you would need an Sd of over 10000cm2 for a driver with 16mm xmax in a 600mm wide open baffle.

In other words it would take 20 OB12s to get that sort of output.

To put this into context, according to Unibox (and spl_max) you would need 4 SW12s in a 200L sealed box to hit 105dB at 20Hz. I'm just having trouble figuring out how 4 OB12s in an open baffle can even come close.

Brian,

Thanks for the explanation. I understand how servo can help a driver and amp combination extract the most from each other, and how Q lowers the amount of EQ necessary. But neither servo nor Q manipulation can do anything for the basic mechanical limits of the driver, and it is these limits which determine the overall SPL capability as far as I know.

Am I wrong here?

I have little doubt that the quality of the bass would be top notch, but as someone who does listen to deep bass at highish volume (I like organ music and reggae ) I'm still trying to get to grips with this aspect of their performance.

[Davey]

Steve,

Ah, sorry.  Misunderstood your comparison.

I'm sure their listening room at RMAF was a long way from free-field conditions and there are plenty of other variables at work also.  However, I feel fairly confident in saying the laws of physics were not broken. 


Brian,

Max excursion is not max excursion?  Q, power requirements, etc, shouldn't factor into the maximum SPL calculation.  Only two parameters.....Sd, and Xmax.  (I'm talking about just the driver itself.)

The high-Q servo driver coupled with the appropriate servo amp does look like a very cool an interesting combination.  Next project maybe.

Cheers,

Dave.

[dvenardos]

Oh man, you just saved me a lot of grief.
Thanks. 

I am working on my second one just today...  can't wait to see what 2 can do!


So you don't have the driver currently, correct?
If this is the W version, you should drill out the screw holes out before you glue your flatpack up (if its not predrilled, that is...)


-Tony

[rythmik]

Steve,

Ah, sorry.  Misunderstood your comparison.

I'm sure their listening room at RMAF was a long way from free-field conditions and there are plenty of other variables at work also.  However, I feel fairly confident in saying the laws of physics were not broken. 


Brian,

Max excursion is not max excursion?  Q, power requirements, etc, shouldn't factor into the maximum SPL calculation.  Only two parameters.....Sd, and Xmax.  (I'm talking about just the driver itself.)

The high-Q servo driver coupled with the appropriate servo amp does look like a very cool an interesting combination.  Next project maybe.

Cheers,

Dave.

Dave,

We can take the max excursion limit example to the extreme say we want to drive 15mm excursion at 80hz. That is not practical because that amp can be 5,000W and  it will destroy the driver in a second at anything lower than 80hz. The Max Excursion is really not just a number. A lot of IB and OB drivers are plagued with two things: 1) constant bottoming (at subsonic frequencies) and 2) no sufficient passband output. My explanation connected these two together.  Actually it is quite easy to modify Siegfried's spreadsheet by adding correcting factor related to Q.

Steve,

I understand what you are saying. The 105db Danny mentioned was the peak from an actual music playing.  It is not the max from 20hz single tone output.  Although the front row position does mimic more like near field results.  I calculate that four 12" in OB will have about the same output as one 12" in sealed with the same excursion limit, which is about 101db with 14mm of one way excursion at 20hz. What can be gained from the driver itself is the SW12-16 driver has a voice coil winding width of 2.2", about 0.6" longer than SW12-4. That translates to additional output. In short, Servo does not change the physical world. To take advantage, we need to think outside the box about what parameters are most helpful in squeezing out additional output. It is all very subtle, 20% here and 20% there.... 

BTW, one single A370 can drive up to 4 SW12-16. I haven't seen anyone going to that extreme. 

Brian



 

[TomS]

...
BTW, one single A370 can drive up to 4 SW12-16. I haven't seen anyone going to that extreme. 

Brian

Brian,

So are there already 4 sets of connections for servo loops and drivers if using a single 370PEQ with 4 OB drivers?  That is, if for some crazy reason someone might want 4 OB drivers per side and only 2 amps 

Tom
 

[zybar]

...
BTW, one single A370 can drive up to 4 SW12-16. I haven't seen anyone going to that extreme. 

Brian

Brian,

So are there already 4 sets of connections for servo loops and drivers if using a single 370PEQ with 4 OB drivers?  That is, if for some crazy reason someone might want 4 OB drivers per side and only 2 amps 

Tom

To be on the safe side, I went with two amps for my four drivers.

It not only provides more power, but a lot more flexibility for setup and fine tuning.

George

[TomS]

...
BTW, one single A370 can drive up to 4 SW12-16. I haven't seen anyone going to that extreme. 

Brian

Brian,

So are there already 4 sets of connections for servo loops and drivers if using a single 370PEQ with 4 OB drivers?  That is, if for some crazy reason someone might want 4 OB drivers per side and only 2 amps 

Tom

To be on the safe side, I went with two amps for my four drivers.

It not only provides more power, but a lot more flexibility for setup and fine tuning.

George

Yeah, really just having a Tim Taylor moment...

[Davey]

Brian,

You keep wanting to bring real world limitations into the analysis.    And rightly so, and I understand.

However, there is no need to modify simple SL's spreadsheet.  He already has another spreadsheet which can provide the basic information plus much more real world, limitation, Q-related, etc, information.

http://www.linkwitzlab.com/dipole_spl_limit.htm

Cheers,

Dave.