Ricci's CKRAM Subwoofer and Files

[Ricci]

 

The CKRAM is a compact, single 21" subwoofer, that is designed to be flexible enough to suit a variety of applications, but is primarily intended for sound reinforcement or Pro use.

Dimensions:  (24" / 610mm) Width x (30" / 762mm) Height x (28" / 711mm) Depth

Drivers: Most of the good pro audio 21" drivers will work well in this cab. 

The Lavoce SAN214.50 is a good budget option. 

The Eminence NSW6021-6 is what the CKRAM was designed with during modeling. It or the B&C 21Ipal are top tier options. 

Other drivers that should work well include 18 Sound 21NLW9601, 21ID, B&C 21SW152, 21DS115, etc...

Design: The CKRAM is a hybrid 6th order design in some ways similar to the SKHORN and SKRAM designs but notably different in others. The CKRAM has a multi-port back section which allows for adjustment of the port tuning to suite the application. The front section is in effect an extremely short, symmetric, horn/slot/small vented chamber. The front assembly is intended to be removable and the CKRAM can also be operated in a standard vented alignment. 

Vent Tunings: Note that the CKRAM can be operated in 6th order or standard vented operation. In either case the vent tunings remain the same. 

All vents open = 30Hz

3 vents open = 25Hz

2 vents open = 21Hz

1 vent open = 15Hz

 

 

 

 

 

 

 

 

 

 

Horn Response Models: These should be close. The inputs for each vent configuration are included. The match is quite close to a much more detailed Akabak model. Use the semi inductance specs for the driver you are simulating! 

NSW6021-6 semi inductance

 

This is the Akabak script if anyone wants to dive that deep. 

ckcomplx.aks

[Ricci]

Here are the print files. 

As usual 18mm and 12mm, void free BB ply is recommended. Can be built with or without CNC. CNC programming will have to be developed on a case by case basis for your shop/machines/software. Modify as needed for your situation. The basic dimensions and overall layout are what is critical. The hardware, assembly, materials and other processes can and should should be adjusted to suit the situation. For example the front panel assembly that covers the driver should be modified if cut on a CNC for the inner disc and outer round panel to "lock" the braces into place and screws should be used along with glue. 

Free use of the plans to build subs for yourself is fine. Use of them to make a finished commercially sold product is not. 

NOTE about the prints!

These are extremely detailed due to being designed in Solidworks and the plans from which my personal cabs would be built. The simplified layout drawing is really all that should be adhered to to build this sub. The bracing and front panel assembly can be simplified to suite your own ideas or the tools available to build the cabs. Just make sure it is solid! Any or all of the hardware can be deleted or substituted or modified to suite your needs. Think of the plans as a chassis guideline that can be modified to taste. You don't have to put the holes in your braces or use the handles, a roundover on the cab edges, add a cutout for a plate amp, etc...

DXF's will be added later...

 

CKRAM print.pdf

[Ricci]

Design Notes/Goals: I started on the CKRAM long ago and got about 85% done before hitting a crossroads of analysis paralysis. I left the project in the dust and got busy with other things for a long time. After I picked it back up towards the end of last year I made some progress towards getting it finalized but was still not sure about the choices I'd made. I left it alone again for a few months before pulling it back up last week. At that point I looked at it with a fresh perspective and changed the final design to what has been posted here. 

The original goals for this stem back to the dual driver SKHORN design. As soon as it was finished a single driver version was planned, but simply cutting the SKHORN in half was never an option for me. What is a nice, well braced , symmetric layout in the SKHORN is not so pretty once the cab is cut in half. That meant a totally new design was needed. Something the size of half of a SKHORN is quite small for a 21" design. I've always found that the smaller the design is the larger and more impactful the tradeoffs become. It doesn't help that I'm overly analytical and obsessive. I quickly reached a crossroads with the design where I had a few paths to take and didn't really like any of them. At that point I designed the SKRAM instead which was easier due to a bit of extra cabinet size to work with which made it easier to fit everything together. The SKRAM was actually how I wanted the smaller half-SKHORN layout to be but I just couldn't get it to fit in the size I wanted without compromises that were too much for me. I scaled it up to a size that would work well and that became the SKRAM. 

Eventually I came back to the smaller, unfinished half-SKHORN design to have a fresh look at it and the design goals. I started with what I considered to be concrete design restrictions and a second wish list of design wants. This is all in addition to the usual goals of smooth FR, low distortion, smooth GD, efficiency, power handling, etc. 

