Note: Telephone numbers exclude international and national dialling codes
Asia:
Australia: 264-270 Normanby Road, South Melbourne, Victoria 3205, Australia
081 560 3171
802 Houston Centre, 63 Mody Road, Tsimshatsui East, Kowloon, Hong Kong
Tel: 368 5161
Tel: 3 646 4522 Fax: 3 646 6727
Fax: 369 4890
iii
ii
&!
i!j
'
i!
~
Canada: 2430 Lucknow Drive, Unit 15, Missisauga, Ontario L5S 1V3 Canada
France: 26 Villa Des Fleurs, 92400 Courbevoie, France
Germany: PO Box 4449, 3300 Braunschweig, Germany
Italy: Via delle Gardenie 33 (Pontia Vecchia Km 33,400), 00040 Pomezia-Roma, Italy Tel: 6 919 7123 Fax: 6 919 7136
USA: PO Box 9004, 18111 South Santa Fe Avenue, Rancho Dominguez, CA 90221 USA Tel: 213 637 7500 Fax: 213 632 5519
Whilst every effort has been made to ensure the completeness and accuracy of the information in this
handbook, no liability can be accepted for any errors or omissions.
Strand Lighting Ltd. reserves the right to amend the specifications of the product at any time.
This document is issued on condition that the information contained herein is not copied, reprinted, or
otherwise duplicated by a third party wholly or in part without the consent in writing of Strand Lighting
Limited. Certain features of the equipment described in this document may form the subject of patents or
patent applications.
The purpose of this document is to describe the hardware of the MX console
range in detail.
SCOPE
This document, together with the associated circuit diagrams, should impart
enough information about the hardware for service and test. A part of this
document is the test specification for the
The reader is assumed to be familiar with lighting concepts and the terms such as
DMX512 etc. Information may be found in the references.
MX range of products.
OVERVIEW
In order to make the circuit design manageable, the hard~are can be split down
into functional blocks.
~
9
'3
-~
-~
;3
~
~
·~
-~
4
Section B describes these functional blocks and describes how these blocks are
related. The inputs and outputs of each block are described, with references to
related documents for further detail. Any relevant assumptions and calculations
are contained here.
Section C describes the signals that pass between the various functional blocks
and what controls each, its reset [power-up] state and any special information
relating to the signal. This is primarily for software reference.
Section D details the memory map.
Section E is the test specification for the MX product range. This covers what is
required of the in-circuit tests and functional tests which are strongly related to
the Built-In-Self-Test
[7] EIA standard - RS-232-C, August 1969, June 1981
[8] EIA standard- EIA-232-D, January 1987
[9] MIDI standard specification 1.0, Version 4.1, January 1989
[10] IEC 65 -General Construction
[11] IEC 335 -Household and similar appliances
[12] BS 800 -RF characteristics
[13] 80C196KC user manual - Intel 270704-001 May 1989
[14] 80C196KC data sheet - Intel 270741-002 January 1990
[15] 8279 data sheet - Intel 290123-002 September 1987
[16] Toshiba8-bitMicroprocessorDataBook, September 1988-TMP82C79P-2
[17] Mullard technical handbook, Book 4, Part 9a - SCN 2692, 18 August 1986.
[18] Philips Components - AN174, Book ICU, December 1988
[19] Philips Components - ANl 76, Book ICl 1, December 1988
[20] Philips Components - NE570/571 Product Specification,
The general design principles are discussed in the System Architecture document
(ref. [2]). This description splits the circuitry up into the processor PCB which
contains all of the processing electronics, power supply, interfaces and
connectors, and the fader PCB, of which there may be up to 4 in a system.
PROCESSOR PCB
The processor PCB is split into functional blocks according to the schematic
drawings. These are each given individual sections, and at the end of each section
and there is a description of the (BIST) test facilities offered by each functional
block.
Processor
General
I
I
2.1.2
The processor is an 80C196KC. This is the latest version of the 80C196, an
upgrade from the 80C196KB. The processor runs with a 16MHz maximum
crystal and a very fast external bus. This leads to the requirement for use of AC
series high speed CMOS in most areas of the processor logic (refs. [13], [14] &
[15]).
