Does anyone think it might be good to take some of this "new crossover..." idea and use it in a modified, low-impact way on 4312As?
By "modified" I mean:
1) Reducing the port area by adding straws(as was discussed a while back) instead of sealing the ports, and
2) adding a coil & cap on the 12" woofer to get a 12dB/octave rolloff.
Would these "low impact" mods improve the 4312 without having to change the woofer or re-do the whole network?
Just wondering.
Tom
1. No. Straw ports are too restrictive. You'd be better off with a vario-vent.
2. Yes but it would impact the midrange driver, necessitating a redesign there too. You would have voltage and current phase problems that would show up either as peaks or valleys. best to do it right. It would be wise to start with another network that has 12 db slopes on the two drivers as a starting point and work from there.
If I was to do anything simple it would be a 2.5 Mh choke on the woofer with a Zobel to tame the impedance rise and allow the coil to work better. A 6 db slop is phase friendly enough and would offer modest improvements at minimal fuss.
It's a good question - not nitpicking. From my point of view, there's a difference between the electrical performance of a filter and the acoustic performance of a driver combined with the filter. In this case, the filters, which are clearly not 4th order electrical, when combined with the drivers result in acoustic roll-off slopes that more closely resemble LR 4th order than anything else. You can see this in the second graph in part 4 of my long write up, where the polarity of the midrange is reversed. I should have been more clear about that when I first wrote this.
Yes. For instance, some systems end up having 2-pole electrical filters summed with sealed enclosures (which are 2-pole mechanical filters) to achieve a 4th order acoustic roll-off. Note that compression drivers are "sealed enclosures".
Very nice! Driver layout can't be stressed enough. Of course in this particular instance the whole design was "by design" and people really liked them (or disliked them). I liked them during their era. They were fun. It's pretty fun to mess around with new filters since they have such a profound effect on the various systems.
Agreed
This whole excercise has been a vivid lesson for me why almost all speakers today keep the drivers arranged in a vertical line. This photo http://www.audioheritage.org/html/profiles/jbl/l100.htm ) shows different layouts of various JBL studio monitors made in the '70s. The redesigned crossover I’ve described above is designed specifically for the L-100A (2nd from left). It fixes the additive and subtractive peaks caused by its particular "goofy" driver layout. For other layouts, the crossover may work OK, but because I haven't had a chance to test it with those layouts, all bets are off.
Very true.
There have been more than a few who have taken various driver complements and changed baffle dimensions and driver layouts who then wondered why the results didn't sound nearly as good as expected due to them not reworking the filters as well.
There are really good points that have been stressed again in this thread and I won't belabor them.
Thanks for starting the thread. It is a great example of the kinds of threads I'd hoped to eventually see occur here. Excellent presentation and no real flak from the peanut gallery. I like it.
No doubt! Even small differences in baffles can make a difference! My modified L100s in the larger cabinets (using a straight-line config on the mid and tweeter, with the drivers offset less extremely compared to a stock L100) had MUCH BETTER behaviour in the 3-5KHz region than a stock L100, even using a stock L100 crossover! All because of the driver LOCATION in the baffle, as the width was the SAME as the original L100!
As they say on "the street": Game recognizes game!
Regards,
Gordon.
I estimated a 5 dB drop from eyeballing the old and new frequency response curves. Both of them are posted in my earlier posts. Look at them and decide for yourself. A problem with such a comparison is that the old frequency response curve was far from smooth and how do you decide which SPL level represents the nominal sensitivity. Is it the peaks or the low points which are about 5 dB lower? The new crossover is much flatter and seems to hover around 80 dB (as measured), with a few higher frequency dips below that.
Please remember that no effort was made to measure the correct overall SPL levels. You would need a known calibration standard to do that accurately. They do, however, show the relative differences in SPL between the old and new crossovers.
I drive these speakers with a moderately powered Denon AVR-1800 home theater receiver. It is rated at 75 wpc RMS when two channels are used. It works fine with the new crossovers. I do increase the volume somewhat over levels that I used to use, but it is not a problem in my family room (about 16 × 22'). Although a larger amp might be great with these speakers, they do not require one.
I was concerned with the possibility of loosing sensitivity when I did this redesign. In my hands, this has not been a problem, and I don't want this to discourage any one from trying this design. I hope this helps.
Look at the frequency response curve of the original crossover (post #2). There are two triangular markers labled 1 and 2, located at 500 Hz and 1,000 Hz respectively. Below the graph itself are two boxes that show the relative SPL of each of those markers. At 500 Hz it is 81.4 dB and at 1,000 Hz it is 81.7 dB.
Scroll down to post #7 and see those two markers at the same frequencies for the new crossover. 500 Hz is now 79.7 dB and 1,000 Hz is 78.7 dB. Subtracting, you get a loss of 1.7 dB at 500 Hz and 3.0 dB at 1,000 Hz.
500 Hz is an octave below the crossover point, where the new crossover rolls-off little of the woofer's response, and 1,000 Hz is where both the woofer and midrange contribute nearly equally, in the middle of the woofer-midrange crossover point. At those two points, the new crossover lowers sensitivity by 3 dB or less. That is better than my previous estimate of a 5 dB loss.
I duplicated exactly the L20 driver offsets and distances for my tower redesign because the acoustic centers are critical for comparing one box design to another. This left the box as the only variable.
I also agree that we should continue to recognize efforts for what they are and learn from them without bigger pictures of " why do that" or " that's not the way it was done". I plan to show how to cheaply and easily tune the L100T crossover to L100T3 specs without buying a fortune in parts. The exercise should show a way to get there without much trouble
Article courtesy DavidF.
[Zilchscan Enterprises, Inc.]
There are currently 1 users browsing this thread. (0 members and 1 guests)