Concrete design considerations...
1.) 12cu ft external or smaller. 
2.) Variable vent tuning
3.) 30Hz tuning with all vents open
4.) Single 21" driver
5.) Room for plate amp, handles, casters, feet, stack lugs, pole mounts as needed. 

"Wish" list in no particular order...
1.) Maximize vent/slot/horn area / minimize air velocity
2.) Straight vent / slot / horn
3.) Thorough bracing / Keep cabinet vibration low especially around the driver
4.) Horizontal driver / Driver stays horizontal regardless of sub orientation
5.) Symmetric loading on the driver cone
6.) Works well with a variety of drivers
7.) Keep construction simple
8.) Minimize weight 
9.) Minimize direct sound and operational noises from the driver. 
 

If you read through the 2 lists above it becomes apparent that some of these are at cross purposes. There is no way to achieve everything without compromise. Do you maximize vent area, or keep a straight vent? More bracing to make sure the cab is dead vs less weight and build complexity? Eventually I prioritized a few things more than others and got to a place that I thought was a reasonable set of compromises for a sub this size. The major design considerations all had to do with the front section of the drivers output. Since the sub design had to shrink so much space for everything is at a premium. I knew that the back chamber air volume would be at a premium. I decided to shrink the front volume as much as possible while still achieving something useful. I quickly realized that I would not be able to do a proper "horn" section like the SKHORN and SKRAM subs due to the space premium. I looked at straight slots and expanding vs contracting. The physical size of the driver was setting limits on the size of the slots and the air speeds due to the reduced area were getting higher than I would like. They also put the driver into orientations that were not ideal. Eventually I settled something more similar to a BP chamber than a horn or slot. It had the advantage of keeping the driver horizontal and placing the output radiators symmetric about the center of the cone, which should result in a more even pressure distribution on the cone. It also shortens the average path length for the driver energy to exit but also allowed a larger, quicker expanding area at the exit. At first I was going to use rectangular or triangle corner exits. Before deciding on a circular / triangular profile. 

At that point I was almost done with the design and had settled on this arrangement for the front section. It was at this point that I started to think about going from a hatch system with internal driver to a removable front section with a standard front mounted driver. Originally I had a hatch for driver access and the front section was non removable and solidly attached to the rest of the cab. It would've worked but the hatch was a very tight fit and it compromised the bracing in the center of the cab and complicated things. Positives for moving to a removable front assembly were: The driver hatch could be eliminated, which simplified the internals and bracing. Driver access would be easier. Also the cab could be run in a normal vented configuration without the front assembly. Negatives were that it complicated the front assembly, reduced the driver excursion clearance somewhat and reduced the size of the bracing and possibly the solidity of the front section. I decided the positives outweighed the negatives so the design was updated one more time to what is seen here. 

This cab is 24x28x30" which is quite small for a 21" design. It is actually slightly smaller than a half of a SKHORN but the design ended up much different. For comparison an Othorn is 24x36x36 and a SKRAM is 24x32x36. This cab shaves 4" off of the depth and 6" off of the height compared to a SKRAM and does so with a significant weight savings as well. HIL is always in effect so this cab does give up efficiency & sensitivity to the larger SKRAM and OTHORN cabs. There is simply no way around that. I chose to focus most of the CKRAM performance into the deep bass rather than using the front volume to turbo charge the mid and upper bandwidth like with the SKHORN and SKRAM. The gain of the front volume is probably only going to be 1 or maybe 2dB compared to a regular vented cab if I get lucky. Testing with and without the front assembly should be interesting. This tradeoff was acceptable to me in order to keep smooth response and powerful output in the deep bass from a cab this size.  

[peniku8]

Interesting approach, I like how you ended up with almost exactly the same total system volume as I did. I just couldn‘t get below 320L, no matter how much I tried.
Could you attach a FR graph? I‘m too lazy to copy the inputs over to HR

Will we see the typical bandpass FR with this type of front chamber? You can basically look at the driver from the sides, I wonder if you‘ll be able to see that in an upper bandwidth polar pattern.

Anyways, now you‘ve got two projects without cabs. I‘d build you one if I lived in the states!

[klipsch]

I was thinking it would be great to have a functional speaker stand.  The dimensions of this ckram would work quite well.

Looking forward to a FR graph.

Thanks for sharing Ricci!

[Droogne]

23 hours ago, Ricci said:

How did you arrive at the S1 and Par parameters? 

S2 is with the added cone surface (~350cm²)?

