uploads/19/Martin_Audio_F2B_Model_1.pdf" rel="nofollow - uploads/19/Martin_Audio_F2B_Model_1.pdf

uploads/19/Martin_Audio_F2B_Model_2.pdf" rel="nofollow - uploads/19/Martin_Audio_F2B_Model_2.pdf

uploads/19/Martin_Audio_F2B_Model_3.pdf" rel="nofollow - uploads/19/Martin_Audio_F2B_Model_3.pdf

uploads/19/Martin_Audio_F2B_Model_4.pdf" rel="nofollow - uploads/19/Martin_Audio_F2B_Model_4.pdf

uploads/19/Martin_Audio_F2B_Model_5.pdf" rel="nofollow - uploads/19/Martin_Audio_F2B_Model_5.pdf

uploads/19/Martin_Audio_F2B_Model_6.pdf" rel="nofollow - uploads/19/Martin_Audio_F2B_Model_6.pdf

uploads/19/Martin_Audio_F2B_Model_7.pdf" rel="nofollow - uploads/19/Martin_Audio_F2B_Model_7.pdf

uploads/19/Martin_Audio_F2B_Model_9.pdf" rel="nofollow - uploads/19/Martin_Audio_F2B_Model_9.pdf

uploads/19/Martin_Audio_F2B_Model_10.pdf" rel="nofollow - uploads/19/Martin_Audio_F2B_Model_10.pdf

What would be a good modern driver for this design?

It says "alternative driver", no guarantee, take it with a tablespoon of salt. There are also warnings on several pages urging the prospective builder to evaluate driver clearances before attempting a build. You have been warned. 

the L15-554K does fit easily. the other two RCF I mentioned don't.

It may well do externally but a PD1550 will not fit this plan, including the driver rebate you have about 155mm depth on this drawing, a PD1550 has a total depth of 160mm, I'm convinced this plan is not totally accurate to the original or the stretched ones.

 

No it's not an issue, it's a good plan and while you are doing it you can get the rear chamber volume correct for the driver you are choosing, though I an not convinced it is 100% accurate to the original it is a very good starting point to make something very special.

There is another trick... and that's to route out a few mil in the access hatch.

How realistic are the quoted response figures ?

Has anyone tried extending this to 1220mm height, to fit across 2x modern subs?

Looking at drivers, sure it would not be difficult find drivers that match the important TS parms, and adopt chambers accordingly.

djeddie wrote: Well the outside dimensions match my 'extended' F2B's. And as I've said on this forum before, try them with the PD1550!

Might be good for <150hz, but would it really excel above 200hz?

No, nothing of particular interest...... i just acted as a strawman in this case.....

Jimmer wrote: Flat upto 250

Can concur with that.

2 drivers @ 4ohm

2 x F2B, 4 drivers @ 8 ohm

Once I get my sim parameters sorted, im happy to sim a number of different drivers. I will also go about increasing the depth of the rear chamber to accommodate more drivers.

Here is a link to the sketchup drawing https://www.dropbox.com/s/1quejvp5zda1b7s/hornresp.skp?dl=0

Any one interested in making this cab as 2x singles?

Robbo wrote: No point in building singles when the original is very compact in the first place.

Actually this is one of the reasons I never utilized this cab.

A double for me needs to be at least 1200H, so it can lay across the top of 2x subs.

Which is why THL828 is great size, even though it can't do what this cab will.

One question, the current drawing leaves about 3mm of clearance behind the driver, is there going to affect performance? 

Current sketchup drawing showing 10mm clearance for driver and new method of finding hornresp inputs.

New hornresp inputs

1 x F2B at 4 ohm Black=PA100-375 Bass, Grey = MB15N405

2 x F2B at 8 ohm Black=PA100-375 Bass, Grey = MB15N405

Link to sketchup drawing https://www.dropbox.com/s/wmyzdnscotz79p0/hornrespv2.skp?dl=0

JBL always recommended 76mm (three inches for the metrically challenged) minimum gap behind their 2226 for adequate cooling, FWIW.

