AudioCircle
Industry Circles => GR Research => Topic started by: HAL on 6 Nov 2012, 01:16 am
-
With help from Danny , jparkhur and Dave, here is a full range driver speaker product. The pictures are of the driver and baffles from jparkhur for the Dayton PS220-8 driver as a test driver and three of the 8” GR-Research servo woofers. There is a separate cabinet for each of the new servo amp to drive the 3 servo woofers.
The reason behind the project is to see if using new impulse response correction techniques we can improve the sound of any full range driver. The Dayton 8” full range driver is a test case, as it is a very good sounding driver.
My part of the project is the supercomputer digital crossover. The idea is that I measure each Dayton full range driver and do impulse response correction and crossover for each driver. This makes logisitics interesting as each driver has to be cataloged with the data. They will be sold as pairs with the digital crossover for a specific speaker.
The digital crossover will have two different stereo output stages. One with be passively coupled from the DAC to the power amp to full range driver. This will help protect the full range driver from extended low frequency excursion. The other will be a DC coupled output stage for extended bass response to drive the servo amps.
The digital crossover and DAC’s will not use any linear phase filtering. The input will be either via asynchronous USB directly from a PC music server or via a 24bit/192KHz A/D converter. The processing rate will be 32bit/192KHz floating point math for all four channels.
The digital crossover prototype will only have A/D conversion until the full production crossover is completed. There will be software for a PC Music Server to run digital volume control and input switching between the A/D and USB inputs.
Working on final pricing on the digital crossover system and driver pair.
Will post in the thread progress as it is made along the way.
Presenting The Monolith speaker.
(http://www.audiocircle.com/image.php?id=70432)
(http://www.audiocircle.com/image.php?id=70433)
-
Can you post more pictures of the cabinets? The picture is somewhat dark.
-
Most people won't be able to afford a supercomputer...
-
Here is a better picture of the Monolith baffles and one of the amp enclosures:
(http://www.audiocircle.com/image.php?id=70435)
-
Most people won't be able to afford a supercomputer...
The DSP board that I am using is about 27 times the throughput of a Cray 1 supercomputer from back in the day. Today that amount of floating point throughput is available at what I would call a reasonable price.
-
This may help.. Jon Parkhurst
(http://www.audiocircle.com/image.php?id=70442)
-
And this... More to come... from The Monoliths......
(http://www.audiocircle.com/image.php?id=70443)
Yes= three 8's is similar to two 12's from Danny.. Order both...
Yes= this is fun and bottom end sounds great... waiting on the top to finalize... .....its like Xmas... early
-
They're multiplying! :)
The rest of the servo drivers arrived today to complete the build.
Had a death in the family and will be on travel this weekend and part of next week. More to come after that trip.
-
They're multiplying! :)
The rest of the servo drivers arrived today to complete the build.
Had a death in the family and will be on travel this weekend and part of next week. More to come after that trip.
Hope all turns out well and may the force be with you.. We will continue on for you as I will be selling the rights to Disney for 4 billion dollars....
-
The Daytons are very nice. I like them a lot more than my Fostex 206's. Great dynamics and not nearly as harsh. with the Servo 8's these should just kill.
Dave
-
Hope all turns out well and may the force be with you.. We will continue on for you as I will be selling the rights to Disney for 4 billion dollars....
That might be hard without the lightsaber! :)
-
The DSP board that I am using is about 27 times the throughput of a Cray 1 supercomputer from back in the day. Today that amount of floating point throughput is available at what I would call a reasonable price.
What is "reasonable"? That term is pretty subjective, especially in this industry :green:
Scott
-
I will tell everyone once the unit is completed and have final pricing.
-
Who designed the "crossovers"?
-
I will be designing them once I have full measurements and Danny has a chance to look at the results.
To help avoid confusion, this will be a HAL product with parts coming from all the principles listed in the original post.
Once this design is done, we can work with other full range drivers when there is interest.
-
HAL-
This has been the topic of endless debate.... but what do you see as the advantages in the full range speaker that over shadows their weaknesses?
-
Are the crossovers only sold through you?
-
HAL, interesting concept. I notice that the servos are set up on a completely open baffle along with the full range driver; is the output on the full range driver low enough (or the servos together high enough) to not require the same H-frame baffle design Danny has been using in the V1/super-v/super-7, or are you simply increasing the gain on the amp to compensate?
Also, any thoughts on isolating the vibrations in the baffle from the servos on the full range driver, as I was under the impression that with the amount of movement the servos could deliver, transmitting vibrations could be an issue?
JP, very nice work, as usual! I really like the look, and engineering the amplifier box to be a counterweight to the driver loaded baffle looks to be a most elegant looking solution.
Danny, is the output of the servos in this configuration enough to work with the top half of the V2, or would the servos need a shallow H-frame as has been discussed in other threads? I'm pretty sure if the answer was "yes", at the very least Tyson would be looking at his piggy bank with a hammer in hand. :green:
-
Has a prototype for these speakers actually been made/heard yet?
Cool concept, how tall are they?
-
This was the format for one of the Feastrex speakers at RMAF a few years ago, OB with 2 15" woofers per side and a custom digital x-over. I think it's a good option for augmenting a full range driver for sure. The Feastrex is one of the best speakers I have ever heard, but the drivers were the field coils at about $50k/pair.
Also, Omega is coming out with a line of DIY drivers... their oem drivers are as good as anything out there regardless of price. The Dayton you're using is one of the few full range drivers I haven't been able to listen to yet...
Good Luck!
-
Here is a better picture of the Monolith baffles and one of the amp enclosures:
(http://www.audiocircle.com/image.php?id=70435)
Hi HAL and all Audio Circle members.
This is a project that I will follow very closely,
because it's the kind of project that I think is close to what I want.
By the way, in the future do you think all crossovers will be done via computers ?
I hope not, for me anyway !
Guy 13
-
The DSP board that I am using is about 27 times the throughput of a Cray 1 supercomputer from back in the day. Today that amount of floating point throughput is available at what I would call a reasonable price.
Amazing vs. just not too many years ago.
The TI C67 series does the 2100 Mflops you mention for around $10 or less.
I2C so it's easy to interface to the A/D D/A chain.
Analog Devices has similar parts.
I remember doing filters with an Analog Devices Multiplier.
Stuff like Code Composer makes this a whole lot easeir.
Sounds like a very fun project.
You could reduce computational demands by pre-processing the
driver and sub audio into seperate files then maybe you could get this
into a part that already has a USB section and not have to jack
around with an A/D. A lot of ARM cores can nearly do this sort
of thing.
Just some late night rambling. Good luck.
-
Heh, well, 35 years in computing is a looong time... an iPad 4 is almost equal to 10 supercomputers by that measure: http://www.barefeats.com/ipad402.html :)
I am curious why you are using a DSP board rather than doing it in the server? (I had assumed it would be the latter.)
-
Please refer to driver properly as
full extended range. If truly full range it certainly wouldn't need (3)! woofers. My Bob Brines mass loaded transmission line FTA-2000 with a single $400+ Fostex F200A (not your typical Fostex driver for sure) is rated 27 - 20,000 Hz, 109 dB peaks in room. With Behringer DEQ2496 for baffle step and speaker/room EQ and EnABL treatments it is about as close to full range, general use speaker available (not saying they're perfect).
Lumping bass units with the extended range unit is the absolute #1 mistake made by speaker developers. In room standing waves cause massive sound pressure variations based on frequency and location (think waves in a bath tub). Have you considered a swarm/DEBRA type of solution (search here at AC) as per Floyd Toole, et al?
-
Thanks for the feedback and well wishes!
Will try to answer the questions in order as best as possible.
The advantage of this idea for a full range driver in an extended range speaker is to have a single point source with flat response for the vocal range and up. Open baffle to get rid of box coloration for the full range driver and open baffle servo's for extended low frequency response with less room loading for bass mode excitation.
The digital crossover software is capable of being run on the Music Server in realtime. The issue is getting all the D/A converter channels time aligned properly. This is still an option that can be explored, but for this product will be a stand alone crossover chassis to feed the amps and speakers.
Adding a SWARM subwoofer setup is always possible with The Monolith speaker. The concept was used at RMAF with the Super-V's last year with very good results. Keeping one set of subs in the main speaker will help with keeping the center image stable as they will be crossed over at the low end of the vocal range.
Yeah, I am old and use old references for signal processing. Worked on a project one time using a CRAY supercomputer at a facility. Cool machine!