The processor includes all the MCS-96 I/O, but with the extra features of 2 extra
PWM outputs and the Peripheral Transaction
response. The two extra PWM outputs are used to drive the LCD contrast and
backlight intensity.
Seiver which allows faster interrupt
1/0 Facilities
In order to reduce costs, all on-chip I/O facilities are used in this implementation.
The input and output ports are used to control specific I/O devices . The HSO is
used to generate timing synchronisation signals for the dimmer analogue
multiplex signals and a test signal for the audio processing circuit. The HSI is
used for time-stamped inputs and other interrupts. The
reading the fader values and for testing parts of the circuit
ND converter is used for
Issue 6
-81 -
Page 10
FUNCTIONAL BLOCK DESCRIPTION
1X33213
2.1.3
2.1.3.1
Memory
General
The 80C 196KC bus is extremely fast, therefore in order to run without wait states,
the memory has to have a very fast access time. The INST pin is used to expand
the total address space available to the application code. Two 64K x 8 EPROMS
(27C512) are used and these are configured to be word wide to get the best
performance from the processor. Timings are dependent on the memory decode
delays, and these are set by the decode devices (see below) . To run without wait
states, the EPROMS need to have lOOns max. access.
For the 12/24 way versions, 1 wait state is inserted, as these cannot stand the cost
penalty of lOOns devices. In order to meet the timing requirements with lO0ns
parts, a pair of tri-state buffers, IC's 30 & 33, were added to reduce the output
enable delay
The analysis provided in the System Architecture Document (Ref. [2]) shows that
32k x 8 SRAM is required as a minimum. The design allows for a pair of these
devices to be used so that expansion is possible without laying a new PCB.
Because of the high speed bus, it was not possible to economically design 0 wait
state RAM into the system. Therefore 16 bit RAM with 1 wait state was
implemented, with 2 devices fitted from the outset. Critical elements of the code
use internal registers to speed operation . When the protected RAM area
( controlled by the GAL programming) is write-protected,
states.
time.
it is READ with 0 wait
2.1.3.2
1/0, including the memory card, is run in 8 bit mode. The processor's
All
buswidth is dynamically changed by the memory decode GAL's using the
buswidth pin.
Memory Map Decoding
In order to provide maximum flexibility and cost effectiveness, a pair of
GAL20V8-15 devices (IC28, 31) are used to decode the memory map.
The devices generate the necessary wait states for the READY pin. See the
section on memory map for exact memory mapping. The processor provides two
output port pins to allow
of wait states. On the 48 way version lOOns ROM's are fitted and the processor
switches these into 0 wait state by setting the ROM WAIT pin low. It defaults to
high - 1 wait state - at power up.
An extra 3 to 8 line decoder, IC 32, is used to generate the peripheral enables to a
high resolution so as not to waste address space. All peripherals run with 3 wait
states. 1 wait state adds 1 CLKOUT cycle to a given memory access.
it to page the code part of ROM, and control the number
-82-Issue 6
Page 11
1X33213
FUNCTIONAL BLOCK DESCRIPTION
2.1.4
I
Memory Card Interface
The memory card interlace is derived from a Mitsubishi application note
MDS-MEMCARD4l7-9-88-250. The circuit allows the card to be inserted
while the system is powered up, and also allows the power to be switched on and
off while the card is inserted. All lines to the card are buffered by tri-state devices,
these being controlled by the card detection signal (/CD) and /RESET gated
together. This ensures that:
a)
b)
The /CD signal is schmitt triggered with an RC input network to ensure reliable
control.
The power to the card is also controlled by a transistor VT 13, enabled by the
above signal.
The signals are not enabled until the card is in place (because of the /CD
contact and switch-on delay).
During
MX power up/down sequences the lines are tri-state .
Data is controlled by a bidirectional tri-state buffer, IC38, a 74HC245. Its
direction is controlled by the
gated with a decoded /CE_CARD signal.
I
The full address range of the card is catered for by a page register IC41, a
74AC573 latch. This has a tri-state output controlled in the same way as the
above data and address buffers. Before reading or writing to the card, this page
register is loaded with a high order 8 bit address by a uniquely decoded memory
address. This solution provides 255 2kbyte windows onto the card giving a
0.5Mbyte possible max. addressing range allowing the use of larger cards in the
future.
