Designing a custom bass cab - third iteration, 2x8"+15" compact neo 3-way...

[alexclaber]

Update - things have changed quite a lot, not least of which because I'll be acquiring a 5-string. The latest design is down at the end of the thread.

 

---------------------------------

 

I'm not quite sure how I got so involved in this, blame greenboy...

 

Anyway, a week or so ago I downloaded WinISD Pro - http://www.linearteam.dk/default.aspx?pageid=winisdpro - and starting kicking around various drivers, in an attempt to design a compact cab that maintains the clean accurate sound of my Acmes, sacrifices the extreme LF extension (not necessary as a 4-stringer) for greater sensitivity and output, increases max SPL across the frequency spectrum (particularly high mids and treble) and reduces the weight.

 

In particular I've been looking at the neodymium drivers from BMS Electronik - http://www.bmspro.com/

 

Alex

[Rocky McDougall]

What frequency response range are you trying for?

[alexclaber]

That's a slightly tough one to answer - it's more a question of juggling a whole bunch of different woofers, cab volumes and tunings, and comparing the sub 100Hz response curves, the power handling and max SPL and the group delay.

 

The Acmes are -3dB @ 41Hz, -6dB @ 31Hz, which is greater extension than anything else on the market. The -6dB figure is irrelevant to me as my bass stops at 41Hz and only 3dB down at 41Hz is more bass than I need at stage volumes with the benefit of boundary reinforcement.

 

Alex

[saxofunk]

Gosh. Wish I was in the market for a couple well loved Acmes.

 

Best of luck to you Alex. The information you share on forums tells me that if someone can pull off a well done homemade cabinet, it's you. Let us know when you're going into mass production.

[Paul K]

I've also been toying with winISD to build a ported speaker cab out of a 16 or 18 inch dia. floor tom and a neo speaker. Still can't find a 4 ohm 12" neo to do it with, and I don't want to use two 8 ohm speakers. Very interested in how yours turns out.

[alexclaber]

First attempts were with the BMS 12N630 - http://www.bmspro.info/index.php?show=item&usbid=10285&id=5047865 - which goes very low, requires hardly any cab volume and has so much Xmax that it'll handle serious power even right down at the bottom. Unfortunately on modelling it I found that the calculated sensitivity was closer to 90dB@1W1m which is the same as an Acme 10".

 

The woofer is in a 40 litre enclosure tuned to 48Hz. The woofer has a highpass 4th order filter @ 30Hz and lowpass 1st order @ 800Hz.

 

The logic was that the relatively high woofer enclosure tuning allied to the low Fs woofer would give a huge amount of power handling in the 40-60Hz region, and the hump in response could be flattened by the high pass filter in my power amp.

 

This allows the woofer to handle full RMS power (600W) right down to low E (41Hz) apart from a dip to 500W at 65Hz. The mid is rated at 200W thermal but can handle 1800W RMS whilst staying within Xmax. The lowest max RMS SPL for the mid and woofer in this arrangement is 118dB thermally limited - max peak SPL will be much higher.

 

Being a true woofer (if not subwoofer) this speaker doesn't go high and requires some kind of midrange driver. For that I decided to use the BMS 8CN252 - http://www.bmspro.info/index.php?show=item&usbid=10283&id=5059989 - which is a nice speaker. 95dB@1W1m, 200W AES, 4mm Xmax, and response from 70-3000Hz and then the coaxial tweeter takes over up to 20kHz (crossover @ 1900Hz). I decided to put that in a 7 litre enclosure tuned to 170Hz. The mid has a highpass 1st order @ 250Hz. The mid is padded back by 5dB to balance the output.

 

Also by porting the mid enclosure this gives a really big hump in response and power handling, but the response is then flattened by the passive high pass filter. The mid would share the midrange with the woofer for more output (giving a broad hump in on-axis output) and the improved resolution of a smaller driver. The lack of thermal power handling (200W AES) for the mid would be a non-issue because midrange output from bass guitars is rarely sustained being predominantly part of the attack/decay phase and not in the subsequent sustain/release portion of the envelope.