-
Has a prototype for these speakers actually been made/heard yet?
Cool concept, how tall are they?
Yes the prototype is done and there are actually a couple out there. I have been listening to mine minus the most important part... Super computer. I am really looking forward to the digital xo
-
Cool :) What's the approx height on them?
-
Cool :) What's the approx height on them?
Good ear height sitting. Middle of driver is ~ 37 inches
-
Sounds great, thank you :)
-
HAL, interesting concept. I notice that the servos are set up on a completely open baffle along with the full range driver; is the output on the full range driver low enough (or the servos together high enough) to not require the same H-frame baffle design Danny has been using in the V1/super-v/super-7, or are you simply increasing the gain on the amp to compensate?
Also, any thoughts on isolating the vibrations in the baffle from the servos on the full range driver, as I was under the impression that with the amount of movement the servos could deliver, transmitting vibrations could be an issue?
JP, very nice work, as usual! I really like the look, and engineering the amplifier box to be a counterweight to the driver loaded baffle looks to be a most elegant looking solution.
Danny, is the output of the servos in this configuration enough to work with the top half of the V2, or would the servos need a shallow H-frame as has been discussed in other threads? I'm pretty sure if the answer was "yes", at the very least Tyson would be looking at his piggy bank with a hammer in hand. :green:
JJ-yes the top driver, currently a Dayton 8, will match up on output with the lower servos. You have all the gain control you need and as well each driver will be individually measured before setting up the SC. You will have the ability, once HAL has everything tight, to use different drivers in the top section to appeal to your preference of music/driver selection . I like this on many points, but allows me to individually mix and match my tastes easily and economically. JP
-
The Feastrex is one of the best speakers I have ever heard, but the drivers were the field coils at about $50k/pair.
Also, Omega is coming out with a line of DIY drivers... their oem drivers are as good as anything out there regardless of price. The Dayton you're using is one of the few full range drivers I haven't been able to listen to yet...
All of those can be implemented into this design.
-
Lumping bass units with the extended range unit is the absolute #1 mistake made by speaker developers. In room standing waves cause massive sound pressure variations based on frequency and location (think waves in a bath tub). Have you considered a swarm/DEBRA type of solution (search here at AC) as per Floyd Toole, et al?
Actually this implementation using open baffle servo controlled woofers is the best possible way to add low bass to a wide band driver. The output of the lower woofers cancel at 90 degrees off axis and they do not load the room like a typical woofer. The room is loaded very evenly and without variations.
The open baffle design also does away with panel resonances that typically plague horn loaded or transmission line designs. So all of that honky sound and buzzing that so many full range driver implementations have are also not present in this design.
And this is not a room EQ system. Room EQ's done in the digital realm simply do not work. You are only correcting for a single point in space when you do that. Move a little in any direction and you might as well start over. This is a correction for the driver. And since the drivers (wide band drivers) are mounted in a small open baffle then they also cancel at 90 degrees off axis. So their effect on the room can also be limited through their lower ranges. Like with all other speakers typical room treatments must be used to handle the upper ranges. They also lack the coloration and disruption of surface reflections (big baffles) that give the playing from the wall sound that really hurts imaging and sound stage depth.
Another big advantage is that this is one of the first true audiophile level (top level) DAC's to use digital correction. Prior to this most digital EQ's (like the Behringer and DEQX systems) are a real bottleneck to the concept. Some like the Behringer really suck the life out of the music.
-
Those look fantastic -- a lot of potential.
-
Are the crossovers only sold through you?
Yes they will be when completed.
-
Has a prototype for these speakers actually been made/heard yet?
Cool concept, how tall are they?
The speaker baffles you have seen are the units for testing.
The digital crossover prototype hardware is in work. Waiting on the first unit with the USB audio input.
The concepts we are using have been around for awhile, these are new software tools for trying the techniques.
The speaker measurements will be made at 24bit/192KHz. The crossover signal processing will work at that rate.
-
Sounds interesting. What kind of timeline are we looking at until completeion Hal?
-
I am waiting on news of the timeline for the crossover board shipment. It is out of my hands at the moment. Will keep everyone posted when I know.
-
The digital crossover software is capable of being run on the Music Server in realtime. The issue is getting all the D/A converter channels time aligned properly. This is still an option that can be explored, but for this product will be a stand alone crossover chassis to feed the amps and speakers.
The judicious use of threads goes a long ways toward helping with that problem.
Assuming your have some reasonable facsimile of a RTOS (not Windows).
Think latency. You only have to be real time after you fill up the buffers.
You may have considered this but 2 or 3 seconds to spool up can give
you a lot of preprocessing to handle the synchronization issues.
I had to deal with some of these issues when I worked on a DLP project
at TI. If old hat, excuse the redundant information.
-
No RTOS for this application. To make is easy for people it will be running on Windows OS. So far Windows 7 x64 has done well. Need to try Windows 8 x64 to see how it works out.
The development software also runs on Windows OS.
-
If you don't load it down with crap, the W8 kernel should be more deterministic.
If you're on Windows, big buffers will further add to predictability.
I think Mac OSX is probably better but that is a issue to some.
However a true RTOS is best but understand the motivations.
If you're doing a D/A card anyway, a FPGA or DSP would be great so
you can hide your secret sauce. :wink:
-
Wired and playing the last two days. Just waiting for the breakin to finish
(http://www.audiocircle.com/image.php?id=70551)
[/img]
-
Look great. :thumb:
-
Hi jparkhur and all Audio Circle members.
Anxious to read your comments...
Very nice looking speakers.
Guy 13
-
Morning, We still have to wait on the major puzzle piece as HAL has stated.. The xo is the most important part here...flattening the top end and putting in "Best".....
Jon P
-
Looking good Jon!
Will get my drivers hooked up and burning in on Friday.
The wait for the XO is ongoing. It is in work at this time. Waiting on an update.
-
This looks interesting. I've heard a few single/"full range" driver speakers in my time and they do some things quite well. The trade offs have always always prevented them from serious consideration.
Reading this thread, along with some of the others currently active, has made me ask a few questions that I'm not quite sure how to ask. These questions concern the general design philosophy of speakers, single driver speakers, and the strengths and disadvantages of these various approaches. Some of them may sound critical, but I assure you they are not - I am merely seeking clarification and a better understanding.
Is it OK to ask here, or should I start a new thread? If so, should I go to a different area? (Enclosures and The Lab come to mind)?
-
If Danny is good with it, I would say post the questions here. Should be an interesting discussion.
-
This looks interesting. I've heard a few single/"full range" driver speakers in my time and they do some things quite well. The trade offs have always always prevented them from serious consideration.
Man, I feel the same way. There were just too many compromises or short comings. But I think this approach (or this solution) will remedy the short comings and even excel in those areas where the wide band drivers fall short.
Reading this thread, along with some of the others currently active, has made me ask a few questions that I'm not quite sure how to ask. These questions concern the general design philosophy of speakers, single driver speakers, and the strengths and disadvantages of these various approaches. Some of them may sound critical, but I assure you they are not - I am merely seeking clarification and a better understanding.
Is it OK to ask here, or should I start a new thread? If so, should I go to a different area? (Enclosures and The Lab come to mind)?
Go ahead and fire away.
-
OK, here goes:
First, (and this is the one I'm most hesitant to ask as it may sound negative) - What makes this a ""Full Range Driver"?. It seems closer to a 2 way design with a driver that covers a wider range for the most critical region (say somewhere in the 150/200 Hz range on up) and very good bass drivers, with a very sophisticated crossover/digital system handling the separation between the drivers. :scratch:
Another of the issues being raised pertains to cabinet vibrations when placing a speaker on top of a subwoofer causing unwanted vibrations which in turn introduces problems to the soundstage.
So, question number 2: Why is this worse then tying them together in the same frame? Aren't vibrations vibrations?
As far as a single driver full range speaker goes: I can remember going into the Emerald Physics room at one of the shows (think it was T.H.E Show Newport) and they were playing a simple piece - Iz singing Over the rainbow. Just his clear voice and a Ukelele. I was astounded by the quality and how natural it sounded (conversely: came home and threw it on my setup and realized how far I had to go - mind you, it's better since then :green:).
But then they threw on a different piece with more range and dynamics........ left the room very quickly thereafter.
What is the frequency range that should be considered anyway" I know that 20-20k is usually the standard that is used.
