_ _.  _. . .  . . . _ _  .  . . . .
G B 3 Edge Hill
 _ _.  _. . .  . . . _ _  .  . . . .

The Logic-8 Module

Logic-8 was designed to be an interim Module to provide a Repeater within adequate control, prior to the introduction of a more comprehensive facility. However, it was recognised that some Installers may not wish (or have funds) to proceed beyond Logic-8, once their Repeater is on the air and is providing a satisfactory service to its Users. Therefore, further modularity has been applied within the Logic-8 Card so that only two of the four sub-systems are essential, as a short-term expedient, while the remaining two mechanisms are options which would only be used if the Repeater is to operate as a long-term Logic-8 installation.

In practice, even the two options are sufficiently simple (and inexpensive) that most Installers would implement them for a realistic set of "user-friendly" on-air characteristics from the outset. Furthermore, within the circuits for each mechanism, the Installers are at liberty to select component values and make adjustments, both on the bench and then on site (though not on air, remotely), to alter these characteristics to satisfy any specific user-requirements. In most cases, the selections and adjustments relate to the onset and durations of the significant timed events; the adjustments being achieved by a jeweller's screwdriver applied to a multi-turn potentiometer (e.g. the light-blue items between the DILICs in the photo below).

The Logic-8 Sub-System PCB

LM .1 Logic-8 Sub-Systems

Logic-8 operates on the basis of two essential sub-systems; the Access-Hang (AH) control and the Periodic Callsign (PC) generator, and two optional mechanisms; a Time-Out (TO) control and an end-of-Over Pip (OP) generator.

The Access-Hang timer determines when the Repeater is "up" to its users. It is activated 'from cold' by a user-access tone-burst to engage the through-audio coupling and to sustain this condition to 'bridge' the short loss of Input during the turn-arounds between consecutive user overs.

Independently, the Periodic Callsign generator operates at preset intervals which are adjustable up to about 15 minutes (max.), to deliver the Identification.

The optional Time-Out control is a watch-dog timer to prevent a single user from sending an uninterrupted over of excessive duration. It also guards against excessive spurious responses under "lift conditions" when the Repeater may receive abnormal co-channel signals.

The optional "Pip" tone generator signifies when a user has dropped carrier (by delivering an Over-"E") and has the further sub-option that Installers may engage a feedback path to generate a sequence of Quiet-Time pips, if required.

All four sub-systems operate entirely with logic signals to or from the ATIC module, so the minimum Logic requires just one Logic-8 Card and one ATIC, coupled together. As noted earlier, the ATIC also couples the Receiver and Transmitter radios together through the CAIRO-8 interface.

LM .2 Optional Periods

Each of the four mechanisms uses a pair of 555 timers, from the single 556 IC-package. Two mechanisms (AH and TO) use the partner-discharge arrangement while the other two use the first 555 of the pair in a conventional configuration; one as an astable with a "long" cyclic period (PC) and the other as a monostable with a "short" period (OP), to trigger the second 555 into its monostable action. The periods are pre-set but adjustable because each stack is implemented with fixed and variable resistors and a range-setting capacitor.

It is useful to consider that any of these timing mechanisms could have been designed as counter-divider derivatives from the 50Hz signal - CLOCK - that is available from the ATIC. However, the 556 approach is intrinsically easier to comprehend and it offers Installers greater flexibility in their choice of the major periods, both in the final on-site implementation and during the earlier commissioning stages in the constructor's workshop. For example, if the Time-Out option is implemented, some Keepers may wish to allow the Users a "generous" period, between about 4 and 6 minutes, while other Keepers may wish to be more "restrictive" with an interval between about 1 and 3 minutes. Then again, some Keepers might wish to be generous in the early years, when the Users are small in number, but gradually reduce this generosity, over a period of several years, as the user-base increases and diversifies. Such adjustments, which are less straightforward in the counter-divider approach, are easily implemented in the 556-approach.

