A place to discuss member's DIY audio projects & post pictures/schematics. NOTE: There is a limit of 2MB per attachment, and a maximum of 3 attachments per message. If you need to post more than 3 attachments, just add another message.
January 14th, 2016, 5:42 pm
Here are my DIY OB planar hybrid speakers.
The OB planar section are a BG NEO3 tweeter and NEO10 midrange in a sidewing style open baffle. The design was inspired by some work done by GR-Research in OB planar speakers.
The bottom is a 2x12 open baffle H-Frame with two 12" GR-Research servo drivers with the Rythmik Audio HX800 amps. This is a section of a larger 6x12 open baffle H-Frame servo sub array I have been working on.
The system is run with the HAL MS-3 Music Server driving a Danville Signal dspMusik 2x8 digital crossover. Did measurements with Audiomatica CLIOwin and Room EQ Wizard for the system integration.
Since BG was bought by Christy Digital no more planar drivers are available, so turned into a one of a kind speaker system. Sounds really good in my setup for a full digital replay system. It does have A/D inputs for analog sources like my turntable system.
Attached is a picture of my room with the speakers and digital crossover prototype setup.
- Attachments
-
![WP_20151203_005.jpg (412.03 KiB) Viewed 205541 times WP_20151203_005.jpg](./download/file.php?id=375&t=1)
January 14th, 2016, 6:10 pm
Very nice Hal.
A question about your subwoofer to OB crossover point. In general, doesn't one keep subs to under 80Hz or so? Can the H-frame baffle subs perform well above this? [Maybe because of the servo, they sound good enough to use to a couple hundred Hz? Doesn't the Neo10 lose usable output below about 250Hz?
On paper, it would seem that there is an issue. How did that work out.
BTW, I'm very interested in your crossover and measurement knowledge, if you're willing to share it.
REW and a Dayton calibrated USB mike are in my tool bag, but not much in the way of experience....
Stuart
- Attachments
-
- BG_Neo_10.pdf
- (212.74 KiB) Downloaded 1057 times
January 14th, 2016, 7:08 pm
OY VEY!
January 14th, 2016, 8:38 pm
Stuart Polansky wrote:Very nice Hal.
A question about your subwoofer to OB crossover point. In general, doesn't one keep subs to under 80Hz or so? Can the H-frame baffle subs perform well above this? [Maybe because of the servo, they sound good enough to use to a couple hundred Hz? Doesn't the Neo10 lose usable output below about 250Hz?
On paper, it would seem that there is an issue. How did that work out.
BTW, I'm very interested in your crossover and measurement knowledge, if you're willing to share it.
REW and a Dayton calibrated USB mike are in my tool bag, but not much in the way of experience....
Stuart
Stuart,
The subs are with the main drivers, so no problem running them up higher in frequency to crossover to the NEO10. The servo subs go up pretty high at -6dB to crossover to the NEO10 without any issue. There is no localization issue with the subs and mains being separated. I do time align the subs to the main baffle with the digital crossover.
Since you are working with line array's, the measurement process has to be modified to take into account the line array length. The farfield measurement distance should not be closer than the vertical length of the array. REW use to offer gated time measurements to do the individual measurements needed to design the crossover. You need frequency response measurements for each of the driver arrays and the subs individually to design the crossover. Also only measure one channel at a time.
Be careful to start with the volume low for the measurements as to not blow drivers by mistake. I have to be very careful since the BG drivers are no longer produced and hard to get spares.
The drivers response is heavily influenced by the baffle design, so the response plots on the drivers from the mfg are just a starting point.
If you do speaker measurements, I am sure folks will post insights into the crossover design needed.
January 15th, 2016, 9:56 am
Hal,
I meant more of a sound quality issue, were the subs fast enough in transient response to keep up with the Neo10?
Allegedly, the line array of 6 Neo8s on my drawing board will reach about the same lows as your single Neo10, about 250-300Hz. My assumption was that a fast mid-bass driver is needed to bridge from a ~12" sub to at least that frequency. That's why I've included the Dayton RS180-4 in an array of four from ~60Hz to 1kHz, where power handling of the Neo array is improved, and presumably distortion.
Is that not needed with the right (servo?) sub arrangement?
As to individual measurements, in the baffle, should these be preformed outdoors, to avoid room effects? I better read up on "gated time measurements" since I am clueless.....
Thanks again!
Stuart
January 15th, 2016, 10:51 am
The 2x12" servo subs easily keep up with the BG NEO10 planars. They blend well for sound quality. These can work up to 250Hz at the -6dB point to crossover to the NEO10's. I have heard 3x12" servo arrays with larger line arrays and they are about the best low frequency reproduction I have experienced besides the original Infinity IRS Reference III's with the 6x12" servo arrays.
