Finding his photographic feet, one photo at a time
I was taking some photos of phones and stuff this afternoon, and Penny decided she wanted to join in. As ever, her joining in involved wandering into shot, lying down and licking her bits.
In phone exchange news – it’s all working! I can make calls through my ATE 10 line PAX. Hurrah! Video will follow just as s0on as I manage to find my youtube password.
This is why I haven’t been taking photos or updating my blog this month.
What you are looking at, is the beginning of the telephone exchange I’m designing and building from old GPO components and circuit techniques. I’m designing it from scratch, picking and choosing circuit elements from various telephone exchange designs from the 1930s-50s.
Tonight, I managed to build the first two line circuits (there are enough relays in the image above for four line circuits, but better to start small!) After a little bit of debugging (the connection tabs on the back of those relays are really confusing!) all the bits of wire have the voltages I’m expecting on them! Yay!
Some of my prototyping methods are a bit vintage as well. These are copper coated nails, which are used to make an impromptu wirewrap board. This particular bit of prototyping is the subs-multiple.
The stuff I’ve built doesn’t really do much at the moment (apart from click a bit when you pick up a phone) and the next stage is to build the linefinder. That bit is much more interesting as it spins round and makes nice raspberry noises.
I wonder if my video camera still works
(click to get a 0.5Mb PDF of the appropriate pages)
I’ve done some reading around on this, and happened to find an appropriate section of “Telephony – Volume 1″ by J. Atkinson (Volume 1, Pages 432-433) which describes the alarm fuses I’ve got on my AT&E PAX.
It doesn’t quite marry up with the information I got through the THG mailing list, so I’m not sure what to trust more – I’m inclined to go with the book as it’s probably more reliable than memory – although I may check with the list to see if it’s a known error in the book, or if it was a case of falible memory.
Perhaps I should check my physics textbooks next, perhaps I’ll turn up a formula for calculating the fusible current for a given diameter strand of copper wire…
The AT&E 10/2 PAX (10 line electromechanical phone exchange) I’m restoring has some interesting fuses in it.
They’re designed so that when the piece of fuse wire blows, the flag pops up to let you know which fuse has blown, and a strip of metal pops down to make contact with a bar running the length of the fuse board. This bar is connected to an alarm circuit. If a fuse blows, the alarm circuit activates and a light comes on to tell you about it.
The idea is similar to the fuses I’m used to seeing on the UAX/PABX kit we’ve got at the railway, although they’ve just got two sprung contacts instead of cool little flags.
I’ve done a reasonable amount of googling for every search term I can think of, but can’t find any information about this particular type of fuse (let alone a source for replacements!)
I assume that the colour of the flag (red, blue or black) denotes the value of the fuse, and whilst there are part numbers stamped on the ends, there are no values. The markings are as follows:
| Fuse Colour | Horizontal Slot | Vertical Slot | Makers Mark? |
| Blue Fuse | L27317 | 31/1 | U.C.A |
| Red Fuse | L27318 | 31/2 | U.I.C |
| Black Fuse | L27319 | 31/3 |
I assume the marks in the Horizontal Slot column are model numbers, but I’m not sure what the “31/n” bit is about, especially as it doesn’t match up with the last three digits of the model number.
So – has anyone got any ideas where I could get some more of these fuses? Has anyone got any ideas what value fuse wire I should attempt to repair (or replace) them with?
Ian came round after work today, to help me unload this from my car (where it’s been since Sunday morning when I bought it) as I can’t lift it on my own!
It’s an ATE 10-2 PAX, which is a 10 line electromechanical phone exchange (lots of buzzing, clicking and whirring bits in there!) – designed for small business/factory use, it was built by the Automatic Telephone & Electric Co Ltd, of Liverpool.
It’s missing its power unit, and only one of the line fuses is intact, but otherwise it appears to be complete.
My plan is as follows:
First things first though, it’s been sat in a garrage for years and is filthy, so my first job is to give it a clean!
I’m trying out a new black and white post processing technique. It seems to have worked quite well and given lots of contrast without sacrificing detail. I think I like it!
In phone news:
That last one is the one which is causing the most head scratching, but the first one is the most interesting!
Today I made a trip over to Cheltenham, to visit some chaps driveway as he was getting shot of a load of old telephone equipment. This is one of the (many) items I came back with.
I’m told it’s some kind of directory enquiries console, but I’m not 100% sure if that’s right. Still, I’m sure it’ll be useful for something!
I also came back with a couple of relay sets, and a bunch of uniselectors. I *should* have enough bits to build most of the guts of an electro mechanical telephone exchange now! I can have a phone exchange which clicks and whirs! Hurrah!
Isn’t it a lovely colour! It could do with a bit of a clean, and the mic side of the voice bridge doesn’t seem to work properly (despite me swapping out the mic element) – and various bits of it aren’t original – but it’s interesting for other reasons.
Feel free to wander off if you’re not as into the phone thing as me ;-)
I bought this GPO 706 Phone off ebay, as I’ve been looking for a 706 MK1 for a while and I noticed that this 706 had a metal dial (which appears to be a “Dial No 12“) so stood a good chance of being a MK1, possibly an early one at that.
Taking the lid off shows that it is a MK1 (the MK2 had a PCB) but there’s something odd about it.
