I have gone on about this before, but here it is again with a bit more detail and info that may be of interest to others planning on a similar adventure.
As most of you know the JBL 4355 was the largest example of the JBL 43XX series of monitors. It was an updated version of the 4350. It used slightly different components than the 4350 with a more sophisticated crossover design. The 4350 used the 3107 crossover and the 4355 used the 3155 crossover.
http://www.jblproservice.com/pdf/Net...%20Network.pdf
http://www.jblproservice.com/pdf/Net...%20Network.pdf
Shortly after deciding to tackle the 4355 project I discovered that the 3155 crossover that I planned on copying used several tapped autoformers in lieu of standard inductors. I called JBL parts seeking these custom parts and was informed that I was SOL. In my search for a solution Giskard generously offered to contact G.T. (The original designer of the 3155 network.) and together they worked out the L-R networks to create an equivalent circuit to the original 3155 with autoformers.
Below is a schematic of the new 3155. I added the 16 ohm and 8 ohm L-pads as per the original 3155, other than that I built this design exactly as suggested matching L, R, and C values to within 1%.
Now the 3155 crossover is a passive 3 -way design meant to be used in conjunction with an active crossover to complete the 4-way system.
I have used a 24dB/octave and a 12dB/octave active crossover to complete the system. I also tried the system with a 12dB/octave low pass section on the twin woofers and allowing the 2202 woofer to run out naturally being limited by enclosure tuning.
I was not satisfied with the sound quality of my Rane AC22 crossover as to my ears at least it is not as clean and musical sounding as I would like. Unfortunately every combination that I tried with the 12dB/octave crossover sounded thick and mid bass heavy. Changing crossover frequency between 100Hz and 400Hz would change it slightly but not fix it. I even tried a natural run out for the 2202 and crossing the 2235Hs at 100Hz. Still not right. Shooting the system with my RTA confirmed this showing a broad bump centered between 200Hz. and 300Hz. I try to avoid EQ if at all possible, but I was thinking that a notch filter was in my system's future.
Then Giskard came to the rescue again! He posted:
http://www.lansingheritage.org/html/...de-monitor.htm
If you read the last column of the second page it clearly addresses my problem. It was the mutual coupling of the woofers and the 2202! I scratched my head a bit, thought about getting a 5234A with the appropriate card in it, and then tried a fix with equipment that I have here. I used the low pass section of the Rane AC22 to feed the 2235Hs. The crossover frequency is around 275Hz. With a parallel output of my preamp I fed an Audioarts 12dB/octave crossover with it's high pass section set to around 275 Hz. Bingo!!! The asymmetrical slopes help create the needed notch and the steep slope on the 2235Hs helps prevent them from contributing too much mid bass.
Reply With Quote