Pay on Answer (POA)
BOXES COIN COLLECTING No's 700 & 705 & 735
Why did POA and 'pips' come into existence in 1958 ?
When the A and B button system was nearly 35 years old, subscriber trunk dialling (STD) was introduced in 1958 there arose the need for a coinbox system which could deal with the new periodic metering for both local and trunk calls.Before STD a local call had a fixed fee however long it was. In order to keep the coinbox mechanism as simple as possible so that capital and maintenance costs might be reduced, a post-payment system (known as pay-on-answer) was adopted; coins were cashed on insertion and not held in suspense. A complex piece of apparatus at the local telephone exchange, known as the coin-and-fee check (C and FC) relay-set, compares the meter signals from the call-charging equipment with the coin signals received from the coinbox. The C and FC unit controls the opening of the coinslots, gives the customer audible warning when his money has run out, and cuts off the call on expiry of a short period of grace.
Who designed the POA ?
Page 19 of Telephone Boxes (1989) by Gavin Stemp ISBN 9780701133665 says that (Douglas) Scott was the industrial designer responsible for the new universal pay-phone equipment which replaced the ...button A Button B in all the ... (K2 and K6) kiosks designed by his namesake (Sir Giles Gilbert Scott,RA)
How it worked
The coin telephone was connected by the usual pair of cables to the telephone exchange, but instead of going straight into the SUBs line equipment it went through an extra piece of equipment called the coin-and-fee check (C and FC) relay-set, this allowed not only dial pulses to be signalled from the payphone to the PSTN but also coin pulses when coins were inserted.
Dial digits were sent as usual by breaking the line 1 break for digit 1, 2 breaks for digit 2 ...10 breaks for digit 0.
Coin pulses were sent by passing the line through coin pulsing contacts, in parallel with a 5k resistor
Sending coin pulses (In detail)
The coin pulsing clock work sender has 4 additional contacts to the coin pulsing contacts
So 5 contacts all together
Coin Pulsing contacts - "CP" on fig 4, "Coin pulsing element" on fig 6 above and CP Contacts on fig 5
These contacts are driven by a cam following 1 of three shapes cut out of metal
You can see below this mechanism can send 1,2 or 5 pulses.
The cam and the whole of the coin pulsing contacts are moved left to right depending on what coin is inserted so that the cam follows the appropriate metal shape.
Sequence of Events
Line polarity Reverses opening coin slot, After called Party answers
Subscriber pushes in 10p coin. (which is to be represented by 2 coin pulses, 5p is 1 pulse)
The coin raises the clockwork arm of the coin pulser mechanism against gravity and a spring
If the coin is pushed in further, it disappears into the slot and the coin hits the mask contact buffer and opens mask contact.
The clock work arm then begins to fall.
Next to operate is the contact marked A, the contacts close.(CP.ON3 closes)
Next the B contact closes ( POEEJ says CP.ON1 short-circuits the receiver during coin pulsing. As the introduction of relay SU increased the direct current flowing through the receiver, the current was reduced to its normal value by the 10kΩ resistor switched in by GS3 to minimize the click when CP.ON I operates.)
Then for the first time the CP contacts open, starting to send coin fee pulse 1
Then Contact C closes (POEEJ says CP.ON2 is required to ensure that the line is closed by a precise 5kΩ coin-pulsing condition and to prevent interference with the coin pulsing by the gravity switch, the dial, or by the variable resistance of the transmitter. It was found, however, that its operation before commencement of coin pulsing, i.e.while the line current was high, caused a click to be heard by the called subscriber. Hence it is timed to operate approximately 50 ms after the commencement of the first coin-pulse, by which time the exchange equipment has disconnected the transmission path and thus prevented the called subscriber hearing any resulting click. The caller is, of course, protected from clicks by the CP.ON I receiver short-circuit.)
Then CP contacts close again - ending coin fee pulse 1
So now all the contacts are closed apart from the Mask Contact
Now the CP contact open again, starting the second coin fee pulse
Next the CP contact closes ending coin fee pulse 2
Now contact A opens - POEEJ says CP.ON3 initiates the full break at the end of a coin train.
