I will tell you guys about a true to life complaint that telecom dealt with when I was working for them. This is quite Hilarious but totally true. A farmer rang and complained that his phone did not ring but he always knew when someone was giving him a call because his DOG WOULD BARK.
The technician called out to check the problem found that the outside cable was shorting on the wire fence. When ever someone rang him his poor old dog would get a 100 volt charge through him because he was chained to the fence lol.

Oh ok, I wonder if he put up more than the usual fight about getting chained to the fence.
It looks like early next week I will get these parts and go for it.
Cheers, Brek.

That is actually a true story .

I don’t doubt it, apparently in the early days of Edison’s DC distribution,
horses used to keep getting shocked from wet roads, and that caused some problems for the people riding them

Inspired by Art I went back to the H-bridge circuit I had previously looked at and tried it in an emulator program (http://www.falstad.com/circuit/).

It is made with discrete transistors not ICs but it is supposedly self oscillating. I played with the values to adjust the frequency and got it 'working' as a ring generator. The idea is to switch 24V+/- alternatively across the bell motor (1.4H and 1000R) in series with the 2μF capacitor which gives the effect of applying 48V AC. Below is the circuit and 'scope traces. The output is a square wave as expected but the waveform through the inductance is quite interesting.

I'm not presently in a position to build the circuit but it would be interesting if someone (Art?) could try it to see if it actually worked.

Andrew

Hi Guys,
I made a pretty big mistake in reasoning when I described the logic circuit,
since I’ve been working with 595’s, I had decade counters confused with shift registers.
I have soldered the pair of 4017 decade counters to the board already,
but sure the phone ring timing can be worked out with them.

Andrew, There is already a way I can think to streamline the first oscillator,
but to tell the truth, the H-Bridge part of the circuit wasn’t made just for this,
and it does need to remain a variable oscillator (with stepped up voltage).
Plus it is good, since I’m pretty sure I can, to get my own going.

I will hit dead end still when you pick up the phone, to detect the change in impedance,
but best look at that problem at the time it naturally arrives.

ps. I think I’ll have a development over the weekend despite having to rethink it
74595 would be an easier design, but I think 4017 is cheaper.

Hi again Guys
I have completed the circuit with timing GTC described: 400ms on, 200ms delay, 400ms on, 2000ms delay, repeat.
Is there a longer ring at the start? I seem to remember, but might be only in my imagination,
and I don’t have a POT here to test.

Mine is different. There are three rings, and that is controlled by my router which allows customisation of other signals as well since it provides the simulated POT line so you can use old phones.

What I seem to remember is a longer ring first like this:
Riiiiiiiiing.....ring..ring.....ring..ring.....ring..ring..... and so on.
Cheers, Brek.

What I seem to remember is a longer ring first like this: Riiiiiiiiing..... ring..ring..... ring..ring..... ring..ring..... and so on.

I recall that and also the short jingle when the handset is hung up.

‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾
A valve a day keeps the transistor away...

I’ll have another video up soon. It turns out more convoluted than I thought,
considering one programmable micro would replace five of the chips in the whole circuit.
This schema is just the duration timing part.
Image Link

It still somehow seems appropriate.
I do think I’ll do a micro controller version as well for pic 16F628A (still pretty popular micro).
Cheers, Brek.

I do think I’ll do a micro controller version as well for pic 16F628A

Can't speak for choice of particular PIC, but look forward to seeing the design.

Cool

Here is the video: https://www.youtube.com/watch?v=bvitOCPqqW0

Brad, Maybe that’s right. The idea I had for extending the first ring won’t really work anymore if it is true unfortunately, but another thing a micro makes easy of course.

I think maybe the pulses should be sharper like made with a dynamo. Not necessarily produced with a 50% duty cycle digitally.

Relay, that principle will work. It will be automatically tracking the resonant frequency
of the series LC circuit created by the phone coil and cap in series,
and with DC blocked by the capacitor the wave would be the singe decaying ring from the phone coil (I think).

The simplest Tesla coil drivers work the same way with a single transistor, are able to lock to,
and track the resonant frequency of the secondary coil which varies with various factors (the obvious one.. moving things around it).
There’s a capacitor in the circuit formed with the Earth to complete the LC circuit for the secondary coil.
Image Link
It’s like 1/4 of the same circuit!

As previously mentioned back in the days of electro-mechanical ring generators small manual exchanges and some PAX installations often used a pole-changer which was a tuned relay or vibrator that ‘oscillated’ at about 16Hz and alternately switched 24–48V +/- to the line to give a square wave of 48-96V PP. A photo (courtesy of http://www.museumoftechnology.org.uk/) shows a pole-changer made by Siemens Bros c1920.

Larger exchanges had motor-generator that generated the 16-20Hz ring, the 33Hz (later 133Hz) Dial Tone, Ring Tone and the 400Hz tone that was used for Busy and NU. Cams geared from the central shaft operated spring sets that interrupted the signal at the required cadence/interval. These came in various sizes and ranged from heavy to extremely heavy (if you have ever had to move one). Another photo (courtesy of http://www.museumoftechnology.org.uk/) shows an example of a small ring generator made about 1950.



The ring cadence on a 2000-type SxS exchange was a sequence of 400ms ON, 200ms OFF, 400ms ON, 2sec OFF whereas the US Automatic Electric (SxS) and Bell/Western Electric (SxS & Panel) exchanges was 2sec ON, 4sec OFF.

In theory the ring power was only switched to the called line at the start of the sequence so that there was not a long first ring however on pre 2000-type or if the exchange was experiencing heavy traffic the ring could switch in at the wrong point in the sequence resulting in a long first ring. I don’t know if the Bell/Western Electric exchanges had an intentional long first ring.

As mentioned by Clive, ring trip on a pre 2000-type and 2000-type SxS exchange was achieved by having the ring go through a slugged relay (H-Rly) in the Final Selector. The ring was imposed on a DC feed. The H-Rly was ‘slugged’ so it did not ‘chatter’ on the AC but when the handset was lifted looping the line, there was a DC path of about 200R and the H-Rly operated disconnecting the ring. For experimental purposes this sort of ring-trip can be achieved using a 12V 300-400R DPDT relay, a capacitor and a 1N4007 diode. Below is a circuit that I adapted from the Link PAX design. The first image shows the circuit in the ringing state and the second image shows when the line is looped to allow a DC path. My change from the original design was to use the second set of contacts to make the relay hold on the 24V DC as soon as it operates.



An alternative ring-trip using a transistor is taken from a very useful document, TELECOM DESIGN TRICKS, free from Cambridge Electronics Laboratories (http://www.camblab.com/lit/lit.htm). This allows easy interfacing with a microprocessor.



Andrew

Yes it would be a little easy with a micro (again) you could use an ADC and look at the ringer circuit indirectly with that.
I do plan to get to this, then the talk circuit would come together, and maybe the dialling as well.

"When ever someone rang him his poor old dog would get a 100 volt charge through him because he was chained to the fence lol."

Tallar Carl, was this in the Tamworth area in the 2nd half of the 80s?