OB "omni" experiment

Status
Not open for further replies.
Jump to Latest
K

Kjeldsen

Member
Joined 2003
kjeldsensforbindelser.dk
#1
I have begun my experiment, very much inspired by Graaf and Tinnitus's flooders and OB because I have never tried it before.

I will have one Markaudio CHR-70 and four 10" peerless woofer per speaker. I have made prototypes, but still haven't had any sound yet. The prototypes only have two woofer, and are only ment to give the thumbs up or down for further experiments.

First shot of the not very professional prototypes. They are made of scrap wood and corrugated.

An externally hosted image should be here but it was not working when we last tested it.
 
Last edited:
#7
Project gor a little delayed. I don't use analogue output from my minidsp, and forgot that the outputs were hidden away in the case. I have build a new case, but now need an angled USB cable. I also thought I had two NAD amps. I knew one of them periodically don't play in left channel (got for free). But there was a lot of hum and I couldn't test with this amp, so only one speaker have been testet. Since I can't tweak the miniDSP untill I have the angled USB cable, I have only the tone controls to use, on my NAD amp.

Suprisingly I was able to get a pretty flat respons from 35 to 300 Hz, with Bass eq, Bass control and loudness on. I also tweeked with eq in foobar. I sounded horrible, but prooved that I can get plenty of low end with these Peerless woofers.

I have one measurement of woofer section with two 10" Peerless woofers. Final system will have four per speaker. I will post later today (European time)
 
#9
OB sub.jpg
 
#14
There is potentially a major issue when mixing omni and dipole. Your dipole has output from the rear of the driver that is out of phase with the front output of the driver.

Your omni output is generally going to be in phase to the front and rear. At some point, the front or rear wave of the dipole is likely to conflict or be out of phase with the omni drivers output, causing cancellations, especially as yu approach the crossover region, where the drivers are covering the same frequencies.

A better mix with omni is probably a bipole or monopole configuration for the second driver.
 
#15
Generally speaking, dipole + monopole = cardioid by mutual cancellation of the back wave, when the drivers are on the same vertical plane, in phase and as close as possible.

This can be a good association for low frequencies radiation when thinking of the room boundaries. But of course, it's not omni anymore.
 
#16
dlneubec said:
At some point, the front or rear wave of the dipole is likely to conflict or be out of phase with the omni drivers output, causing cancellations, especially as yu approach the crossover region, where the drivers are covering the same frequencies.
Click to expand...

maybe thats what actually makes it tick
the key is probably how much of the upper omni drivers output is phased out, the frequency, and the xo, definately not 'textbook'
if radiating 'only'180 degr, this might be the real reason it works at all
 
#17
tinitus said:
maybe thats what actually makes it tick
the key is probably how much of the upper omni drivers output is phased out, the frequency, and the xo, definately not 'textbook'
if radiating 'only'180 degr, this might be the real reason it works at all
Click to expand...

I just don't see anything good that can happen. You'd have driver cancellation though a portion of both the stop band and the pass band of both the omni and dipole drivers, across the same frequencies. This cancellation would have to cause a huge dip in the power responce, which would then peak again as the dipole driver rolls off and omni takes over. Then the power response is likely to drop off again as the omni driver becomese more directional at higher frequencies. I'm afraid I can't see how that kind of roller coaster power response would ever be a good thing. Voicing would likely be a nightmare.
 
#19
Kjeldsen said:
Why, its also in OB (the full range omni). It's simply a 2 way OB tilted back.
Click to expand...

Your classic dipole output is a theoretical figure eight when looking from above the speaker, which signifies the cancellation that takes place so the sides of the speaker when the out of phase output from the front and rear of the driver meet. Your plan is to have one drive oriented this way. Let’s call the front radiation positive and the rear negative. This is a good thing, since the nulls at the side reduce the influence of side walls from a reflection standpoint, but the rear output requires the speaker to be out from the front wall to avoid reflections arriving at the listener too quickly.

Now you want to add a second dipole driver facing up so the figure 8 is turned 90º. Let’s say the positive is up and the negative down. This turns the dipole null for this driver into a horizontal plane which is aimed at the listener. You have just created a null with this driver toward the listener. Add to that, the positive output from the up facing driver is heading equally front to rear, as is the negative portion that is aimed at the floor. The rearward output is conflicting with the output of the woofer in both directions, creating further nulls or deepening the ones you already created by turning the mid dipole driver facing up. The positive mid driver output heading toward the rear wall is cancelled by your negative, rearward woofer output through the crossover region. Similarly, your negative mid dipole output is heading at the floor and cancels the positive front output of the bass through the crossover region. Add to that, as mentioned, dipoles like to be away from surfaces in the direction of their rear output, usually the front wall. However, your midrange dipole is turned so it is probably on a 2-3’ from the floor and its rear output is quickly reflected back up, potentially creating further complications.

You have such a complicated driver interaction I can’t see how it doesn’t end up with a totally wacked out polar response in all directions, a varying power response as your move through the crossover region. Let’s say you crossover at 400hz, 24 db/octave slope. At 200hz you are 27-30db down and perhaps far enough down that the cancellation is acceptable. The other direction takes you up to 800hz before one driver is that 27-30db down. That suggests perhaps a 2 octave range of significant cancellation on certain axes.

As mentioned, the up facing full ranger is also going to begin to beam at some point, depnding on its size, etc. just because of the naturally reduced ouput off axis. Combine the natural off axis roll off with the dipole null you have created toward the listener by using it dipole and you end up with very little energy at the high end headed toward the listening position.

I applaud anyone for experimenting, I just think this one is fraught with a huge amount of problems and I don’t see how they can be worked out.

Perhaps a potentially simpler solution, if you want to try and get some dipole mixed with some omni characteristics might be to place both the woofer and mid on the same baffle or plane, so that their dipole radiation pattern match one another and then experiment with tipping the entire thing backward. The backward tilt would tend to aim the dipole null toward the floor, reduction floor bounce, though ceiling bounce would certainly be increased. The side null would still be amied at the side wall as well. At some point the mid driver would sit far enough behind the woofer acoustically that it might take some active delay to get them to work together. Also, the further you tilt it back, the faster the top end of the mid driver will roll off and your on axis high end will start to reduce substantially. Still, just a 30º tilt back might be enough to give you some omni effect.

I wish you goo luck with whatver direction you decide to head in.
 
#20
My fault that i didn't post the drawings from my original thread.
Ter version seen here, is just to see if I can live with OB sub, nothing else.
My design will be more like this,
236923d1314126791-dipole-monopole-ob1.jpg


It will not be dipole, but probably more like cardioid.

I will be able to tilt or not to tilt, and anything in between. Simply to be able to easily tweek. Because I now this is an unusual speaker, I will not have just one possible angle, but as many as possible.

I really appreciate your comments. Could I make it less complicated by having u-frame separate for all woofers og maybe just for two in pairs. And maybe somehow isolate the full range (it has very wide dispersion up to 10 khz as I remember)
 
Last edited:
Status
Not open for further replies.
Top Bottom