The servo corrects the response of the 12" driver to stop it after the signal stops. The servo sense coil is wound with the drive coil for very little response lag.
The gated time measurements allow you to make in room farfield measurements to minimize the effect of the room. Outside is better for low frequencies, but hard to do with large speakers. If you can keep the room reflections within the first 4-5mSec out of the measurement, it will be good down to 1/T = 200-250Hz. If you need to crossover lower, then measurements outside will work correctly as long as the distance from surfaces is far enough.
Another technique called ground plane measurements can work as well for low frequencies.
January 15th, 2016, 11:52 am
HAL wrote:The 2x12" servo subs easily keep up with the BG NEO10 planars. They blend well for sound quality. These can work up to 250Hz at the -6dB point to crossover to the NEO10's. I have heard 3x12" servo arrays with larger line arrays and they are about the best low frequency reproduction I have experienced besides the original Infinity IRS Reference III's with the 6x12" servo arrays.
The servo corrects the response of the 12" driver to stop it after the signal stops. The servo sense coil is wound with the drive coil for very little response lag.
The gated time measurements allow you to make in room farfield measurements to minimize the effect of the room. Outside is better for low frequencies, but hard to do with large speakers. If you can keep the room reflections within the first 4-5mSec out of the measurement, it will be good down to 1/T = 200-250Hz. If you need to crossover lower, then measurements outside will work correctly as long as the distance from surfaces is far enough.
Another technique called ground plane measurements can work as well for low frequencies.
Properly designed servo subs are a great way to lower distortion and improve transient response. Ultimately the linear volume displacement of the drivers used sets limits. Once Xmax is exceeded the distortion rises dramatically. One of the Velodyne engineers told a friend who wrote extensively on subs that the biggest problem in designing a servo sub was to maximize the correction capabilities of the servo without destroying the driver. The offset to the complexity and cost of a servo implementation is more drivers in large sealed boxes so that the drivers stay in a linear region and exhibit good transient response. Both are valid approaches and you must pick what works for you. The servo yields great response within the limits of the driver and takes up less room. The multi driver/box approach ultimately yields higher output for a given maximum allowable distortion but certainly tests the WAF. Open box subs are virtually imposssible to design for high SPLs at very low frequencies (below 30 Hz) due to front to back cancellations unless the baffles are VERY large. In this case using a servo to drive the speaker harder to get more output at frequencies where there is cancellation does not work because as you push the driver harder both the cancelling front and back waves increase the same amount. It is like trying to fill in a high Q suckout in a room with electrical boost at the cancelloing frequency. Good luck with either one.
January 15th, 2016, 12:16 pm
The Carver Amazing loudspeaker got the job done with open baffle woofers. He used woofers with very low free-air resonance and small motors together with an equalizer.
I recall auditioning them at Soundscape in Baltimore. They definitely reached down to the "bottom" and were detailed. The midrange and treble were push-pull line array planars similar to what you guys have been chatting about.
January 15th, 2016, 12:58 pm
SoundMods wrote:The Carver Amazing loudspeaker got the job done with open baffle woofers. He used woofers with very low free-air resonance and small motors together with an equalizer.
I recall auditioning them at Soundscape in Baltimore. They definitely reached down to the "bottom" and were detailed. The midrange and treble were push-pull line array planars similar to what you guys have been chatting about.
I first heard the prototype of the Carver Amazing speaker at the apartment of my friend Larry Klein who used to write for Stereo Review. Bob was there for setup and his usual enthusiastic comments. As I remember the speaker was quite clear as you would expect from a dipole planar. However, although there was sufficient bass in the range above 30Hz and in some cases too much, there was no low bass. Bob achieved this bu having woofers with a very high Q that caused a boost near resonance but that did not help at real low frequencies. If you had room gain at low frequencies it would help and also the fact that the baffle was large in the vertical dimension gave more output in the mid bass but the resonse essentially dropped like a rock below 30Hz. Also remember that Bob went on to build sealed box subs having very low VAS that you could stuff into a small box if you had enough power. They wern't the best out there but the shock factor of low bass coming out of a small box moved the industry in that direction.
January 15th, 2016, 1:39 pm
Then there is Longwood Gardens with the massive concert organ that can do 16-Hz. It made my pants legs shutter. WOO-HOO!!
Powered by phpBB © phpBB Group.
phpBB Mobile / SEO by Artodia.