The status signals from the card are all fed to the DUART's input port for reading
by the processor. Mitsubishi cards give 2 pieces of information that other cards do
not generally give. These are:
The WP output. This tells the system, when high, that the card's write-protect
switch is on and the card cannot be written to. The card protects itself, even if
writes are attempted.
MX uses 64Kbyte cards initially.
/RD signal and it is enabled by the above signal
Issue 6
The BO, B 1, B2 outputs. These tell the system what type of card is connected. The
codes are as follows:
-83-
Page 12
FUNCTIONAL BLOCK DESCRIPTION
1X33213
BO
RAM 0
OTP ROM 0
Mask ROM 1
In the MX system, these are connected to IP4, IP5, IP6 respectively.
2.1.5
Test Facilities
The memory card interface incorporates the facility to allow the card battery
voltage to be read. The processor reads this via a multiplexer that is powered from
V _RAM (See 2.2.5). The card voltage should be read in the A/D's 10 bit mode
and is O.K. if above 2.60V. (Mitsubishi figure).
2.2 PSU
2.2.1 5V Switching Circuit
B1
0
1
0
B2
0
0
0
The heart of the power supply is the switching regulator, the STM L4963 (IClO).
This chip fully implements a buck converter and includes all the necessary power
handling circuits. It is ideal for MX because it incorporates power fail and reset
circuitry also, with a minimum of components.
2.2.1.1 Input
The circuit runs off a 15V nominal supply, this may be 23Voffload and its worst
case load must be>= 15V. The input is filtered with a 2200µF cap. and contains a
1.35A polyswitch and l.5kW/5W transzorb with a maximum clamping voltage
of approx 43V. This provides over/reverse voltage protection. The polyswitch
will hold LOA continuously at max. operating temperature (35°C)
temperature rise of 15C. A twin hole bead, Ll6, is wound with two windings in
phase opposition to attenuate mains-borne asymmetric RFI.
2.2.1.2
Power Fail
The power fail element of the L4963 (pin 7) has a threshold of 3.6V falling and
4.4 V
to raise the thresholds to 1 OV and 12- l 2.5V respectively. This is a low impedance
divider to swamp the internal divider to reduce tolerance errors. The low going
threshold is set to allow 1 V before the L4963 falls out of regulation. In practice
+ internal
rising~ A precision external voltage divider is placed around pin 7 of the chip
-84-
Issue 6
Page 13
1X33213
FUNCTIONAL BLOCK DESCRIPTION
this was measured as 8V, but the spec. gives a worst case 9V. This 1 V fall requires
a 2000µF capacitor minimum to provide the System Architecture's requirement
of a 2ms warning period, assuming a constant current load. The 2200µF capacitor
provides enough reserve. The lOV level provides for an allowable supply droop
of approx. 30%, more than is required by the design spec. because it was thought
at design review that the board should work with a 240V transformer on a 220V
I
supply and should therefore work to 198V.
Power fail is also connected to the enable of the decoder driving the source
transistors for the LED's. This disables the LED's at the instant of power fail and
therefore extends the available warning period to approx. 4ms.
2.2.2
1
+9 I -9V Regulator
MX requires both positive and negative rails for D54, RS232 and analogue signal
buffers. The ideal voltage is around 8.5V to give RS232 signals of sufficient
I
amplitude, without making the D54 driver dissipate too much power. At least
8.5V is needed on the positive rail to allow for voltage drops in the D54 driver
which must be able to output 6V (nominal +20%).
The solution was to wind a pair of secondaries on the main inductor, creating a
I
pair of flyback converters to generate a nominal +/-14V. This is then regulated
down to + and -9V by a pair of 7809 regulators. The stacking configuration of
regulators is used as it was possible to provide a pair of isolated secondaries
without extra expense over a split secondary, and this lowers the component
count.
14V is required to maintain regulation with the worst case input of lOV and
lowest load current. The current needed is approx 40mA on the negative rail and
90mA on the positive rail (the asymmetry is caused by D54 signals). The devices
incorporate thermal shut-down protection and over current foldback protection.
78 series regulators are used as power dissipation could be a problem in the worst
case, particularly on the positive rail.
2.2.3
2.2.3.1
2.2.3.2
Issue 6
Battery Back-up
Back-Up Power Source
The CMOS RAM in MX is supported on a 3.6V Ni-Cad cell.
