As soon as I get a real day off or pay someone to cut them out. All I do is work anymore and its making me insane! The B&C's are just sitting in the boxes next to the pinball
I've already got it lined up to have them textured and hardcoated
mRgSr's Build uses the 18 Sound 18lw2400 1200W AES 2400W program 18" Driver.
See Post 1 in this thread. This is a very good driver at the price.
I originally modelled the speaker using the Precision Devices PD1850 / PD1851.
The Void Acoustics V18-1000 18" 1000w AES seems like a good cheaper alternative for UK / European builds (If you can find a pair to buy!)
Comparative sim posted here
http://www.diyaudio.com/forums/subwoofers/190635-th-18-flat-35hz-xoc1s-design-6.html#post2605590
Regards
Martin
See Post 1 in this thread. This is a very good driver at the price.
I originally modelled the speaker using the Precision Devices PD1850 / PD1851.
The Void Acoustics V18-1000 18" 1000w AES seems like a good cheaper alternative for UK / European builds (If you can find a pair to buy!)
Comparative sim posted here
http://www.diyaudio.com/forums/subwoofers/190635-th-18-flat-35hz-xoc1s-design-6.html#post2605590
Regards
Martin
This are the only pics I have at the moment. Yes it's made out of chipboard, the bracings are "little" modified so it was an easier and faster build because it was only intended as a test box. I really liked this TH-18 design and wanted to give it a go. A have a smaller 15" TH but wanted to hear a bigger, more PA oriented TH 18" design 
The sub in front of the TH is the LAB.
The sub in front of the TH is the LAB.
Attachments
Hi SG3525
Very interesting to see you test results.
The roll off at the LF is of some concern, and the deep null around 250hz is interesting.
Playing around with the parameters in Hornresp My first assumption is that the Inductance Value, Le quoted at 1.7 mH must be considerably more at lower frequencies.
Have you any way of measuring the efficiency of the speaker?
Regards
Martin
Very interesting to see you test results.
The roll off at the LF is of some concern, and the deep null around 250hz is interesting.
Playing around with the parameters in Hornresp My first assumption is that the Inductance Value, Le quoted at 1.7 mH must be considerably more at lower frequencies.
Have you any way of measuring the efficiency of the speaker?
Regards
Martin
Attachments
SG3525, the other point I noticed with your test results, is the lowest Impedance peak is at a higher frequency.
You have a peak at 23 Hz but Hornresp predicted the lowest peak at 16Hz. The other peaks are at the predicted frequencies.
I wonder if this low frequency is affected by other parameters, such as any Hi Pass filters in the test system.
Regards
Martin
You have a peak at 23 Hz but Hornresp predicted the lowest peak at 16Hz. The other peaks are at the predicted frequencies.
I wonder if this low frequency is affected by other parameters, such as any Hi Pass filters in the test system.
Regards
Martin
Martin, the roll off is not a concern and happens with all TH's of this efficiency.
It starts to roll off (the system looses its efficiency) from ½ WL just like an ordinary horn and it’s -3dB point is around 1/3WL.
If you lower the efficiency of the TH the roll off is lowered in frequency also. In stacks this will be fully recovered. HornResp doesn't 'see' the changing of the Le of the driver downwards.
It starts to roll off (the system looses its efficiency) from ½ WL just like an ordinary horn and it’s -3dB point is around 1/3WL.
If you lower the efficiency of the TH the roll off is lowered in frequency also. In stacks this will be fully recovered. HornResp doesn't 'see' the changing of the Le of the driver downwards.
this is why i always double le vallue from 1 khz value for subs.
i discussed this with david mcbean.
this is agreement with martins post 308.
so you lose a bit in the 40 hz range,but the overal fr looks verry nice indeed.
how does it sound SG3525?
Doubling or even tripling the value is a possibility. The problem however is when you double the value it isn't high enough for the lowest octave of the sub range and too high for the highest octave of the sub. I would prefer to see some sort of curve of which you can change the values. But the roll off is not just the function of the changing Le value. It's also 'old fashion' horn principles. I agree that the curve from SG3525 measurements looks actually pretty good.
another one on the wishlist for david mcbean
(would be verry diffecult to implement i think )Just done some EQ work. In the measurement below the filters are set at 32Hz and 100Hz both 24db/BW and one EQ at 38Hz +5db Q:2,8.
Hm... It's pretty god at lower levels but at higher levels the "chuffling" noise from the B&C driver really becomes annoying. It's probably the cooling system or distortion from the driver when over Xmax, maybe a combination of both, still need to do some more testing. I'll have to try it with a stronger driver and see how it goes.
Attachments
I have taken to analyzing the shape of the response curve using the Hornresp Loudspeaker Wizard. The Le value can be varied upward to get an idea of how the response changes. Unfortunately varying the Le value has a greater effect than any minor tweeks that I have tried with the shape of the TH18 horn.
I'm very pleased to see some real measurement data, thanks SG3525.
The bandwidth can be increased by reducing the mouth size considerably, but as said you loose the efficiency.
My objectives were to design something that was loud and powerful. Built for 'rock and roll' if you know what I mean!
My other goal was to achieve a higher frequency than 100 Hz for the HF so that mid top cabinets to match don't need to be so large.
Regards
Martin