Maker Pro
Maker Pro

How to electronically put a telephone line "off-hook"

L

Linden

Jan 1, 1970
0
Hi,

I know how to detect the ring signal and create a logic signal for a
microcontroller, but then how do I "pick-up" the phone line to stop the
ringing and connect with the caller ?

Perhaps with an optocoupler as a relay but what does the relay connect
up to the line to stop the ringing ?

Help appreciated, cheers !
 
T

Tom Biasi

Jan 1, 1970
0
Linden said:
Hi,

I know how to detect the ring signal and create a logic signal for a
microcontroller, but then how do I "pick-up" the phone line to stop the
ringing and connect with the caller ?

Perhaps with an optocoupler as a relay but what does the relay connect
up to the line to stop the ringing ?

Help appreciated, cheers !
In the USA putting 600 ohms across the line will produce an off-hook
condition. The audio can be pulled off through a capacitor. Surge
suppressors like MOVs should be used also.
Tom
 
L

Linden

Jan 1, 1970
0
Tom said:
In the USA putting 600 ohms across the line will produce an off-hook
condition. The audio can be pulled off through a capacitor. Surge
suppressors like MOVs should be used also.
Tom

Oh ok thanks, is that why I see 600 ohm 1:1 transformers in circuits ?
So once the line has 600 ohms across it there is a circuit between me
and the caller and I can then do DTMF or audio ?

To hang up I just remove the 600 ohm resistor ? Cheers!
 
D

Dbowey

Jan 1, 1970
0
Linden posted:

<< I know how to detect the ring signal and create a logic signal for a
microcontroller, but then how do I "pick-up" the phone line to stop the
ringing and connect with the caller ?

Perhaps with an optocoupler as a relay but what does the relay connect
up to the line to stop the ringing ?
What are you going to use as a "telephone?" You indicate that you want your
equipment to auto-answer an incoming call, but then what happens?

A good answer to your question can't be provided without more information.

By the way, your telephone line isn't 600 Ohms impedance, and tripping the ring
with a 600 Ohm resistor is a marginal approach which, depending on how it is
done, can degrade the transmission path.

Don
 
E

Externet

Jan 1, 1970
0
Hi.
To seize the telephone line you need ~200 Ohms across the conductors.

The 600 Ohm transformers are to sort of near match the AUDIO impedance
; NOT the DC load to seize the line.

To inject to or extract audio from the telephone line you should use a
8 Ohm audio output transformer winding in SERIES with ONE of the telco
conductors.
The primary of any higher impedance can source/inject audio to any
other circuits.

L1----------------8-------
RD 200
L2----------------SW------

RD= ring detector circuit in parallel to line
8 = one transformer winding in series to one line
SW= seize/hangup switch or electronically controlled relay
200= resistor, ½Watt


If the transformer resistance is in the neighborhood of 200 Ohms and
capable of conducting 50 milliamperes, it can be used as both the DC
resistive path and the audio coupling.

To hang up you disconnect the DC current path anywhere it be and leave
only the ringing detector in circuit.

The ringing is stopped by the telco central office when DC current
flows in the loop.

Miguel
 
T

Tom Biasi

Jan 1, 1970
0
Dbowey said:
Linden posted:

<< I know how to detect the ring signal and create a logic signal for a
microcontroller, but then how do I "pick-up" the phone line to stop the
ringing and connect with the caller ?

Perhaps with an optocoupler as a relay but what does the relay connect
up to the line to stop the ringing ?

What are you going to use as a "telephone?" You indicate that you want your
equipment to auto-answer an incoming call, but then what happens?

A good answer to your question can't be provided without more information.

By the way, your telephone line isn't 600 Ohms impedance, and tripping the ring
with a 600 Ohm resistor is a marginal approach which, depending on how it is
done, can degrade the transmission path.

Don

600 ohms will usually provide enough line current to easily provide off-hook
condition. I kept it simple but the OP can look here:
http://www.hut.fi/Misc/Electronics/circuits/teleinterface.html#audioint
 
E

Externet

Jan 1, 1970
0
Linden posted:Hi.
Here at the bottom of this page is the RD part of my circuit below,
plus the whole enchilada if you want to do it in another way:

www.midcom-inc.com/tech/tn88.htm
Miguel's circuit below seems its what I want I believe ?
L1----------------8-------
RD 200
L2----------------SW------

RD= ring detector circuit in parallel to line
8 = one transformer winding in series to one line
SW= seize/hangup switch or electronically controlled relay
200= resistor, ½Watt

Miguel
 
L

Linden

Jan 1, 1970
0
Thanks Miguel, I have something similar to your ringdetect but my
ringdetect signal doesnt stay low for the duration of the ring it
oscillates low/high in sync with the ring.

Does your circuit give a steady output ? Maybe I have to integrate the
input or the output first ?

Cheers,
Linden.
 
