RS28f/RS225 Omni Experiment {The Catboxpanlids 3000's!} Omni deflector @ 70%

[Raptor550]

So I read a lot on here but that doesn't seem to keep me from starting a project half-cocked. It has been several years since my last build and I have the bug again.

I have always wanted to make an Omni speaker, hearing the H.O.S.S. years ago only cemented that intuition that I would like omnis. dlneubec is a hero to me for his innovation and cabinet finishing skills.
I have also always wanted to use the rs225's in a project so I have kind of mashed the two ideas together.

Thoughts.

• Mark K's MTs & the M8ta designs tell me that someone smart thinks that these two drivers can work together.
• I am very strapped for cash but I want this to sound like something that wont leave me with upgradeitis. So I chose two drivers that are high value for re-using in other projects in case this one is a flop.
• To save money/add flexibility I am using my computer as an active crossover
• JRiver Media center's DSP studio with OX & slope , EQ, ParamEQ, delay, and attenuation per driver. (I am looking for OS wide solution so I can watch movies via youtube Netflix etc on my speakers. Allocator VST? sound easy? Any other solutions?
• Minisdp boards look like a good alternative for flexible active setups.
• I am using an old Yamaha rx-v2095 off craigslist ($120) because it has 6 (possibly 8?) channels of 100w RMS amplification
• 5.1 Creative X-Fi pro usb soundcard w/ ASIO out. Cant get 5.1 optical to work
• I have some baffles cut and ~10"x48" sonotube and precision ports ready for cutting and construction. My ears are wide open for enclosure sizes & port sizing. I have my own Ideas but I want to see what other people come up with and let the experienced speak.
• It would be nice to have a digital room correction on top of my final DSP desighn. again, and software solutions?

I plan on making deflectors out of hard curing plaster over shaping foam in a upsidown teardrop shape or ikea aluminum bowls.
Until then; a prototype...

BEHOLD!
the Catboxpanlids 3000's!



(insert gushing subjective stereophile review here. Perhaps talk about veils lifting everywhere)

DSP:

XO is 1250@24db rs28f
XO is 1300@24db rs225

I read a little overlap can sound pretty good with steep crossovers somewhere.
I am a little bit worried about the crossover being in such a critical location and would have preferred a 3-way with the rs52 or rs100 if I had more money. I am banking on a little extra in the low end from the rs28F over the rs28A. How concerned should I be with the rs225 firing up & how its motor is hung?

I could not wait to get real enclosures so I scrounged what was around me to make a very temporary enclosure. My Girlfriend has access to tools at her school (she is currently building a boat) and I am waiting for her to cut some countersunk holes in the baffle rings. In the mean time the catboxpanlids (She named them) sound reasonably good with the DSP and boxes.





Really I am posting these for the LOL's but I did actually model both boxes in baffle diffraction simulator. First time using the application so it was good practice.

This is what the OX measured with my crappy mic before eq. 6 measurements; woofers, tweeters, and combined for both boxes



right now I have the tweeters lowered -8db

Q1 -6db notch at 10.040k
Q1 -2db notch at 1140hz
Q1 +5db notch at 100hz

all tweaks are iffy because I only have my 2 laptop mics to go off of so I cant do any fine adjustments. I was going to buy a MiniDSP UMIK-1 or Dayton EMM-6 but they are out of stock. I was going to take some more measurements with the dEQ changes but look like my housemate is napping.

[Hank]

Somebody tell me I'm having a bad dream. No doubt too much tequila last night at my saloon.
OK Raptor550, you get the "Different" prototype award for 2013!:T

[Jonasz]

Love this thread, keep the craziness going! :P

[JonMarsh]

Interesting... vewy interesting... would have been confusing if you'd started the post a few days earlier!

Didn't know True RTA was still around. Where you're at right now, I wouldn't invest too much time tweaking by instrument until you can get some kind of mic to work with. Certainly looks like you're having fun! :W

[Raptor550]

So Jon, you should know. In my mind you ARE professor brown and when you talk to me I feel like its literally the Doc. One of the upsides for everyone who is consistent with there avatar. You are right, there is not a lot I can do until I get real measurement equipment.
Dayton UMM-6 or the MiniDSP UMIC-1. Any reason not to get the cheaper $75 Umic-1? The Cross Spectrum UMM-6 is $90 but I don't know if that would be saving me any more value. Thoughts? Also, I have seen REW and that look like a good software package. Is there any free crossover software? I have not found anything better then JRiver, but there is too much lag doing a system wide loopback thru the program.