In many cases, we will declare the value of the stack resistors and capacitor required for the typical, minimum period so that Installers may simply use two such capacitors in parallel to achieve longer durations, if required. This principle has the further advantage that for the construction and test phase, a small-value capacitor can be installed at first, to avoid unnecessarily long waiting-times, after which the required capacitor(s) may then be substituted. Unless a very elaborate counter and comparator scheme was used, designs based on the 50Hz signal would impose unnecessary constraints on the various periods which, in reality are determined more in response to local "custom or request" than by License demands or "best practice" dictates.

LM .3 Access-Hang Control

The essential Access-Hang control - AH-556 (Fig. 8) - is a partner-discharge pair. The latch OUTput of the Slave (555-B) determines when talk-through is valid by asserting (^OPEN^ = 1) the Through-Audio coupler on the ATIC. (It is also ORed with a control signal from the Beacon Callsign sub-system to assert the PTT.) This Slave latch becomes set when TONBST goes low at the TRigger input, to remove the DIScharge clamp on the stack. Normally, this would allow the stack to ramp-up, but the Master (555-A) holds the capacitor discharged for as long as the Receiver Squelch is open, and ^SQLCH^ is high ('1') as a consequence. (Here, the master is a digital Schmidt with its TR/TH inputs working to its internal resistance ladder.)

Only when ^SQLCH^ goes low ('0'), as a user completes his Over, does the Master release the stack to begin its ramp towards the Slave's THreshold voltage. This level will be reached in the preset Hang-Time (typically set at 5 secs.), for ^OPEN^ to go low to disable Through-Audio (and PTT). However, if, as is normal, the next user transmits within this period (with or without Tone-Burst) the ^SQLCH^ signal causes the Master to DIScharge the stack capacitor again to prevent the Slave from completing its intended monostable action. Thus, the OUTput remains high (^OPEN^ = 1) to span all user turn-around breaks of permitted duration. (This Access protocol does not require a period of user-speech, so has no dependence on the ATIC's VOX signal.)

Fig. 8 : Access-Hang Partner Discharge Pair; AH-556

LM .4 Periodic Callsign

The second essential mechanism, for the periodic Callsign, uses the PC-556 pair (555-C and 555-D, Fig. 9). The 555-C runs as an astable with an almost 50% duty cycle, and an overall period which is adjustable between about 4 and 15 minutes (max.), as the periodic interval. When its OUTput makes the mark-to-space (1 to 0) transition, after 50% of the full interval has elapsed, it TRiggers 555-D into giving a short, fixed-period monostable pulse of about 60 mS (^KICK^). This pulse sets a simple latch (BCON) in the Callsign generator, for the Morse-code sequence to ensue. Meanwhile, the 555-C completes its discharge period and re-commences the charge period before it re-triggers 555-D again, for a Beacon cycle that runs in perpetuity, to ensure compliance with the License, regardless of what may happen, elsewhere in the Logic-8 scheme, in response to user activity.

Fig. 9 : Beacon Callsign Period Astable Pair; PC-556
This Fig is currently in error for not showing the ^open^ and Zenner at 555-C

For the convenience of Installers, the long-period Stack (at 555-C) has been devised to meet the general astable criteria of the 555, while also allowing simplified calculations to be made in the selection of the capacitor. In summary, the capacitor's value, in µF, is also the mid-range period, in seconds. Thus, for a 5-minute Beacon period (i.e. 5 x 60 = 300 seconds), the stack will require a Capacitor at the nearest preferred component value, i.e. 330 µF.

The usual, rigorous time-constant calculation for an astable operation, should be " t = 0.693 x (Ra + 2 Rb) x C ". However, with the main "R" (Rb = 510K (or 470K)-fixed + 500K-pot = 1M) being so much greater than the discharge pull-up (Ra= 1K), this relationship simplifies to just; " t = 1.4 x R(b) x C ". Then it may be noted that the mid-range adjustment of the 'pot' yields a combined value of about 725K, so this further simplifies to " tmid = 1.4 x 0.725 x C " or just " tmid = C ". The approximate range limits are then; " tmin = 0.7 C " and " tmax = 1.4 C ", respectively.