I know modern drivers often have extra cooling between the chassis and suspension, so that will help, but my gut instinct is that only 3mm is still way too small a gap.

madboffin wrote: I'd recommend rounding off that sharp corner in the first section of the folded horn, otherwise there's a risk of chuffing noises. The original box used a shaped piece of wood there - see my note further up the thread.

Thanks for the idea, I have updated the drawing with a 1cm radius bevel, see the image below. 

As far as potential rattling, based on my past experience I don't think it will. I have 6 double punishers and 4 xtros that have ~100 gigs under their belt with 0 rattles, cabs are still solid, despite being made of 15mm ply. The F2B being made of 18mm has to count for something, also, I have drawn joins so the panels can easily be screwed and glued together. The cab should be rock solid, especially around those cavities.  

madboffin wrote: If you search the forum there is quite a lot of information on these boxes that might be helpful, although it will take a bit of finding. I used to work for Martin Audio and built some of the original boxes. By the way -  did you know that the late lamented Dave Martin grew up as a farmer's boy, somewhere near Bunbury?

I have done a bit of reading, a lot of interesting posts, interesting how accurate peoples descriptions are despite there being no publicly available plan, even in posts from a decade ago. 

Sketchup: https://www.dropbox.com/s/bzn9hki99oc8si8/v0.1b.skp?dl=0

changing to 45 degrees will make little difference acoustically. and you could make that part out of a piece of 2"*3" rather than three seperate pieces.

the offcut might be just big enought to make the triangle for the corner opposite too!

if you make all the parts in red out of solid timber that baffle is going to incredibly well braced. you could make that 6mm baffle piece even thinner to wrangle 2mm or 3mm extra. I assume the original design had flares expaning the letter box in the vertical dimension in that first few centimetres of the throat.

maybe make the 18mm baffle piece smaller and removable with the driver. that way if you ever want to use a different driver with different sized cutout, bolt pattern, or want to increase excursion clearance above 18mm, you can just make a new baffle. fix it into the solid timber points with eight large threaded inserts and should be solid.

you need to be careful rebateding that much out of the panel behind the driver. going from 18mm to 8mm reduces the stiffness by a factor of about 11. it's not a massive area but it could be enough to reduce the Q of the rear enclousure and affect the system efficiency. How about using an aluminium plate for the access panel - aluminium is about ten times stiffer than plywood.

its an extreme, but for an example system - if Qmc falls from 20 to 5, system efficiency goes down from 46% to 35%, which reduces thermally limited max output by ~2dB

I would normally want to stiffen a panel in the rear chamber, not weaken it. The pressures are quite significant.

8mm aluminium would be good and act as a heat exchanger.

also you need to place some recessed bar handles - only place they can really go is on the rear chamber. fit them with gaskets and mounting plates to get them totally airtight and stiff. and even more heatsinks!

And?

  Just recently, I bought 4 of the F2Bs unloaded and now I am debating what to load them with. I d like to ask you guys to help me out to have a good idea of what would be the best ( most cost effective as well of course) option.

 I have two pairs of PD156s which I could use in two of  them , they will just about to fit with a little modification to the mounting area in the box. What do you guys think how will this sound ? 

 I also have a pair of RCF L15P-200/a   which i m thinking to fit in the third one and maybe get a pair of these to the forth F2b.

 I understand that the above may not be ideal and of course all boxes with the same matching drivers are the real deal. This may be on the radar next year but for now i need to work with what I have. 

  Whats do you guys think ? 

Also a good reason for 1/4 pack.

