TXID1620 Repeater Voice Identifier (10mins) RCS4600s

How it works ...                            correct to January 4 1998

Firstly let us divide the circuit up into 2 main sections.
The Timing Engine and the Voice/REC.Play Section.

The TIMING Engine.
This section uses the most frustrating, versatile, cantankerous, handy,
ornery, unpredictable timing I.C. ever devised, now I know you purists
would suggest using an XR2240, the rolls royce in timing IC's, but as
this whole project is expected to get most parts from local junk boxes,
the XR2240 is not a common item. We will use the 555, an excellent
device for detecting lighting storms many miles (kilometers) away or
false triggering from the fridge/frypan etc. Why choose this item? well
it's abundantly available, small, convenient voltage and most of all it
is very cheap, and with a little bit of careful design we can make it
99.9% successful and those are quite acceptable expendable military
odds compared to the cost and fiddling around of an alternative.
When all the common discrete components are inserted "EXCEPT the IC's"
and you have used sockets of course, insert the 555, IC3 (area 10 I)
and then put the +12v DC ON. By adjusting the trim pot RV4, a 10k pcb
mount trimmer (11M), do a minor adjustment to set the clock rate at
about 64 Beats Per Minute (we will call this a binary minute), with the
correct values of capacitance C18 and the 2 resistors R12, R13 this
will be very close with the wiper of the trimmer at 12 o'clock.  The
diode D3 is a by-pass of timing so you have one resistor set for ON and
the other resistor for OFF. You will notice the mandatory o.1uF disc
ceramic capacitor (C19) on the pin 5 of the IC to ground, another is
close to the power rails of the IC4 (C20) to try and remove any RF or
noise near the 555 and the main counter chip alongside.  LED 3, the LED
nearest the 555 should now be blinking at the desired speed, 64bpm.
Turn the DC power OFF now please.

As LED3 was blinking correctly along at 64bpm then the next item to
insert and test is IC4 the 4040, so with the DC power still  OFF,
insert the 4040, (area 14 I) be careful as this is a cmos device, so
beware of static, and soldering irons with poor earths, if your
girlfriend is helping you build this, inform her that static from nylon
underclothes can interfere with this procedure, so please remove all
offending garments (well it was worth a try), now back to the 4040  a
12 stage Binary divider counter, around 10meg maximum count speed, so
to do 64bpm is a yawn.   The 4040 is one of the most useful/powerful
IC's to come out of the 70's , these days they are all 'B' (Buffered)
series, thank goodness, or even more stringent installation procedures
would be required.   (harken back to VK2XOF Mark's pre-amp warnings).
Power up the board again and if all is well LED8 will blink every 64
pulses, and if you set up 64bpm then each LED8 blip is a MINUTE, use
this to fine tune the trimpot RV4.

The 4040 is a binary counter, that is, there are 12 stages of divide by
2 from /Q1 to /Q12, each successive stage requires 2 'bumps' to move
along to the next. If a pulse comes into Pin 10 (known as the clock) of
this IC then, lets say it moves up a 1/2 a spot, another pulse from the
555 now comes in and moves the blip along to /Q1 which is Pin 9, so
therefore Pin 9 would blink every 2nd pulse from the 555. Binary is the
sweetest form of mathematics and this procedure is now carried onto
/Q2, meaning it requires 2 /Q1 blips to move it to /Q2, and so on,
resulting in the value of 4 blips from the 555 to make it move off the
/Q2 pin, LED8 is connected to /Q7 thus requiring 64bpm from the 555 to
'energise' this pin with a 'ON' command for the next minute, when
another binary minute has elapsed it will 'push' the /Q7 along to /Q8
representing the 2nd minute, this is displayed as LED9 and it is Pin 13
of the 4040.  Pin 13 (representing 2 binary minutes) is an important
pin in this project, keep in mind.  After more pulses have occured and
the /Q7, 1 minute LED has blinked 9 times, we are in final countdown
mode.     This is displayed as 1,2,4,8,16,32,64,128,256,512,1024,2048.
The binary timer is clocking the registers up inside and 'moving' them
along, /Q7=1min, /Q8=2min, /Q9=4min. /Q10Pin14=512 'blips' so what this
means after 512 blips this pin is now active, and we are now waiting
for the earlier /Q counters to fill up another 128 blips, thus making
now /Q10=8mins (LED10) and /Q8=2mins (LED9) both 'high' 12v, this
represents 512+128 blips or 640 binary seconds, and at 64bpm, this now
represents 10 minutes exactly ! now you see our 'binary' minute in use.