[Ricci]

21 hours ago, peniku8 said:

Interesting approach, I like how you ended up with almost exactly the same total system volume as I did. I just couldn‘t get below 320L, no matter how much I tried.
Could you attach a FR graph? I‘m too lazy to copy the inputs over to HR

Will we see the typical bandpass FR with this type of front chamber? You can basically look at the driver from the sides, I wonder if you‘ll be able to see that in an upper bandwidth polar pattern.

Anyways, now you‘ve got two projects without cabs. I‘d build you one if I lived in the states!

I wish we were closer I'd take you up on that! Especially after seeing those shop pics. 

Honestly I have a lot of educated guesses about the top end behavior of the front section but little in real data. It should be predictable in the bass range. Outside of that who knows...Disc diameter is 20.5" so there shouldn't be any cancellation issues until well outside of the sub range. 

Here's a couple of FR images.

Akabak sim with the front section separated into 8 waveguides each with 12 or so segments and the 4 vents all modeled separately. The comparison FR is from HR. The two are really close other than Akabak exaggerating the resonances up top quite a bit. It'll be closer to HR in reality. Driver is the NSW6021-6.

Here is the HR response with and without the Edge data added. 2 volts at 1m in halfspace sim. 

Vent tuning comparison. 

Driver Comparison. NSW6021-6, SAN214.50. 21DS115-4, 21IPAL. 

[Ricci]

5 hours ago, Droogne said:

How did you arrive at the S1 and Par parameters? 

S2 is with the added cone surface (~350cm²)?

This question has a rather complicated  and long answer.

The simplest way I can explain it is that the S1 value is the minimum cross sectional area of the air volume in front of the driver cone at the center dust-cap. Technically there really isn't an S1 as typically defined since the driver radiation expands in 360 deg in that axis. The driver is placed in an OD rather than ND arrangement but the exit is closer to an ND arrangement. In some ways it is also similar to a traditional ported bandpass chamber. I modeled it all three ways and surprisingly there is little difference between them. Mostly due to the very limited size and "tuning" being well out of the sub band. 

The S2 is the area of a circular cross section centered at the cone / dust-cap joint. 

S3 is the minimum area of any circular cross section seen by the driver as the output is considered to expand from a central point. 

S4 is the area at the point that the bracing and support structures end. 

S5 is the area of all exits at the outer surface including the round over. 

All distances are actual from measurements inside the model. 

I design backwards from most. After a preliminary rough sim that gives something to aim for in about the size that is available to work with, I design the speaker. The initial sim part is quick and easy. The physical design of the speaker drives the inputs back into the simulation after this. This is a much more difficult much longer process. The dimensions, sketching and values derived are all done by 3D modeling and analysis. It's an iterative feedback looped process until a final design is completed. 

[peniku8]

I think baffle gain will be more. Like I've suggested in my PM, my sims were much closer when I placed two cabs in Edge to replicate the half space environment. I simply made the cab twice as high and placed the mic in the center.

I like how you treat the 21DS115 as 2 Ohm driver. That beast is so damn efficient!

[Ricci]

20 hours ago, peniku8 said:

I think baffle gain will be more. Like I've suggested in my PM, my sims were much closer when I placed two cabs in Edge to replicate the half space environment. I simply made the cab twice as high and placed the mic in the center.

I like how you treat the 21DS115 as 2 Ohm driver. That beast is so damn efficient!

Since the 21DS115 is 2.2 ohm DCR I usually do consider it as 2ohm nominal. Voltage sensitivity is a tricky subject since speaker cab impedance is all over the map. Still useful but not as much as many make it out to be IMO. With that said I was trying to maintain 95ish dB at 30Hz from this cab with nominal 1W voltage. The size makes it difficult. 

The Edge sim was for a single cab. I threw it together quickly. A mirrored cab shows about 1dB of gain at 40Hz another 1dB over the single cab by 70Hz and an additional 1.5dB 100-200Hz.  

[Ukko Kari]

Awesome work as always Josh, refreshing to see out of the box thinking.

[SME]

Very interesting.  How well do you think it will work to limit output of high-harmonic distortion and noise?

[Jesal]

Excellent job Josh on packing this much deep bass for such a small size. Just what the Doctor ordered. Two nearfield corner table CKRAMs coming right up 😱

[klipsch]

How well would these integrate with the skrams and skhorns if they are all overlapping the same frequencies? Asking for a friend... 

[Ukko Kari]

7 hours ago, klipsch said:

How well would these integrate with the skrams and skhorns if they are all overlapping the same frequencies? Asking for a friend... 