I don't personally think ythe F2B needs redesigning, to the extent it now becomes something different. It's a good box, it is what it is. And I applaud people like jimmer who use it for what it is. 

those drivers won't be terrible but you would do a lot better getting eight matching:

http://www.loudspeakerdatabase.com/RCF/MB15N405" rel="nofollow - http://www.loudspeakerdatabase.com/RCF/MB15N405

or

http://www.loudspeakerdatabase.com/Beyma/15P80Nd" rel="nofollow - http://www.loudspeakerdatabase.com/Beyma/15P80Nd

have a think about bolt patterns and whether it will be easy to upgrade to other drivers in future.

those drivers won't be terrible but you would do a lot better getting eight matching:

http://www.loudspeakerdatabase.com/RCF/MB15N405" rel="nofollow - http://www.loudspeakerdatabase.com/RCF/MB15N405

or

http://www.loudspeakerdatabase.com/Beyma/15P80Nd" rel="nofollow - http://www.loudspeakerdatabase.com/Beyma/15P80Nd

have a think about bolt patterns and whether it will be easy to upgrade to other drivers in future.

All sims are of a single f2b, (2 drivers). The hornresp sims are 2 x 8ohm drivers in parallel @ 2v (1watt total), the AKABAK sims are not referenced to any voltage. All sims use the ATC PA100-375 BASS as a reference, the ATC is grey in the hornresp sims, blue in the AKABAK sims. 

Kappa 15LF Hornresp

Kappa 15LF AKABAK

KAPPA PRO-15LF-2

KAPPA PRO-15LF-2

pd.1550

pd.1550

MB15N405

MB15N401

MB15N401

Seems hornresp sims FLHs quite well, the sims don't look that far off AKABAK.

More sims to come, ask here or PM me if want a driver simmed. 

L15P200 AK-II

L15P200 AK-II

Cannot stop thinking about the added distortion caused by cone flexing under high pressure.. at least symmetrical cone loading would be a good idea with these.

It wouldn't be my pick but someone on the previous page asked about it, I don't think you need to spend much money to find a better driver for an F2B. 

It seems like everyone have ignored Lev here.

I was wondering the same thing.

Substantially flat to 40Hz? 

HOW?

Using constant power, you will see the efficiency curve.

It does not, but I think it is likely an artifact of the simulation, one of the options for the simulation is 'mesh density', lower values increase accuracy at higher frequencies. Mesh density was low for this as simming multiple large cabs with high mesh density takes a loonnnggg time, hours... Here is a plot for 1 cab in red, 2 in blue. The one cab sim is 2v @ 4 ohm, the two cab sim is 1.41v @ 2 ohm, measured from 1m. 

MarjanM wrote: It seems like everyone have ignored Lev here. I was wondering the same thing. Substantially flat to 40Hz?  HOW?

You don't have to back up your claims if you use poorly defined terms like 'Substantially flat'. My punishers are substantially flat to 10hz ... what ever that means. 

Having said that, the -10db point is below 50hz for 2 stacked cabs. I can see how people find these to be usable as a one way bass solution. 

Tonskulus wrote: Impedance is not flat, so real input power vs. frequency is not constant either using constant voltage source as is the case when measuring typical response curves.  So it is not uncommon to see much different response curves when using constant power instead of constant voltage, maybe that is the case here as it says "power response".  Who knows, I would like to see the impedance curve. Using constant power, you will see the efficiency curve.

 

Maybe its because i'm not an electrical engineer but I do not find AKABAKs labelling intuitive. There are two types of source in AKABAK 3; Potential Source and Flow Source. I have been using Potential because to my understanding, voltage is a measurement of difference in potential. I have tried flow, and it gives wacky results, I suspect this is constant power rather than constant voltage. 

