I will try to bring the Dayton bass drivers back home from the upcomng ski trip to Utah 🤞
Currently thinking through the front baffle layout and shape, position of internal braces, back panel and amp box layout, fixings layout (threaded inserts).
I am thinking to double the current wall thickness by adding 12mm MDF and a 1mm butyl sheet between.
The front baffle will be a new additional 18mm MDF panel on top of the existing 12mm one. About 60cm long (3/4 of speaker height).
Currently thinking through the front baffle layout and shape, position of internal braces, back panel and amp box layout, fixings layout (threaded inserts).
I am thinking to double the current wall thickness by adding 12mm MDF and a 1mm butyl sheet between.
The front baffle will be a new additional 18mm MDF panel on top of the existing 12mm one. About 60cm long (3/4 of speaker height).
For the front baffle:
I knew rounding/bevelling improves directivity, so I am thinking this sort of bevels. But I do not know what drives the bevel dimensions and angles and how? Tweete diameter? Baffle width? Crossover f?
I did not know the tweeter-mid driver CtC distance is important. I thought they just need to be as close as possible. With this layout I get CtC distance of 12.0cm which is roughly 1.2WL of the 3.5kHz crossover, which is possible for these drives. The exact 1.2WL for 3.5kHz is 11.76cm.
The distance from drivers mounts to bevel start is 5mm everywhere. Distance driver mount to baffle edge for tweeter is 2.5cm at the sides and 3cm at the top. How do I calculate better bevel distances, if any?
The panel width is 15.4cm
I knew rounding/bevelling improves directivity, so I am thinking this sort of bevels. But I do not know what drives the bevel dimensions and angles and how? Tweete diameter? Baffle width? Crossover f?
I did not know the tweeter-mid driver CtC distance is important. I thought they just need to be as close as possible. With this layout I get CtC distance of 12.0cm which is roughly 1.2WL of the 3.5kHz crossover, which is possible for these drives. The exact 1.2WL for 3.5kHz is 11.76cm.
The distance from drivers mounts to bevel start is 5mm everywhere. Distance driver mount to baffle edge for tweeter is 2.5cm at the sides and 3cm at the top. How do I calculate better bevel distances, if any?
The panel width is 15.4cm
I for sure understand the itch to buy new drivers 🙂
Here is a study of how to do bracing:
https://www.somasonus.net/box-construction-methods
It is rather long though.
You already started with okayish cabinet made from particle board. Better than MDF. I suggest you avoid MDF. It is cheap and thats it.
HDF is okay for front. Bevare both is VERY dusty so wear mask during work.
For testing bracing you could use some scrap wood and adhere with double sided tape, before final glueing.
As for the fronts, what I like to do is simulate response and then test using card board and painters tape. What you want is a smooth surface.
From above picture I would guess tweeter response would improve (less ripple) if top has more surface before facet. Because as is now three of the sides have same distance to center.
Here is a study of how to do bracing:
https://www.somasonus.net/box-construction-methods
It is rather long though.
You already started with okayish cabinet made from particle board. Better than MDF. I suggest you avoid MDF. It is cheap and thats it.
HDF is okay for front. Bevare both is VERY dusty so wear mask during work.
For testing bracing you could use some scrap wood and adhere with double sided tape, before final glueing.
As for the fronts, what I like to do is simulate response and then test using card board and painters tape. What you want is a smooth surface.
From above picture I would guess tweeter response would improve (less ripple) if top has more surface before facet. Because as is now three of the sides have same distance to center.
Go check out the link Rokytherman provided. Look down about 1/3 of the way on the page. You're looking for a bunch of comparison pictures.
He's trying to test various "CLD" methods. Let's set aside the term CLD for now. What he is doing is sandwiching stuff together to make a laminate or composite panel structure. Which means his panels have different thicknesses, and he isn't controlling for that variable.
So it takes a bit of back and forth looking at his data to see this: total panel thickness is the most important variable in shaping resonance response. It doesn't matter if that thickness comes from being solid plywood, or laminating foam/honeycomb between thin sheets of MDF.
This isn't surprising at all. Think about a metal I beam. It is shaped that way to make the beam stiffer/stronger while minimizing extra material. This is the same reason composite aircraft are built with foam or honeycomb sandwiched between thin layers of carbon fiber.
Long ramble, short recommendation: Skip the butyl rubber layer, just use polyurethane construction adhesive and a notched trowel to get an even layer and glue in 1/4 inch MDF/plywood. Your total panel thickness should then be 3/4 inch. Normal cabinet construction thickness. Then isolate your driver chambers. Then add any braces you want to the long span panels.
He's trying to test various "CLD" methods. Let's set aside the term CLD for now. What he is doing is sandwiching stuff together to make a laminate or composite panel structure. Which means his panels have different thicknesses, and he isn't controlling for that variable.
