GPO CANDLESTICK

Let's take a Candlestick to Bits

This article by Sam Hallas advises on how to dismantle a GPO Candlestick Telephone No. 150 to show its construction and also passes his own opinionated comments on the design and choice of material.  

Telephone No. 150

The pillar style of telephone, better known as the candlestick, appeared in the 1890s.  The design was smaller and neater than other contemporary designs, which usually had a large wooden box, usually fixed on a wall, containing the bell, magneto and induction coil.

The convenience of the candlestick was achieved by separating the transmitter, receiver and switchhook from the remainder of the parts which could be mounted out of the way in a wooden box, or Bell Set.  Even though telephones with handsets became available about the same time, they remained bulky and heavy.  The candlestick remained popular for many years because it could easily be carried about the room.  

What we have here is a British Post Office Telephone No. 150, introduced in 1924 as a BPO standard design.  Like many, this one was remanufactured from an earlier Telephone No. 2.  It's dated 1928.  Before we take it apart, let's have a look at its construction and materials.

The telephone stands on a cast steel base plate.  A good choice of material, adding weight to give stability.  The base accounts for a good deal of the surprising weight of the whole telephone.  Originally it would have been fitted with a hard rubber ring to protect surfaces from scratching.  As rubber deteriorates with age I presume that any ring this one had has long since perished and fallen off.

The curved base and tubular shaft are two steel pressings.  The shaft screws onto the base.  The steel gives structural rigidity without raising the centre of gravity, preventing toppling.  The black finish is stove enamelled.  The switchhook fork is brass with a black oxidised finished.  It is made from two pieces of strip joined at the body and spread at the far end.  I can't see any benefit here of using brass instead of steel, except that it allows the decorative finish.  The machining involved is minimal.

The shaft cap and transmitter support are a single brass piece, turned and machined to shape.  The superiority of brass at machining means that it's an obvious choice.  The transmitter is hinged using a separate brass piece not visible on this picture.  The transmitter housing is black Bakelite - the wonder material of the age - durable, heat and chemical resistant, with high electrical insulation.  The receiver cap is in brown plastic, probably a version of Ebonite, but the housing is brass with a brown coating, described as 'enamelled'.

Dismantling Main Body
A single captive screw holds the base plate.  Once removed the routing of the cords is evident. 

The transmitter housing which supports the internal chassis is now loose so you should not let it slide out of the stem yet.  To allow removal of the internal chassis the dial cord must be disconnected.

Note the wire colours before you start.  They don't always match those in the N250 diagram.

Push plenty of slack in the desk and receiver cords through the grommet in the base and slide the internal chassis and transmitter housing out.

Terminal Block
The terminal block is a rat's nest of wires and in my opinion is a weak point in the design.  The desk cord has a lashing loop at its end which ties it and the receiver cord to the chassis.  Both these cords have been replaced and you can see now that although the receiver cord is braided cotton outside, it's newer PVC inside.

Before we proceed further with dismantling the phone we need to disconnect the various cords after untying the lashing loop.  Again make a note of the colours just in case.  Typical GPO!  The nuts are 3 BA (and the threads it appears).

I used a box spanner to loosen them.  Washers tend to get lost over time and I thought there were a few short.  4BA washers from the spares box fitted fine.

The way the terminals are constructed is quite cunning.  It avoids the need for a separate insulating support for the screws.  Each hole is over-size and fitted with a fibre collar.  The screws have fibre washers either side of the chassis making them insulated.  Neat, eh? Note that the physical layout of the terminals matches the N diagram, even though the stamped letters don't.

Transmitter Support & Housing
The transmitter support is held on the to main chassis by two 4BA screws above the hook switch lever shown here. There are three tapped holes, and I see why there were only two screws as the third hole is almost unreachable.  Once they are removed it can be slid away.

This is a Transmitter No 22, which holds a Transmitter Inset No 13.  Earlier models had a solid-back transmitter.

To remove the transmitter case from the hinge assembly, first remove the horn.  This is the usual way of securing transmitter caps on telephones from the Bakelite era.  Using a pointed instrument (here I'm using a blunt bradawl or even a well made drawing pin) press firmly into the hole at the base whilst twisting the cap anti-clockwise.  Once free the three fixing screws are visible inside.  They're brass with a black oxidised finished, which is odd since they're entirely invisible and needn't be decorative at all. In fact they could well be steel.  When you remove these you might want to note which way up the case goes.  The three retaining lugs are not symmetrical and you'll want the hole in the lid to be at the bottom when you re-assemble it.

