Folded Horns .

[mal421]

Many years ago I owned a pair of Martin horns (B115's I think) . If my memory is correct these had Celestion drivers in them . Now , 25 years later , I would like some more but I would change the drivers as the originals were only 300w . My question is this . What parameters determine the volume of the sealed chamber into which the driver is mounted ? Or , given the volume of the chamber what determines the most suitable driver ? I have found lots of information regarding tuning reflex systems and lots of folded horn designs I even have copies of the plans for the Martins I want but have so far not found an answer to my questions .

[Rob_Beech]

A driver is chosen for a box, sometimes a box is chosen for a driver. Most of the time, particularly with horn loaded designs the type of driver is critical. A slight change in driver will ruin the sound and with the right (or should that be wrong) signal can destroy the driver very quickly at a level much lower than its rated power. This is down to the tuning of the box etc.

 

You need to find the full details of the original drivers and see what is available very close to these values. IMO the power rating means nothing here. if the driver is more efficient it will be louder at a given power input. to stay with Martin the W8 are only rated at around 400watts, for a 2x12 1x6.5 and 1x1 box of 90kg some may think this makes the cabinet rubbish only being rated at 400watts, I can certainly say otherwise.

[fogg]

I know a guy that has the 2 x 15 version of the martin bins that you are talking about, I think he used 2 Fane collosus drives in them, I think they are rated at 400 each. They seem to sound nice and throw the sound a LONG way!

[SteveB]

I would have to disagree with Rob and state that any driver will work in a horn and its output enhanced (within the operating bandwidth of the horn) over the same none loaded driver. However to obtain optimum performance the driver should be matched to the horn. I’m not sure exactly when the Martin horns you mention were designed, but the main problem that I had when I first started building loudspeakers was the distinct lack of information that now comes as default with drive units.

 

Coming back to the horn and rear chamber, the volume of the latter is used to increase the resonant frequency of the drive unit above that of its free air resonance. This is a technique known as reactance annulling. I don’t know who was the first person to come up with the idea, but I first came across it in a paper by Daniel Plach of the Jensen Manufacturing Company.

 

To obtain the maximum power transfer from a driver to the horn, the driver impedance should be a conjugate of the horn impedance. If it is, the resistive components and the reactive components are both equal in magnitude but the reactive components are opposite in polarity and cancel out.

 

If the cut off frequency of the horn is set below the resonant frequency of the drive unit, the horn reactance is positive and increases as the frequency goes down while the drive unit reactance is negative and increases as the frequency decreases. By matching the horn cut off to the resonant frequency of the driver it is possible to annul the reactive components, hence the term.

 

As I mentioned above, the rear chamber, which can be considered to place a stiffness in series with the suspension system of the drive unit, can be used to increase the effective resonant frequency of the drive unit to match the horn.

 

If you wade through the early posts at PSW LAB sub forum I think you will find that Tom Danley used reactance annulling in that particular design. Bill Fitzmaurice (Mr Tuba) however deliberately uses a larger rear chamber because he claims that this enables a better low frequency response than using the annulling technique. To some extent this is true, but that extent is because below the cut off frequency of the horn the drive unit behaves as though it is in a sealed box with a volume the same as the rear chamber. If you look at the published responses of the Tuba horns the sensitivity drops off by about 10db at a fairly high frequency response. This is augmented by a peak where the horn is ¼ wavelength long. At the claimed low frequency response of 30ish Hz the sensitivity is not much greater than a front loaded box. Both the above horn designs have their fanatical advocates so there must be some merit in both approaches.

 

Where does this leave you? Well as I said any drive unit that will fit in the cabinet will work. If you are determined to optimise the performance I can email you the equations but to be honest it would be easier and more accurate to model it in Hornresponse. Another thing to consider is whether any of the above was considered when the horn was originally designed? Once the basic horn had been drawn up may be that was the space left to put the drive unit in. Drive units have come a long way since “many years ago”. I wouldn’t think that it would be difficult to match the original performance.

 

Steve.

[Dan Appleby]

Mal - heres a couple of things to consider...

 

If you have a perfectly good horn-loaded sub and you replace the driver in it - which is of the same size and rating but not identical to the original - say from a different manufacturer, the chances are it won't sound as good - to do with something called the 'Theile Small' parameters of the driver I believe - that's why an enclosure (not just the horn) is often designed 'around' a driver.

 

So you might want to do a bit of research into the old drivers, see what you can find out. There are various formulas etc you can use to calculate certain parameters of your enclosure, such as volume, and port size and length etc.

[Rob_Beech]

I stated that you need to choose the right drive for the right box (or the other way around) you said the same thing, which part are you disagreeing on?

[mal421]

Thankyou all for the interesting information . I've managed to find that the original drivers were probably Martin B38 . This means that the Celestions were retro fitted . So far I have not managed to find out anything about the Martin B38 but the suggested alternatives are JBL E140 and Gauss 4583A . As we are going back almost to pre Thiele Small days the amount of information available is pretty sketchy apart from physical dimensions and power ratings the only other information is the resonant frequency which is 29Hz for the Gauss 4583A which would tie in nicely with why the Fane Colossus works because that has a resonant frequency of 30Hz . I will keep digging and will certainly read the articles mentioned .

[SteveB]

You stated that;

 

“A driver is chosen for a box, sometimes a box is chosen for a driver. Most of the time, particularly with horn loaded designs the type of driver is critical. A slight change in driver will ruin the sound and with the right (or should that be wrong) signal can destroy the driver very quickly at a level much lower than its rated power. This is down to the tuning of the box etc.”

 

With a horn although the correct choice of driver can improve the performance a slight change in driver parameters shouldn’t “ruin” the sound.

 

If I remember correctly the Martin 115 was the ubiquitous bass cabinet of the late 1970s and early 1980s, easily recognised by the final part of the flare being wider than the rest of the cabinet, I would guess that it was a completely different design from the 215 because (again if I remember correctly) there was not much difference in the overall cabinet size. There were two versions of the Gauss 4583, the A and the F the A had a slightly higher resonant frequency and a 400W rated power compared with the 300W for the F. I built some horns using the 4582 which Gauss advertised for use in horns.

 

Completely irrelevant to the topic, but back around 1980 the RCF drivers were considered to be JBL copies.

[Rob_Beech]

I see, yes I think it is the use of Slight Change rather than ruin. as the WRONG driver will not sound good, although this is true of any box design. whereas a slight change as you say wouldn't "ruin" it.

 

Rob