 

---------------------

 

I then went through some other cab designs due to concerns expressed by various knowledgeable individuals and came full circle to this 3-way compact (same size as a Low-B2) 12" plus coax 8" cab, with some significant tweaks:

 

The big thing is that I replaced the 12N630 with the 12N620 - http://www.bmspro.info/index.php?show=item&usbid=10285&id=5047866 - a very different kind of woofer. This woofer handles 500W AES instead of 600W, but has a much smaller Xmax of 5.5mm. The big plus though is the gain in sensitivity to 95dB@1W1m. Of course, you never get something for nothing - you lose a whole bundle of extension but I don't play 5-string and don't believe the lowest note fundamentals are critical live with a 4-string - though they are nice...

 

The enclosure is up to 50 litres. The tuning frequency is down to 45Hz. In combination this gives an 'extended bass shelf' type of alignment (where a woofer that is rolling off higher up is put in a larger box than typical and then tuned low and the port resonance causes the response to rise again and then plummet). Downside is the worse group delay (time lag on certain frequencies) but it's pretty low so hard to hear.

 

The woofer has a 2nd order lowpass @ 1000Hz. The BMS 8CN252 in a smaller 3 litre enclosure tuned to 200Hz, 2nd order highpass @ 400Hz. The crossover point is something to tweak with but I suspected that the overlap of the drivers could provide benefits in the increased midrange output on-axis equally a clearer sound off-axis (where most people are).

 

Alex

[alexclaber]

This cab is 93dB@1W1m within a few dB and then -6dB @ 40Hz.

 

Thermal power handling 500W RMS below 200Hz, 200W RMS above 200Hz. Mechanical power handling (excursion limited) drops to 250W @ 65Hz but it's generally pretty high.

 

Group delay is an important issue when it comes to good bottom - unfortunately the extensive use of porting to maximise power handling causes some possible issues here.

 

Threshold of audibility for group delay:

 

8 kHz 2 msec

4 kHz 1.5 msec

2 kHz 1 msec

1 kHz 2 msec

500 Hz 3.2 msec

 

However in the bass region group delay has to be in the 20ms+ region to be considered muddy.

 

Alex

[alexclaber]

The next challenge of course, is juggling box dimensions. Basically your port has to be able to move enough air without the air having to go above 5% of the speed of sound (otherwise it gets noisy) and this is dependant on the air the woofer can displace (Xmax x Sd = Vmax).

 

This determines the cross-sectional area of the port(s). To maintain the same tuning, the length of the port increase in proportion with the area of port. So if your woofer can move a lot of air, then you need a port that is big in area, which demands a port which is long in length, which makes the port take up a lot of cab volume.

 

And of course, not only does the volume of the port have to fit into the cab, but also the length of the port has to fit in the internal dimensions and have room to breathe.

 

Because the porting for the 8" works in the low midrange area it has to be positioned at the front and as close to the 8" as possible to maintain a near-point source. The 8" needs to be in its own sub-enclosure so it can be tuned accurately and doesn't get battered from the high pressures in the woofer enclosure, thus it would seem logical to make the woofer porting run between the main enclosure walls and the 8" sub enclosure walls.

 

Comme ca:

 

http://www.thehumanzoo.com/alex/Speaker%20Plots/8+12%20Front.JPG

 

The vertical dark rectangular areas are the ports for the main enclosure, whilst the small triangular dark areas are the ports for the midrange enclosure.

 

Alex

[Rocky McDougall]

Alex, I have thought of building a cabinet with a adjustable port. Maybe one tube sliding within another tube so the length could be tailored exactly. What are your thoughts on front or rear firing ports?

Rocky

[getz out]

Front firing ports look louder, but they do not have an audible difference from rear firing ports. However, if it looks louder, it usually is louder, right?

[alexclaber]

Originally posted by Rocky3840:

Alex, I have thought of building a cabinet with a adjustable port. Maybe one tube sliding within another tube so the length could be tailored exactly. What are your thoughts on front or rear firing ports?

Rocky

A popular approach is to use a plastic tubular port, cut it too long and then trim it down until you find the right frequency. Go too far and it's easy enough to swap with a longer one.

