I have a T-mobile@home analog phone by my desk and there are times when I want to turn off the ringer and use a visual alert instead so I designed a circuit to flash an LED when the phone rings. The circuit consists of a 1N4003 diode in series with an 8K resistor and a white LED.
I used leftover/recycled components for my build, but here are some comparative Digikey part numbers:
D1 .. 1N4003FSTR-ND
LED1 .. C535A-WJN-CS0V0231-ND
R1 .. P8.2KW-1BK-ND
Disclaimer: I do not claim that this circuit will work, nor do I recommend that anyone try this circuit. There are high voltages involved in phone circuits that can be dangerous if handled improperly. It is also possible to damage your phone system by connecting anything other than FCC approved equipment to it.
Step 1: Simulating the Design
Linear Technologies has an excellent (and free) circuit simulator called LTspice that can be downloaded from
For the T-Mobile@home phone lines with reference to pin 2 (TIP), pin 3 (RING) is 48V DC when on-hook with about 90V AC superimposed ring voltage. For this circuit, the diodes are reverse biased until the ring voltage occurs at which point we see half-wave pulses that will forward bias and light up the LED. The one troubling flaw in this circuit is the the reverse-bias voltage on the LED is well above its spec. This may cause a failure over time, but for now the circuit is working fine.
Step 2: Pictures of the Build
Here's a picture of the build. I used a 4-parallel LED module rather than a single LED for the simple reason that it was in my bin of recycled parts.
Once I had tested the circuit, I used some .75" heat-shrink tubing to enclose the circuit. I injected hot glue into the heat shrink tubing to keep everything in place.
Step 3: Circuit Demo Video
Here's a short video of the visual alert circuit in operation (I left both the audio and visual alerts on for demonstration; normally I would have the audio turned off).
Step 4: Back to the Drawing Board
As mentioned in step 2, the original circuit has a troubling flaw which is the the reverse-bias voltage on the LED. I thought about this for a while and found that it can be fixed by adding one more resistor to the circuit. As can be seen in this LTspice simulation, there is no longer any reverse bias voltage on the LED. In this new circuit R2 should be a 1W resistor, but R1 only dissipates a few mW (and only when the phone rings) so it can can be a 1/8 Watt resistor.