My understanding of the theory is that:
Its not obvious however to what extent I should care if:
It seems to me that for ease of construction, having to make a hole sufficient for the basket or tweeter magnet system - but not necessarily tidy, or even exactly the right size - has a lot of advantages. One might even use a jigsaw, for example, and any untidiness around finishing veneering or PVC wrap is coverred up too.
Troels G seems respected despite many designs having stepped baffles. GR Research seem to do OK, though I suspect a combination of a recessed tweeter and a surface mounted driver can yield a closer driver spacing.
I know there will be diffraction issues from the cabinet edges and no practical roundover is likely to make much difference. Would we just better not obsessing about it, maybe add a bit of felt? Don't see that very often these days.
- low bass, it doesn't matter for those frequencies - a few mm (or even edge reflections from the cabinet) don't matter much
- frequencies in the midrange and up where the driver is not beaming may be impacted by an effect from the driver's own flange, and also any other driver that is 'nearby' - especially if a bass or midbass has a deep flange
- can impact midrange and also tweeter
Its not obvious however to what extent I should care if:
- I have a (sub)woofer cabinet below the mid/tweeter and there's a cabinet break/step?
- I have a stepped cabinet (Troels G does this quite a lot it seems) - I'd expect so but maybe time alignment trumps it
- I have a largeish coaxial driver and upper midrange upwards is already bouncing around like anything - none of them (or full range) seem to have smooth trebble on axis, albeit off axis suggests that sound energy is more linear
- I have a surface mount driver with a tapered flange (I'm thinking GR Research and perhaps other Peerless India customers [guess!], but access to a 3-printer could probably make an adapter for arbitrary small drivers)
It seems to me that for ease of construction, having to make a hole sufficient for the basket or tweeter magnet system - but not necessarily tidy, or even exactly the right size - has a lot of advantages. One might even use a jigsaw, for example, and any untidiness around finishing veneering or PVC wrap is coverred up too.
Troels G seems respected despite many designs having stepped baffles. GR Research seem to do OK, though I suspect a combination of a recessed tweeter and a surface mounted driver can yield a closer driver spacing.
I know there will be diffraction issues from the cabinet edges and no practical roundover is likely to make much difference. Would we just better not obsessing about it, maybe add a bit of felt? Don't see that very often these days.
I assume you mean 'flush mounting', especially for tweeters? I don't know the technicalities, but Zaph Audio has an article on why this matters and there are measurements to demonstrate the slightly spiky frequency response of surface mounting a tweeter.
Certainly, flush mounting can be a pain, particularly if you don't have a Jasper Circle Jig. However, flush mounting something like a a little Vifa XTSC90 is still a pain. Would this issue put off DIY-ers? Yes.
There are some woofers which can be flush mounted and are designed to be so, like the Peerless SDS series or GR Research woofers, but it's rare to find surface mountable tweeters. Vifa's TC-7 full range comes to mind and has been used as a tweeter in some successful designs.
I haven't tried felt but it would look a bit naff unless hidden by a grille.
Geoff
Certainly, flush mounting can be a pain, particularly if you don't have a Jasper Circle Jig. However, flush mounting something like a a little Vifa XTSC90 is still a pain. Would this issue put off DIY-ers? Yes.
There are some woofers which can be flush mounted and are designed to be so, like the Peerless SDS series or GR Research woofers, but it's rare to find surface mountable tweeters. Vifa's TC-7 full range comes to mind and has been used as a tweeter in some successful designs.
I haven't tried felt but it would look a bit naff unless hidden by a grille.
Geoff
By looking at response, Zaph has shown that there is interaction. The response (including polar response) is one thing. In addition you have the delay of the reflections and other aspects of secondary sources which we now know can make a difference. It's not always something that we can audibly separate so it may sometimes be hard to identify, but it can present a limit to a refined design.
Last edited:
I'm not disputing that its not ideal, just wondering what the relative issues are. I don't recall ever seeing a discussion with coax or full range. And many tall thin full range seem to have the driver equidistant from top and sides. Go figure!
I must confess that I do not care too much about flush mounting or not in my DIY speakers...
In fact, many years ago, an Audio friend and I built the exact same speaker - one flush mounted, the other not. These were 2-way bookshelf speakers, and we did not noticed significant audible difference between flush and non-flush.
But it's me, OK ? 😉
T
In fact, many years ago, an Audio friend and I built the exact same speaker - one flush mounted, the other not. These were 2-way bookshelf speakers, and we did not noticed significant audible difference between flush and non-flush.
But it's me, OK ? 😉
T
Not everyone considers baffle position in the same light. Some don't consider it at all. Having been there, I would be duly careful about surveying all the existing designs trying to glean specific information.
For example? Eg are you asking what design choices create the issue or how they are audible, it's not a small topic.
Coax speaker are very problematic. It takes a lot of engineering and even more refinement of the crossover. On the other hand, they are very conveniant for many low cost tasks and theoretically can be a ideal point source. That are the positive sides.
So you find a lot of lower quality coax speakers and only a very few, really taking advantage ot the principle.
So you find a lot of lower quality coax speakers and only a very few, really taking advantage ot the principle.
Tall skinny usually means a TL, usually an ML-TL. The distance of the driver from the closed end of the pipe is a critical part of the design. Dead centre is most often seen in an ML-Voigt or Metronme.
Side to side, sometimes th ebox is so slim relative to the driver size, maximizing the internalk distance to the sides (ie reflective surface) than to distribute the diffraction affects of the shape of the enclosure.
