Here is the schematics: (from Matilo)
And my own sketch to systematize the functions. I see no reason for shorting the capacitor in off hook pos. and I see no reason for shorting the voice path in on hook pos. :)
dsk
Now they are working (pictures : https://goo.gl/yzC7jA (https://goo.gl/yzC7jA) )
and after some testing I got the counter working, it steals energy from the line so I have let it use 5 minutes to charge the capacitor.
How does it work?
The capacitor uses lots of time to charge when the phone is on hook. (5 minutes)
When I go off hook and moves the dial, a contact in the dial connects the counter, and the capacitor powers the counter for about 1 second, the next digits will just be dialed, but since the capacitor is far from charged, the counter will not count more, so 1 step on the counter pr time you dial out.
dsk
(My line voltage is 24V and the capacitor is rated 35V)
I did test on a 48V line with a 470uF (400V) and 330k and it works. No great science, but tested and it works OK.
I tried to measure the REN load, and at 25Hz I came to an impedance of 3370 ohms, this should give a REN load of about 2.1
dsk
It is plenty of room for the components needed for using the phone with a working counter.
On the third picture, you may see the resitors and diode, directly on the terminals. The capacitor is just insulated and "fixed" with brown tape. I have protected the capacitor with a Zenerdiode, but this has not been tested more than a few minutes, I will tell if that makes problems. It will draw a current from the line of approx 1/6 to 1/3 milli-amp in on hook stage depending on voltage of the line.
dsk
Nice strange lookin phones.
Dial center for the phone:
Here is a suggestion for how to get it adapted to the British & Socket system.
The 3k3 capacitor may be replaced by a 1 uF 200V (or higher) capacitor.
The zenerdiode is only to protect the capacitor. I hav use a 33V Zener and a 35V capacitor, and it works well.
dsk