Then the Mask contact closes -Causing current to flow in the loop again
next contact C opens (CP.ON2 opens allowing current through the speech circuits again)
Then contact B opens (CP.ON1 opening un mutes the receiver again)
Mechanism is now at rest
Shane's video of a POA Phone in action
Interfering with POA
Some extracts from the THG discussion Group
In the seventies, lots of pubs had a pay-on-answer phone box as the only line. As the landlords were usually tenants, they had to insert coins to make a call. They sometimes had 746 to make emergency calls and answer incoming calls. PE said see http://www.samhallas.co.uk/repository/n_diagrams/4000/N4562.pdf
They usually had customers that were telephone engineers and after a few drinks "on the house'' they would arrange to come back with a recall button and a resistor and fit it to the 746 so the landlord could simulate inserting coins by a certain number of presses of the button. This was followed by a verification check of a flash recall by the use of the switch hooks. This was supposedly an ant-fraud device but if the other info was
A 5k resistor and a diode in series across the dial impulsing springs was a much more elegant solution
That wouldn't work on any Strowger that I've played with as there was another security feature in place that if a coin pulse was received before a charge pulse, as soon as the remote end answered and a charge pulse was received, the call would disconnect with the caller getting NU. I must see if this is the case next time I'm in a working museum as maybe this was just a London (director) thing?
known about flashing a button through a resistor, they must have known about giving a flash disconnection as well!JH said
More elegant using the dial
Coin and fee check on TXE4
After 01-965 in Harlesden and 01-997 Perivale? both went TXE4a you didn't need the line break after a coin pulse. Following the 1st series of coin pulse(s), the system would wait with the audio bridge broken, then around the same time you'd expect to hear NU on a Strowger (1-2 seconds maybe?), the audio would re-connect and that was that. Next time you coin pulsed, there would be no break in audio BUT if after the third pulse you did do a line break, you were back to square one with the break in audio restored following a pulse (or series of). I don't know why this was but thought it was a mistake at the time. Does anyone know if the line break after coin pulse security feature was restored with the introduction System -X?
AT5716 required the dis after a coin pulse train. Later versions did not (60665 and 61486 for UAXs.) TXE 4 and later did their own thing. I am not aware of any POA boxes omitting the dis, but possibly a mechinism could be so adjusted.
The Demise of POA
To the GPO, the POA coinbox system had many limitations, of which the 2 most affecting profitability were the lack of flexibility in setting tariffs, and the poor reliability of the system. The user difficulties mentioned above also affected profitability by discouraging people from making high-tariff calls. The basic difficulty over tariff-setting was that the unit fee had to be equal to a single-coin value ; and whilst fine adjustments can be made by altering the timing of the meter signals in the network, the unit fee is a significant parameter because a very large proportion of calls last for only one meter period. In times of high inflation, unprofitable tariffs were retained in deference to external pressure on GPO to avoid the large percentage increases in unit fees necessitated by the available coins of the realm. The combination of a complex relay-set in the telephone exchange and a coinbox mechanism containing several hundred moving parts is an obvious reliability hazard; perhaps less obvious are the difficulties of managing a system where the responsibility for the major parts is divided between the essentially separate internal and external maintenance staffs. The difficulty of attributing faults to the appropriate part of the system is probably the largest contributory factor causing recurrent call failure.
Full cash-boxes had always been a problem, both in rented payphones and in Public Call Offices. It is by no means trivial: a full cashbox generally causes coins to jam, often resulting in a visit by an engineer. Service can often be restored by shaking the coins down and clearing the jam ; but unless the cash-box is emptied immediately, the trouble will quickly recur. A number of attempts had been made to detect and report coin jams in POA boxes, with only limited success.
Taken from Pay-on-Answer Coinboxes - An extract from J R Lewin article in BTEJ April 1982
Payphone Test equipment
Coins 10p A MAX,A MIN and R MIN
Payphone Test Weights
I have little original information on this era of pay phones, but I am interested in them because I remember them and used them a lot.
it can be found on page 343 called "The pay-on-answer Coin-Box System"
That describes both the box and exchange kit in good detail.
There is also an article later on in the same journal vol 54 part 3 October 1961 on page 176 describing "a pay-on-answer coin-box for subscribers' installations"
Luckily Bob Freshwater's site has lots of great information on them.
Try the following links
The Exchange kit was called "Coin and Fee Checking Equipment", there is a picture of the equipment used to test this kit at http://www.aeolian-hall.myzen.co.uk/test219a.jpg