Thus, the range of periods with a 330µF or "5-Mins" capacitor, lies between about 4 and 7.5 minutes, allowing an exact 5-minute period to be found (by adjustment of the pot), whatever the tolerance discrepancy in the actual capacitor component that is used. Similarly, a pair of 330µF capacitors in parallel provides a nominal period of about "10-Mins", that is adjustable between about 7.5 and 15 minutes; the latter being significant as the License upper limit. But in this context, please note that it is permissible to use a 470K (E12 series) resistor instead of the 510K (E24 series) as shown, but this may result in a stack which cannot be adjusted to a full and exact 15-minute period, if the two selected capacitors are on the low side of their batch tolerance. Some Keepers may wish the Beacon period to be adjusted to operate exactly at this 'License-limit', in which case a 510K resistor should be used. (A 560K may not be used because it will take the impedance of the entire stack outside the criteria for reliable 555 astable operation.)

LM .5 Beacon Generator

The issue of the periodic Callsign commences as the 60mS pulse (^KICK^, from 555-D) sets a simple S-R latch (74_00 NANDs) to assert BCON, Fig. 10.
This releases the reset inhibit on a 74_90 IC which is configured as a ÷5 counter on the 50Hz CLOCK (from the ATIC) to generate 100mS clock pulses (63% duty-cycle) as the basic "dit" period of 12wpm Morse. At the same time, BCON also releases the reset inhibit on a 74_393 IC which is configured as a ÷256 (÷28) counter which, in principle, runs from 0 to 255 (maximum). The counter's outputs are used to assert the 8 address-lines (A0 - A7) of a 74S287 PROM (e.g. [639-951]) to access each location in turn, for one 12wpm "dit" epoch.

Fig. 10 : Beacon Generator and Glue Logic

When the Callsign identification sequence has been delivered, the PROM produces a low output (Q4) to reset the BCON-latch and thereby reset to all-zeros, the ÷256 and ÷5 counters. Hence, the PROM "rests" with its first location accessed, ready for the next periodic Beacon sequence to be initiated by the PC-556 pair. (If required, the constructor may configure the 74_90 as a ÷10 counter so that the basic "dit" period clock (into the 74_393 IC) runs at 200mS for Callsigns to be delivered at 6wpm.)

LM .6 Morse Sequence

The Callsign is generated as a Morse Code sequence from a pre-programmed 1K (= 256 x 4) PROM (74S287). Each consecutive location is accessed in turn, by the ÷256 counter, for the resulting outputs at Q1-Q4, to be a sequence of four simultaneous waveforms. Typically, Bit-0 (Q1) carries the required Morse sequence, encoded as a '1' for a "dit" sound and a '0' for a silence of 'dit'-duration. In the full sequence, three consecutive 1's (dit-periods) constitute a "dah" and three consecutive 0's constitute an inter-letter silence. Considering this aspect alone, the Bit-0 (Q1-output) sequence from the first forty-five consecutive locations (of this and any other UK Repeater; "GB3...") would be;

0 0 0 0 0

 1 1 1 0 1 1 1 0 1

 0 0 0

 1 1 1 0 1 0 1 0 1

 0 0 0

 1 0 1 0 1 0 1 1 1 0 1 1 1

 0 0 0

 1 ....

  _ _ .   _ . . .   . . . _ _  

see note *

  G   B   3    
  Dah dah dit   Dah di di dit   Di di di dah dah    

n.b. : Exaggerated spacing and underlining used to help visualise the morse character bit sets

The sequence of leading zeros is included as a commencement delay so that the second and third bits (Q2, Q3) in each PROM location, can be used to assert the required ATIC tone generator "taps" for the Callsign. These selections must be made first, to allow appropriate settling times, so that the morse sequence is transmitted in its entirety. The delay is also needed to allow sufficient lapse time, after BCON has engaged the Transmitter, to ensure that the receiver squelch, of any user who may be monitoring the Repeater during its idle times, has been opened by the resulting "blank carrier" transmission. The commencing delay, together with a delay at the end of the active PROM sequence, is also required when the Over-Pip option is implemented. These silent extensions ensure that, if the beacon commences just as an over-pip is due, or when a pip-sequence is in progress, the beacon Morse and pip(s) are suitably separated. An interlock to the Over-Pip logic, ensures that the beacon sequence takes priority, but suspended pip(s) will resume when the beacon delivery is complete.

