Ringing an Old Telephone


I’m trying to control the ringer in an old telephone (see attachment below). I thought these would be hooked up to electromagnets, in the same way as old school bells, but I think the circuit is a little more complicated. There is one black wire coming out of the left electromagnet, and one red wire coming out of the right electromagnet. Not sure if all the parts are working, but I wouldn’t mind replacing them with some new ones if I knew how to drive those electromagnets to ring the bells.

To the right is a large capacitor, and in the middle is an inductor. I don’t actually need this to function as a phone, I just want to be able to control the bells, but I’m not sure how to power the coils that control the clapper.

I found the following wiring diagram online if that helps.

Again, if I could just get some guidance on how to power those magnets to ring the bells that would be great.


Unlike a door bell that has a contact that makes and breaks to generate a pulsing magnetic field you have to feed a telephone with an AC square wave. You need to use a h-bridge circuit to do this, the sort that is normally used for driving motors.
You need something like 40 to 80V to drive this, and an Arduino or 555 timer generating the square wave at about 10 to 30 Hz depending on the sound you want.

Okay cool. If I'm understanding correctly, then I feed the electromagnets with 40-80v, and use an Arduino and H-Bridge to swap the polarity at a particular frequency? What kind of power draw do you think these magnets have?

Not sure, you could measure the resistance and work it out. In practice you will draw less than this measurement due to you keep swapping the polarity. I would expect it to be in the 100mA region though.

I think I'd use a low power 240 : 15+15 mains transformer backwards ( so giving 240/30 :1 step-up ie 8:1) driven by a class B amplifier of a couple of watts rating. Buy one, or I'll knock up a schematic is you're interested.

Drive that either by a sine or square wave. At those frequencies you could even use continually varying
PWM from an arduino....

The brit system used 25 Hz from a motor-generator in the old days. You could tell if the exchange was busy, because the frequency audibly dropped slightly during the 'on' bit of the pulse.....



In their Port-o-rotary project Sparkfun engineers used a 55V boost converter and a 55V H-bridge to ring a rotary telephone. Schematic is shown on the project page.

Well as I remember US phones were 20Hz @ 80 - 100V (it was shocking) >:(
P.S. Not sure about waveform but think it was AC square.(+/- 50).

ISTR ringing current was limited at the exchange to 30 mA or so.

10 mA should be plenty to ring the bell.

POTS was more like 48V, with ringing added on top of that.

Cool, I think I have a good idea of what to do from here.

Option one is to send a 90v ac @ 20Hz on the ringing line, and should cause the bells to ring. The caveat being that it's relying on an inductor and capacitor from the 1940's.

Option two is to pull out the old parts, and build the ringer circuit listed on SparkFun's website.

Even though SparkFun's circuit has it's issues, it seems like the more reliable solution would be to integrate new circuitry to control the ringer.

The caveat being that it’s relying on an inductor and capacitor from the 1940’s.

Inductors should be fine even at that age. Capacitors less so but they are not essential for the operation of the bell. They are only fitted to reduce interference from the inductors.


Arduino + H-Bridge, using a mains transformer wired backwards and sending about 60vac directly to the electromagnets controlling the bells' clapper.

Because I'm using the Arduino, I'm also able to vary the amount of dc power going into the transformer using pwm, and change the frequency that the bells ring at.

Thanks for the help!

I'm also able to vary the amount of dc power going into the transformer using pwm, and change the frequency that the bells ring at.


If I remember rightly from taking an old phone apart decades ago the two electromagnets have
permanent magnets inside them and are driven in antiphase. At rest the clapper sticks to one side
and can be manually flipped to the other. So the whole setup is a bit like a bipolar latched relay.

Actually, the capacitor was necessary since it blocked the DC component (the ringer was 90vac @ 20 hz superimposed on the -48v battery voltage). If you didn't block the dc component, then the system would see the current as "off hook" (someone picked the phone up to answer it) and would stop the ringer voltage and connect the line to the incoming signal from the office. Fun trick to play on someone was to put an old NE-2 neon lite across the tip and ring - it would not ionize with the -48v line, so the phone was considered "on hook" (hung up) and people could use the phone to dial out with no problem (the line drops to about 6-12v dc when the normal phone network is connected (off hook)), however, an incoming call with the higher ringer voltage would ionize the NE-2, drawing enough to tell the office someone had answered the phone which would kill the ringer. People could not figure out why they were not getting incoming calls (or at best one single "ding" from the ringer) and people on the other end were confused why it might ringback once to them, then act like someone answered the phone but would then disconnect. Single neon bulb (with no current limit resistor) would drive everybody nuts (we used to have all sorts of fun things we would do with phones when I was younger ... heh heh heh)