I know for a fact that I can no longer "Hear" above 15k, and when my tinnitus/ringing in my ears is acting up (somewhere around 10-13k), I doubt I can hear anything above that.
Yet I KNOW I can sense (feel) anything above 0 Hz - from earthquakes to subsonic things (rocket lifting off), and that if a speaker doesn't go above what I can actually hear, things sound wrong. The decay of cymbals being the one I notice most easily.
So, next question: What is the upper limit on this design? (I figure the "subwoofer section of this speaker will be very good :green:)
I've got more questions, but I think I can tailor them better after hearing back on these. :thumb:
-
This looks interesting. I've heard a few single/"full range" driver speakers in my time and they do some things quite well. The trade offs have always always prevented them from serious consideration.
Same goes for 95% of multi-way speakers. :green:
-
I'll try to hit this with one thought at a time.
First, (and this is the one I'm most hesitant to ask as it may sound negative) - What makes this a ""Full Range Driver"?. It seems closer to a 2 way design with a driver that covers a wider range for the most critical region (say somewhere in the 150/200 Hz range on up) and very good bass drivers, with a very sophisticated crossover/digital system handling the separation between the drivers.
What we are doing really is making a two way design out of it. But we are still using this driver as a wide band driver just like everyone else. We are just supplementing it in an area below what it is capable of playing anyway.
The real key though is what the digital correct will do for the response. Just have a look at the factory specs on the driver that we choose to test this with: http://www.parts-express.com/pedocs/brochures/295-346-dayton-audio-ps220-8-brochure-9530.pdf
It's response has a 14db swing from 400Hz to 4kHz. And when mounted in a narrow baffle the baffle step loss will further increase that as output levels below 6 or 700 hundred Hz really drops off.
Now what if the entire range was +/-.5db or less?
It could go from being a speaker that burns a hole in your head to a very accurate speaker.
-
Another of the issues being raised pertains to cabinet vibrations when placing a speaker on top of a subwoofer causing unwanted vibrations which in turn introduces problems to the soundstage.
So, question number 2: Why is this worse then tying them together in the same frame? Aren't vibrations vibrations?
Imagine mounting a speaker over a large drum. Now imagine the difference if it were mounted to the frame of the drum verses mounted on the playing surface of a drum.
It is not hard to envision the differences between this:
(http://gr-research.com/images/av4gray3.jpg)
And this:
(http://www.audiocircle.com/image.php?id=70180)
Now see the before and after. Before:
(http://www.audiocircle.com/image.php?id=69668)
And after:
(http://www.audiocircle.com/image.php?id=70234)
Comments from the owner:
Well I removed the AV-2's from the subs and placed them back on the stands I made when I built the speakers. I listened to the same pieces that I had listened to before and I could not believe the difference. Before I used the 1/4" rods the anchor the AV-2's to subs as I tried Spikes but the AV-2's would rattle on the spikes during some low passages so I firmly anchored them with the 1/4" bolts with brass tubing to make it look finished.
Danny! you were right and I just cannot believe what that little difference makes.
Panel resonances are not your friend. They exists in all speaker to some degree. Minimizing them in any way possible is a huge plus.
-
Lumping these thoughts together...
As far as a single driver full range speaker goes: I can remember going into the Emerald Physics room at one of the shows (think it was T.H.E Show Newport) and they were playing a simple piece - Iz singing Over the rainbow. Just his clear voice and a Ukelele. I was astounded by the quality and how natural it sounded (conversely: came home and threw it on my setup and realized how far I had to go - mind you, it's better since then ).
But then they threw on a different piece with more range and dynamics........ left the room very quickly thereafter.
What is the frequency range that should be considered anyway" I know that 20-20k is usually the standard that is used.
I know for a fact that I can no longer "Hear" above 15k, and when my tinnitus/ringing in my ears is acting up (somewhere around 10-13k), I doubt I can hear anything above that.
Yet I KNOW I can sense (feel) anything above 0 Hz - from earthquakes to subsonic things (rocket lifting off), and that if a speaker doesn't go above what I can actually hear, things sound wrong. The decay of cymbals being the one I notice most easily.
So, next question: What is the upper limit on this design? (I figure the "subwoofer section of this speaker will be very good )
I've got more questions, but I think I can tailor them better after hearing back on these.
Some full range drivers do some ranges really well. And many people love them just for those vocal ranges, and that is all they want to listen to. They don't care that the top end is really hot, or that they have no bass.
Some people can't get past those things and their favorite music may cause the shortcomings to just be too glaring.
If all of this works out like I think it will then these are going to be real hard to knock. There will literally be no shortcomings. Even if the customer picks a favorite full range driver that rolls off some above 15kHz or so. That too can be gained back to some degree as those ranges will simply see more output to increase the output where it was naturally rolling off. So if a driver has some reasonable output to 20kHz then it could actually be corrected to be flat to 20kHz. And the peaks that a driver may have up top can be attenuated to a much lower level. So those problems can go away as well.
-
If all of this works out like I think it will then these are going to be real hard to knock. There will literally be no shortcomings.
My biggest issue with a single driver (assuming it is decent to begin with) is related to SPL and transient impact. The maximum SPLs are rather limited and there isn't enough impact in the mid bass. I don't see any design that truly uses one single driver overcoming these limitations, even with subwoofers. There just isn't enough surface area....
-
My biggest issue with a single driver (assuming it is decent to begin with) is related to SPL and transient impact. The maximum SPLs are rather limited and there isn't enough impact in the mid bass. I don't see any design that truly uses one single driver overcoming these limitations, even with subwoofers. There just isn't enough surface area....
That's true, but in this case the mid-bass is being covered by three 8" woofers with servo control.
-
That's true, but in this case the mid-bass is being covered by three 8" woofers with servo control.
Whats the x-over freq looking like? Didn't look like those drivers would extend up into the mid-bass but if so maybe that will work out... of course then you have more integration issues like with a conventional speaker as well vs. using subwoofers at ~60 Hz and lower which aren't too hard to integrate.
-
Imagine mounting a speaker over a large drum. Now imagine the difference if it were mounted to the frame of the drum verses mounted on the playing surface of a drum.
It is not hard to envision the differences between this:
(http://gr-research.com/images/av4gray3.jpg)
And this:
(http://www.audiocircle.com/image.php?id=70180)
Now see the before and after. Before:
(http://www.audiocircle.com/image.php?id=69668)
And after:
(http://www.audiocircle.com/image.php?id=70234)
Comments from the owner:
Panel resonances are not your friend. They exists in all speaker to some degree. Minimizing them in any way possible is a huge plus.
Maybe I need to rephrase and clarify the purpose of my question.
I can easily see by example of setting a speaker on the surface of a drum is a bad thing. I would never consider doing so. If you were to place a speaker on top of a truly inert subwoofer cabinet (I'm thinking of something similar to your sand-filled cabinet plans that I've seen posted), it seems that there would be very little, if any, vibrations present.
However, if I take a piece of wood or MDF and support it on either side with another piece of wood or MDF (the support columns in this case), and then mount 4 drivers (3 Subwoofers and the full range driver) in that piece of wood or MDF, it seems that there is a greater potential for vibrations unless additional bracing is used.
(not sure if I'm painting a clear, or for that matter accurate, picture :oops:)
I guess another way of putting it is that the piture I see of the test enclosure would be somewhat inclined to vibrate w/o some additional bracing (which I readily admit could be a completely wrong conclusion without knowing more info)
-
I agree the pic at the bottom is far more ideal but there's a lot of people out there that want a full range loudspeaker, think bigger is better, and don't want to place subwoofers. It's not like subs are a new idea, but I think many people are only starting to see that bass coming from symmetrically set-up speakers in a rectangular room just doesn't work out so well. Certainly better with the dipole design of many GR products, but I think many people are going to start preferring smaller main speakers with multiple subs, it is technically a better solution and doesn't require huge and expensive main speakers.
-
I guess another way of putting it is that the piture I see of the test enclosure would be somewhat inclined to vibrate w/o some additional bracing (which I readily admit could be a completely wrong conclusion without knowing more info)
This one is 1.5" thick with a brace down each side. And a large section of the baffle has four large holes cut through it and drivers mounted in it.
(http://www.audiocircle.com/image.php?id=70551)
So what you really end up with is a frame work as much as a baffle. There really isn't enough baffle left for there to have any resonance issues. If it were a solid baffle with no holes then it would be pretty easy to start a resonance at some frequency range relative to its size.