 

One port sliding within another would mess with the cross sectional area and also cause turbulence.

 

It doesn't matter where you put the port unless it's on the back of a huge cab where the cab rear acts like a room boundary - but that would take a serious wall of cabs. Best to put them where they fit!

 

And I agree with Maury - front ports definitely look (and feel) louder! But whatever...

 

Alex

[getz out]

Originally posted by C. Alexander Claber:

And I agree with Maury - front ports definitely look (and feel) louder! But whatever...

I knew you missed me, you old, married fart!

[alexclaber]

Originally posted by getz76:

I knew you missed me, you old, married fart!

Takes one to know one!

 

Alex

[alexclaber]

Now here's where I intend to make the big weight saving: Composite panel construction!

 

Basically I'm looking to get hold of some foamcore plywood, which typically has 3mm hardwood ply facings with stiff and light foam in between. Available in 1/2", 3/4" and 1" thicknesses, all with the same facing thickness, and a weight per 4'x8' sheet of 23, 26 and 29lbs respectively.

 

The joy of this is that in a piece of plywood most of the stiffness (though not the ultimate strength but it's stiffness that we need to maximise) is provided by the outer plys. So if you remove the inner plys and replace them with something light that maintains the outer spacing you keep most of the stiffness and lose over half the weight.

 

This means that this Acme Low-B2 sized cab with 500W 12" neo woofer and 200W 8" neo coaxial mid/high could weigh in at under 30lbs. Oh yeah...

 

Alex

[Rocky McDougall]

The aircraft industry uses sheets of plywood (honeycomb interiror faced with thin veneers) for custom furniture. Very light and very stiff and strong.

Rocky

[alexclaber]

There seem to be two challenges with the composite panels - the first is actually getting hold of a single sheet of the stuff! The second is the joints - not as simple as glueing plywood together.

 

Problems with this design:

 

1. Similar sensitivity and power handling to the Acmes - less max low SPL but more max mid/high SPL.

 

2. Complicated! 3 way passive crossover required and relatively complex construction with sub enclosure, slot ports and corner ports.

 

3. Expensive - 12N620 and 8CN252 cost 189 and 229 EUR respectively, that's over $500. Hmmm...

 

4. Interaction of 8" and 12" in midrange area will cause unpredictable response (both frequency and polar) due to difference in phase response.

 

Pros:

 

1. Cleaner and louder mids and highs.

 

2. Very light ~ 30lbs

 

Not good enough for the time and money outlay. Need rethink...

 

Alex

[getz out]

I don't know... with minimal sacrifices, you could get a good package in the Epifani UL112, right? Might be the way to go... you won't get everything you are looking for, but it will get you close. Just a thought.

[alexclaber]

I think you're right if I were to stick with that kind of design. But on further pondering, I'm realising that my needs have changed since I was a non-singing bass player.

 

What I really need is a cab that I can use as a bass (and possibly vocal) monitor whilst I'm singing at gigs but also have enough power to play without PA support at smaller venues.

 

And having modelled a whole pile of drivers, here's the one to do it - the BMS 15CN680 - http://www.bmspro.info/index.php?show=item&usbid=10282&id=5060110

 

http://www.bmspro.info/photos/bmspro_info/463765328.jpg

 

Notable features of this driver are:

 

Fs (resonant frequency) = 38.5Hz

Pe (power handling) = 500W RMS

Xmax (maximum linear excursion) = 5.5mm

SPL (sensitivity) = 96dB@1W1m

Net Weight = 5.2kg

 

It's a coaxial design with 1" compression tweeter in the middle that handles 80W RMS (450W peak) which fires through an 80deg by 60deg waveguide (which should also help the dispersion of the woofer) and the recommended crossover point is 1.2kHz.

 

For maximum bottom it needs a 100 litre enclosure but I've tried various sizes and I think even in as small as 60 litres it'll give sufficient bottom.

 

Alex

[TimR]

What would the effect of having smaller but multiple ports have?

My Trace-Elliot 1x15" has two ports. I think the idea is that you can fill one (or both?)to tune the cab to a different frequency. I've never played with this idea, because I've only just remembered about it reading this tread, but it might work for Rocky's tunable cab idea. Maybe you can have alternate tunings for 4 and 5 strings.