* The last " 1 ... ", in the above sequence, is either a first 'dit' (in its own right) or the first of three consecutive dit epochs constitutinng a 'dah', as the commencing Morse character of the repeater-specific letter-pair. The shortest (UK) Callsign, in terms of this PROM coding method, would be GB3EE requiring only a further 4 active locations, while the longest might be GB3JQ (or QJ, JJ, QQ) requiring a further 28 active locations. (GB3EH is one of the shorter UK callsigns in requiring a further 11 locations.)  We recognise that PROM coding of callsign sequences is a widespread practice and we make no extravagant claims for our approach in relation to this issue.

But for reasons of the variable length issue, PROM Bit-3 (Q4) is used to label as active, all contiguous locations which have encoded Morse within them. It is used to perform a reset action to disable the counter-divider chain when the first 'un-used' location is encountered at the end of the beacon sequence. It is this fourth bit which makes the PROM technique applicable to any Callsign, regardless of its precise "length" when translated from Morse into a bit sequence. Bit-3 (Q4) should be '1' in all active locations, then '0' for the remainder - in practice, the first 2 or 3 unused PROM locations.

LM .7 Tone Bits

Two further PROM bits (B2, B1) determine which one of four (of the six) ATIC tones will be used for the Callsign since the output codes (at Q3, Q2) can be "00" = 1750, "01" = 875, "10" = 1K or "11"= 500 Hz. This provision allows the Installers to use one tone for the Callsign itself and a different tone for any extensions which might follow; e.g. the locator ID (e.g. Maidenhead) or the CTCSS group letter. By this means "GB3EH---IO92FC---A", for example, can be sent with "GB3EH" at 1KHz, the locator extension at 875Hz and the CTCSS-"A" extension at 500Hz. The change of pitch occurs between locations which coincide with the long silences ("---"). The PROM location sequence for this would resemble the table (addresses and values given in Hex);


Installers might also like to consider that the longer the Beacon sequence, measured in terms of the number of active PROM locations needed to encode it, then the longer might be the periodic interval between issues so that regular listeners to the Repeater, when it is idle, are not too frequently subjected to its intrusion into their tranquillity. However, this consideration must be balanced against the need for travellers and occasional users (perhaps with scanning transceivers) to discover that there is a Repeater here which is "alive and well"!

LM .8 The Time-Out Option

The optional Time-Out watch-dog mechanism uses another partner-discharge pair of timers (TO-556; 555-E and 555-F, Fig 11). The single CR-stack, at the TR/TH inputs of the Slave (555-F) is chosen for an adjustable but "long" Time-Out period, e.g. 2 - 5 mins. This will commence when the Master (555-E) releases its DIScharge clamp on the capacitor, every time the SQLCH signal goes low as the Receiver re-opens to new user carrier. However, each time the squelch closes again, between user overs, SQLCH goes high to cause the Master to discharge the capacitor again, thus re-triggering the Slave for its latch OUTput to remain high (SHUT = '1'). Hence, this monostable only runs to completion when uninterrupted user-input exceeds the permitted period. Whenever it does, SHUT goes low ('0') to immediately RESet the Slave (555-B) of the Access-Hang partnership. As noted already, the AH-partnership now requires the Squelch to close, and then re-open with an accompanying Tone-Burst, to re-initialize the protocol for re-Access.

Fig. 11 : Time-Out Partner-Discharge Watch-Dog Pair ; TO-556

LM .9 Over-Pip ("E")

The optional Over-Pip facility uses the OP-556 pair (555-G and 555-H, Fig. 12) to generate either a single Pip ('E') at the end of a user-over, or else a sequence of Quiet-Time pips which span the Hang-Time period, if a simple feedback path is enabled. This is not a partner-discharge pair, but a conventional pair of cross-coupled monostables each with its own CR-stack.