The next stage is to 'decode' or take these 2 lines /Q8 & /Q10 over to
IC5 a 4082  2 package QUAD AND GATE.  The 4082 is known as an 'AND'
gate, meaning that for an output +12v high to appear on the output of
this IC at Pin13 we have to have all 4 inputs (Pin 12,11,10,9) at +12v,
ok meaning that for the output to go high we have to have Pin 12 AND
Pin 11 AND Pin 10 AND Pin 9 all high.  These Pins are held at "fake"
ground by 100k resistors, these resistors basically tell the IC to
ignore any stray voltages 'floating' around'.   By having the 100k
resistors to ground, the real count outputs from the 4040 and the
"don't care" floating voltages on the pins do not accidently "trigger"
the inputs to the 4082.  "fake ground" is the easiest way to describe a
100k resistor in the circuit to 0v, it means that it will reuire a
'real' +12v from the 4040 to the inputs of the 4082 to action anything.

The procedure is often used on the other side of the coin, to put a
100k to +12v means for the circuit to ignore any random "low's", these
are referred to as "pull up" or "pull down" resistors, safety only.  In
practice this is known as essential and good house keeping, and well
advised policy, these procedures will eliminate most spurious problems.

If the 4040 now has the /Q10 & /Q8 high, representing 8m+2m=10mins then
these lines go into the 4082 for the next stage. /Q10 & /Q8 are share
bridged at XX and YY respectively, so as /Q8 (2mins) comes from the
4040, share bridged at YY and joins on the 4082 at Pin 11 AND 12. (AB).
/Q10 (8mins) does similar to XX & joins the 4082 at Pin 9 AND 10 (CD).
If the 4082 has all 4 inputs high, then we have a high output on Pin 13

Now you may ask why did this techo. use a 2 x 4 input AND gate instead
of a 2 input AND gate, you are correct. An engineer would do this , but
a technician would know from past experience, an idiotic salesman would
sell the item saying that it is flexible and you can vary the timings.
OK, leave aside the 555 variance, by 'bridging' the inputs of the 4082,
this means we can CUT the inputs at XX,YY and allow 4 separate outputs
from the 4040 over into the 4082 and by following simple /Q1 to /Q12
logic we can go from Q1 bridged 4 places up to
/Q12=A, /Q11=B, /Q10=C, /Q9=D, thus your simple timer now can go out to
2024+1024+512+256 binary seconds = 3840 blips and if you are running at
64bpm this represents exactly 60 minutes in real time = 1HOUR you could
use this circuit as a complete repeater shutdown or a host of other
operations, now you see why it was chosen, (it is also a garden variety
IC in the Dick Smith Catalogue, handy ???).
This 4082 can 'see' the 4040 as permantly connected for the 10min mode,
but with a little bit of "track cutting" or you can 'hard wire' connect
anything from the 4040 to the tops of the ABCD, thus allowing the 100k
resistors to remain always connected, but do not forget to cut the
existing /Q10 & /Q8 lines from the 4040 coming over, or chaos.

Right, now the 4082 has gone high, +12v up through R22 a 3k3 resistor
into the base of Q1 (any npn silicon transistor here), we, through
exhaustive testing and the fact the Dick Smith catalogue fell on the
floor at the right page have chosen a BC547 (but any BC107, anything
will do) is ready to conduct.  A popular 2N2222 will also work here.

Relay 1, is the relay which provides the closing of the contacts for
the ISD1416 chip to operate, putting a real 0v ground negative on Pin
24, which uses a "pull up" 100k resistor to +12v to stop spurious
operation. When the relay contacts close and putting a solid Ov onto
this line, thus emulating the /PLAY E Push Button (Yellow/White) PB2.

