ADDING A "FACILITIES" SOCKET TO A WESTMINSTER ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ If you are running a Westminster as a packet rig, you may after a while find it a bit inconvenient with all the connecting leads etc. to speaker & mic sockets. If you look on the rear panel of a Wessie, you may find a small rectangular metal plate held on with two 6BA screws. If you remove this plate, the hole that is left will take a 15-pin "D" connector. [This hole is DESIGNED to take a 15-pin "D"; it's where a "facilities" connector usually goes, to operate a selective-calling unit]. With a suitable socket mounted in this hole, you can cable up all sorts of useful things since you have more pins than are available on the mic connector. I wired duplicates of all the lines from the mic. socket, plus a +12V line controlled by the Wessie on-off switch; this is handy for running the TNC. I am working on using this socket for various other things (software-controlled channel-change to switch between the packet frequencies from the keyboard). Having the extra pins available makes things so much more convenient. The D connectors will easily handle 2 or 3 amps, so no worries about overheating. SETTING DEVIATION ON FM WESTMINSTERS ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ OK, so you have got your Wessie crystalled up, and you are ready to venture on the bands. STOP RIGHT THERE! before you fire it up, you must set the deviation, or you will make enemies of your neighbouring packet users. FM Westminster Modulation. ~~~~~~~~~~~~~~~~~~~~~~~~~~ "FM" Westminsters actually use Phase Modulation. This has the strange characteristic of producing a deviation which is a linear function of the modulating frequency. High modulation frequencies result in a greater deviation. Hence there is a need for a "roll off" of the audio response in the modulator if "linear" modulation is to be obtained. This roll-off is achieved by means of a L-C network in the modulator, and a suitable choice of coupling components. *** Attempting to bypass the LPF network will more than likely result *** in horrible unbalanced modulation! Do not try to bypass the LPF *** unless you know what you are doing, and incorporate a roll-off *** network. There are also two diodes in the audio path; these are designed to act as peak-clippers. Now the last thing that you want with packet is for the waveforms to be messed up by clipping! Setting up. ~~~~~~~~~~~ Connect the TX to a dummy-load, and the TNC to the audio input. Set the deviation pot. (RV1 on board 10 inside the Wessie; this is a preset pot usually coloured red) fully counter-clockwise. Set the TNC to key the radio and transmit either a continuous "mark" or a continuous "space" tone. Attach a scope to the junction of D2, R11 and C7 of the modulator board. Increase the audio level from the TNC until you start to see a trace on the scope, keep on going till no further increase in amplitude is noticed. This is the point at which clipping has started. Alternate between "mark" and "space" tones whilst gradually reducing the audio level from the TNC until the audio level seen by the scope begins to fall. [Note that there is a fair chance that you will notice a difference between the amplitude of the "mark" and "space" tones; set the level so that the strongest tone just begins to show a reduction] then turn the TNC audio level back by approximately 10% We now have the level of signal from the TNC set so as not to bring the diodes into clipping-mode. Next, get yourself a scanner, or other radio able to listen on the packet frequency; equip it with an antenna and a suitable connection to put the scope on its audio output. Listen to the local packet activity, and adjust the scope for a comfortably-sized display. Get the TNC to send something into the Wessie (remember we are still on the dummy-load; there will be enough leakage from this to be heard on the monitor radio). With the TNC transmitting, slowly turn the deviation control (RV1 on the modulator) clockwise until the scope displays your signal as having approx 2/3 of the deviation of the rest of your local packeteers. You will now have set the complete audio chain to give less deviation than the local users; and you are ready to go on the air and see what happens. Remember that if you over-deviate, nobody will like you, whereas if you under-deviate by a bit, nobody will mind! At least with the settings like this, you can connect to somebody without causing too much interference. Final "tweaks" of the modulation pot can be done on a quiet channel (late at night?) by connecting to yourself via a friend(!) and monitoring for rejects. Do not advance the modulation pot. by more than 10% from the point established by the tests, or you will start overdeviating. Remember, overdeviating reduces your communications efficiency by increasing the likelihood of your packets being rejected by their intended destination. IDENTIFYING PYE WESTMINSTERS ============================ Quite a lot of people are interested in buying Pye Westminsters etc. for use as packet nodes. There seems to be a degree of confusion as to how you identify the different versions; I have often heard people at rallies say 'But how do you know what frequency it covers?', or 'How do you know if it is AM or FM?'