Back-Up Circuit
The back-up battery is switched into the
diode D18, dropping approx. 0.6V.
-85-
V _RAM supply by a simple silicon
V _RAM must be at least 3V so that the
Page 14
I
FUNCTIONAL BLOCK DESCRIPTION
memory card, ifleft connected, is not discharged by the AID multiplexing circuit.
The battery is charged at a current of approx . 3mA. This current is above the
normal trickle charge as the operation of the product is somewhere between
standby and cycle. V _RAM is connected to the 5V rail by VTl , the circuit
ensuring that V _RAM is always at 5V during normal use, giving maximum noise
immunity.
A prototype PCB yielded a l.6µA back-up current at 25C.
2.2.4
Memory Protection
As part of the Architecture, MX incorporates comprehensive memory corruption
protection . The RAM chip select (active low) is driven from a 74HC00, IC21,
which is powered from
will operate down to 2V. /RESET is gated by the HCOO so as to guarantee that
upon reset the memory chip select becomes inactive high. The L4963 ensures that
RESET falls before the 5V rail goes out of tolerance.
V _RAM and only consumes approx. 1 µA on standby and
1X33213
I
I
I
I
I
I
I
]
I
~I
:-i
I
I
I
I
2.2.5
Test Facilities
MX is equipped with a test connector to allow voltage measurements to be taken
at one place. All power rails and internally generated references are available on
this connector, as are the two processor test inputs. The test inputs are deliberately
not buffered (or debounced) so as to retain the minimum amount of hardware
between the processor and outside world for the PCB functional tests. The power
supply may be turned on and off by use of the
turns off the L4963 and hence the other rails, allowing remote supply cycling and
making a memory retention test easier to perform.
The power supply circuit also permits the processor to sample the relevant signals
to perform its internal diagnostic tests. There is a particular problem with this
approach where the battery voltages have to
be powered from the V _RAM supply so that V _RAM does not attempt to power
the rest of the board via the multiplexer protection diodes. The worst case standby
current is 5µA for the 4051 and this is an acceptable drain . One gate of IC21 is
used to disable the multiplexer in reset.
Unfortunately, the processor
could result in a 0.3V ( 10%)
fed back to the multiplexer via 470R as well as a 100k resistor. This allows the
processor to check its input leakage current by comparing the two voltages. The
voltage read on CAL_ VOLTS can
leakage (within the power supply reference tolerance of +/-2% ). Having read the
AID has a worst case DC leakage of 3µA and this
drop at the AID. Hence the battery back-up voltage is
be assumed to be accurate regardless of
/INH signal. When held low, this
be sampled. The multiplexer has to
-86-
Issue 6
Page 15
:r
1X33213
FUNCTIONAL BLOCK DESCRIPTION
value of BATI_ VOLTS and taken the difference, CARD_ VOLTS can be
. measured, and since the input resistors are equal in each case, the difference
directly added to arrive at the absolute card voltage.
This method saves a buffer IC and associated components, and since these
voltages are read infrequently, has no performance implication. Since the card
voltage and back-up voltage are of similar magnitude, the method is accurate.
The difference should be measured each time the card voltage is read, as
temperature fluctuations
will be significant as far as input leakage is concerned.
2.3
2.3.1
I
Comms
DUART
The Duart used in MX is the S CN2692. This is functionally similar to the familiar
2681 chip, but is fabricated in CMOS. It provides the majority of the I/O
associated with the various comms links, analogue and digital. The two
asynchronous link requirements are:
1 MIDI-full duplex, fixed baud rate: 31250, 1
2 RS232 -full duplex, variable baud rate up to 19200, 1
MIDI was allocated to channel A, RS232 to B. Unfortunately the two baud rate
range(s) are not achievable with one clock source because of incompatible
division factors. The solution is to supply 4MHz divided from the CLKOUT of
the processor (8MHz) to the internal timer/counter of the chip. The timer/counter
can be set to give exactly 31250 baud. The RS232 links are supplied by the
DUART's own crystal oscillator at the obscure frequency of 3.6864MHz. The
Duart provides its own fixed dividers from this frequency to generate all the
required standard baud rates accurately up to 38400 baud.