So by chain of events the transistor Q1 conducts, thus allowing the +12
through the coil down through the collector and down to earth via the
emmitter, the coil operates whilst the base is momentarily held high.

Note a Test Push Button is in the area 22M, this item bridges out the
transistor Q1, and emulates the conductance of such, hence the Relay
will turn on as long as the push button is depressed, it will also
latch up Relay 2 until it is extinguished by the eom marker on the ISD
or turning off the +12v power supply.  At this point we 'borrow' the
operation of this 10 minute timer to also operate Relay 2.
Relay 2 has the +12v waiting through the coil, through the diode D7 and
down to the collector of Q1. so as soon as Q1 conducts it allows both
relays to operate.  The diodes D2 and D6 are normal back emf protection
extinguishing diodes, very essential unless you want 80v+ running
around like a bull in a china shop, (Flemmings right or left hand rule)

The diodes are essential for IC and transistor protection, no argument
of this point.   In relay 2 we have on the board side, a set of
changeover contacts for the purpose of 'latching' up the relay, and
leaving the other external side for 'other operations' .  A sneaky
thing is done here so we can get a Green idiot LED to indicate Relay 2
'NOT ON. I take the +12 from the main supply, through the coil, and
because transistor Q3 is NOT conducting and Q1 is not conducting then
is has nowhere to go, so because the N/C contacts of relay 2 have
nothing else to do, I take this point over to the base of Q4 (via a 3k3
res) to provide bias to turn the GREEN LED On. Wastefull, yes, but it
gives you a clear indication of the mode at any moment, bear in mind
this item maybe some 50+ miles away and your local techo is telling you
the state of play for fault finding, obviously when the relay operates
you loose this circuit and the RED LED is across the coil of the relay
giving you actual conditions.

Let us assume that the opto isolator, a 4N25/28/35 etc. is in the 'off
state'. With NO conduction on the output, it will present NO voltage to
base of Transistor Q2 (another npn BC547) so the collector being fed by
a 3k3 from +12v will be 'alive' with nowhere to go, thus it will go
around to the base of transistor Q3 (npn BC547), and having a 'hot'
base, Q3 will conduct. So with Q3 conducting, as soon as the relay
operates it will continue conducting from +12v > the coil of relay 2
(blocked by diode D7) and to the collector of Q3, down to Ov ground of
Q3 emmitter.  The gain of this circuit would not 'pull the relay in -
only 'hold' , therein is another safety element.
This relay will stay up until something open circuits this chain.

One of the unique characteristics of the ISD1416 is when it has reached
the 'END OF MESSAGE' (eom) Pin 26 is briefly short circuited to Ov.
With this in mind RECORD LED blinks on briefly because it has +5v
present, through R2 (a 1k) then through the LED1, then onto Pin 26 of
IC1. What we are going to do is to 'parasite' this conduction and allow
an opto isolator (LED>Photo transistor package) to conduct for the same
brief moment.    WARNING THIS IS ALL +5V : NO +12v operation here or
$20 is wasted. Simply, I wish to separate the +5 circuitry from the
+12v and the best way was an opto-isolator.
Now the eom conducts to Ov, thus the opto conducts, so R29 goes high
with +12v, it goes to the base of Transistor Q2, it turns ON, and by
turning ON it allows the +12v from R30 to be dumped to ground, thus not
allowing any more base voltage/current to Q3 and so Q3 turns OFF.
By turning Q3 OFF it open circuits the negative holding the Relay 2 ON
and the relay drops out, so end of message = you loose relay 2.
This combination of Q2, Q3 is a classic Stage 1 Radio Trades,
transistor inverter (pre 7404 or 4001 era).
This allows the relay to stay up during the message for promotional
reasons or to allow for the ISD chip to take OVER what's going to the
repeater for audio transmission. This same relay could drive another
relay to video switch in a 3 second station ID if required every 10mins
This feature can be used for a host of other functions, this relay
represents the actual length of the message with COMPLETE Isolation
from the ISD1416 voice integrated circuit.