[alexclaber]

The way to think of a ported cab is as a mass on a spring:

 

The volume of air inside the cab acts as the spring - the greater the volume of air the softer the spring.

 

The mass of air within the port represents the mass - the bigger the port, the greater the mass.

 

However, the greater cross sectional area of port, the stiffer the spring is, simply because there is more cab air meeting port air.

 

Consequently a short narrow port can be tuned to the same frequency as a long fat port.

 

Unfortunately the air speed inside the port has to be kept below 5% of Mach to avoid port noise - and the narrower the port, the greater the air speed.

 

What this all adds up to is that multiple ports act like multiple masses on the same spring.

 

I remember those TE cabs - they had two circular ports and you could leave both open for 45Hz tuning or close one for 30Hz tuning. What happened here is that by closing a port you halved the mass on the spring, stiffened the spring slightly due to the smaller port area but then softened the spring slightly due to the marginally increased cab volume (from the unused port). Great idea, shame the new cabs don't have it.

 

Alex

[Paul K]

Your idea of using composite panel is brilliant. Hope it works! A small piece of this composite panel might be just the trick for my floor tom shell cab: use the composite to cap the ends (or at least the end without the speaker) I'm thinking of retaining the lugs and rim as a way to attach the front and back of the cab to the drum shell. Would make the joints in your design a non-issue.

[alexclaber]

Paul, check out this BMS 15" - it'll fit in your 16" or 18" floor tom and should blow away any 12" speaker. I think there's a 4 ohm version of a near identical model and there should be a 4 ohm one of these coming out.

 

I'd like to also use the pre-stressed panels as in the Acmes but I can't yet think of an easy way to do it.

 

Currently being bothered by crossover and EQ related thoughts... There's a lot to be said for a cab with such inherently flat response as an Acme - and it's not like I need more SPL now I'm singing. Wild goose chasing, am I?

 

Alex

[Tom Capasso]

I'm sure you'll have fun constructing something like this. My fear would be that all the charts and specs would get you close, but not close enough. That might mean port changes (not too bad), or something more design-upsetting.

 

PA systems attempt to provide full range through one system. Maybe for your needs you could stick with your current rig for bass and get a small PA thing (or even a full range combo) to use with your amp for rehearsals - something like the Carvin acoustic guitar amp?

 

Best of luck

Tom

[Paul K]

Well, sure, you could just buy something. But what fun would that be?

[Paul K]

I finally scored a drum shell for cheap on eBay. It's a 16x22 dDrum kick drum; bigger than I'd hoped for but does have those adjustable kick drum legs up front. It'll be stable, and the legs can be extended enough to tilt the drum/cab back. Yes, it barely fits in the back of my Stealth, but relocating the CD changer would make it an easier fit. I'll temporarily be using the EV series II 15" speaker out of my Peavey Hernia for the prototype stage. If I dig it, I'll probably spring for an emminance basslite 15. It's 4 ohms, which is what I need.

 

Another advantage of the kick drum instead of the 16" floor tom is that it has the tom mounts in place. Might be a cool platform to plant (via velcro?) a small and light head. (Not appropriate for my soundtech ps800 and preamp.)

 

Anyway, C.Alex: any luck on obtaining the composite panel? If so, tell me how you scored it. This project is moving slowly, but it is moving. And I think it's gonna look way, way cool.

[alexclaber]

On to iteration 3 (at least the third I've written up here, probably the 30th in reality!):

 

Eminence Kappalite 3015LF in a 100 litre enclosure tuned to about 37Hz. I'm having the 3015LF firing into a side chamber from a slanted baffle - partly to keep the cabinet width/height down, partly to acoustically filter out the highs. 96dB sensitivity; -3dB @ 53Hz; -6dB @ 40Hz; -10dB @ 31Hz; max excursion @ 60Hz and 31Hz; 450W RMS, 900W peak to hit Xmax; 2nd order low pass filter at 400Hz.