In the mean time Jenisa & I have been experimenting with enclosure shapes.
Here are some sketches from our brainstorming sessions:






I hope to put up some more concept Ideas for you folks, and I am honored Hank, I have never won a speaker award before :W

Right now I am torn between the orientation of the tweeter up/deflector or head on. We may move the drivers to a monopole design from the above sketches later so let us know if there's something that looks cool. For now sonotube is $6 so that's what we are going to use in the mean time
VS Mayfield-Acoustics-Model-1-loudspeakers1.jpg Image not available

[Raptor550]

Cone deflector construction

Let me know if this is correct:

• If sound starts beaming when a cone is half the width of a frequency of sound,
• the rs225 has a ~7" radiating surface.
• So, rs225 will start beaming when the the wave reaches 14" at 965.13hz?
• Therefore I need a deflector that matches the beam width?

Today I thought of an idea to make cones; paper model the basic shape and then pour plaster on and sand smooth. The design will be using the same shape as the Fontanas with the deflector also being an upfireing tweeter enclosure. I did the first cones, but will this be enough or will i need to go bigger? I can stack these on a bigger shallower cone to get a curved deflector like Duvell. Will this be a wide enough deflector? The cone is 90 degrees so upfireing vertical waves should deflect to the horizontal plane.

The cones use a thick weight paper and I have double layered them to make them more rigid for when they take the plaster.



I have heard that deflectors may even be unnecessary and the H.o.s.s. and Mentors went monopole for the tweeters, I am still on the fence and I have half a mind to make the tweeter enclosure a rotatable fixture.

[Raptor550]

diffuser tops update

So Minor update, I bought a mic, I went with the mini dsp umic-1. With shipping the Dayton one was cheaper and it has a cooler box but I realized this seconds after pulling the trigger, oh well, at least the umic is gun metal which I hear is one of the more sonically accurate colors, so no worries ^_^.

I carved a hemisphere(ish) out of foam with a sharp non serrated kitchen knife. I then mashed it on my table rolling it to round out the facets. It was all by eyeball so its not very even, but I think I can get it more accurate slathering plaster and lathing that with a guide while its still damp.





Right now they look like ice cream cones or Flintstones deflectors.

[Raptor550]

Deflector Simulations

I wanted to model out some deflection shapes in cross-section using a particle simulator I had on my computer. A disclaimer; this is a particle simulator, not a sound wave simulator. So it will not take into consideration cancellations, sound wave passing thru solids, or wrapping around edges. It can show you only bounce radiations and time delays (because of distance traveled in bounces)

I have a generic tweeter putting out three tones/disruptions.

• Purple = 90° near omni low tones
• Orange = 45° dispersion mids
• Green = 10° beaming highs

The exact frequencies are relative to the size of the driver vs wavelength of a given frequency. Beaming occurs somewhere around 1 to .5 x the F width. (I need clarification on which it is because I have read both.

I can kind of visualize this stuff already because I play with particle sims often, but it helps to have it out of my head to analyze further and perhaps it may give you guys some food for thought.

Sphere
The first shape I modeled was a sphere or ball. The interactions would cause lots of interference i would imagination. If you moved the ball closer those low and high green angles would come into a more even spread but with more bouncing. How would this effect efficiency? balls seem bad; am I wrong?



Bullet
A bullet fairs better but still points the highs downward an down pointing egg will probably be somewhere in between a sphere and a bullet.



Funnel
A funnel seems good if properly distanced away for distance/beam ratio; that might take a complex simulator tho. It looks like the lows in purple that don't bounce get further ahead.



Cone
The best choice to me look like the cone.



Funky things happen if the cone is too close. The sound gets split by f to angle. The lows radiate omni, the mids radiate tiled upwards or at the ceiling if at 45 degrees already meanwhile the highest notes got directly out.



Is this accurate to how sound actually reflects?

[Raptor550]

Need bass advice

I am using a sonotube for the rs225 it is fixed 10" diameter. For volume changes I will be adjusting how high the tube is. I am hoping if I can get some sound wisdom from you lovely folks I can cut the tubes tonight!
Here are curves for tuning/enclosure size alternatives.



Right now I am trying to balance these three things:

1. Low Extension
2. Tweeter Highth
3. Power handling/room response

My sitting ear level is 3' (+1-4")
In the red are sealed and maximally flat ported enclosures. Max flat is at 50L but that will put the tweeter at 4'5" which I think will be too high for sitting position
a 40L tube will put the tweeter roughly at 3'8"
a 30L tube will put the tweeter roughly at 3'2"

I was reading in Zaphs' mantas:

Normally, I'll tune lower to get a more gradual rolloff that works better with room response and puts power handing where I need it. It's basically a type of tuning that is a combination between sealed and vented alignments. Note that this is not an Extended Bass Shelf (EBS) tuning. That kind of tuning implies an oversized enclosure volume, with a dip above the tuning frequency. You get the lowest anechoic F3 with that configuration, but in a real room you get an overbearing one-note bass at the tuning frequency, a huge midbass dip, and crappy power handling.

