Hi all,
I'm going to post bits of this in other forums but, in view of being a newbie I wanted to sketch out the basic idea here so you can all tell me how silly I'm being.
This all started because I got hold of a couple of QUAD 405-2 based amps on ebay. I've spent the last year slowly modding them and (yes) they work. One is now driving my front pair, and the other (with a doubled-up output stage for low impedance loads) is planned bridged to drive the active sub.
I also made the mistake (possibly) of buying my sub driver on a trip to England a year ago so I'm stuck with it whatever you say, but hope to get it to do a reasonable job. It's a Kenwood KFC-W3085 12". The Thiel Smalls are sketchy and I don't have the equipment or expertise to measure them myself, so I'm planning to stick to the recommended enclosures in the brochure.
Using a 2 cu ft. sealed enclosure the response peaks at about 60 Hz. and rolls off (straight) by 18 dB to 20 Hz.
The idea is to tailor the electronic low pass to balance out the rollof and give me a flat response with a sensitivity of 71dB/W@1m from 60 to 30 Hz. A sketchy graph shows that I can do this with a 15dB/8ve 30Hz low pass network. This gives me a -3db cutoff at around 20Hz and falls by 20dB/8ve from 60Hz. This is a slightly higher than I would have liked as my front pair is good down to 43 Hz so I will ultimately also need a stereo high pass there.
So to the bridging (I'll post this to the electronics forum as well):
The Quad 405-2 boards have unbalanced inputs (I have signal, GND, output, + and -) so I don't see how I can do a classic bridging job by inverting both the input of one channel unless I build an inverting stage for one channel and a non-inverting stage for the other, feeding the sub signal equally to both.
Is there any reason why I can't do this as part of the filter network? Do I have to take the halved input impedance of this stage into account (because of connecting two networks in parallel)?
Thanks in advance for your help!
I'm going to post bits of this in other forums but, in view of being a newbie I wanted to sketch out the basic idea here so you can all tell me how silly I'm being.
This all started because I got hold of a couple of QUAD 405-2 based amps on ebay. I've spent the last year slowly modding them and (yes) they work. One is now driving my front pair, and the other (with a doubled-up output stage for low impedance loads) is planned bridged to drive the active sub.
I also made the mistake (possibly) of buying my sub driver on a trip to England a year ago so I'm stuck with it whatever you say, but hope to get it to do a reasonable job. It's a Kenwood KFC-W3085 12". The Thiel Smalls are sketchy and I don't have the equipment or expertise to measure them myself, so I'm planning to stick to the recommended enclosures in the brochure.
Using a 2 cu ft. sealed enclosure the response peaks at about 60 Hz. and rolls off (straight) by 18 dB to 20 Hz.
The idea is to tailor the electronic low pass to balance out the rollof and give me a flat response with a sensitivity of 71dB/W@1m from 60 to 30 Hz. A sketchy graph shows that I can do this with a 15dB/8ve 30Hz low pass network. This gives me a -3db cutoff at around 20Hz and falls by 20dB/8ve from 60Hz. This is a slightly higher than I would have liked as my front pair is good down to 43 Hz so I will ultimately also need a stereo high pass there.
So to the bridging (I'll post this to the electronics forum as well):
The Quad 405-2 boards have unbalanced inputs (I have signal, GND, output, + and -) so I don't see how I can do a classic bridging job by inverting both the input of one channel unless I build an inverting stage for one channel and a non-inverting stage for the other, feeding the sub signal equally to both.
Is there any reason why I can't do this as part of the filter network? Do I have to take the halved input impedance of this stage into account (because of connecting two networks in parallel)?
Thanks in advance for your help!
My primary concern is how you are going to achieve a 15dB/oct low-pass filter.
You are going to have to build an active filter anyway so you might as well incorporate phase inversion.
You appear to have ignored room gain, which will mean that your anechoic -3dB @ 20Hz will be overblown. You can ignore room gain though if you are playing into a very large open space or have a leaky room such as serving hatches open to other large rooms.