Then on the front, a pair of 8" midrange speakers, one above the other, to give enough sensitivity and power handling. Precision Devices make a nice neo model with pretty smooth response between 200Hz and 3500Hz, 150W power handling and 4mm Xmax. These each be in individual 5 litre sealed enclosures made from 220mm diameter plastic pipe backed with 1/4" ply. 97dB sensitivity, 300W RMS thermal limit. 2nd order high pass filter at 400Hz.

 

Then mounted between the midranges, in a D'Appolito confuguration, four 5"x2" horn piezo tweeters in a vertical array 8"Hx5"W. That should give 200W power handling total and 97dB sensivity average (series/parallel wiring).

 

Overall cab dimensions are 900Hx400Wx450D (mm) - the narrow width gets it in the boot and through doorways comfortably and stacks neatly with the Acmes if they're ever needed for big gigs without SR - will be using edgemount 4" casters for an easy load-in.

 

What I'm looking forward to with this is the way the 8" mids will actually carry the majority of the tone, with the 15" just filling out the bottom and keeping the 8"s running as clean as possible by minimising excursion and power compression, and then the piezo array filling in the treble snap from 3kHz upwards. These 8"s are much flatter than most I've found - they're within a 2dB range (not plus/minus) from 200-900Hz, then the centre of that range ramps up 2dB through 1kHz (presumably as the polar pattern narrows) and then they stay within a 2dB range until 3.5kHz (at which point the tweeters take over). Or to put it another way, a very smooth example of plus/minus 3dB 200-3500Hz. The D'Appolito spacing of the midranges causes mutual coupling below 850Hz, further flattening the response and raising sensitivity, and the tall narrow front baffle should help control vertical dispersion and improve horizontal polar pattern.

 

I'm looking into minimising the weight by going to a spaceframe with stressed panel construction, using 1"x1" balsa for the frame and 1/4" plywood for the panels. That should drop the weight below 50lbs. Doing this wouldn't be feasible for production cabs because it gets incredibly complicated and adds hours to the build time - but it's a fun challenge to design the strongest most efficient structure whilst still making it possible to build, possible to tune the port when completing the build and possible to maintain (no good being unable to swap drivers in the event of a mishap).

 

Parts costs will be around £400 - not cheap by any means but I imagine a cab like this would retail at an unpalatable price... The 15" woofer is as good as anything in any bass cab, the 8" mids are far more expensive than any equivalent bass cab speakers, the piezos are a bargain, the balsa block and 1/4" ply are pretty cheap, the crossover has some expensive components like the 14 gauge series inductor on the woofer's low pass filter, and the hardware will be solid and quite pricey - like the 4" heavy duty casters.

 

I do have a small stockpile of speakers and parts that I've been collecting over the past year so I think I'll build a prototype cab with these parts, and basic high pass filtering on the midranges, see how it sounds and see how the unusual construction works.

 

Alex

 

P.S. Paul, no luck on the composite panels here in the UK but I'm pretty sure they're available in the USA via the power of google - try searching for foamcore ply, or failing that okoumé ply.

[Paul K]

I have searched, and can buy a whole sheet of 3/4 inch thickness composite panel for the likes of $180. (I didn't bother asking about shipping...) Haven't had any luck with local boatyards willing to sell me a scrap or two of a piece 2'x2' in size. After the holidays I'll search more in the local lumber yards for a composite type material, then probably settle for some 1/2 inch ply.

[alexclaber]

$180?!! Definitely not worth it.

 

1/2" looks the way to go!

 

I'm very hopeful about the combination of balsa structure and 1/4" skin - the skin will be fixed to the balsa along all edges and corners but will be left separate elsewhere. Then I'm going to gently hammer in wedges to pre-stress the skin outwards and the frame inwards and glue it once all the wedges are solidly in.

 

I wish the Kappalite 3015LF came in 4 ohm impedance. As it is I'll have the choice of 550W (one channel) or 2000W (bridged) into 8 ohms but it precludes getting full power from any standard heads.

 

Alex

[TimR]

Bump

[alexclaber]

Getting the fine details of the design sorted, spec'ing out wood cuts and specific components. Based on my sums I'm looking at the cab weighing exactly 50lbs - not bad considering the size and amount of components.

 

Alex