So the important issue then really becomes stability and anchoring the speaker well with floor spikes so that the speaker isn't allowed to rock as it would if just placed on a carpeted floor. Of coarse this is a common issue with any speaker and not just this one.
I think many people are going to start preferring smaller main speakers with multiple subs, it is technically a better solution and doesn't require huge and expensive main speakers.
This is very true.
-
My biggest issue with a single driver (assuming it is decent to begin with) is related to SPL and transient impact. The maximum SPLs are rather limited and there isn't enough impact in the mid bass. I don't see any design that truly uses one single driver overcoming these limitations, even with subwoofers. There just isn't enough surface area....
This is also a concern for me. Placed in a large room and playing loud, a full range driver just doesn't cut it. The x mas looks pretty decent on the full range. How will these play at high SPLs?
-
This is also a concern for me. Placed in a large room and playing loud, a full range driver just doesn't cut it. The x mas looks pretty decent on the full range. How will these play at high SPLs?
Fellows when you take away the burden of a driver to have to cover 200Hz and down then you are taking away 90% of the burden. When 200Hz and down is taken away there is no concern of reaching the drivers X-Max. Its limitation will be thermal and not mechanical.
-
The reason many like full range drivers is because of the lack of crossovers or minimally the movement of the crossover between drivers out of the midrange. Hence crossovers 200Hz or lower on the low end with subs or even on the high end >7 kHz with supertweeters where they are less noticeable.
-
It sounds like this involves some heavy digital EQing. How does this effect phase?
-
Fellows when you take away the burden of a driver to have to cover 200Hz and down then you are taking away 90% of the burden. When 200Hz and down is taken away there is no concern of reaching the drivers X-Max. Its limitation will be thermal and not mechanical.
Wow, you're not joking when you say the new 8" handles the mids when crossed at 200Hz. If you already mentioned being crossed over at 200Hz, then sorry for the last post.
-
It sounds like this involves some heavy digital EQing. How does this effect phase?
This is not EQ. This is impulse response correction. A totally different idea for driver response correction, and phase correction is part of it.
-
I like full range speakers.
Heard them, really like them, Sam.
but, I prefer to have eggs and ham.
:D
:green:
-
I don't like full range speakers.
Heard them, don't like them, Sam.
I prefer to have eggs and ham.
:D
It appears you didn't read the entire book. He likes them at the end. This coming from someone who is reading as I type. Rather listening to it. ;)
-
I don't like full range speakers.
Heard them, don't like them, Sam.
I prefer to have eggs and ham.
:D
So why are you here? :lol:
Do you also go to country bars and tell the othe patrons you don't like country music?!?
I myself like them, but they have their place. And this is an interesting design that looks like it could improve the low frequency extension.
-
This is a very cool project.
How many amps are needed to run this setup? Will it be offered as a kit someday?
-
You will need one stereo amp for the two full range drivers. The three servo drivers per side need one of the special servo amps that Danny sells.
-
I don't like full range speakers.
Heard them, don't like them, Sam.
I prefer to have eggs and ham.
:D
Those N3 TL towers of yours are full range speakers IMO. To me, 25 to 30 hz is effectively full range for music and the TL's do it.
-
[/quote]You will need one stereo amp for the two full range drivers. The three servo drivers per side need one of the special servo amps that Danny sells.
Cost is almost the same for two amps and two 12 inch as compared to three 8 s and two amps.
-
Question that came to mind regarding the digital crossover - will this tech be able to adjust the input somehow to compensate for the dispersion pattern of the full range driver in the upper frequencies? In other words, in addition to the rise in the high frequency response that will be corrected out, will this technology be able to compensate in some fashion for the higher frequencies tendency to beam, relative to the dispersion pattern of the lower frequencies the full range driver plays?. . . or is this still in the realm of the driver designer to address physically alone?
-
Question that came to mind regarding the digital crossover - will this tech be able to adjust the input somehow to compensate for the dispersion pattern of the full range driver in the upper frequencies? In other words, in addition to the rise in the high frequency response that will be corrected out, will this technology be able to compensate in some fashion for the higher frequencies tendency to beam, relative to the dispersion pattern of the lower frequencies the full range driver plays?. . . or is this still in the realm of the driver designer to address physically alone?
Loss of off axis response or beaming effects are a direct effect of the size of the diaphragm. There is no changing it with the input signal.
-
Will the digital crossover/correction also feed the servo subs? Does that mean you won't use the eq in the sub amps, and have complete eq control over the bass too?
-
Will the digital crossover/correction also feed the servo subs? Does that mean you won't use the eq in the sub amps, and have complete eq control over the bass too?
All of the digital correction will be for the full range driver only. Correction will be unnecessary for the lower woofers.
-
The digital correction would only be for the full range driver.
The second stereo XO channels would be to feed the servo subs with time delay. This would insure the time delay through the XO to both the sub and full range driver is the same. The servo amp would operate as normal with all the controls.
Ok, I just cannot type fast enough. Danny beat me! :)
-
Thanks for the answers :)
-
I like full range speakers.
Heard them, really like them, Sam.
but, I prefer to have eggs and ham.
:D
Does it read better now? :thumb:
-
Hi Danny and all Audio Circle members.
I don't remember where I got this sketch,
but what do you think about it?
Do you think it will work taming the Dayton PS220-8?
Guy 13
(http://www.audiocircle.com/image.php?id=70589)
-
TT asked for this I think.. to see proportions.. That is a MFW 15 behind it.. This shoudl help.. Small foot print. Big on sound!
Jon Parkhurst
(http://www.audiocircle.com/image.php?id=70592)
-
Hi all Audio Circle members.
Small it is, but will it be BIG on sound?
Sure hope so !
Guy 13
-
Hi Danny and all Audio Circle members.
I don't remember where I got this sketch,
but what do you think about it?
Do you think it will work taming the Dayton PS220-8?
Guy 13
(http://www.audiocircle.com/image.php?id=70589)
The first part of the circuit is for baffle step correction. The value of the inductor depends on the baffle width and the resistor level depends on how hot the top end is. Typically inductor values around 1.0mH to 1.5mH are used, and resistor values between 3 ohms and 5 ohms. A 12 ohm resistor presents a huge amount of attenuation of the top end.
The LCR or notch filter is used to reduce a peak in a give range. The values used will attenuate a very narrow range centered at 3.5kHz. That is right where that driver has a pretty big peak. So that's good.
-
I like the whole idea regarding this speaker, in that, experimenting with new ideas pushes the state of this art/craft/industry. :thumb:
Woofer questions.
Why the U-shaped cabinet for the dipole woofers vs. the typical H- shaped cabinet that seems to be preferred?
How will the sound be different 1 vs the other?
How was the length of the legs in the U determined to be appropriate for this setup?
Lastly, in the V series the front of the H frame is in the same plane as the coax drivers where in this setup, the plane for the drivers they are bolted to is the same, how is time alignment affected by this?
-
Mechanical time alignment can be done with signal time delays in the digital crossover. The time step size is 5.2 microseconds. This means that the signal can be electrically altered in 0.071 inch steps between the full range driver and servo woofers.
You can align them just about anywhere you like electrically with the time delay capability.
-
Mechanical time alignment can be done with signal time delays in the digital crossover. The time step size is 5.2 microseconds. This means that the signal can be electrically altered in 0.071 inch steps between the full range driver and servo woofers.
You can align them just about anywhere you like electrically with the time delay capability.
I understand.
Does your plan include this alignment for this driver above this woofer in its current state, or will it be available to the end user to calculate and make the adjustment in their environment?
I personally believe that it should be done in the crossover because all the parameters for each driver, woofer, and "box" are known at the time the crossover is being designed. Am I correct?
-
Yes, it will be time aligned by me in the XO before shipping.
It will be plug-in and listen.
-
The new driver burn-in system is running with the two Dayton drivers.
Will give them a few day's of work out and start some measurements.
-
looking forward to hearing more.
too bad we have to wait...
-
Yes, I am in the same boat as everyone.
Waiting is always the hardest part! :(
-
Mechanical time alignment can be done with signal time delays in the digital crossover. The time step size is 5.2 microseconds. This means that the signal can be electrically altered in 0.071 inch steps between the full range driver and servo woofers.
You can align them just about anywhere you like electrically with the time delay capability.
I assume that you have factored in the additional processing latency on the full range so when this offset is
zero you automatically add in the correct sample delay so the sub output is delayed to coincide with the
full range calculation output.