RESETS: We have 2, the first one is when the +12 DC power is turned ON
and we want to clear all the count stages in the 4040, this is done by
putting a +12v on Pin 11 of the 4040, the circuitry which does this is
C21 and R15.  They have selected values so as C21 fills with a
controlled 'bottom' R15 it will take about 1/4 to 1/2 second to charge
up, during this time the + is impressed onto Pin 11 of the 4040 via D4.

The other 4040 reset is when you have reached 10 minutes (etc) or 640
blips) then the output splits from Pin 13 with one track going to R22
but the other going to a larger resistor R21, which goes to the 'other
side' of the 4082 package. Same rules, as all inputs were at 'fake'
ground , as soon as the input goes high, the AND gate then outputs Pin
1 (of the 4082) to high and this in turn will also hit Pin 11 of the
4040 to reset.
Now the juggling act occurs, we want the 4082 to pulse Transistor Q1
long enough to 'operate' the relay, but short enough to reset the 4040
without scrambling the entire /Q register and causing a premature
dropout of count. This is accomplished by C22, When it is hit with the
initial 4082 o/p  'I am ready' pulse, it 'charges' up the capacitor and
allows a constant +12v to present itself through R21, thus keeping a
slightly 'delayed' +12v on the inputs to the 'other' side 4082, and
because it being slightly delayed in getting the 'go' pulse because of
the charge time required, then the discharge time will also be
effected, thus after about 1/4 to 1/2 sec a very solid +12v will now
leave Pin 1 of the 4082, go via D5 and up to the 4040 Pin reset. Both
diodes D4 and D5 are required to 'r/s' the 4040 but not interfere or
back feed with each other.  When it comes to fault finding later on in
life, always remember the axiom, "a circuit is inversely reliable to
the number of electrolytics used", these days capacitors are far better
quality, but in the case of any fault, check your electro's ...

The Amplifier section. We have chosen a LM386 monitor operational
amplifier, simple, cheap, effective and it works from other well proven
projects. Bernie Orbell went over my values and has used this IC to
oblivion with great success, so usual care in inserting the IC, whilst
this is not a sensitive cmos device, mechanical care is of most
importance.  The supporting components are R11, C10, C14, C15, C16, C17
and of course an 'el cheapo' 8 ohm car speaker whatever, connected to
the monitor output points.  Turn the power on and put your finger on
the trimpot RV3, listen for the 50/60Hz mains hum at the speaker, if
the hum is NOT present, turn the trim pot fully up and try again or use
a screwdriver as an aerial to listen to rf scrunch.  As soon as you
hear the hum assume it works ok for the next stage, so return the trim
pot to the 12 o'clock position and turn the power OFF.  You will
notice, unlike most magazine projects, all the trim pots are connected
correctly with max sweep being clockwise.

The VOICE RECORD/PLAY Section is possibly the most technical on the
board and the least spoken about because what's inside the ISD1416
really does not matter, as we cannot tailor or do surgery then the
insides are of little consequence, suffice to say it is a 16 second
endless loop cassette, reasonably thick tracks are feeding this IC with
a 470uf at the area 2D to support minuscule surge currents at around
80Hz. (according to Bob Parker a mate of mine.)  The AGC according to
the good book says a 470k and a 4.7uF are needed to ground, but BP with
many years of experience with this device says the book is wrong, use a
1meg and 1uF, now who do you believe ISD engineers who designed it ? or
a real world technical useage from Bob Parker, We will use 1meg/1uF.
After ensuring all the voltages are correct on the IC socket, ie. no
evidence of +12v ANYWHERE, and checking the +5v power supply is ok, now
turn the power OFF, CAREFULLY insert the cmos ISD1416, remembering the
rules/guidelines on handling cmos regarding friction/static/grounding.
The ISD chip is in and if all the voltages were correct, then when the
power gets turned ON, no smoke escapes from the IC, one can assume that
all is well.

