In putting together a wooden radial horn with a sand cast bronze adapter, the rectangular dimension of the adapter exit is slightly smaller than the rectangular entry of the wood horn. Not by much , but being rectangular instead of square, there is the option of matching it perfectly on the bottom and top with a (≈3/64") step at the sides, or rotated a quarter turn with a more even (≈1/32") step all the way around.
Though the mismatch is too small to make much of a difference in the audible range I"m still interested to know which one would create the least turbulence - especially in a case where the differences might be larger. In the imagination I can see ways either one might be worse.
Does anybody here know which would be least detrimental to performance ? Which would you choose?
TIA!
Though the mismatch is too small to make much of a difference in the audible range I"m still interested to know which one would create the least turbulence - especially in a case where the differences might be larger. In the imagination I can see ways either one might be worse.
Does anybody here know which would be least detrimental to performance ? Which would you choose?
TIA!
I built a hornspeaker not so long ago. You are wise to be concerned about the dimensional tolerances at the throat, as the horn will progressively amplify any minute turbulence on the way towards the mouth. It may be measurable, maybe not. May affect the sound, probably not. But then, horns by nature are difficult to build and the throat is everything.
If it was me, I would go for the latter option of having an equal mismatch on all four sides. It would then be relatively easy to gently sand away the wood at the throat for a perfect fit.
What is important for your build is to get the driver exactly centred in the throat to minimise said turbulence.
If it was me, I would go for the latter option of having an equal mismatch on all four sides. It would then be relatively easy to gently sand away the wood at the throat for a perfect fit.
What is important for your build is to get the driver exactly centred in the throat to minimise said turbulence.
It's actually the other way around, the adapter is the smaller dimension so it's an opening step from the driver's point of view. If need be I could smooth the transition with a little filler.
Part of my interest is in which would have the most effect in a radial horn, the vertically straight bell of the of the relatively small surface area of the sides or the much more complex contour of the top and bottom's much larger surface area. In this case, it could be lined up perfectly for the top and bottom but not for the sides, or as mentioned before, a smaller deviation all the way around.
Thanks for the quick reply, The first horn is a good fit but I was holding off doing the second in hopes of a little tech talk before committing to drilling the bolt holes.
Part of my interest is in which would have the most effect in a radial horn, the vertically straight bell of the of the relatively small surface area of the sides or the much more complex contour of the top and bottom's much larger surface area. In this case, it could be lined up perfectly for the top and bottom but not for the sides, or as mentioned before, a smaller deviation all the way around.
Thanks for the quick reply, The first horn is a good fit but I was holding off doing the second in hopes of a little tech talk before committing to drilling the bolt holes.
Thank you for correcting my misunderstanding, I reckon the use of filler on all four sides is the way to go. A radial horn? - quite a bit of bouncing around before exiting the mouth. My 'imaginarium' tells me that an equal amount of turbulence would most probably cancel itself out (more or less) and that (maybe) an unequal amount of turbulence could become worse - somewhat like a butterfly wing flap in a hurricane. But then a hornspeaker is the sum total of its own inaccuracies, and more than likely will sound perfectly ok.
A smooth transition at the throat is important and if I were you i would match the top and bottom surfaces and then make the sides smooth by filing and sanding the brass adapter, that is if the opening angle is narrower on the adapter side. If the adapter has a steeper angle, filler in the horn is the way to go. If you want to us it as-is, i think that having the step on the sides will be better, than it will act as diffraction slot and at secure the horizontal pattern control of the horn.
BR,
Anders
BR,
Anders
I don't think there will be any real "turbulence", probably better to think of it as a discontinuity for the sound wave.
I suspect it won't be audible either way, you could do a test for fun.
If you only have one horn/adapter then unfortunately you need someone else to swap it over so results aren't biased.
Or do you have a measurement microphone?
Personally I would file the adapter or fill the horn, as Anders says, just to make it look correct, even if it's not audible.
Best wishes
David
I think you are right.
Semantics apart, it is hard (for me!) to define what is good or what is bad 'turbulence'. I guess 'good' turbulence makes a horn sound good, and 'bad' turbulence well, maybe not so good.
By way of analogy, we have two massive sycamore trees outside our croft. Every winter there are violent storms that rip out branches. The wind balls up into a fist and punches so hard that something has to give. At the base of each tree are numerous small branches which, according to a tree surgeon pal of mine, cause turbulent vortices of moving air that helps to dissipate the winds punch. The branches lost are an evolutionary measure to save a tree from being uprooted. This is why pollard-ing and excessive pruning of trees is not a good idea. It may look pretty, but it is the side branches causing vortices in the wind upstroke that help a tree stay firmly rooted in the ground.
And I am quite fascinated by the idea of introducing what could be considered benign turbulence into the mouth of a horn for my next build by way of suitable finishing techniques and materials.
Kindness tapestryofsound
Thanks guys. Taking a long look at it I agree the larger upper and lower surfaces might well have more influence than the sides as well so pulled it apart (I had tacked it together with mono silicone - which by the way works really well!). Trying again though, it was a bear to find the same position I'd happened on yesterday (It seems both apertures are slightly out of square) so went with the nearest I could find and wiil clean it all up with putty.
Fired the first one up tonight. Wow, radial sure is different ! (compared to any 360˚symmetrical shape I've tried)
Thanks again!
Fired the first one up tonight. Wow, radial sure is different ! (compared to any 360˚symmetrical shape I've tried)
Thanks again!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.