I haven't removed the cords from the transmitter housing as the ends are damaged and they'll be difficult to get back under the screws.

The hinge assembly is oxidised brass.  It will have to be split if you want to re-coat the metal.  I was unable to release the screw.  A more determined restorer might succeed!

Receiver
The receiver cap unscrews.  Mine had a crusty coating round the thread which had to be scraped to allow it to unscrew.  The coating washed off readily in warm water.  The receiver internal chassis can be slid out of its housing.  You can see that the shape of the receiver echoes the original Bell 'butterstamp' pattern which was that shape because it contained a long horseshoe magnet.  Here too the steel frame forms the poles of a magnet.  The receiver shape also provides a convenient grip for the user.

If you remove the diaphragm it is important to slide it off sideways.  This avoids distorting it.  The tags on the new cord are slightly too big for the hole in the end of the receiver cover.  They need to be bent a little to allow them through (and back again on re-assembly).

Don't polish the receiver cap too much.  It's black Bakelite with a brown coating.  Too much polishing will let the black show through.
I haven't dismantled the receiver further because unless there is anything wrong with it there is no point.  You can see that the new cord has been lashed very neatly and I didn't think I'd manage to re-tie it as well.

Dial
The dial is held in place by a screw which is located underneath the front of the dial, through the steel base pressing.  Take it right out, twist the dial anti-clockwise and pull.  It's interesting to note that this type of fixing predates this telephone and was retained right up until the time when dials became obsolete.  There can't be much wrong with a design that doesn't need to be changed in that length of time.

Furthermore, sources tell me that early models were fitted with a Dial No. 8.  The N diagram quotes a Dial No. 10.  This telephone has been fitted with the much more modern Dial No. 21 with its trigger action.  A triumph for the Post Office's design compatibility.

I'm not going to dismantle the dial further as dial maintenance is a separate subject, so best covered elsewhere.

Hookswitch and Cradle
The design of the hookswitch mechanism has some good and bad points.

First the good points
There's no need for a separate spring to raise the cradle when the receiver is lifted.  The lower contact of the springset doubles as part of the double-make contact and lifting spring.  See the close-up in the centre.  The receiver cradle is easily detachable.  Depress the steel rocker with the Bakelite peg fully and pull the cradle out.  You can take the whole pivot assembly out by removing the cotter pin, but unless it's badly seized, there isn't any point.  A drop of light oil is all that's needed to keep it working sweetly.

Now the bad points
The contact set is very difficult to remove - see the picture below.  The method of connection by extending the contacts back to the terminal screws makes it very awkward.  It's not self-contained: the contact spacing relies on a peg fitted in the chassis.  The fixing screws also hold the contact assembly together.  As you can see in the picture they tend to fall apart once the screws are removed.

Summary & Commentary
The internal chassis in the telephone is evidently inherited from its predecessor, the Telephone No 2.  It was a good idea when that telephone was designed, before automatic working added the complication of a dial.  By the time of the Telephone No. 150, placing all the terminals in the confined space inside the shaft was less than ideal.  It is a well-known fact that the most common faults on telephones related to the cords.  So making the terminals more accessible would have been a good idea, but would have involved a total redesign of the chassis.  It was probably not considered worth while in the aftermath of the Great War and with other developments in telephones looming.

Improvements in design of contact assemblies seems to be a lesson learned.  Later designs, such as dial springs and relay contacts, used separate fixing screws from the screw holding them together.

The GPO candlestick Telephone No. 150 lived a somewhat charmed life.  By the time it appeared GEC, Siemens and Ericsson were already developing self-contained Bakelite telephones with handsets.  This culminated in the adoption by the Post Office in 1929 of the Telephone No. 162.  Nonetheless the 150 remained in service for a surprising length of time.  The Post Office's policy of inter-changeability of parts means that this 1928 vintage telephone is still in full working order, despite having a new dial, transmitter insert and cords.

Acknowledgements
First I must thank the retired British Telecom engineer who kindly donated this telephone to my collection. 
Other acknowledgements are due to Peter Walker and Bob Freshwater, Ian Jolly and Ron Sewell for historical and technical information.  Finally to Laurence Rudolph for scanning his diagram N250.

Pictures and text are all Copyright of Sam Hallas

Last revised: November 05, 2022