Ideally your calculation chain on the full range would take some multiple of 1/192KHz to complete.
Just saying.
-
They do look pretty sweet. Keep it rollin' fellas.
-
I assume that you have factored in the additional processing latency on the full range so when this offset is
zero you automatically add in the correct sample delay so the sub output is delayed to coincide with the
full range calculation output.
Ideally your calculation chain on the full range would take some multiple of 1/192KHz to complete.
Just saying.
Yes,
The software development code lets me benchmark the times in the processing chains and then I can apply the correct offset for the other channels.
-
This project is interesting as I have been thinking about playing around with OB. Have a couple of questions.
The Qts of this driver is listed as .30. I was under the impression that higher Qts of .6-.8 was better for OB/
How much does he circuit posted by Guy 13 change the sensitivity of the driver?
-
This project is interesting as I have been thinking about playing around with OB. Have a couple of questions.
The Qts of this driver is listed as .30. I was under the impression that higher Qts of .6-.8 was better for OB/
Higher QTS values are important if you are trying to get bass from an OB driver. Since this speaker is using the servo subs to generate bass there is little need for the Dayton driver to have a high QTS.
Best,
Ed
-
This project is interesting as I have been thinking about playing around with OB. Have a couple of questions.
The Qts of this driver is listed as .30. I was under the impression that higher Qts of .6-.8 was better for OB/
How much does he circuit posted by Guy 13 change the sensitivity of the driver?
A higher Qts is more ideal, but not an issue in this application since the full range driver won't see any really low frequency information.
-
A higher Qts is more ideal, but not an issue in this application since the full range driver won't see any really low frequency information.
If you aren't sending any low frequencies to the full range, why use an 8"? Aren't there 4-5" drivers out there that would image better and react faster than an eight?
-
Scott,
One of the ideas is to start with the Dayton full range driver and then try other units.
jparkhur is making inserts for the 8" mounting hole that will let smaller drivers be adapted to the baffle.
Since I have to measure every driver to do this, the concept lends itself to being flexible.
-
I was wondering that as well, but had decided to just wait until we had more measurements. Sounds like it could be amazingly versatile.
-
As everyone has certain drivers they like, the master plan is to allow the end user to pick and choose the upper driver themselves. The Dsyton 8, as has been said, shows a good starting point and proof of concept. You could choose to have an 8 inch hole cut or stick with one specific to your driver. I made blanks for myself and have put in a Fostex FF125WKeN and 85WKeN, the later has great top end but in OB can't really get down far enough. The 125 is good and I threw my TB. 1808 in also. This has a higher Q to start with but still has 14 db of play at the top that would need to be taken care if. Think of it as the Sizzler, you pick all the pieces and make the best loudspeaker for yourself. As well, it will be flat across the board :)
-
It's going to be hard enough to get an 8" driver to play down to 200Hz in an open baffle. I am sure it will require some gain in the 400Hz and down area to get there. A smaller driver would be even tougher.
-
:duh: I should have thought about that... open baffle
-
The Dayton 8" drivers should be about ready to try some measurements. Got them running in a bit more now and then will get them mounted in the baffle.
The new combo A/D and D/A arrived and is setup for measurements.
Will use the techniques my speaker measurement teachers have taught me and see what happens! :)
-
After doing some discussions with the software tool developer, from a comment in this thread, it maybe possible to just use the MS-1 Music Server as the digital crossover system for The Monolith speaker. This does mean a sophisticated four channel USB DAC setup, but that is being worked for trials. This will be a very good DAC system as well. This all has to be tested, but it is in work.
This idea would be for a PC music server only operation with a customers stereo amp and The Monolith speakers. The 8" subs will need the new servo amps from Danny.
Is there any interest in a system like this? If an analog input is needed, then the dedicated digital crossover system is needed.
Please post your thoughts.
-
This idea would be for a PC music server only operation with a customers stereo amp and The Monolith speakers.
This is a dumb question since it's stated, but this won't work on Macs?
-
The software runtime tools do not run on a MAC.
We are working on a possible Foobar2000 interface. It does not run on a MAC.
-
After doing some discussions with the software tool developer, from a comment in this thread, it maybe possible to just use the MS-1 Music Server as the digital crossover system for The Monolith speaker. This does mean a sophisticated four channel USB DAC setup, but that is being worked for trials. This will be a very good DAC system as well. This all has to be tested, but it is in work.
This idea would be for a PC music server only operation with a customers stereo amp and The Monolith speakers. The 8" subs will need the new servo amps from Danny.
Is there any interest in a system like this? If an analog input is needed, then the dedicated digital crossover system is needed.
Please post your thoughts.
Anyone have any feedback on the idea?
-
That should be an all digital system - no analog input - keep things simple and costs low ...
-
Back to previous questions. Why the U shaped baffle instead of an H shaped baffle? Is the U shaped baffle as good as an H shaped baffle? Are the sides of the U shaped baffle the correct size for these woofers?
-
Back to previous questions. Why the U shaped baffle instead of an H shaped baffle? Is the U shaped baffle as good as an H shaped baffle? Are the sides of the U shaped baffle the correct size for these woofers?
For good low end output the H frame is better. In fact the U shape can't really be very deep. If it gets too deep then it gets hard to support the sides and the sides will tend to resonant.
But the number one problem with an H frame is that everyone wants to see all the drivers on one plane. I don't know how many times I have been asked if it would be okay to build a V-1 or Super-V with the drivers on one plane.
This speaker will have the look that everyone wants.
-
Look who has the drivers on Black Friday Sale.
http://www.parts-express.com/term/black_friday_2012?pg=1
$94 each
-
That is what I payed for each of mine from PE. Got these a few weeks ago.
-
But the number one problem with an H frame is that everyone wants to see all the drivers on one plane. I don't know how many times I have been asked if it would be okay to build a V-1 or Super-V with the drivers on one plane.
This speaker will have the look that everyone wants.
guilty.
-
For good low end output the H frame is better. In fact the U shape can't really be very deep. If it gets too deep then it gets hard to support the sides and the sides will tend to resonant.
But the number one problem with an H frame is that everyone wants to see all the drivers on one plane. I don't know how many times I have been asked if it would be okay to build a V-1 or Super-V with the drivers on one plane.
This speaker will have the look that everyone wants.
That is why they invented grill cloth.
Dave
-
^^^^^
What he said! :)
-
Truth of the matter is: speaker that looks good will probably out sell a speaker that sounds good. Fantastic design and great photographs can sell $10k speakers everyday (jamo 909).
-
By a very strange coincidence the German Klang + Ton DIY magazine presents a bassreflex speaker with the Dayton PS-220-8 in its September number.
The constructors choose the same path as I in their correction of the frequency response a BSC and a Notch filter. Contrasting my approach
they choose to put the notch in the signal path with paralleled inductor, capacitor and resistor. As I simulated my circuit I also did so with the German constructors':
First the effect of the Klang + Ton circuit as simulated:
(http://img845.imageshack.us/img845/5857/lcserialnotchfordaytonp.jpg)
Then the simulation of my circuit:
(http://img690.imageshack.us/img690/6792/lcparallellnotchfordayt.jpg)
Please observe the difference in scale.
The Germans have worked with real units, I made my circuit from the published Specification Sheet.
And last, to please Richard, Klang + Ton did think the the Dayton speaker sound better with a Tube amp. :)
/Erling
I constructed the correction circuit mentioned and pictured above for Richard and there is a thread over at the Open Baffle circle: http://www.audiocircle.com/index.php?topic=107397.0
It was made only from specs and published frequency curves so it is a bit of a theoretical excersize. If someone is testing it please publish impressions.
Regarding Danny's objections I think my simulation answers his remarks. With Open Baffles there is no baffle width effect. I could recommend 'The Edge' program for simulating
baffle effects beeing it OB or box speakers. It is a freeware program downloadable here: http://www.tolvan.com/index.php?page=/edge/edge.php , certainly very useful in all design processes.
/Erling
-
Parts are arriving today to build the master clock splitter and drive circuit to drive two DAC's simultaneously. Will be breadboarding it after a trip to the Post Office.
The baffles are in the work room and hope to install drivers today. After that some measurements to see what the driver response looks like.
-
(http://www.audiocircle.com/image.php?id=71489)
-
Jon,
Looking good!
I started installing drivers this evening. Should be done by this weekend.