For the most bass I would say green is the best, but keeping zaphs words in mind perhaps I should use the yellow trajectory? The pink has the smallest enclosure. I need your guys advice!

[PMazz]

My ideal would probably be 35 liter tuned to 25Hz. Nice roll-off, ~100w power handling, F3 ~40Hz F10 ~25Hz, 106dB max output.

[Raptor550]

Thank you very muchI think ill go with that, that will probably look close to the yellow line. This is a noob question, How do you tell what is good power handling, and max output??

[cjd]

Remember that a dome tweeter does not radiate like the particle sims you've posted... at least not quite. It is frequency dependent just HOW they work, too (note off-axis tweeter responses)

Very cool project though.

I tend to tune lean, and I make sure to include the likely series impedance. I almost always push box size up as high as is sensibly usable - I believe the result is audible as well, but not everyone does.

C

[PMazz]

Thank you very muchI think ill go with that, that will probably look close to the yellow line. This is a noob question, How do you tell what is good power handling, and max output??

Here's my sim. Excursion in red, impedance in gold. Input is 100W. As you can see, excursion limit (7mm) is reached with 100W power applied reaching 106 dB output (theoretically). Disregard excursion graph below tuning because your driver just exploded!

[cjd]

I am a little unsure about this because when using Pete's suggestion, I'm getting the same results for SPL, etc. at 100W, but hit that excursion graph at only 50W. Also it seems there is some rumble filter included in this model, but only for the frequency response! I'm using Jeff B's Woofer box model and circuit designer (all active sections turned off) - if I use a 4th order LR rumble at 24Hz I get a similar response graph but still see excursion limits hit at 50W.

I'd be tuning a single RS225-8 a little differently as mentioned, but now with numbers!

52L tuned to 21.5Hz - that's a 3x20" port in a box with interior dimensions of 10x14x24. Excursion is comfortable up to ~40W.

I wouldn't mess with porting at 35L personally and would rather do 35L sealed, which is an F3 of 56Hz, F6 of 39Hz, F10 of 28Hz. You're not gaining that much porting here far as frequency extension goes, but in my experience the sound quality will differ subtly.

I've verified the tool against measured results on the Ansonica recently, since I discovered I messed something up on the saved file I had for Unibox and it was being weird - now I wonder about this too.

This is part of the design process though - evaluating your options, weighing the trade-offs, and choosing a direction.

[PMazz]

If I run your vented numbers (52L), excursion doesn't occur until ~65W. There is no rumble filter in the sim. I downloaded the modeling program you used and couldn't get any graphs but SPL to work. I input 100W into available power but must be missing something.

I also prefer a sealed alignment but I always have.

Edit: I needed to add the Analysis Toolpack to Excel. The excursion shows 40W as you stated. I don't know which one is right tho.

Edit 2: I ran the numbers in Unibox and it mostly agrees with Jeff B's sim.

[cjd]

Jeff B's sim is definitely accurate. I've verified it vs measured data.

[Raptor550]

Cone deflector/ tweeter mount construction

I am using "Subwoofer simulator by Isaac MCN", I could not find were to up the watts, but I did up the voltage and the closer to 106 db I got I was able to see -22hz as an Xmax danger zone. I will be using software XO so should I put a steep 48db/octave highpass at 23hz to protect my driver?

Casting the tweeter mold:

1. Place RS28f faceplate in a sandwich bag.
2. Tape to wooden mount Ring to tweeter
3. Create clay wall to keep plaster out of the tweeters driver volume
4. Vaseline bowl, tweeter faceplate, tape along the faceplate edge (no Vaseline on the tape that is on the wood mount, no Vaseline on the wood mount)
5. Pour FGR

Trimming the cone molds fiberglass.



The tweeter mounts should have a nice flush mounting now as my lovely assistant demonstrates.



We used a sandwich bag with a small pinky sized hole in to squeeze the frosting that will cement our two halves with a low water mixture. You have to move fast to sandwich the pieces; in the time it took me to take this picture and place the cone the goop was too dry to stick and we had to knock it off and redo the ring.