This board has been designed with a fool proof (where have I heard that
before) safety device installed. Look carefully around schematic area
on the map 3F and area 11K. What we have here is a safety 2 pin setup.
The 2 pins on each area are shorted out underneath, it is time to
cut/scratch the join and make sure you have soldered in the 2 jumper
pin devices, these jumper pin settings are the same as those on the old
hard drive DS lines and nowadays master/slave settings. This board has
2 sets as mentioned earlier, put the shorting pin across the 'REC'
SAFETY at area 3F, thus allowing the RED Recording Push Button to work.
By being about 6" or 150mm away from the board, having rehearsed your
15 second speech hold the RED Button down for the length of your talk.
You will notice that the RED Record LED (LED1 area 3K) illuminates to
indicate 'on air', now to hear your master piece, simply press the PLAY
(yellow/white PLAY E Push Button) , this plays the message entirely.

A hint here is to allow about 1 or 2 seconds pregnant pause before your
speech, thus bringing to the attention of the listener that 'some thing
is coming', anyhow whatever turns you on. REC record and play your talk
This should boom out through the monitor amplifier IC2. The volume of
this device has NO effect on the final 'injected' volume to the
repeater. The output comes from the IC1 through to RV2, a 10k trim pot.
This has a screened cable going over the board and up to Relay 2 N/O
(normally open) contacts, your Normally Closed contacts have the
continuous pilot tone or general chatter or nothing connected to it so
as when the relay is in the power OFF position the audio goes through
the relay and out from the common to your tranmitter audio modulator,
our ISD chip when in the chatting mode has already 'tripped' the relay
1 and relay 2 in so our speech comes out of RV2 and up to the relay and
again out on the common, and behold, the announcement occurs etc.
So RV2 effects the repeater announcement volume and RV3 is the monitor.

Now you have your special announcement in the ISD, you can do the
following, use a tape recorder from the 'line out' download a safe
working copy, or reload at the 'line in' section, (Line in info at the
bottom of this article) there is a little mod to do.  And lastly remove
the SAFETY PIN Jumper and put it back up near the 64bpm clock, it makes
no sense going out to the repeater site with the new whiz bang repeater
ID, and then accidently touching the red record button. If that LED3 is
not blinking then assume the safety 2 pin jumper is in the wrong spot.

Summary, this board could have been designed smaller, big deal, Dave
VK2GIO and Mark VK2XOF must be wondering when the next bus is coming
with all the people to be let out with this amount of space available.
Bear in mind this board follows all the military design rules, All
diodes point north, all capacitors point north with the negatives -
pointing south, and of course all the IC's point West>East , this does
cost space but the benefits on assembly and fault finding from poor
soldered joints are immense, and this is done this way so the earliest
of amateur entrants can build this project with a minimum of questions
being asked of me.  The components have been laid out in 'sections' so
as fault finding can be reduced to a minimum of decision between faulty
components or excess/minimum solder.  All the LED's on the board are
totally useless in the actual operation of the unit, BUT, they sure cut
down on a lot of searching and testing of the relevant sections.
The 12+ positive regulator can be bypassed if you can guarantee the 12v
supply coming in, this project is designed to sit on our main 15v p/s
looking after the 'power bricks' up the mast, and so the 15v is pushing
the friendship for reliability, hence the inclusion of this regulator.
If you are going to use a 12v Regulator then locate the pc track
between D1 and D2 (area 21 O), known on the overlay as 'KK' CUT this
please and this now puts the regulator 'in circuit'. With this TO220
7812 regulator soldered in correctly, this will ensure correct voltage
to the general purpose IC's and relays.  Very cheap insurance indeed.

* Line in modification. You will have to choose if you want to load the
repeater ID from a microphone or off a tape recorder. If you are going
to use a tape recorder line procedure, then look at the area (5 F) and
locate resistor R4. You will notice another hole, move the southern end
of the resistor to this hole and re-solder, so as to put the 'Line in'
pot RV1 in circuit. An auto loader was developed off an IBM computer
for this purpose many years ago for other applications using your sound
card, *.wav files etc..., this is the correct way, but what the hell
this is a repeater voice ID, not the Boston Pops Orchestra.

That is all for now, if I can answer any FAQ's then I will add them
here, thankyou. Bob Barnes VJ2R   send an email or +61 2 9587-3491

This project is meant as a simple effective 10 minute repeater timer,
an intelligent PIC16C84 version will follow later on in the year.

end of file