-
(http://www.audiocircle.com/image.php?id=71489)
Jon, looks great! Has the frequency response started to even out for you? I noticed on mine that the 3K-4K peak started to mellow with time and actually came down about 1.5dB with just a long break in.
I have both generations of these drivers and the second definitely is smoother than the first.
Dave
-
D- yes they have.. I am waiting for HAL/R to send me the measurements on the driver out of the box OB. I would like to see a real room response curve with his tools.
I am liking them, but they will get better after the XO is done...
Jon P
-
The final parts to build the dual USB DAC digital XO have arrived. Will be working on finishing the pulse preamp splitter this weekend. I have a MS-2 prototype Music Server up and running. Will see if the software likes Windows 8.
Driver installation is ongoing. Should get some time this weekend to do some far field measurements on the Dayton drivers. They should be pretty much ready to go.
I am going very slowly on the driver install. My ususal install method usually has a hole in a driver by the time I am finished. Did I mention I have 10 left thumbs when it comes to mechanical installations.
-
Did I mention I have 10 left thumbs when it comes to mechanical installations.
:D Hey Hal...they have a class library for that!
That's an old tired software cliche guys. 8)
-
LOL! :thumb:
-
Finished the prototype of the dual clock driver circuit. The clock driver circuit logic level output for both channels are correct driving 75 ohm loads.
Once I get the O'Scope warmed up will see how the 24MHz clock looks into the 75 ohm loads. The Evo clock is here and will get that measured as well. If that is all good, then we have a dual USB to I2S Bus driver for the digital crossover. The first I2S Bus DAC is working. The second I2S Bus DAC is being built for the setup.
More driver installation today.
-
Here is the front end for the two DAC's to run the speakers.
M2Tech Evo Clock and two Evo's. The Evo I2S Bus outputs will drive two separate DAC's.
(http://www.audiocircle.com/image.php?id=71771)
Putting the clock splitter in a chassis with 75 Ohm BNC connectors to tie the system together.
Both Evo's will be driven from the MS-2 prototype Music Server.
-
Hal, did you also consider using a multichannel DAC? This would be simpler than using a clock splitter. The exaSound e18 is an 8 channel DAC designed for audiophiles. It will handle up to 384 kHz and DSD. The best thing is that you can use the additional channels for more subwoofers, if desired.
(http://www.exasound.com/portals/0/Images/exaSound/side-view.jpg)
(http://www.exasound.com/portals/0/Images/exaSound/back-view.jpg)
-
That would be easier for implementation.
The DAC's I have been working on do not use the ESS DAC chip. I do not like the way the ESS DAC's sound.
Will let everyone hear this idea and decide for themselves.
-
Way late to the game, but I'll add my two cents.
"Full range driver" is a misnomer, as no driver is truly full range. For that reason, folks have started to refer to them as "wide-banders."
The advantage of wide-banders, as other have alluded to, is that like coaxial drivers, they approximate a point source. They also tend to be punchy, with good transient response, and relatively easy to drive. The down side (in addition to their not truly being full-range) is that they often have annoying peaks (e.g., the Lowther "shout").
Recognizing that there is only so much bass output one can wrestle from them, some manufactuers are now adding woofers below them; I refer to this arrrangement as "augmented wide bander." One of the difficulties of this arrangement is finding a woofer that will "keep up" with the wide-bander; I daresay Danny's servo subs are an excellent choice.
As for the wide-banders themselves, not all are created equal. Some are awful, some are superb (the Voxtive being the best I've heard by a long shot), and most fall somewhere in between those extremes. In my experience, many speakers (not just those based on wide-banders) suffer due to the choice of driver. Certainly a good cross-over and well-designed cabinet (or lack thereof, i.e., open baffle) can optimize the performance, but one cannot make a silk purse out of a sow's ear. I am very intrigued by Rich's approach to digitally correct certain characteristics of the Dayton, though I have no experience with this driver, nor with the correction.
I recently reviewed a very well designed "augmented widebander" which uses a Tangband on top, and six 10" woofers below in a diploe, slot-loaded configuration. (http://www.dagogo.com/Print-Article.asp?hArticle=1115) The dipole configuration necessitated a separate amp and Eq but IMO, this is actually a benefit, not a deficit. Danny has taken this approach for some time now, and I am continually surprised at how few designers have caught on to the benefits.
I am delighted that Rich, Danny and their colleagues are building the Monolith, as the more such speakers there are on the market, the better for the consumers (i.e., us). I especially applaud Rich for the novel elements of his design, and eagerly await the opportunity hear it. Perhaps if I'm nice to him, he'll let me review it. :)
Regards,
Larry
-
Larry,
The cool thing about this design, that Rich is working on, is that any wide band driver can be used. The customer can choose anything they want. The Dayton wide band driver was just chosen as an inexpensive test unit to prove the concept.
-
Danny:
Very impressive, except that you've left the choices to neurotic, "can't make up my mind" audiophiles. :)
But won't the cross-over points (and perhaps slopes), notch filters etc. have to be optimized for each driver?
Larry
-
Once things progress, each speaker and pair will be measured individually and the electronics set to those drivers. Yes.. very accurate and specific.
Jon
-
Aha, so it's not just the impulse that is being customized (for want of a better word), but the x-over as well. Not to rain on your parade, but as you undoubtedly know - and as Danny can certainly attest to - measurements are a good starting point for x-overs but the final arbiter must be listening, which is a slow process.
My suggestion (not that you asked :) ) is to identify the best driver (as defined by sound, price, etc.) and optimize the speaker accordingly.
Larry
-
Yeah Larry, regardless of what driver they choose the phase and frequency response is corrected to a very high level. And any of those drivers can have amplitude correction allowing a crossover point to the lower woofers to be right at 200Hz.
-
Thanks!
-
That would be easier for implementation.
The DAC's I have been working on do not use the ESS DAC chip. I do not like the way the ESS DAC's sound.
Will let everyone hear this idea and decide for themselves.
I realize this is off topic, but I just saw that there is a new review (http://www.audiocircle.com/index.php?topic=112169.0) here at Audiocircle of an exaSound E20 DAC. It is the two channel version of what I posted earlier.
-
Intersting, but I think we can do better for $2500.00 for the DAC's. :D
-
Intersting, but I think we can do better for $2500.00 for the DAC's.
obvious next question is: what's the list of candidates?
-
Something I built. This is going to be fun! :D
-
My pair of The Monolith's are complete:
(http://www.audiocircle.com/image.php?id=72400)
Here is a set of measurements of the Dayton 8" driver on-axis far-field installed in the baffle as in the picture.
(http://www.audiocircle.com/image.php?id=72402)
Next onto the digital crossover work.
-
Hi HAL and all Audio Circle members.
Even if you are not a professional photographer,
I am sure you can do a better picture than this one.
This picture does not do justice to your beautiful work of art.
Please try again and at the same time,
can we also see the back side?
I would sure like to compare side by side your speakers with my GR Research V1.
Guy 13
-
Guy,
Where the speakers are located is dark and hard to get a good picture even with a flash. That picture had to be corrected to look that bright. The better pictures are earlier in the thread that Jon and I took.
The Monolith back is simply raw drivers coming out with connectors ready for wiring.
If I take them outside for more measurements, will try taking more pictures.
-
Guy,
Where the speakers are located is dark and hard to get a good picture even with a flash. That picture had to be corrected to look that bright. The better pictures are earlier in the thread that Jon and I took.
The Monolith back is simply raw drivers coming out with connectors ready for wiring.
If I take them outside for more measurements, will try taking more pictures.
Hi HAL and all Audio Circle members.
Thanks.
Look forward to see better pictures.
Guy 13
-
This may help.. Jon Parkhurst
(http://www.audiocircle.com/image.php?id=70442)
Jon
The top view shows a very shallow "U" frame, would the bass response be further improved by a deeper "H" frame? I was thinking a variable width "H" frame, narrow at the top to maintain the 200Hz upper limit and wider at the bottom to assist bass response. Imagine your side / support panels with a slight "A" shape, +/- 100mm for the bottom woofer and narrowing to nothing just above the top woofer.
Shawn ... Thinking outside the box.
-
Yes, with working with a couple more variables... greater H frame would be a lower response. I think Danny has talked about that somewhere in one of these threads. You have to be careful of the negatives of adding to much to the back as it may cause bad response issues.
JP
One of the secondary objectives was to have a small baffle speaker with huge potential. Some people can't have a 14 inch front baffle speaker for various reasons.