Here are the impedance curves 

bit of ripple but get better in a stack:

http://www.loudspeakerdatabase.com/Radian/2216Neo%20" rel="nofollow - http://www.loudspeakerdatabase.com/Radian/2216Neo

http://www.loudspeakerdatabase.com/Radian/2216Neo" rel="nofollow - http://www.loudspeakerdatabase.com/Radian/2216Neo

at 750usd each the radian option doesnt look tempting tbh., compard to 15ds115 or 15nmb1000?

http://audioroundtable.com/PiSpeakers/messages/17535.html" rel="nofollow - http://audioroundtable.com/PiSpeakers/messages/17535.html

http://www.pispeakers.com/Hornsub.html" rel="nofollow - http://www.pispeakers.com/Hornsub.html

If taken into account that the share one horn, S4 must be around 5000. 

With S4=5000 you get a way better response nearly flat. But also this would mean splitting the horn in two parts is not possible without making the cab wider?

Is there any theory behind it that would back this up, I think they share one mouth from S3-S4 so S4 musst be the total front area or at least more than half.

Also if I set S4=5000 i get way better results in the simulation with one or two drivers..

I doubt that but also think that just simulate it as a halve horn is not perfekt.

Would totally appreciate it if you are right because that would mean splitting them in halve and build singles would work so no reason to build doubles. Never the less from experience I am skeptical 

 

with S4=2750

with S4=5000:

Obviously it wouldn't make sense to build Single cabs with a mouth area this big. But making them a bit wider to fit existing cabs would also improve the output, the bigger S4 the better results i get in the simulation. 

Please correct me if I'm wrong, still learning  

two half cabs is exactly the same as one full cab.
if you change the shape of the half cabs you have changed the system so it's going to have different response.

the F2B was designed to be use in multiples.

Isn't that what Danley does with the Jericho cabs and, for h.f., the paraline lens?

If I'm understanding things correctly (and I'm by no means any sort of expert!) I think you are probably right to be paranoid.

Slot loading, presents asymmetric cone loading, which causes the diaphragm to flex significantly. I know this for certain, because powerful strobe lighting was a thing, in the early 90s UK rave scene! Also, I used to periodically rotate the drivers in my 215s through 90 degrees, to stop the diaphragms getting bent too far out of shape. Interestingly, the drivers were mounted with L clamps, so perhaps this practice was foreseen/intended by MA?

My theory (and please do tell if I'm talking shite!) is thus - the flexing of the cone diaphragm is intentional and key to the performance of these cabinets. It has two effects, which fall completely outside the scope, of established, conventional, speaker design principals:

1) The flexing somehow divides the 'effective' circumference of the driver, such that it behaves like a smaller driver, enhancing HF extension. Of course the narrow throat, permits a lower tuning, in a smaller cabinet also. Hence the remarkable, seemingly impossible, bandwidth and efficiency of F2B and similar slot loaded cabinets.

2) The flexing, along a single axis, forces cone breakup into a state, where 2nd/even order harmonics dominate, such that the massive amounts of HD, thus produced, enhances both output and tone, in a way that sounds pleasant. Unlike most other designs, well tuned slot-loaded FLH actually sound better when spanked hard!

With this in mind, it can be seen, that the tuning of the flare is absolutely critical. Because to avoid growling or similar, the HD produced by diaphragm flexing must be kept within the pass-band. And the energy from out-of-band HD, acoustically filtered, has to go somewhere...

Faithful reproduction it isn't...but if it sounds great, who cares!

Incidentally, having read some of your previous posts, about what you are aiming to do with your rig, I think you will find F2B very interesting. 

Its worth me pointing out, that F2 was pretty much the default rig, in the early 90's UK breakbeat hardcore/jungle mega-rave scene. That was IMO, to to this day, the largest single-genre dance scene, that ever existed, with single-arena events of 10K-50K attendance not uncommon, as well as a huge number of smaller events, every weekend. From the mid 90s on, the scene split and fragmented musically, resulting in the multi-arena/festival type formats that are more common today, in the UK at least.

Anyways, eventually Jungle morphed into what we now call Drum and Bass, so I think you can probably see, how that's all relevant, to what you are trying to do 

not sure about this idea.