 

🤣

[Ricci]

On 3/7/2020 at 3:12 AM, SME said:

Very interesting.  How well do you think it will work to limit output of high-harmonic distortion and noise?

Not sure yet. Likely to be better than a direct radiator but not as good as my other designs since the driver is so close to the exits and visible from some angles. 

[Ricci]

On 3/7/2020 at 8:55 PM, klipsch said:

How well would these integrate with the skrams and skhorns if they are all overlapping the same frequencies? Asking for a friend... 

Should be fine. They are more similar than different. This one just has much less path length and gain from the front section to keep the size down. 

[Tahoejmfc]

On 3/4/2020 at 8:47 AM, Ricci said:

Only concern of mine is the front and possible vibration.  I am thinking of making this smaller box for another 21SW152 I have... I will need to mix it with my original SKRAM and see what happens. Maybe it will hit different and fill in what is lacking in my SKRAM.  Not sure if I will need to mess with the time delay for the different subs in combo and the distance differences in the design. I have never really felt much of a chest pounding with the SKRAM, but gosh damn it hits and is loud. It's pretty incredible how the SKRAM hits but with very little air movement out of the ports. 

CKRAM print.pdf 110.13 MB · 2 downloads

 

[Ricci]

I think you are the first person who has said that the chest pound is missing. Is it possible that what you are looking for is occurring in frequencies that are above the sub band? What some guys would refer to as the MBM range (100-200Hz) Are you using them outdoors, at home, big venues? Is it possible that you have acoustics issues contributing? 

[peniku8]

Well if he is using the cab at home it's probably the room. I have a null at exactly 63Hz in my room, so I also feel like the SKHorn is lacking something.

When there are nulls, there are peaks somewhere else. If I stand in the 63Hz peak it feels like I'm in a fist fight with Mike Tyson.

[klipsch]

2 hours ago, peniku8 said:

Well if he is using the cab at home it's probably the room. I have a null at exactly 63Hz in my room, so I also feel like the SKHorn is lacking something.

When there are nulls, there are peaks somewhere else. If I stand in the 63Hz peak it feels like I'm in a fist fight with Mike Tyson.

Haha exactly. 

Using the room simulator in REW really helped me when I was designing / building my room. 

I originally had my first row of seating in a massively bad null area. Moved it forward 12 inches, got better... Moved it 18 inches and all nulls were basically removed. Moving the speakers did nothing in the simulation as that spot in the room was just a bad room node spot based on dimensions. Adding different treatments helped that original location's response, but why start with something poor like that if you don't have too? 

[peniku8]

1 hour ago, klipsch said:

Haha exactly. 

Using the room simulator in REW really helped me when I was designing / building my room. 

I originally had my first row of seating in a massively bad null area. Moved it forward 12 inches, got better... Moved it 18 inches and all nulls were basically removed. Moving the speakers did nothing in the simulation as that spot in the room was just a bad room node spot based on dimensions. Adding different treatments helped that original location's response, but why start with something poor like that if you don't have too? 

I have not come across that feature yet. So you'd say it's reliable? Did it predict the actual room response/dispersion well enough?

[klipsch]

2 hours ago, peniku8 said:

I have not come across that feature yet. So you'd say it's reliable? Did it predict the actual room response/dispersion well enough?

It certainly did enough for me. The in-room measurements match quite well to the simulation too. Sadly I found the feature after I built and finished the riser for the second row. 

[kipman725]

cool design! it reminds me a bit of my own subs.  Mine differ in that its more of series design, with a port meeting an isobaric driver in the chamber with the cover plate on it.  My simulations results (just reflex) where not matching in the upper bass, I hadn't thought about modeling it as a 6th order bandpass:

https://www.diyaudio.com/forums/subwoofers/345807-isobaric-sub-based-pyle-plpw15d.html

Looking at the plans for the CKRAM I don't quite understand how your driver cover attaches, as it appears to me that the design is saying to screw into the end grain of the ply?

[Ricci]

2 hours ago, kipman725 said:

Looking at the plans for the CKRAM I don't quite understand how your driver cover attaches, as it appears to me that the design is saying to screw into the end grain of the ply?

The primary mechanism will be 8 bolts into the baffle face with inserts, or screws could be used. 3 other screws would come into the end grain of the cover struts. 1 each through the 2 sides and the top. The holes would be pre-drilled to avoid splitting. A 4th would be pre-drilled and run at a 45deg angle into the vent panel. The main holding force would be the fasteners into the baffle. The other 4 screws would mainly be to keep those 4 struts from vibrating. 

As shown here.