. This will hopefully give some enlightenment to anyone who is contemplating getting a cheap packet node. WESTMINSTERS ~~~~~~~~~~~~ There are two basic versions, dash-mount and boot-mount. Dash-mount ones are single units with a front panel; boot-mount ones have a 'signal unit' which is the radio proper, connected by a thick multiway cable to the control box (which just contains switches, potentiometers and indicator lights). Westminsters come in AM and FM versions, both 'VHF' and 'UHF'. VHF ones can be AM or FM. UHF ones are all FM. VHF ones can be dash or boot mount; UHF ones are almost always boot-mount. Not being familiar with the UHF ones, the following discussion will relate to the VHF radios only. NAMING ~~~~~~ All Westminsters have a serial number plate on the case (or they did when they were manufactured...). This will tell you a lot about the radio that lurks inside the box. All serial numbers begin with the letter W. The next part of the serial is the power output (in watts). Usually 15, but there are some 'W30' versions - these are 'high power' and use a valve PA. If you come across anything with 'W20' it is a 'Whitehall'; an interesting beastie with both AM and FM capability in the same box. After the power designation is a mode indicator, either AM or FM. Do not buy an AM version and think of converting it to FM. It's just not worth the effort, time or money. There is then a final letter indicating dash or boot mounting (D or B) You will then have somewhere on the plate, frequency information; this is either engraved as TX and RX frequencies in MHz, or uses the Pye designations as follows: 68-88MHz'E' band 79-101MHz'P' band 88-108MHz'D' band 132-156MHz'B' band 148-174MHz'A' band A number after the frequency range indicates how many channels are fitted (1, 6 or 10) Note that there are quite a few about with 'splits' so they may transmit on 'B' band and receive on 'P' band! Beware! Finally, channel-spacing. Three versions exist, designated as follows: 12.5KHz'S' 20/30KHz'V' 40/60KHz'N' 'S' and 'V' versions are the ones to go for; 'N' models have the wonderful ability to monitor 2 channels at the same time... You can change the channel-spacing by replacing the crystal block filter on the 1st IF board and the L-C filter in the 2nd IF, but its usually not worth the cost (unless you can get another scrap Wessie for next to nothing). Now you have identified the beastie you want, go and find one. Expect to pay between 5 and 25. If you pay more, you are probably being taken for a ride. Do not worry about appearance; remember, these radios are usually fitted to water-board, council, electricity-board vehicles, and they are designed to take a pounding that would destroy the more familiar amateur radios. The most filthy examples can come up looking pristine after a going over with white spirit and a toothbrush. PYE WESTMINSTER - CONNECTIONS ============================= Westminsters have a curious power connector; this is to allow use in both positive and negative-earth applications. The 7-pin plug needs to be wired correctly, with the right straps, before the radio will consider working. Unless you get it 100% right, the radio will stay resolutely silent. The wiring is as follows: Positive earth:- Connect power to socket SKA: +ve Pin 1 -ve Pin 3 Link the remaining pins: 2 to 5 4 to 7 Ignore pin 6. Negative earth:- Connect power to SKA: +ve Pin 1 -ve Pin 7 Link the remaining pins: 2 to 3 4 to 5 Ignore pin 6. The non-earthed line should be fitted with a 5-amp quick-blow fuse. Note that the radio is reverse-polarity-protected; the internal relay is fed from the on-off switch through a diode, so it will not power on in the event of the supply being reversed. Speaker connection ~~~~~~~~~~~~~~~~~~ On the back of the radio (or the control box in the case of boot-mount radios) there will be a length of two-core cable. This is frequently colour-coded brown and blue (mains cable). Do NOT be tempted into putting a mains plug on the end and plugging into the nearest 13-amp ring main; these leads are the SPEAKER connections! Any speaker with an impedance between 3 and 8 ohms will do. Whatever you do, do not short the speaker leads either to earth, or to each other; if you do, then the audio output transistors are more than likely to take offence at such treatment, and refuse to have any further involvement in proceedings. Microphone connection ~~~~~~~~~~~~~~~~~~~~~ This is via a locking DIN plug either on the control box or the radio. Any suitable dynamic microphone between 200 and 1000 ohms impedance will do. Note that if using a genuine Pye microphone, there are two types, one designed for AM sets, the other for FM. Using an 'AM' mic. on an FM radio will result in excessively toppy speech (the frequency responses of the microphones are different) FM mics on AM radios give muffled, woolly speech. This of course doesn't matter if you are using the radio with a TNC, because you will be