You are going to have to build an active filter anyway so you might as well incorporate phase inversion.
You appear to have ignored room gain, which will mean that your anechoic -3dB @ 20Hz will be overblown. You can ignore room gain though if you are playing into a very large open space or have a leaky room such as serving hatches open to other large rooms.
roomgain etc.
Aha.... Told you I was a newbie! Sorry, 18dB/8ve.
OK, please tell me how that works, I just assumed the speaker response should be as flat as poss. It will play into a normal sitting room (around 100 m3). How should the curve look?
Thanks.
Aha.... Told you I was a newbie! Sorry, 18dB/8ve.
OK, please tell me how that works, I just assumed the speaker response should be as flat as poss. It will play into a normal sitting room (around 100 m3). How should the curve look?
Thanks.
Hrrm! I get a Frg of 86 (whatever that means) for my ca. 8m long room, but it's fairly open into ther rooms also so don't know how this effects it. The natural rolloff of the speaker is around 10dB/8ve from 60 Hz. Does this mean I'd be better off without correction than guessing the wrong factor?
Frg
Thanks for that. I now get 21.5, however, still don't know what the numbers mean. I'm getting very cold feet about trying to correct the speaker characeristics in this way, so am wondering about this box instead: http://www.marchandelec.com/wm8.html
Mistakes would be less final! All I have to do then would be to build in a low pass that matches the cut-off of my front pair... much easier!
What do you think?
Cheers!
Thanks for that. I now get 21.5, however, still don't know what the numbers mean. I'm getting very cold feet about trying to correct the speaker characeristics in this way, so am wondering about this box instead: http://www.marchandelec.com/wm8.html
Mistakes would be less final! All I have to do then would be to build in a low pass that matches the cut-off of my front pair... much easier!
What do you think?
Cheers!
wm8
I thought so!...Only problem now is the number of ops in the signal path. Must try to limit them by building a croosver network into an existing feedback loop somewhere. I also need to turn the unbalkanced signal into a balanced one between the signal processing and the power amp ( for bridging).
Cheers
I thought so!...Only problem now is the number of ops in the signal path. Must try to limit them by building a croosver network into an existing feedback loop somewhere. I also need to turn the unbalkanced signal into a balanced one between the signal processing and the power amp ( for bridging).
Cheers
Can I attend class also? 🙂
I've seen this calculation posted, but never understood how to use the mathematical result. I suppose one would calculate this for a reason, but I don't yet know what for (we,, yes, for room gain), but seem to be missing that final part.
Jennice
I've seen this calculation posted, but never understood how to use the mathematical result. I suppose one would calculate this for a reason, but I don't yet know what for (we,, yes, for room gain), but seem to be missing that final part.
Jennice
As a rough approximation, you simply set the -3dB point of your sub (which must be a sealed type, to roughly match the room gain slope) to that frequency which the formula above tells you.
Bingo!
Penny suddenly dropped about what Frg means! Sorry about slowness! In fact if 21.5 is the correct frequency I'm not a million miles off with my original corrected slope. Th X axis is logarithmic so I only have exact positions for 15, 30, 60 and 120 Hz, but it's somewhere in the ballpark!
Only problkem is whether theory gets anywhere near reality in an oddly shaped room with open doors. lol
Welcome Jennice. I hope you're enjoying this as much as I am. I find this phase much more fun than getting out the soldering iron and circular saw 🙂
Penny suddenly dropped about what Frg means! Sorry about slowness! In fact if 21.5 is the correct frequency I'm not a million miles off with my original corrected slope. Th X axis is logarithmic so I only have exact positions for 15, 30, 60 and 120 Hz, but it's somewhere in the ballpark!
Only problkem is whether theory gets anywhere near reality in an oddly shaped room with open doors. lol
Welcome Jennice. I hope you're enjoying this as much as I am. I find this phase much more fun than getting out the soldering iron and circular saw 🙂
- Status
- Not open for further replies.