The 555-G is TRiggered by a gated and a.c.-coupled BUSY signal, from 555-A, and runs for an adjustable monostable period of about 1 to 3 seconds. At the end of this period, the a.c.-coupled output triggers 555-H for an adjustable monostable period of about 60mS (being a 20wpm "dit" epoch). The 555-H OUTput, ^PIP^, is then used to gate open a tone-path on the ATIC's audio coupling bus.

Fig. 12 : Over-Pip or Quiet-Time Pips Generator Pair ; OP-556

LM .10 Quiet-Time Pips

For the optional Quiet-Time pip sequence, the 555-G must be re-triggered each time the 555-H completes its single-pip monostable period. This is done with a feedback path of NAND gates, which OR the BUSY signal with the 555-H ^PIP^ output. A further two-input NAND gate is then included (as a controlled inverter) so that its second, but normally tied-high input, ONEPIP, may be grounded to disable this pip-sequence option. The switch would be a DIP-type, banked with those needed to select the pitch at which the pip-tone is delivered.. The 555-G partner is further inhibited, at its Reset input, by the ^OPEN^ signal from 555-B, so that the single pip (or the pip sequence) can only ever occur when the Repeater is genuinely "up". Likewise, the BUSY signal is also used to inhibit the 555-H partner, at its RESet input. This control will suppress any pip which might become due for issue just as a user commences a new over, and will also suppress the pips which might coincide with a Beacon callsign transmission, whilst also causing them to resume (for a full Hang period) when the Beacon delivery is over.

LM .11 Interlocks

To ensure that the Logic operates correctly, regardless of whether or not the optional mechanisms are implemented, the scheme includes some additional interlocks.

The logic output, BUSY, from the 555-A latch is used to trigger the optional OP-556 pair into issuing an Over-Pip (or the sequence of quiet-time pips), Fig. 12 above. To ensure that such issues never clash with the delivery of a periodic callsign when one is in progress, the RESet input of 555-A (Fig. 8) uses BCON, from the Beacon Callsign mechanism, as a second inhibit signal. This ensures that BUSY encodes all the conditions which make the Transmitter busy, and will only go high when there is neither user audio nor beacon tones to relay. If a user was to finish an over during a Beacon Callsign epoch and the next user delayed his reply, as if to let the Repeater "have its say", then the pip(s) will occur after the callsign has been delivered in full. Conversely, if the next user replies without waiting, the pip(s) are suppressed. The RESet input of 555-B is tied high with a 4.7K resistor so that it will accept the secondary control signal, SHUT, from the optional Time-Out mechanism (TO-556 pair, Fig. 11). This forces ^OPEN^ to go low, as the means to close-down the Repeater if continual user or, more likely, spurious Repeater Input keeps the CR-stack from ramping-up to its hang-time completion. Once this has occurred, the Squelch must close and then user Input must be accompanied by a Tone-Burst, for TONBST to trigger the 555-B slave again.

LM .13 Glue Logic

Fig. 10 also shows the ancillary logic - the 'glue' - which is required to invert or combine the various Logic-8 control signals into appropriate outputs to the ATIC module. The through-audio path, which couples receiver audio to the transmitter, must engage whenever the Repeater is OPEN, as determined by 555-B of the AH-556 pair. Inversion of ^OPEN^ yeilds EAR. The transmitter should be keyed (KEY = PTT) both when OPEN is true and also when BCON is true. The NAND-and-Inverter pair of gates achieves this.

The level at which tone(s) are injected onto the ATIC's audio coupling, when the callsign is delivered, is considered to be dependent on whether the Repeater is idle or is in use. If it is totally idle, in the sense that there is no user activity whatsoever, then the "high"-level coupler is selected and the "low"-level coupler is disabled. Conversely, if it is OPEN and there is actual user talk-through in progress (as determined by VOX), then the "low"-level coupler is selected and the "high"-level one is disabled, so that the callsign does not intrude on those users' exchanges. Another NAND-and-Inverter pair of gates achieves this. Please note that the H and L couplers are operated with mutually exclusive signals so there is never a "loud" level, in the Logic-8 scheme.