E

Externet

Jan 1, 1970
0
Hi.
To make an steady output DURING the duration of the 20 Hz AC ring
cadence, you have to replace RD with a full wave rectifier, and its
output filtered.
The size of the filter capacitors will determine the duration of the
steady output by integration. It may even last between rings.
That will be about 200 Volts DC, that must be current limited to drive
an optoisolator to feed the microcontroller at logic levels.
Miguel
 
L

Linden

Jan 1, 1970
0
Hi.
To make an steady output DURING the duration of the 20 Hz AC ring
cadence, you have to replace RD with a full wave rectifier, and its
output filtered.
The size of the filter capacitors will determine the duration of the
steady output by integration. It may even last between rings.
That will be about 200 Volts DC, that must be current limited to drive
an optoisolator to feed the microcontroller at logic levels.
Miguel

Thanks, i've setup your RD circuit without the sidactor and 33V zener.
D1=1N4148, OPTO=4N25 but my /RINGDET just never goes low enough.

i.e Vcc=5v, /ringdet goes from 4.95V down to only 3.5V on each ring so I
think the opto LED doesnt have enough forward voltage across it ??

-Linden.
 
L

Linden

Jan 1, 1970
0
Linden said:
Thanks, i've setup your RD circuit without the sidactor and 33V zener.
D1=1N4148, OPTO=4N25 but my /RINGDET just never goes low enough.

i.e Vcc=5v, /ringdet goes from 4.95V down to only 3.5V on each ring so I
think the opto LED doesnt have enough forward voltage across it ??

-Linden.

All the ring detect circuits I have seen around the place look similar,
but for the life of me I cannot get a 5 - 0v swing on the ouput of the
optoisolator. Usually only 5 - 3.5V during the ring. When putting a PNP
transistor on the output to give proper logic levels it further reduces
the swing from 5V - 4.5V during the ring and this is not enough to
switch the transistor. Ive tried a couple of opto's now (H11G1 and 4N25)
and get similar results.

It switches properly using DC input instead of the telephone line
though! So obviously the line is not turning on the opto LED enough ??
Frustrating for an amateur !

- Linden.
 
E

Externet

Jan 1, 1970
0
Hi Linden.
The optoisolator output should have a NPN transistor collector and
emitter pins, no biased base pin.
The emitter must connect to microcontroller ground, the collector to
the microcontroller {low when off hook} sense pin AND to a weak pullup
resistor to +5V. (about 100K) If this resistor is much smaller, it
may cause the behavior you describe.
->Do NOT have any direct metallic connection to the telephone line.
Everything must be transformer-coupled, capacitor coupled or
opto-coupled.
->There are logic output opsoisolators you may try if the above
somehow does not work for you.
->The current to the optocoupler led should be sufficient to saturate.
->Or, canibalize a defunct modern telephone/anwering machine and look
at its circuit for your country telco system.
Miguel
 
L

Linden

Jan 1, 1970
0
Externet said:
Hi Linden.
The optoisolator output should have a NPN transistor collector and
emitter pins, no biased base pin.
The emitter must connect to microcontroller ground, the collector to
the microcontroller {low when off hook} sense pin AND to a weak pullup
resistor to +5V. (about 100K) If this resistor is much smaller, it
may cause the behavior you describe.
->Do NOT have any direct metallic connection to the telephone line.
Everything must be transformer-coupled, capacitor coupled or
opto-coupled.
->There are logic output opsoisolators you may try if the above
somehow does not work for you.
->The current to the optocoupler led should be sufficient to saturate.
->Or, canibalize a defunct modern telephone/anwering machine and look
at its circuit for your country telco system.
Miguel

I still dont think it is driving the LED hard enough to saturate... I
had the exact same setup (optocoupler with NPN output, with 100K pullup
to 5V) no mcu connected yet but it still didnt go low enough.

Changed optocoupler to a 6N138 with darlington transistor output, found
a circuit online and it works perfectly... I should measure but upon
looking at the specs the avg forward input current for the 4N25 is
60-100mA while the 6N138 is about 20mA.

Thanks again for your help, lets see how I go with the rest hehe.

-Linden.
 
R

Rich Grise

Jan 1, 1970
0
All the ring detect circuits I have seen around the place look similar,
but for the life of me I cannot get a 5 - 0v swing on the ouput of the
optoisolator. Usually only 5 - 3.5V during the ring. When putting a PNP
transistor on the output to give proper logic levels it further reduces
the swing from 5V - 4.5V during the ring and this is not enough to
switch the transistor. Ive tried a couple of opto's now (H11G1 and 4N25)
and get similar results.

It switches properly using DC input instead of the telephone line
though! So obviously the line is not turning on the opto LED enough ??
Frustrating for an amateur !
Your meter is reading the average from the pulses from the LED
in the opto rectifying the ring signal, 20 pps. A bridge before
the LED would fix this, pretty much.

Cheers!
Rich
 
Top