I am about to launch into a new 3 way project and would like to hear views/comments etc. form the forum members before I commit to the driver buy.
- 3 Way initially active with Minidsp (Nanodigi + 4xKardas Toneboard + Rotel RMB1066) then maybe passive later
- Box size - Re-use Akai SW125 with extra bracing and sound deadening panels (bitumen) plus 25mm MDF baffle level with box edge - external 350 x 560 x 295mm (13-7/8" x 21" x 11-5/8") - approx 40litres net volume
- Drivers - SB WO24P-4 + Morel EM1308 + Morel ET448
- Crossovers - doing some modelling in Boxsin 2.0, Vituixcad and Jeff Bagby's PCD I get good consistent result with the published data (obviously will measure and re-do) with very good phase tracking with all LR2 electrical at 700hz and 3600hz plus a tank/notch on the woofer at 3500hz and a notch on the tweeter at 4500. Mid out of phase. The crossover point are also fairly consistent with the couple of papers that Jeff Bagby produced on the EM1308 and ET448 which gave me some confidence I was on the right track.
- Alignment - for the WO24P-4 40l is pretty much spot on for a sealed box and also can be used as a BR tuned to 28hz giving a very slight lift at 75hz, F3/6/10 are 52/38/28.5hz and 38/28.5/24hz respectively according to WinISD.
Attachments
Last edited:
"Box size - Re-use Akai SW125 with extra bracing and sound deadening panels (bitumen) plus 25mm MDF baffle level with box edge - external 350 x 560 x 295mm (13-7/8" x 21" x 11-5/8") - approx 40litres net volume"
If you're going to that much effort, just build the enclosure from scratch the way you want it rather than try to kludge something together from what I guess is a pretty average enclosure.
If you're going to that much effort, just build the enclosure from scratch the way you want it rather than try to kludge something together from what I guess is a pretty average enclosure.
Brett, thanks for the comment, the Akai is pretty much the box that I would have built any way, I like the old Monkey Coffin style, and the dimension are pretty much spot on for volume and baffle size. Using the Morels which have a profiles face plate which is supposed to reduce defraction effects will help.
These were bought as empty boxes for pocket money (wood in the UK is very expensive at the moment), combined with my limited wood work skills and lack of serious tools, they will give me at least a decent test box and I can remake later if necessary. They are actually 18mm ply with some bracing already so happy to use as a first try and then optimise.
These were bought as empty boxes for pocket money (wood in the UK is very expensive at the moment), combined with my limited wood work skills and lack of serious tools, they will give me at least a decent test box and I can remake later if necessary. They are actually 18mm ply with some bracing already so happy to use as a first try and then optimise.
Your z-axis values look puzzling to me. I'm assuming you're using minimum phase values for modeling. In that case, if the tweeter is the design axis its offsets would be 0,0,0. The mid and woofer Y-axis values along with the mid's Z-axis look normal, but your woofer's Z-axis is positive. User dcibel just posted this exellent summary how to perform minimum phase simulations in VituixCad (attached).
@ernperkins. Thanks for the observation. I have set 0,0,0 to be the baffle position on the tweeter axis, the tweeter dome sticks out a little as does the mid dome (both "measured" using side on photo views scaled) and the woofer is obviously rear of the baffle. I have taken the tip of the dome as the acoustic centre for tweeter and mid and the junction of the dust cap with the cone for the woofer, is this correct?
Vituix uses +ve away from the listener in Z (confirmed by the opposite convention used in Boxsim and PCD to give the same results).
As said, I am currently using manufacturer's data sheets as a first approximation traced in Vituix, I believe generally SB and Morel graphs are not far off reality. However, once I have the units I will measure and redo any crossovers with that data.
Just need to know if there will be any big issues using this combo of drivers and box and there is half a chance of getting close to the sim'd curves/phase.
Will have a read of the pdf later, thanks.
Vituix uses +ve away from the listener in Z (confirmed by the opposite convention used in Boxsim and PCD to give the same results).
As said, I am currently using manufacturer's data sheets as a first approximation traced in Vituix, I believe generally SB and Morel graphs are not far off reality. However, once I have the units I will measure and redo any crossovers with that data.
Just need to know if there will be any big issues using this combo of drivers and box and there is half a chance of getting close to the sim'd curves/phase.
Will have a read of the pdf later, thanks.
Only one comment, I am not an expert on the rest. I hope you already have the nanodigi, because minidsp has just discontinued it.
Oon
Oon
Thanks again for the comments - nanodigi is up and running so all good there.
DSP Crossover project
IMG_8415a by A H, on Flickr
DSP Crossover project
IMG_8415a by A H, on Flickr
Conceptually, it looks like a great project. You have picked out nice drivers, and they seem to match up well.
I can not comment on your cabinet because I am unfamiliar with the Akai SW125. The difference between a box with problems and a box without will often come down to bracing and sound deadening/absorption. If you add bracing and sound deadening as you say, it becomes a different box.
Have you tried to use VituixCad2's full simulation capability to model the diffraction response through +/- 180 degrees horizontal and vertical? It is not difficult to do, and it can reveal problem areas before you start. I have found the simulations to be close enough to actual measured responses.
You might find, for instance, that you get a better response by placing the mid/tweeter closer to one side rather than in the middle. Then you would have to balance the difficulty of modifying the Akai box versus the benefit. But at least you would have the information at the start of the project.
Best of luck - Jim.
I can not comment on your cabinet because I am unfamiliar with the Akai SW125. The difference between a box with problems and a box without will often come down to bracing and sound deadening/absorption. If you add bracing and sound deadening as you say, it becomes a different box.
Have you tried to use VituixCad2's full simulation capability to model the diffraction response through +/- 180 degrees horizontal and vertical? It is not difficult to do, and it can reveal problem areas before you start. I have found the simulations to be close enough to actual measured responses.
You might find, for instance, that you get a better response by placing the mid/tweeter closer to one side rather than in the middle. Then you would have to balance the difficulty of modifying the Akai box versus the benefit. But at least you would have the information at the start of the project.
Best of luck - Jim.
Thanks Jim for the comments. Will have a go at defraction modelling in Vituix but Boxsim has this built in as well and off axis does not seem too bad plus the Morels have a contoured faceplate that is also supposed to reduce edge diffraction. Offsetting the mid and tweeter is not an issue as I will be making full new front baffles anyway (the Akai’s are set back 45mm).
The Morels look to work nicely together as the tweeter dome is about 5mm behind the mid which takes account for the timing offset nicely. Glad you feel the driver choice looks promising.
The Morels look to work nicely together as the tweeter dome is about 5mm behind the mid which takes account for the timing offset nicely. Glad you feel the driver choice looks promising.
Quick update -
Drivers, wood for the baffle and some safety capacitors (39 uF for the mid and 12uF for the tweeter) plus a few other bits bought, should be delivered later this week or next. Have some time off next month so hopefully will get some progress.
Drivers, wood for the baffle and some safety capacitors (39 uF for the mid and 12uF for the tweeter) plus a few other bits bought, should be delivered later this week or next. Have some time off next month so hopefully will get some progress.
Stuff is arriving! Starting to get exited.