-
JP
I really meant to say variable depth "H" frame, keep the current constant front baffle width and taper your side panels in from bottom to top. This would keep the top woofer operating in it's current 200Hz optimized configuration. The bottom two woofers would see progressively deeper "H" frames to augment the deep bass response. The best of both worlds, unless of course I am completely missing the servo OB basics. ( quite probable since I have never seen this done before : )
Shawn ... Dreaming outside the box
-
Yes, what you asked and what i attempted to say are the same thing. Your are correct...
JP
-
From the test I ran on these a little deeper U frame was recommended for the next version.
An H frame can be used as well if it is no more than 5" deep on each side, but it looses some curb appeal.
-
What's new Hal. We haven't heard much about this for a few weeks.
You in the midst of IIRs and FIRs or some other fun exercise?
I was just getting ready to start work on a new processor and
was curious about how your project was going.
-
The basic crossover is done. Waiting on the prototype servo amps to finish the measurements.
Once that is done will try the correction coefficients. The four channel DAC is waiting on the pulse splitter to be completed.
Been on vacation with family for two weeks. Will get more time starting next week.
-
The Rythmik Servo amps for the 8" servos in The Monoliths have arrived. Will be wiring them up on Monday to start the next part of the measurements.
-
Are these the new style amps? Can you post a picture of one?
-
Yes, they are the new style servo amps.
Will check with Brian to make sure it is ok to post pictures.
-
Here is a picture of the new servo amp prototype from Rythmik.
Brian explains the difference between the prototype and production unit:
" the input jacks will be changed to one RCA plus one mini-XLR. The reason for minXLR is due to space limitation."
The black finished wood cabinet is from jparkhur for the amp.
(http://www.audiocircle.com/image.php?id=74224)
-
Looking good Hal. The IEC shown appears ungrounded is that so ?
charles
-
Yes, the prototypes here are ungrounded.
-
What is the height, width and depth of the new servo amps? And weight? I am sure Danny will have info shortly but have a customer building some speakers and would like to know in advance. Thanks.
-
What is the height, width and depth of the new servo amps? And weight? I am sure Danny will have info shortly but have a customer building some speakers and would like to know in advance. Thanks.
My measurements of the servo amps here show 7.75" wide x 10.25" long x 3.5" deep.
-
what are the measurements of the baffles?
thanks,
drmike
-
Have the Monoliths setup with the servo amps and running without the digital crossover to make more measurements.
I did not see the request for dimensions, so will take baffle size measurements today. The baffles were made by jparkhur.
-
what are the measurements of the baffles?
thanks,
drmike
what are the measurements of the baffles?
thanks,
drmike
what are the measurements of the baffles?
thanks,
drmike
It's full of 8 inch drivers......
-
thanks HAL,
drmike
-
A better picture of the Monolith speakers setup for measurements.
(http://www.audiocircle.com/image.php?id=75477)
-
A better picture of the Monolith speakers setup for measurements.
(http://www.audiocircle.com/image.php?id=75477)
Great project. My PS220'sawait.
Dave
-
Here is a picture of the new servo amp prototype from Rythmik.
Brian explains the difference between the prototype and production unit:
" the input jacks will be changed to one RCA plus one mini-XLR. The reason for minXLR is due to space limitation."
The black finished wood cabinet is from jparkhur for the amp.
(http://www.audiocircle.com/image.php?id=74224)
I wonder if he plans on making a PEQ version of this amp.
-
Hal looking good mate. What are those thingies on top of your amps ?
charles
-
impressive setup
-
Hal looking good mate. What are those thingies on top of your amps ?
charles
Charles,
The components behind the Monolith setup are my friend Bob's Parasound Halo system. They are solid glass spheres. He is an amateur magician and he likes those for decorations. Looks very cool in his setup!
The original lava lamps are outside the camera range. Retrovibe feeling in the room! :)
-
Cool. Lava lamps, where's the bong ? I have mellowed as well now using Himalayan salt crystal lamps which actually affect sound stage and imaging. Similar to ion generators, oh my !
charles
-
He has a large pipe and pipe tobacco collection to complete the look! :)
-
Interesting process. I've got a question though.....There's talk of doing freq and phase impulse correction specific to the driver. The measurement posted a page back is an effectively un-gated in room measurement (if I'm reading the notes on the bottom of the graph), and AFAIK the speaker is on the floor. It's obvious there's a ton of room influence there, which would be baked into any eq done based on that curve. Will the finished units be based on better measurement conditions, as well as gated to "get rid" of the room influence?
-
Yes. I am using another measurement suite that I built to make the time gated high resolution measurements needed for the impulse response correction.
-
:thumb:
-
So, how is the bass on this speaker? Kick ass fast? Powerful? Deep? In room frequency response curve? Is it enough low bass that you are not wanting for a sub? How loud will it play? Are the 8s faster than the 12s? What do you think of making the baffle a V (basically a spread U), so the back braces extend at an angle away from the front baffle....say 10 inches wide front baffle and wings maybe 18 deep inches that extend back at an angle? The full range driver would not have the wing beside it......the wings would not go all the way up.
-
I feel no need for a sub at this point. The three 8" servo's did nicely with both full scale orchestra and pipe organ in my initial listenings. Still breaking in from what I heard from the beginning to the end of that day's trials.
The CLIOwin setup was having issues, so going back over for more trials and measurements this weekend. The measurements I posted earlier are just the 8" full range driver in the baffle without the servo's.
With no crossover on the full range drivers they did pretty well at higher volume levels. That is really pushing the driver as it was full range. The servo's were running a LPF. Still making amp adjustments.
No comment on other baffle designs.
-
Here are the two HAL DAC's being used with the MS-2 Music Server for the digital crossover. They are driven by the Master Clock setup with the M2Tech Evo's. The DAC's are driven by the I2S Bus output of the Evo.
I asked Gary Dodd to put them in his chassis with purpleheart faceplates to match the tube buffer I have.
Each stereo DAC has a different output stage optimized for the specific application.
(http://www.audiocircle.com/image.php?id=76120)
-
Interesting process. I've got a question though.....There's talk of doing freq and phase impulse correction specific to the driver. The measurement posted a page back is an effectively un-gated in room measurement (if I'm reading the notes on the bottom of the graph), and AFAIK the speaker is on the floor. It's obvious there's a ton of room influence there, which would be baked into any eq done based on that curve. Will the finished units be based on better measurement conditions, as well as gated to "get rid" of the room influence?
I just checked and the measurements from CLIOwin are 2.6mSec gated time frequency response plots above. It is the Dayton driver on the baffle on axis response being driven full range.
The next measurements will be gated time as well with the new measurement setup.
-
From a meter away I typically look at a 4ms window myself.
-
Will reset the CLIOwin time gate and see what I get.
-
Here is the same on axis measurement for the Dayton driver with a 4.0mSec gated window.
For some reason the time window indication on CLIOwin 8.51 is not reading out correctly.
(http://www.audiocircle.com/image.php?id=76197)
-
From a meter away I typically look at a 4ms window myself.
Care to elaborate why this particular measurement? :scratch:
-
Care to elaborate why this particular measurement? :scratch:
It is a base standard. And the time window is just about right for that distance.
-
Here is the same on axis measurement for the Dayton driver with a 4.0mSec gated window.
For some reason the time window indication on CLIOwin 8.51 is not reading out correctly.
(http://www.audiocircle.com/image.php?id=76197)
Are you talking about the data on the bottom of the graph? According to the data (I don't have Clio so I might be reading it wrong) your gate length is 5458.65 ms! The fr looks pretty rough for 1/6 octave gated.......even for a fullranger :scratch:
-
Yes, the legend at the bottom of the graph does not reflect the gate that I made on the displayed frequency response graph. The gate was about 4mSec long. I set the low end of the X-axis to 200Hz as the roll-off of the driver in the baffle had droped below the Y axis SPL range.
I did another set of measurements yesterday. Will see if they agree with the first set.
-
I can tell from the measurement that it is gated. If it wasn't then the lower range would be all over the place.
-
Another thing to notice is that the measurement Y axis is 5dB/major division instead of 10dB/major division.
It looks a lot smoother if you go to 10dB instead as a lot of folks do for far field measurements.
-
Any updates on these?
-
The clocking circuit for the two stereo DAC's is now working. Finishing the power supply to complete it.
The digital crossover software for trials is running on the HAL MS-2 Music Server under Windows 8.
Made the measurements to match the output of the servo amps with the three 8" woofers to the 8" wideband driver. Have both servo amps dialed in with the data.
