Open Source Pulse to DTMF converter project/problem, a detailed technical discussion

[ThePillenwerfer]

Thanks for the pointers.

I did wonder about the position of C1.  The 78L05 datasheet says that a 0.33µF capacitor i needed on the input if it is a distance from the reservoir so thought it needed to be there.  It's one of those things that testing would have soon revealed to be a problem.  An LM317 doesn't need capacitors at all so that would be another possibility.

As for current consumption, it can't be more than 100mA as a PSTN line can only supply 40mA and my ATA only 20mA.  These figures have been established by connecting a milliammeter across the line with all telephones on-hook.

I'm pleased to learn that a unidirectional TVS is suitable as these seem more widely available and the datasheets certainly say that they have a much faster response time.

The firmware code is attached, if anybody wishes to look at it.  It has had to be renamed MAIN.TXT from MAIN.C to allow it to attach.

PERSONALLY I'm happy with the circuit as it stands.  The impedance mis-match does causes a reduction in volume but it is barely noticeable, and certainly no worse than if two telephones are being used together on the same line simultaneously.  I appreciate that RoyalBox is having problems so maybe there is some unknown oddity about my set-up which is somehow compensating.  The DTMF volume is easily avoided by not putting the handset to your ear until dialling is completed.  I do however intend fitting TVSs now I know the correct type.

[Alex G. Bell]

Yes, 0.33uF but not 200uF.

The LM317L might be a good or better alternative.  I will look at its specs.

Yes, the line can only supply 100mA at most and if this circuit draws much current it could be starved for power on long loops or ATAs which don't provide much current.  But as I said originally when suggesting the 78L05, it's also possible that the device power dissipation without a heatsink might be exceeded at high input voltages and sufficiently high currents that's why (once again) the current consumption needs to be known.

While the loss imposed by the original circuit may be no greater than the 3-6dB drop produced by a second telephone the loss is nevertheless incremental.  Hence if 2 telephones are in use the loss in the converter will drop the level approx. to that of 3 telephones.  Avoidable loss always should... well... be avoided!  It's a basic principle in telephony.

[royalbox]

I'd be interested to know if anyone gets the chance to try any of these theories out, I haven't myself. Please post if you do. Thanks.