Telephone interface pc-modem

[Number 6]

I managed to get the dialling done acoustically by writing a sine-wave
synthesis program on the PC which responds to buttons pushed like a real
phone.
What I did wrong was that I did not raise the second tone (of the dual
tones used for dialling) by 3dB as somebody suggested.This compensates for
the role-off of the line freq response.It refuses to work without this. The
duration is far less critical.

Thanks for that - now I can dial just by holding the phone next to the PC
loudspeaker - works perfect!
However, I wish to dial directly into the line. I have a telephone 1:1
isolation transformer for this with zeners to block the ringing voltage and
a series capacitor to block dc - I can hear somebody on the line fine but I
cannot seem to 'send' a signal down the line at all.At least the signal is
very weak indeed.
So to take things step at a time - how do I get my dial tones to feed
directly into the phone-line? I tried feeding from the PC speaker via the
transformer but it won't cut the ice.

I also tried this the easy way via a modem by sending ASCII commands. This
works too but again I have no way of sending a signal down the line once the
phone connects - I can dial and it connects at the other end. I can also put
the phone on and off the hook.What am I doing wrong? Surely the signal
strength is not that high - or is it?


Tom

 

[Number 6]

I managed to get the dialling done acoustically by writing a sine-wave
synthesis program on the PC which responds to buttons pushed like a real
phone.
What I did wrong was that I did not raise the second tone (of the dual
tones used for dialling) by 3dB as somebody suggested.This compensates for
the role-off of the line freq response.It refuses to work without this. The
duration is far less critical.

Thanks for that - now I can dial just by holding the phone next to the PC
loudspeaker - works perfect!
However, I wish to dial directly into the line. I have a telephone 1:1
isolation transformer for this with zeners to block the ringing voltage and
a series capacitor to block dc - I can hear somebody on the line fine but I
cannot seem to 'send' a signal down the line at all.At least the signal is
very weak indeed.
So to take things step at a time - how do I get my dial tones to feed
directly into the phone-line? I tried feeding from the PC speaker via the
transformer but it won't cut the ice.

I also tried this the easy way via a modem by sending ASCII commands. This
works too but again I have no way of sending a signal down the line once the
phone connects - I can dial and it connects at the other end. I can also put
the phone on and off the hook.What am I doing wrong? Surely the signal
strength is not that high - or is it?


Tom

I forgot to show you teh interface circuit. I use two of these - one for
input and one for output - is there a better way?
http://www.solorb.com/elect/phone/tap/


Tom

 

[Dbowey]

Tom (#6) posted:

I managed to get the dialling done acoustically by writing a sine-wave
synthesis program on the PC which responds to buttons pushed like a real
phone.
What I did wrong was that I did not raise the second tone (of the dual
tones used for dialling) by 3dB as somebody suggested.This compensates for
the role-off of the line freq response.It refuses to work without this. The
duration is far less critical.

Thanks for that - now I can dial just by holding the phone next to the PC
loudspeaker - works perfect!
However, I wish to dial directly into the line. I have a telephone 1:1
isolation transformer for this with zeners to block the ringing voltage and
a series capacitor to block dc - I can hear somebody on the line fine but I
cannot seem to 'send' a signal down the line at all.At least the signal is
very weak indeed.
So to take things step at a time - how do I get my dial tones to feed
directly into the phone-line? I tried feeding from the PC speaker via the
transformer but it won't cut the ice.

I also tried this the easy way via a modem by sending ASCII commands. This
works too but again I have no way of sending a signal down the line once the
phone connects - I can dial and it connects at the other end. I can also put
the phone on and off the hook.What am I doing wrong? Surely the signal
strength is not that high - or is it?


Tom

I forgot to show you teh interface circuit. I use two of these - one for
input and one for output - is there a better way?
http://www.solorb.com/elect/phone/tap/
Yes, there's a much better way, but be sure to disconnect it from the line when
you are not using it.

Remove the 4.7k resistor, and change the cap from 0.22uF to about 2uF. If you
don't have a 2uF NP cap you can use two 1uF caps in series but with like
polarities connected together.

Although this will fix your transmission problems, it may cause the ring to
trip if you do leave it connected to the line.

Don

 

[Dbowey]

Abooboo posted:

"Remove the 4.7k resistor, and change the cap from 0.22uF to about 2uF. If you
don't have a 2uF NP cap you can use two 1uF caps in series but with like
polarities connected together."

Well, that's what I get for not having coffee yet. Make that two 4uF caps in
series......

Don

 

[Number 6]

Tom (#6) posted:

Yes, there's a much better way, but be sure to disconnect it from the line when
you are not using it.

Remove the 4.7k resistor, and change the cap from 0.22uF to about 2uF. If you
don't have a 2uF NP cap you can use two 1uF caps in series but with like
polarities connected together.

Although this will fix your transmission problems, it may cause the ring to
trip if you do leave it connected to the line.

Don

I assume you are changing the time constant - good thinking. Then I will
have a residual resistance of about 600 ohms (I think that is the resistance
of the winding) + whatever gets referred back from the primary I expect.

Tom

 

[Dbowey]

Tom posted:

Tom (#6) posted:
(snip)

Yes, there's a much better way, but be sure to disconnect it from the line when
you are not using it.

Remove the 4.7k resistor, and change the cap from 0.22uF to about 2uF. If you
don't have a 2uF NP cap you can use two 1uF caps in series but with like
polarities connected together.

Although this will fix your transmission problems, it may cause the ring to
trip if you do leave it connected to the line.

Don

I assume you are changing the time constant - good thinking. Then I will
have a residual resistance of about 600 ohms (I think that is the resistance
of the winding) + whatever gets referred back from the primary I expect.
Changing the capacitor just reduces the 1 kHz. reactance from about 1000 Ohms
to about 100 Ohms, to improve transmission. If your transformer is terminated
at about 900 Ohms on the PC side, the match to the line will be as good as you
need.

Don