And you always have to keep in mind that it is the size of the disruption relative to the (wavelenght of the) frequency you are talking about.
dave
Then there is the other extreme>>> those who mount the drivers behind the baffle! First saw this with a set of Klipsch Cornwall speakers. Even the horn tweeter was mounted in this way.
But for me, most every time I flush mount drivers if not for any other reason it looks less 'homemade'. It only takes a rabbet router bit to accomplish this.
But for me, most every time I flush mount drivers if not for any other reason it looks less 'homemade'. It only takes a rabbet router bit to accomplish this.
Speaker building is the art of taking care of any tiny detail with a physical background and refining the build to the near optimum.
You may take the same well developed set of high quality chassis and build two set's of speakers from them.
One, taking care of bracing and damping the cabinet, good crossover parts & wires, chamfered edges and flush mounting.
The other one with minimal attention to those details, cheapest parts, but same volume.
To compare them, we need high quality source like a CD and a good modern amp. MP3 and old, not really state of the art hardware makes comparisons invalid. If your ears are not totally clogged, you will certainly hear an important difference from doing anything the right way.
The problem is you may not be able to connect any improvement in sound to a certain measure. So the finer detail in the tweeter may be a film capacitor instead of an electrolitc, the flush mounting, the chamfered edges or the right position on the baffle or the sum of all these details.
Same with better bass, is it the bracing, the stronger wire, the air coil or the carefull damping?
You can only attest the whole construction improved audible performance.
It is often a very important point in discussion in some circles, what effort can be spared, as some people want to cheat on everthing. Do less, spent less, get away with it, is considered smart by many.
So yes, if you are lazy, you may ignore points of a construction, but don't expect to get the same result and don't blame the construction for lower performance. By the way, this may often be the cause for totally different options about a subwoofer or speaker plan.
You may take the same well developed set of high quality chassis and build two set's of speakers from them.
One, taking care of bracing and damping the cabinet, good crossover parts & wires, chamfered edges and flush mounting.
The other one with minimal attention to those details, cheapest parts, but same volume.
To compare them, we need high quality source like a CD and a good modern amp. MP3 and old, not really state of the art hardware makes comparisons invalid. If your ears are not totally clogged, you will certainly hear an important difference from doing anything the right way.
The problem is you may not be able to connect any improvement in sound to a certain measure. So the finer detail in the tweeter may be a film capacitor instead of an electrolitc, the flush mounting, the chamfered edges or the right position on the baffle or the sum of all these details.
Same with better bass, is it the bracing, the stronger wire, the air coil or the carefull damping?
You can only attest the whole construction improved audible performance.
It is often a very important point in discussion in some circles, what effort can be spared, as some people want to cheat on everthing. Do less, spent less, get away with it, is considered smart by many.
So yes, if you are lazy, you may ignore points of a construction, but don't expect to get the same result and don't blame the construction for lower performance. By the way, this may often be the cause for totally different options about a subwoofer or speaker plan.
Unless you're gluing the driver basket to the baffle, you still need some board meat for those pesky screws to drive into. Also, with the basket flanges sticking up somewhat above board, there's the opportunity to use thick felt covering the baffle, which would make them flush to the felt.
Hopefully cutting a couple circles in a felt pad and gluing that to the baffle isnt such a turn off in DIY construction effort. Better than rear mounting with a waveguide cut and shaped into the baffle - effort wise, that is.
Plua a router. I did not buy one before all power tool production moved to the country that got me drafted in 1973. That shot 1100 missiles around a country I do respect last year.
Jig saw (sears 1976) or nothing. So far nothing. I've been thinking of putting 1/8" flooring with a driver hole jigsawed in it to make the driver flush. Screwed & glued to the plywood front around the driver rim.
I will use a grill cloth. Bare drivers as so commonly photographed here strikes me as crass. An invitation for the cat to scratch the dancing phase plug.
Lazy me just added a removable false baffle sometimes used for making cloth grills. 😉
Years back I built some three ways and the instructions said recess the HF & mid. I made the baffle in two sections the second was slightly too thick. Cut the holes in the main bit, cut the profiles in the top layer then clamped and bonded the two with PVA. I Packed out the bottom layer cutouts with some veneer offcuts toalign the drive unit surfaces. Clamped & PVA'd the life out of those too.
Seemed to work ok once glue set, it was rock hard and as if it was one piece. I was young and couldn't afford a router etc. etc.
Seemed to work ok once glue set, it was rock hard and as if it was one piece. I was young and couldn't afford a router etc. etc.
I've had better results with speaker DIY going an opposite way.
Tiny details, working on optimum minutia..... just ends effectively being tiny EQs, ime.
So I build speakers looking at the big picture ...always asking myself what would give the greatest marginal improvements.
What do i want, and then what needs fixing the most, to get that..
Details like recessed drivers vs not etc....continue to be sideways adventures...
They can look good, feel good, but seldom really improve sound (other than maybe being an EQ I like)
The better the drivers get, the more important do the details get. You can measure it and you can hear it. If the hardware is low FI anyway, you are basically right, it doesn't care that much. Build in distortion and uneven frequency response don't care for finesse.
If you think tiny details in your work don't care, you may not be the right man for high ende HIFI. Just like some do not care for what they eat, as long as it is enough to fill the belly...
If you think tiny details in your work don't care, you may not be the right man for high ende HIFI. Just like some do not care for what they eat, as long as it is enough to fill the belly...