Once everything was cured I put some in the seam and formed it with my fingers to a point. It dries quickly (3-8 seconds) so Jenisa would squeeze a pea to marble size drop out of the plastic bag and I would swipe it and quickly form it. I used spackeling cement to further refine the shape later.
This is what we have after minor sanding.



Here is what it looks like inside. Once we had them sanded we decided to mix up some more FGR slop and pour it thru the tweeter hole and roll it around to reenforce the internal seam between the dome and cone.



In my opinon, FGR is better then regular plaster because it is less chalky and feels Strong.

Woofer enclosure.
So I did these last night so some of the advice I could not act on in time. I did some remeasuring and with a 36" tube internal I have 41L of airspace, that puts the tweeters at roughly 42". The ports could not be tuned any lower than 26.5hz without buying different pvc.

Nothing out of the ordinary for a sonosub construction. I have to give credit to Jenisa as she made a fit so perfect the woofers are hard to shake out upside-down. She made all her cuts without the woofers nearby, she went off of dimensions I text-ed to her. As good of flushmounts as my best work.



Any good ideas of how to cheaply get rid of that sonotube paper spiral to prep for a piano gloss paint?

[Raptor550]

Disregard excursion graph below tuning because your driver just exploded!

...lol!

[Raptor550]

Body parts are coming together

Construction pictures.

Important things I learned notes.

• Completely cover the tube and MDF in lots of Spackle to cover the spiral
• Use black car primer (its thicker)
• use sandpaper 180 > 220 > 320 > 400 on primer
• Up stroke gets most of your clear off the sponge, one Sloooow single downward stroke top to bottom to even out the clear coat. Your arm will have an up glop-down smooth motion and then you will move 2/3rds of a brush stroke to the right and repeat
• a hair dryer can helP add on paint layers quickly, but too hot and bubbles will form and then you are hosed
• A butane torch is better then a soldering gun for larger wires & metal elements
• LEARN HOW TO WET SAND, IT IS AWESOME

To Do:

1. wet sand the tubes gloss layer, 320 400 600 1000 2000
2. polish and reattach brass feet, belts rings ect.
3. Figure out how I want wiring to come out of the tweeter deflector (I do not look forward to drilling into it)
4. Figure out how the heck to make an adjustable tweeter holder that can rotate for on or off axis and still look sleek with minimal welding skills.
5. Calibrate with miniDSP

[dlneubec]

Cool project! Very creative way to create the diffuser/waveguides.

Here are a couple of suggestions, though it may be too late. I've done a number of omni projects and a lot of experimentation, working with actual measurements of various diffuser/waveguide shapes, positions, etc. The shape that works best, in all my testing was always a concave cone. Your 45º cone for the woofer will probably be fine since it is limited in the frequencies it will reflect. Clearly the lower frequencies, already omni due to wavelength, will simply ignore it, but as the waveleghts get smaller, it will begin to deflect them.

For the tweeter diffuser, I would suggest a couple things. A concave cone shape will yield the best natural frequency response, IMO. You might be able to find a funnel or a plastic wine gless top or something like that you could use as a mold. Second make it so it is suspended much closer to the dome. you will likely have the best results with the tip of the diffuser within 1/8" or so of the dome, just far enough away to that it cannot move into the diffuser when driven. The reason for this is that the tweeter will also be naturally omni in its lower register, but as it begins to beam at higher frequncies/shorter wavelengths, the diffuser will begin to redirect the sound. This is likely to be in the 6khz range or so, I'd guess. Above that is where the diffuser/waveguide begins to control the dispersion. Since the wavelengths in this range get so short, you need the diffiuser as close as you can get it to be effective.

Also, the ring that you are showing to hold up the diffuser will cause all kinds of negative reflections that will both reduce clarity and cause cancellations or peaks in the response. You would be much better off with the diffuser being held up by 3 prongs and no ring around the top, unless the ring is recessed flush with the diffuser surface.

If you have the ability to your own measurements, you might want to experiment with both the tweeter diffuser height, how it it held up and a concave cone shape. It is beneficial, intially, to be able to move the diffuser up and down until you determine the ideal height. Even though you are using the miniDSP to EQ the result, your job will be much easier if you are able to optimize the frequency response through good physical design first and use the miniDSP as the last resort to sort out small variations.

You might look Google the Duevel speakers to get an idea of the optimal concave cone shape. Look at their top line models, but ignore their Planets speakers as a model. I don't believe, based upon my experimentation, that those round balls do anything helpful and are purely for look. One, because round deflectors just don't work nearly as well as concave cones, and two, they are too far away from the driver to do anything positive.

Good luck! It is an ambitious project, but should be very rewarding, especially as you get to share it with your "assistant". They will look very cool as well.