Once the power supply is complete, will assemble the system and give it a listen.
-
Hi Rich -
Seems like you're making significant progress. I hope the listening tests comport with the measurements.
Regards,
Larry
-
Larry,
Yes, making some headway on the speaker digital crossover. The dual DAC clocking circuit has been an interesting challenge. Just about complete.
The good news is that in 22 days I can do this full time! :)
-
Larry,
Yes, making some headway on the speaker digital crossover. The dual DAC clocking circuit has been an interesting challenge. Just about complete.
The good news is that in 22 days I can do this full time! :)
WooHoo!!!
Dave
-
Wow - retiring? You must be really old! :) :)
-
Larry,
Yes, I am retiring at the end of the month. I am not that old! :D
-
One sign of old age is forgetting how old you are. :) :)
-
Here is a picture of the HAL digital crossover project prototype. The MS-2 is the digital crossover driving four channels of clock synched DAC's.
Hope to get some listening and measuring time soon with The Monolith speakers.
This setup will allow testing many ideas past what The Monolith would need, hence the complexity at this stage.
(http://www.audiocircle.com/image.php?id=80672)
-
What happened to this project? Unsuccessful?
-
Still waiting on a major software update from the company that developed it to check out the setup. Once updated the software will no longer work with these DAC's that have the M2Tech Evo's, so will be trying it with a single DAC setup with a different USB async to I2S Bus interface.
No timeframe at this point, as I am waiting on the software to test.
-
Waiting for software - the bane of us hardware guys!
-
Here, I just posted in this thread! :D
-
Here, I just posted in this thread! :D
Is there a question? Not sure what this post means.
-
Still waiting on a major software update from the company that developed it to check out the setup. Once updated the software will no longer work with these DAC's that have the M2Tech Evo's, so will be trying it with a single DAC setup with a different USB async to I2S Bus interface.
No timeframe at this point, as I am waiting on the software to test.
It's been quite a while since the first post on this topic. I like the basic idea of having a speaker custom equalized, but it does raise a lot of questions (as to how well it will work). I wonder if you are any closer to knowing that than when you first started this thread?
Have you done a quick prototype - proof of concept to prove the algorithms? With measurements that show the resulting designed filter improves the sound - and that the speaker doesn't change over time such that the "designed filter" is no better that a generic filter design based on a typical instance of the speaker. (not that that would be so bad).
I am also interested in what you don't like about the ESS SABRE32 DACs sound.
I look forward to your next step and Good Luck!
PS. I like the looks of the speaker!
-
I'll respond to a couple of these.
Have you done a quick prototype - proof of concept to prove the algorithms? With measurements that show the resulting designed filter improves the sound - and that the speaker doesn't change over time such that the "designed filter" is no better that a generic filter design based on a typical instance of the speaker. (not that that would be so bad).
Working with and designing speakers all the time I can tell you that as a speaker burns in the change is not in amplitude. In other words the frequency response stays very constant. Slight changes are usually found in the lower ranges near Fs. But in an open baffle application the full range drivers will be well down in output by that point, so it will have no effect.
I am also interested in what you don't like about the ESS SABRE32 DACs sound.
That DAC is not at the performance level of what Rich has been working with using this concept. I have listened to a couple of the Sabre DAC's as well. It is not bad, but not in the same ballpark of what I have been using. The whole point of this was to use a DAC that is as good as the top level DAC's out there. This is where all of the currently available digital crossovers have lagged way behind.
-
The software only solution for the DSP crossover is not going to occur, since the developer has decided that he has to many other project to take this on. This only occurred last week.
The hardware based DSP crossover version is in build and hope to hear it is shipped before Christmas. That is an update from today.
I need the operational hardware DSP crossover to test the filter that is designed and make measurements with the speakers.
Danny's comments on the difference in DAC's is exactly why I started the project.
-
[snip]
...
Danny's comments on the difference in DAC's is exactly why I started the project.
Which begs the question - what DAC are you using?
-
Will not be revealed until the product is ready.
-
Will not be revealed until the product is ready.
Will it be hardware and software?
-
The four DACs on the DSP board will be hardware.
-
The four DACs on the DSP board will be hardware.
Hi HAL. I'm Dave. I don't care if you let me in our not. I have my own air supply, and it's a really big one.
I know I'm off topic, but I just got up and the caffeine is starting to work..
-
LOL! I am glad the caffeine is working for you. :thumb:
-
The Monolith open baffle speaker with the 8" Dayton full range driver and 3x8" servo woofers is fully operational with the new dspMusik digital crossover. Will take pictures of the setup as it now exists and post.
Took way longer than expected!
-
The Monolith open baffle speaker with the 8" Dayton full range driver and 3x8" servo woofers is fully operational with the new dspMusik digital crossover. Will take pictures of the setup as it now exists and post.
Took way longer than expected!
I look forward to your posts, I was wondering what happened to this project.
-
Here is a plot of the frequency response of the Dayton 8" full range drive before and after correction with the dspMusik crossover. The red graph is before and green is after correction. My measurements are only good down to 500Hz and working to improve that as time goes on.
Looking into the ground plane measurement technique for an improved measurement method.
(http://www.audiocircle.com/image.php?id=116594)
-
Looking at that red graph, I'm wondering how much of that correction you could achieve by simply moving 15-20 degrees off axis?
-
This is at 10 degrees off axis measurement to lower the more peaked on axis response and then correcting for the rest of the response peaks. On axis was very bright. The 10 degree off axis sounded better than on axis and much better after correction.
-
Here is the 8" full range driver on axis without any correction on the same OB U-Frame.
(http://www.audiocircle.com/image.php?id=116604)
-
Wow. That's quite a rise, and an ugly section from 2300-3500 hz. It would be hard/impossible to tame with passive components. I guess you could tame that peak with a notch filter, but the dropout would still be audible. Your computerized crossover seems to be a good solution.
When we would make speakers from driver like this one, the crossover components would run more that the drivers cost. It made me appreciative of the mid woofers like the M-130 that Danny sells.
-
Wow. That's quite a rise, and an ugly section from 2300-3500 hz. It would be hard/impossible to tame with passive components. I guess you could tame that peak with a notch filter, but the dropout would still be audible. Your computerized crossover seems to be a good solution.
When we would make speakers from driver like this one, the crossover components would run more that the drivers cost. It made me appreciative of the mid woofers like the M-130 that Danny sells.
Hi S Clark,
I am sure the extended range drive I bought a few years back
from Common Sense Audio (Audio Nirvana 8"basic 98 USD/pair)
had the same bump between 2Khz to 3.5Khz because it was very
ear piercing and fatiguing.
Well, what did I expected for that price.
I am not a fan of active crossover,
but I guess in this case that would be almost the only solution
to tame that peak.
I would be very curious to see the frequency response curve
of the Omega 7F that I own and like very much,
because it sound pretty linear to me,
but a curve might say otherwise.
Guy 13
Sorry, maybe I did derail this post/topic.
Sorry if I did. :oops:
I also like the Danny's GR Research M130-16
that I have in my N3 speakers.
-
Guy 13,
Well if someone can measure your Omega 7F driver in far field conditions, with a system like Audiomatica CLIO and a calibrated mic, I can make some suggestions.
The dspMusik is not like other digital crossovers in sound quality to me. It sounds like very good analog when being used as a digital crossover and DAC. Once you hear the Dayton 8" full range corrected, I would never go back to the original sound.
The techniques I am using for correction are applicable to any full range driver. For best results, I need a sample of the driver on the correct baffle for measurements to do the design.
-
Guy 13,
Well if someone can measure your Omega 7F driver in far field conditions, with a system like Audiomatica CLIO and a calibrated mic, I can make some suggestions.
The dspMusik is not like other digital crossovers in sound quality to me. It sounds like very good analog when being used as a digital crossover and DAC. Once you hear the Dayton 8" full range corrected, I would never go back to the original sound.
The techniques I am using for correction are applicable to any full range driver. For best results, I need a sample of the driver on the correct baffle for measurements to do the design.
Hi HAL,
thanks for your comments,
I will keep them in mind when I decide to improve the sound of my system.
Rega Apollo CDP, Decware SE84C+ SET amplifier and my Omega 7F extended range driver in Open Baffle Dipole, but maybe latter on I might go with Bass Reflex enclosure due to smaller listening room when back in my home country Canada.
Guy 13 on planet Vietnam, for now...