I don't think that drivers with particularly flexible cones have been used in designs like these. so there has been no intention to encourage greater cone breakup

I'm not sure breakup modes do give a musically nice amount of second harmonic distortion.

If anything slot loading puts a more even load on a driver. This can allow higher maximum powers through drivers as it stops the cone from ripping at points of exceptional pressure and stops the voice coil rocking and touching the sides of the gap.

Slot loading reduces path length differences in the throat which lead to a cancellation if one side of the driver is further from the horn mouth. This produces a cancellation at a wavelength twice the diameter of the driver which is 450Hz for a 15" driver.

If you did want to produce loads of second harmonic then just add some distortion with an effect unit.

a notable thing about the F2B is that its expansion approximates to hyperbolic with T=0 which maximises such throat distortion. there's a useful diagram in Olson showing distortion levels for power/area of an exponential horn but I have never seen the equivalent for a hyperbolic horn. Hornresp can calculate distortion for single segment horns - this suggests changing T from 1 to 0 (and changing the cutoff frequency to keep throat, mouth and length the same) increases distortion by 50%.

also all horns with increasing directivity at higher frequencies are going to concentrate the harmonics more on axis than the fundamental.

a third factor depends on how we think perception of loudness falls off with reducing frequency (big discrepancy between low volume tests with headphones (20db/octave) and high volume tests). second harmonic will be perceived as louder than the fundamental when it is somewhere between approx 1% and 3%. for the F2B this equates very roughly to 50-500 acoustic watts @200Hz.

All good points - thanks both. It looks like I have totally misunderstood the mechanisms here, but I'm happy to admit, I'm way out of my depth...!

I was struggling to see how cone flexing could not cause HD, because essentially flexing is breakup, just in a different axis. However I can see that if the total volume of air in front of the diaphragm remains the same as the cone flexes, the effect will be neutral. As you pointed out, in fact the flexing will probably help reduce 'conventional' breakup by preventing circular ripples from propagating. Also fully agree, that it will probably help the VC remain linear, both by limiting sideways movement of the suspensions, and by limiting excursion, keeping the VC 'in flux'.

I still think, maybe, there is something funny going on, with distortion/harmonics and/or psycho-acoustics, as you intimated. That's consistent with your analysis above, the other clues being the exceptional performance, from such a small enclosure, and also that MA were always very circumspect with their published specs i.e. 'substantially flat' and no curves published. I guess the only way to find out for sure, would be to measure an actual box.

It seems most people, who have actually heard F2B, agree they sound great. 

So I think it would be very useful, if the mechanisms behind the seemingly ingenious loading technique, were better understood, as that would allow others to create new designs using the same concept. That's obviously something well outside my knowledge and resources, but I think it would be great if someone could work it out. Perhaps it would be instructive, to model some of the earlier 215 designs, to see if any trends could be spotted?

Anyways, thanks for your replies, always learning from this forum.

Cheers

Thanks - some good resources!

From what you and Snowflake have said, it does seem possible, that a basic principal behind F2B is to deliberately induce distortion in the throat, which is then acoustically filtered, by the horn, to produce a full and pleasing tone. It is a warm sounding cabinet, and I guess that well-tuned acoustic filtering just sounds natural, because it is.

Interesting too, that there seems to be a bit of a gap in the underlying theory here - it does make you wonder, if MA have some secret unpublished formulae, or if 215/F2B was developed and refined by experimentation. 

couple of surprising things in that paper:
compression ration of 3.66 is a lot higher than recommended in most bass horn design guides. A lot of people believe over 2:1 risks destroying high excursion drivers.
they seem to detect a lot of 3rd harmonic. Could this be due to driver suspension non linearity? The throat distortion and chamber volume variation would create mostly 2nd harmonics wouldn't they?

if the front chamber volume is changing drastically then the LPF of the chamber will change. is this why the fourth and fifth harmonic levels are almost as high as the second and third?