Lastly the 'glue' logic also includes a 74_158 chip with its four, 1-of-2 multiplexers. These use ^BCON^ to determine which ATIC tone "tap" should be engaged (MUX-3, 2 and 1 inputs) and which signal should pulse that tone (at the ATIC's strobe control input). When ^BCON^ =1 then the selection is determined by the PROM, as detailed above, and the Strobe is the morse train. Conversely, when the periodic callsign is inactive, this chip selects the signals which are associated with the optional pip-tone sub-system. Three of these are derived from DIP-switches which Installers use to pre-determine which pitch is used for the pip(s), whilst the Strobe is activated by the ^PIP^ pulse. It should be noted that any one of all six ATIC tones is available to this switch selection allowing Installers to choose one of the tones which is not used in their Beacon callsign/extension sequence.

If over-pip(s) are not required and the OP-556 IC is not installed, it will be essential to set these three DIPS to all-ON or all-OFF. (The fourth DIP-switch determines if a single pip or a sequence of Quiet-Time pips should occur.) It is also useful to note that the periodic-beacon generator controls the chip-select input to ensure that callsign delivery is never corrupted by pip(s), though these will resume afterwards if the circumstances are correct, as noted earlier.

LM .14 Stack Capacitance

Assuming that an ATIC module has already been constructed and tested, it is prudent to construct and test the Logic-8 card on a piecemeal basis. This avoids having some of the more complex interlocks in place which, in the event of a fault somewhere, might inhibit an otherwise working sub-system. In particular, it is convenient to establish, on the basis of a small-value stack capacitor, that each of the two long-period mechanisms; the periodic callsign generator and the time-out watch-dog, are working correctly before setting their stacks to the desired period. This requires the initial use of a "10%" capacitor which is replaced with the target value item, once the mechanism has been tested and given a provisional calibration. Thus, if the eventual capacitor should be 330 µF, for example, install a 33 µF item instead, as a temporary substitute. Meanwhile, calculate in seconds, the intended in-service period, e.g. 5 minutes = 300 seconds, and then adjust the pre-set "pot", against a stop-watch, for a cyclic period which is "10%" of that target, e.g. 30 seconds ± 1 second. When this "10%" period has been achieved, install the target capacitor, noting that the (multi-turn) "pot" should not require more than about two full turns, either clockwise or counter-clockwise, to pull the stack into the desired period. Before making these final adjustments, allow time for the components to reach their operating temperature.

The value of capacitance specified for the long-period stacks, will require an electrolytic. Generally, these have a poor batch tolerance, ±20%, but the item tolerance, ±5%, is satisfactory. Therefore, once a 555 stack has been adjusted for the actual capacitor which has been installed, it should maintain the desired period for a reasonable service lifetime in this application, given that the "long" periods, in Repeater conventions, are not subject to a rigid specification.

When the 556, as a dual-555 package, is used in this project it is usual to implement the Master (or the diagram-left timer) with the A-side 555 (IC pins 1-6) and the Slave monostable (diagram-right) with the B-side 555 (IC pins 13-8), for consistency during set-up testing.

Fig. 13 : Suggested Layout for the Logic-8 Module
(if prototype Eurocard)

LM .15 Construction and Test

To accompany the suggested layout for a prototype Eurocard (Fig.13) we offer the following construction sequence.

First, implement the 7805 (5V @ 1A) regulator to make use of the ATIC's 12V supply. Ensure that this has its two smoothing capacitors (1µF and 100nF) installed in close proximity to the device pins and that the 5V-rail track is correctly live. When all the Logic-8 ICs are in place, this regulator will generate heat and will require a clip-on heat-sink; e.g. 40 °C/W [402-260].