using FM in any case (I hope). The mic. plug is wired as follows; note the non-standard numbering of the DIN plug used by Pye. Looking at wiring side of plug, pin 1 is at 8 o'clock, pin 2 at 10 o'clock, pin 3 at 12 o'clock, pin 4 at 2 o'clock and pin 5 at 4 o'clock. Wiring as follows:- Pin 1 Microphone live Pin 2 Microphone screen Pin 3 PTT Common Pin 4 Earphone audio Pin 5 PTT Make. This is VERY convenient for wiring to a TNC since TX and RX audio are all available at the one socket. WESTMINSTER MODIFICATIONS FOR 4 METRES ====================================== All references to positions of circuit boards are made with the transceiver controls towards you. All measurements are made with an analogue multi-meter of at least 20K ohms per volt. CONNNECTED TO THE NEGATIVE SUPPLY. DO NOT USE A DIGITAL VOLT-METER!! Oscillator multiplier: The oscillator multiplier is located at the top right hand side next to the PA strip. For your information it will be found that the cores of the oscillator multiplier board will when correctly tuned be quite near the bottom of the can. L1 tune for maximum at test point 1 L3 tune for minimum at test point 1 L4 tune for maximum at test point 2 L1 L3 retune for maximum at test point 2 Receive strip: The receive strip is located on the right hand side and runs from back to front of the transceiver. Locate L3 remove the original capacitor, replace with an 8.2pF. Locate L4 remove the original capacitor, replace with an 8.2pF. The modulator driver board: The modulator driver strip is located on the left hand side and runs from back to front of the transceiver. It will be easier to remove this board and solder the caps. underside. Find L3,L4 and solder a 22 pF cap across in parallel with the original cap. Find L5,L6 and solder a 15 pF cap across in parallel with the original cap. Find L7,L8 and solder a 8.2 pF cap across in parallel with the original cap. Find L9,10 and solder a 5.6 pF cap across in parallel with the original cap. Find L11, and solder a 5.6pF cap across in parallel with the original cap. Find L12, and solder a 8.2pF cap across in parallel with the original cap. Tune L1 L2 for maximum at TP1 Tune L3 for minimum at TP1 Tune L3 L4 for maximum at TP2 Tune L5 for minimum at TP2 Tune L5 L6 for maximum at TP3 Tune L7 for minimum at TP3 Tune L7 L8 for maximum at TP4 Tune L9 for minimum at TP4 Tune L9 L10 for maximum at TP5 Tune L11 for minimum at TP5 Locate the PA strip: Find test point 1 on the PA strip. This looks like a small pop rivet on the printed circuit board. Tune L11 L12 for a maximum at test point 1 on the transmit strip. Connect a power meter and 50 ohm dummy load to the antenna connector of the Westminster and tune the PA stage for maximum output. For use on packet radio: To input the audio from the TNC find board 10 and locate the junction of TR3 R16 and R14, feed the audio in via a series circuit consisting of a 68K resistor and a 10nF cap. Deviation is controlled via VR1. Short pins 3 and 4. To output audio to the TNC find board 3 and remove C32 which is a 220nF. Substitute this for a 1k resistor in parallel with a 47nF capacitor. Find board 8 and locate pin 8 and take the audio from here. Ground pin 9. IMPROVING SQUELCH ACTION ON PYE WESTMINSTER ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ This simple mod. will result in a faster-acting squelch action, and reduce the number of packets you miss because the squelch lopped the front off! Squelch operation ~~~~~~~~~~~~~~~~~ The Westminster squelch operates by detecting the high-frequency component of the white-noise generated in the discriminator under no-signal conditions. The detected output is passed to a Schmitt trigger circuit to provide clean transitions between "open" and "closed" squelch states. The problem is that the detector circuit has a time-constant which is really too long for satisfactory packet operation; the time taken for the squelch to open after acquisition of a signal results in the loss of packet headers, and more retries. Improving Squelch Response ~~~~~~~~~~~~~~~~~~~~~~~~~~ The squelch board is number 7. To improve the action, replace the following components: C7 and C8; 0.47uF tantalum bead; replace with 0.22uF 16VW tantalums. C9; 22uF electrolytic or tantalum; replace with 4.7uF tantalum, 16VW. R14; 6.8K; replace with 2.7K. The above changes of values are the ones I have found to give reasonably reliable operation in several VHF Westminsters; note that the squelch action will be noticeably "clicky" for audio work, and this may be annoying if you use the same radio for mobile operation. If you want to be really enthusiastic, you can consider taking the RX audio off from pin 8 of the squelch board to the TNC; this will slightly reduce the distortion on receive, since it bypasses the audio power stages; note though that the front-panel volume control on the Wessie will be inoperative if you do this, so you will have to set the audio level on the TNC input instead. Couple the audio from pin 8 through a 10uF electrolytic; use screened cable. 73 Pete, G6WBJ @ GB7SDN PORTED AND REALIGNED FROM PC TO AMIGA FORMAT BY BOB G0LBQ - N6XDI.