5 The Plan Revisted

So by now I've made a couple of projects and gained a good understanding the ways of the telephone network. Based on what I had learnt what would be nice would be to revisit the Plan 7 and think of a way to make it work properly, as it was designed to work. What I really needed was some kind of method to get connected into something that would detect the dial pules, de-code them and route calls to the desired phone on a telephone network. And the same thing but going in the other direction. Then I could use the extension phone properly in that I could dial out from it and transfer calls to it. What I really needed was a DIY telephone exchange.

So really the point to this blog is that I just wanted to make a few notes about my journey from knowing nothing about electronics and telecom systems to actually building a working telephone exchange of my own. I'm afraid it's not going to be a detailed "how to" type of thing and I'm not going to release any of my coding in the form of a library or some such thing. It's really more about me gaining the knowledge to a get at least enough understanding of the underlying mechanics behind the systems to make it work. I am forever in debt to Andrew Holme and his design without which absolutely none of this would ever have happened. If you want to follow the course of the build in the blog you will definitely need to refer to Andrews design.

I think everyone who has contemplated making their own exchange must have, at some point, looked at the design by L.D.Gunn, his 10 way miniature exchange. It's certainly a very fascinating design but the problem is it's very, very analogue and for so for me, very, very hard to understand. No matter how much I busted my brains over it, I just could not conceptualize how it worked. It was beyond me.* The other school of thought is just to build it and figure it out as you're making it. I find a great way to understand how things work is just to go ahead and make them, learning by your mistakes. I did contemplated this but there were two issues, namely I would have to build it on a "big bang" approach and hope that, once built, it would work as I had no realistic method of testing each of the sub-systems as I went along due to my lack of understanding. The second problem would be sourcing the parts which are probably available but certainly at a cost - I would expect the relays alone would probably be the best part of a decent family curry.

So more digging around unearthed this Andrews design. Worthy of note is how Andrew also had an interest in telephone's and exchanges from an early age. Now the cool thing about this design is that it looked almost do-able. I knew I could work with the solid state relays and opto-couplers as I had gained experience of using them in my other projects. There were some elements of the circuit that I couldn't quite get my head around, the off hook detection with the transistor looked a bit beyond me, but other than that this looked like being do-able. Andrew had used a number of controller logic IC devices to run the exchange Operating System but this being some (at time of writing) some 20 years ago it would be far simpler to replace the whole of the logic side of things with a single Arduino using C to code the exchange OS and a bunch of shift registers for the IO. So the more I thought about it the more I could see no reason why this should not work.

So I'll fast forward here a bit. I did indeed build and complete the exchange and here it is in it's Mk 1 version. 


It had four lines and worked pretty well but I certainly felt there was room for improvement. I had spaced the components out to a considerable degree on the Veroboard so I could give my self plenty of wriggle room in case things needed modifying so it was quite a squeeze to get it into the one box. The ringing transformer wouldn't fit in at all so had to be bolted on in a separate box. 

Eventually I thought it would be better to start again, this time really concentrate on getting everything down to as small a scale as possible and, as an added bonus, see if I could get it work with the house PABX using DTMF. Andrew had suggested this modification in his design using the source line isolation transformer to connect and external line. Also I wanted to use a 4 x 8 segment LED display rather than a single one and delete the LED's though they did look cool as they flashed on and off to display a pulse train. 

*By way of an update, I now understand how some of it works, like the call sensing, the binary counter and routing calls, but the rest of it still just to complicated to follow.

Comments

Post a Comment

Popular posts from this blog

12 Boxing it up

Image
So now I have a working exchange consisting of four lines, numbered 122,222,322 and 422. I made the numbers three digits just for the sake of it really. My first exchange used 11 as the post-fix so I used 2 this time just because. If the extensions are not dialed as above then it will be a false number and a fast busy tone is played to the handset. Here it is boxed up and mounted on the wall. The segment display is used to describe call progress but when the exchange is idling it displays random characters generated every second just to show it's still alive. The red button resets the Arduino and the two black lumps on top are the fuses for the DC feed and the AC ringing. Above it are the junction boxes and batteries used by the Plan 7. Removing the lid shows how the exchange fits in the box. The large single Veroboard shows the exchange itself, so from top to bottom, the powers supply caps and VR's, to the left the call sensing, in the middle the isolations transformers for th...
Read more

11 The Relays do the Exchanging

Image
All that's left to do now is add the relays that will connect each source to each destination phone . As I have four lines I will need a set of eight solid state relays arranged so that one of each of the four source lines can be linked to one of the four destinations line via the bridge relay. I've been using the  Panasonic AQW210EH  throughout with the exception of the ringing relay. As in the bottom left(ish) of the picture, I take four feeds from each of the phone lines and arrange two 18 pin sockets (empty in the picture) to hold the four eight pin relays. At the time the picture was taken I've added copper links to two of the relays so I can dial and link my two phones in either direction, either as source or destination. A more detailed picture shows more copper links added, so we have a relay between each phone and the source and destination lines, each running down a copper strip of the Veroboard, to the right of relays. Each relay will be actuated by a link...
Read more

1 The Beginning

Image
The telephone system has always fascinated me. A couple of my earliest memories about phones date back to the early 1970's, when I was about 9 or 10. My mother came from quite a large family and we used to travel up from London to Somerset a couple of times a year to visit them. On one such visit during a hot summer out roaming the countryside with one of my Uncles, who was maybe five or six years older than me, we came across a derelict coal mine. We ventured in and found what I guess must have been a few Series 300 Bakelite phones and decided to liberated them. Back at his home we took them apart and I remember seeing the governor wheel in the dial mechanism along with a set of driver cogs, fascinated by the way it rotated when the dial was turned. My memory is a bit hazy here but I have a recollection that my Uncle somehow connected two of the phones together with a battery and we talked to each other over the phones. Hours of fun. After a few days with my Mothers family we woul...
Read more