So it takes a bit of back and forth looking at his data to see this: total panel thickness is the most important variable in shaping resonance response. It doesn't matter if that thickness comes from being solid plywood, or laminating foam/honeycomb between thin sheets of MDF.
This isn't surprising at all. Think about a metal I beam. It is shaped that way to make the beam stiffer/stronger while minimizing extra material. This is the same reason composite aircraft are built with foam or honeycomb sandwiched between thin layers of carbon fiber.
Long ramble, short recommendation: Skip the butyl rubber layer, just use polyurethane construction adhesive and a notched trowel to get an even layer and glue in 1/4 inch MDF/plywood. Your total panel thickness should then be 3/4 inch. Normal cabinet construction thickness. Then isolate your driver chambers. Then add any braces you want to the long span panels.
Impressed with PartsExpress. Placed an online order for delivery of the four Daytons to a hotel where I plan to stay soon during a US visit. Was a bit unsure if they will even accept the order not for the billing address. When I asked them to confirm the order they upgraded the delivery at no cost and will time the shipment for my stay dates. Excellent service! And this is on top of the saving of about 60% vs the UK price. Call me pleased!
Last edited:
Here is a dilemma:
I cannot test if the driver is OK while I am in US and back in UK it will be difficult to teturn.
The box has a heavily dented corner. Looks like its 16kg weight had been dropped. The driver looks ok visually and if I press on the cone there are no signs of misalignment.
I cannot test if the driver is OK while I am in US and back in UK it will be difficult to teturn.
The box has a heavily dented corner. Looks like its 16kg weight had been dropped. The driver looks ok visually and if I press on the cone there are no signs of misalignment.
The other drivers have arrived. Cabinet mods next.
Planning the layout of the front panel and box internals using life-size drawing.
Space is tight. Looks like I have to go with 8mm extra MDF on walls rather than 12mm.
Space is tight. Looks like I have to go with 8mm extra MDF on walls rather than 12mm.
The plan for the cabinet conversion is (using valchromat):
This will make the internal volume for LF about 10L, which is suitable for the drivers. The MF volume will be about 3L, so room to add some more material/shape for more MF back absorption
- add a new 18mm front baffle (this will block the port). Front baffle will have 1mm rubber gasket under it. Possbly angled bevels.
- add 8mm panels to all walls internally
- add 18mm internal partition LF/HF
- add internal bracing (3 or 4 braces)
- replace the 12mm back panel with 2x12mm new back panel, with the amp box in it. Back panel will be on threaded inserts and rubber gasket.
- line internal walls with butyl or another (?)
- fill LF and MF chambers with wadding
This will make the internal volume for LF about 10L, which is suitable for the drivers. The MF volume will be about 3L, so room to add some more material/shape for more MF back absorption
Last edited:
I posted a question re internal treatment here: https://www.diyaudio.com/community/threads/cabinet-internal-deadening-lf-and-mf-chambers.422975/
This is the WinISD model of the modified speaker that I could manage. Could the experts please help me to check and understand it?
I have modelled two Epique bass drivers in a shared 10L volume, drivers connected in parallel, with each dual coil connected in series within the driver. Neither the total volume nor coils/drivers connection is clear in WinISD interface (it does not distinguish drivers/coils and does not say if the vol is per driver or total). I assume the volume is total for two drivers and the coils/drivers are as described.
I added the LF/MF crossover, Linquitz transform to boost the response around 30Hz and a cut at 20Hz to limit the xmax below 30Hz.
Now, as far as I can figure the graphs and settings mean that at 36W (total?) to the bass drivers I hit their xmax and that gives me a 3m SPL of 92dB. If that is correct, good. And difficult to believe.
I do not understand the max SPL, max power, apparent power and group delay graphs. Could somebody please explain what they mean and if they show any issues in this simulation. Would appreciate!
I have modelled two Epique bass drivers in a shared 10L volume, drivers connected in parallel, with each dual coil connected in series within the driver. Neither the total volume nor coils/drivers connection is clear in WinISD interface (it does not distinguish drivers/coils and does not say if the vol is per driver or total). I assume the volume is total for two drivers and the coils/drivers are as described.
I added the LF/MF crossover, Linquitz transform to boost the response around 30Hz and a cut at 20Hz to limit the xmax below 30Hz.
Now, as far as I can figure the graphs and settings mean that at 36W (total?) to the bass drivers I hit their xmax and that gives me a 3m SPL of 92dB. If that is correct, good. And difficult to believe.
I do not understand the max SPL, max power, apparent power and group delay graphs. Could somebody please explain what they mean and if they show any issues in this simulation. Would appreciate!
Last edited:
Sorry to ask, but why are you interested in those simulations?
Just make the box, install drivers to FA253 and start measuring, equalizing etc. 😉
Just make the box, install drivers to FA253 and start measuring, equalizing etc. 😉