600ohm passive xover for multiamp

Status
Not open for further replies.
I am planning to build a 600ohm passive xover for multiamping based on a constant impedance LC networks and as it will need custom inductors I was wondering if there is a minmum interest to split costs.

Base module will be 6dB/oct low pass or high pass with selectable cut off freqs (via on board DIP switch) at 75, 150, 300, 600, 1200 and 2400Hz. I might possibly add 4800Hz. A 12 steps discrete attenuator (1200 or 2400 ohm) will serve to set the level for each way.

Modules can be combined to have 6dB/oct pass band filters (eg 600-2400 for a mid using a low pass + high pass) or for higher slope (eg 24dB/oct cascoding modules of the same type). The filter will be-3dB at the cut off so cascoding two low-passes or high-passes with the same freq will attenuate the signal by -6dB.
An other interesting option might be to cascode two low passes or high passes with different cut off freqs (eg first at 300Hz and the following at 600Hz). Basically you can mix and match modules to build the xover as per your needs.

The filter will look as a 600ohm (or 1200ohm) fixed impedance to the preamp and phase will change only for complex or steeper filters.

Clearly you'll need a preamp to drive such a lowish load but I have been rewarded with all the passive xovers I built. My previous filters where not impedance costant and had steep slopes, I found out 6db or 12db sound better with my horns (as many many others did find).

I expect each module to be in the 50-100$ ballpark with the attenuator depending of the inductor price. Alternatively each filter can be terminated by a fixed resistor and you can use a high impedance (50k minimum) potentiometer in the amp. Modules might be installed in the amps as well.

let me know
 
600ohm is a classical value; a 10k network would require larger L, not a big deal probably.

Now, an inductor tapped at (mH) 2600, 1300, 650, 325,163, 81.3, 40.6, 20.3 might be used for the 6dB constant impedance network and also for a Linkewitz-Riley 24dB/oct filter (this would need a second inductor). cut off frequencies would not be exactly 75Hz and higher octaves but very close to those values. clearly PCB would be different and specific for each filter.

attached: schematics of 6dB and LR2 filters and LR2s freq response and impedance. as you see LR2 doesn't sum perfectly flat at cutoff's as I used rounded inductors values and impedance and phase is not that bad.
 

Attachments

  • LR2_sch.JPG
    LR2_sch.JPG
    111.1 KB · Views: 148
  • 6DB_sch.JPG
    6DB_sch.JPG
    92.6 KB · Views: 156
  • LR2_bw.JPG
    LR2_bw.JPG
    156.2 KB · Views: 150
  • LR2_imp.JPG
    LR2_imp.JPG
    169.5 KB · Views: 135
Last edited:
600ohms was adopted for the balanced transmission lines used by the telephone companies.
There is no other sensible use for 600ohms.

Balanced impedance audio gear is not 600ohms. Source is usually in the range 10ohms to 100ohms. Load/Receiver is usually in the range 10k to 10M
~100ohms is a good audio standard for balanced connections since many strip lines and twisted pairs are near this impedance.

50ohms, 75ohms and a few others a bit less than 100ohms are all unbalanced.
 
As the filter will not be buffered it requires a preamp with low output impedance otherwise there would be a sensible insertion loss. There are many tube preamps with output impedance of hundreds of homs, probably 600ohm might be a good compromise.
Higher impedance filters would need to be buffered before the amps.

I follow your line but would reflections be an issue at 20-20kHz? In case you refers to them implicitly
 
In olden times (more than a half century ago) balanced line amps had a 600 Ohm audio output impedance to a 600 Ohm input impedance. The output stage was a small power amplifier.
Now most balanced line amps have an output impedance in the 50 to 200 Ohm range. With input impedances in the 5k to 20k Ohm range.
While the 50 or 75 Ohm impedances are the Radio Frequency Characteristic Impedance which only applies to digital circuits.
 
Status
Not open for further replies.