Next, install the Access-Hang (AH) 556, as shown in Fig. 8, together with its external passive components; CV-pin decoupling and chip supply decoupling, and in this case, the actual stack capacitor for the intended hang period. Temporarily install a 4.7K pull-up resistor on the 555-A reset and use a wandering ground lead to simulate a TONBST pulse on 555-B to observe it firing, at its ^OPEN^ OUTput (pin-9). Fire it again and quickly apply the wander lead to the ^SQLCH^ input of 555-A to observe that ^OPEN^ now remains high for as long as that wander lead is applied, and that it continues to hangover for a short while, when the wander lead is removed. Repeat this test, having first altered the stack pot, to check that the hang-time adjustment is working correctly. Adjust the pot again for a hangover period of about 5 seconds, or longer (say 10 seconds) if the pip generator is to be implemented and is to be configured for its Quiet-Time pip-sequence issue.

Now install the periodic Callsign PC-556 timer and its passive components, Fig. 9, with a temporary 22µF capacitor in the CR-stack. The output of 555-C (pin-5) should be observed (Logic Probe) to be pulsing for about a half-second, at one second intervals, whilst the output (pin-9), ^KICK^, of 555-D should be "blipping" at the same rate. If so, it is expedient to install the BCON-latch 74_00 followed by the 74_90 and 74_393 devices. If all is well, the various outputs will be observed to be toggling, each at twice the rate of previous output. At this stage there will be no reset action and, once the counter-chain has started, it will run indefinitely. When installing the eventual stack capacitor (e.g. 330 µF), expect the first activity in this chain, after power-up, to occur after a full period has elapsed. (In 555 astables, the stack capacitor must charge from 0V to "2V/3", before cycling between "2V/3" (˜ 3.4 volts) and "V/3" (˜ 1.7 volts), thereafter.)

Next, install a DIL socket for the PROM (or a ZIF-DIL), and the 74_158 multiplexer. Insert a DIL-header, into the PROM socket, and use a wire link to couple one of the low-order address line pins, e.g. A2 (pin-7), directly to the Q1 (pin 12). Wire the most significant address line, A7 (pin-15), to a spare 74_04 inverter and return its inverse to the PROM's Q4 output (pin-9). Each time the periodic callsign mechanism now fires, a square-wave should be observed at the ATIC-MUX Strobe output (pin-12) of the 74_158. This will last for 127, 100mS clock pulses after which the A7 address-line signal will reset the BCON-latch. If this is so, and a suitably blown PROM is available, remove the DIL-header temporary wiring and insert the PROM to observe the morse pulse train at the 74_158, pin-12 output (or edge-connector [18]).

When installing the OP-556 for the pip-generator, Fig. 12, use the specified stack capacitors (2.2µF and 1µF) but omit the ^OPEN^ and BUSY signals from the respective 555 partners so that the pair may run continuously in the astable mode. Adjust the two stack resistors until the ^PIP^ output (pin-9) of 555-H "blips" at about a 1-second rate. However, please note that this interim circuit may not always exhibit the astable property because the two partners can sometimes power-up in a mutual deadlock. If this appears to be always the case, after two or three power-up attempts, it will be necessary to include at least the ^OPEN^ signal and manipulate the AH-556 mechanism before the OP-556 pair can be checked and adjusted.

 _ _.  _. . .  . . . _ _  .  . . . .
G B 3 Edge Hill
 _ _.  _. . .  . . . _ _  .  . . . .



Stuff to deal with / check
The Time-Out watch-dog pair (TO-556, Fig. 11) is another candidate for a substitute capacitor during testing; either a 10µF or 22µF.

Install the remaining glue-logic ICs and observe that the various signals, already known to be correct, can be observed at the Logic-8 outputs for the ATIC inputs.

Lastly, at some stage during construction or pre-site commissioning in the workshop .........

Move these to the appropriate pages in the final document.

(Here, I record special thanks to;

Brian, G7NGE for developing the Partner-Discharge 556 timers,

G6WTM who worked with me to develop the prototype ATIC.,

G6WTM who worked with me to develop the prototype Logic-8,

Mark, ?call? for developing the prototype Logic-80 software.