See that slot in the middle? It should have a circuit board mounted vertically in it. Not only is the rectifier board missing, so are the middle three contacts. A close look at the socket seems to suggest they were never present. The way the large resistors are mounted seems to reflect that. It doesn’t really match the N806 diagram, although I did notice one thing that’s interesting.
The N Diagram, Page 1, Note F says:
Early issues have dial brown connector connected to gravity switch spring 4 instead of T3
Which this phone does have. It matches Note F and has a Dial No 21 – so I think it probably is an early MK1.
Another clue that it’s probably an early model is that it’s got a paster diagram inside the case. Now, I’m only guessing here because I can’t find any references that would say yes/no on the paster diagram thing but none of the other 706s I’ve seen had them, but they were all MK2s.
The paster diagram is labeled N98293, but I can’t find any reference to an N diagram of that number online, but I assume the paster diagrams are numbered differently. The paster does match the wiring of the phone more closely than N806 does.
So, this is proving to be quite an interesting ebay purchase!
I’m not 100% convinced that the brown handset and the dial surround are original (I think the green 706s had dark green handsets) and it’s possible that this phone is early enough to have originally had a braided handset cord. I suppose that the paster diagram is probably the real key to working out how this phone was originally configured.
Update – 2010-02-12:
I found a number stamped on the baseplate, it says N1900E69 which means that the phone was built by Ericsson, and is an N1900, or “Etelphone” although some of the details still don’t match. So there’s a bit more digging to be done yet!
This is Jon Peat, British Young Juggler Of The Year 2006, probably one of the most interesting ball jugglers I’ve seen in a long time. He was trying out a new routine on stage at the BoB Juggling Festival this weekend. There are some more photos from the show in this album here.
However, the most productive part of the day for me, was my chat to Tarim about my idea for a telephone based siteswap validator.
The Background:
For those that don’t know, siteswap is a mathematical language for describing juggling patterns. The idea is that you can use a string of numbers to represent the pattern. Each number represents a throw event, with the number determining how many beats later the object is thrown again. Eg a 3 means that the object being thrown at that time, is next thrown 3 beats later.
Some strings of numbers are valid juggling patterns, for example 531 (which is a 3 object pattern) but not all strings are valid. For example, 540 is not a valid siteswap, as the 5 object is next thrown in 5 beats time, but the 4 is next thrown in 4 beats time. This means that they’re both thrown by the same hand at the same time. There’s a collision.
There’s a lot more to siteswap than that, if you’re interested in all the gory details start here with this wikipeida article about siteswap.
It’s handy some times to be able to check if a string of digits is a valid juggling pattern or not, telephone dials have digits, so they can be used as an input device. They also have earpieces, so can be used as an output device.
The Plan:
To build a device which you can plug an old rotary dial phone into, which allows you to do the following.
I had a firm idea about how to do all those steps apart from step 7, getting a microcontroler to play speech isn’t trivial. Talking to Tarim though gave me an idea of how to get around that though by using a cheap MP3 player. Result!
I’ve built a circuit which interfaces to the phone line, and can detect hook state and count dial pulses. I can get audio onto the phone line. I know *how* to make the phone ring, but am lacking the components to do so.
In short, I’m *soo* close!
Hurrah!
It’s only taken me 3 days, but I’ve managed to get my Opera 206 to dump its config out of its RS232 port. It’s designed to be connected to a printer, but I don’t have a printer with a serial port handy so I’ve got it hooked up to my computer instead.
Here’s how you hook it all up:
| RS232 (9 Pin DSUB) |
Opera 206 |
| CTS (Pin 8) | RTS |
| RX (Pin 2) | TXD |
| GND (Pin 5) | GND |
Optional: Connect DSR(Pin 6) and DTR(Pin 4) together on the RS232 plug. (Note, this may be needed for printers, but it seems to work OK without)
Why did that take you 3 days?
Although the above makes perfect logical sense, it’s not what the manual says to do. This is a perfect lesson in “when is it appropriate to trust the manual, and when is it appropriate to think the situation through from first principals”
Just for my reference, here’s the current config of my Opera 206, which looks a lot like the default config:
ispms@super:~$ cu -l /dev/ttyS0 -s 4800
Connected.
Ext Line Number Dialled Duration
Hr Min Sec
--------------------------------------
-----------------------------
Extension Assignment
1 2 3 4 5 6
-----------------------------
Day Service | | | | | |
Line 1 X | | | | |
| | | | | |
Day Service | | | | | |
Line 2 X | | | | |
| | | | | |
Night Service | | | | | |
Line 1 X | | | | |
| | | | | |
Night Service | | | | | |
Line 2 X | | | | |
| | | | | |
Outgoing Assignment | | | | | |
Line 1 X X X X X X
| | | | | |
Outgoing Assignment | | | | | |
Line 2 X X X X X X
| | | | | |
Barring | | | | | |
| | | | | |
Data Protection | | | | | |
------------------------------
Line Programming
1 2
------------------------------
MF : X X
Decadic :
------------------------------
------------------------------
PABX :
Trunk : X X
------------------------------
Ext Line Number Dialled Duration
Hr Min Sec
--------------------------------------
Exception Table
---------------
Ext Line Number Dialled Duration
Hr Min Sec
--------------------------------------
~.
Disconnected.
ispms@super:~$