Quote:
Originally Posted by dagskarlsen
'This is quite complicated, but I have an idea about, instead of that series resistor, just producing heat, a capacitor of e.g 1 µF will do the job, but not use any effect.'

There will always be an effect where capacity and inductance are connected.
Are you suggesting a different value for your master box capacitor (common to all instruments) or are you suggesting using a capacitor to replace the resistor in each telephone? If the former, then the total impedance provided by one telephone and a 1uF cap in the master box (as opposed to a 2uF cap in my earlier post) would be (typically with 59A bellmotor: R=1549 Ohms, Xl = j2032) Z = 4603, which could draw a theoretical maximum of 16.3mA. It would never achieve this, of course, because of line and source resistance, maybe 1000 Ohms or more.
Note that the figures in my earlier post are theory backed up with a practical experiment on a lowimpedance ring generator, and I have not factored in line resistance, assuming 75V RMS at the telephone input!
Now, if four instruments were connected to the modified master box (1uF instead of 2uF), each without the 3k3 resistor, the total impedance presented would be Z = 5871, allowing a theoretical current of 12.8mA. But that current would be divided by four, so each telephone would draw a paltry 3.2mA  far too low to ring a magneto bell.
If you are suggesting keeping the master box at 2uF and replacing the 3k3 resistor in each telephone with a 1uF capacitor (for example) then the total capacitance will always be less than the smallest capacitor used, in this case it would equate to 0.67uF for one telephone, so the total impedance presented would be Z = 7629, allowing a theoretical max. current of 9.8mA. Four telephones would present a total impedance of Z = 4406, with a total theoretical line current of 17mA, or 4.2mA each instrument  again, hardly enough to ring the bell and less than it would be  just!  if 3k3 resistor were used in each telephone instead.
The trouble with capacitive dropping is that there is an element of resonance whereby the impedance changes with combinations of capacitors and bellmotors (a typical 59A bellset has an impedance of 1549 + j2032, or Z = 2555 phaseangle 52.7 degrees. When resonated with a series capacitor of 2uF, the total impedance presented to the line is 1549 j1505 (Z = 2160 phaseangle 44 degrees. This partresonance increases the current flow to a theoretical maximum of 35mA, compared to 29mA without the capacitor.
Resistive dropping offers the consistency that reactive dropping doesn't. I once experimented with 'tuning' a bellmotor with a parallel capacitor to increase dynamic resistance and so reduce current, until it was pointed out to me that I had overlooked the capacitive shunt effect on speech that would occur as a result of wiring a cap directly across the bell in series with the masterbox capacitor!
For less than two telephones, I would forget the resistor anyway  I've always found they work without! If you live near an exchange, this may well hold with four.