% tolerance
% tolerance
r
2.3.2
2.3.3
Issue 6
Reset
In common with the 8279, the 2692 requires a positive RESET signal. This could
have been derived from the /RESET signal, but the signal would be very short ( 16
state times
Therefore a processor port pin generates the signal PERIPHERAL_RESET. The
port pin automatically goes high on reset, software must pull it low to operate the
peripherals.
= 2µs) and the 8279 does not quote a minimum acceptable time.
D/A
The D54 and AMX92 dimmer drives require a D/A converter. The DAC08 was
chosen. It does not have a bus interface, so is connected to the 2692 output port.
The details of this connection are discussed in the section
-B7-
on pin allocations.
Page 16
FUNCTIONAL BLOCK DESCRIPTION
1X33213
The DAC08 provides a Vlc pin to set the operating threshold of the logic inputs to
make them compatible with many different logic families. In this
implementation, the pin is used as an enable, /D _A_ENABLE. The device
guarantees that the logic threshold levels will be 1.4V typically and 1 V worst case
above Vic. Hence by connecting /D_A_ENABLE to HSO.O, the processor has
effective control of the D/ A, with /D _A_ENABLE high the D/ A output is forced
to 0.
The DAC08 is a current based converter and therefore requires a reference
current. In order to provide the user with the facility of being able to trim the
analogue output voltage from the LCD ( to compensate for line drops and dimmer
matching) and implementing the philosophy of removing
system, the reference current is processor derived. The Design Specification
requires a +/- 20% variation. This is achieved by integrating one of the three
pulse-width-modulation (PWM) outputs (PWM_O) over a sufficient time so as
to produce a DC voltage with less than 0.5 LSB ripple. The converter provides a
multiplying function between this reference and the digital input byte, therefore
controlling the effective maximum output current with the PWM value.
The PWM frequency is set to the highest possible with this aim in mind (31250Hz
@ 16MHz), and is integrated using a 3k3 resistor and lOµF cap. The time constant
of the reference network is therefore 17ms since the voltage is converted to
current by another 3k3 resistor (to keep analysis simple) and the DAC08
reference input is a virtual earth. Using the data sheet values for the Yoh of the
PWM output, it will provide approx. 4.8V at the reference current. In order to
provide the+ 20% headroom, the standard PWM output will be 200 decimal. This
is divided by two by the reference smoothing network to give 1.88V, giving a
standard reference current of 0.57mA, peak about 0. 73mA. The recommended
range is 0.2 to 2mA, and keeping on the lower edge keeps power supply current
low and reference smoothing capacitor small. The positive input of the reference
amplifier is connected to analogue OV; no resistor is used to compensate for the
worst case 3µA bias current, as this is only approx. 1 LSB.
all pre set pots. from the
:JI
I
]
Cl
]
~i
·-1
:JI
J
:J
I
:JI
I
:J
]
I
""]
I
:i
J
I
~I
~
I
J
I
I
!
~!
I
2.3.4
The output current is converted to output voltage by the output amplifier.
Analogue Synchronisation Signals
Different synchronisation pulses are needed to generate D54 and AMX192
dimmer control signals. The electrical requirements are different, but the timing
of the relevant control signals is very similar. Fig. 1 shows D54 and AMX sample
waveforms and the control lines that generate them. The HSO's were seen to be
particularly suited to generating these control signals as they must be precisely
timed pulses.
HSO .0 and HSO .1 were chosen to drive ID _A_ENABLE and D54_SYNC
respectively, as they can be driven together in one HSO instruction. This allows
the processor to set up the DIA and generate the D54_SYNC at the same time,
without having to return to service the end of the sync. signal. It also gives a fast
transition between the sync. and the analogue level. The sync. is generated in
form of a current defined by Voh of the HSO output-0.6 (diode drop) and the
referenceresistor(s) . A pairof3k3 resistors were used, resulting (when the DIA is
disabled) in a output sync. pulse of approx. -5.5V, allowing plenty of headroom
for any D54 receiver. The arrangement algebraically adds the
(negative) and the reference current (positive) into a virtual earth- the hardware
does not have to combine sync. and D/ A output with analogue gates.
HSO.2 is used to generate AMX_SYNC and is gated with the TX output of the
processor. This signal drives the RS485 transceiver, IC8. The two signals are
NANDED, so to use TX, HSO.2 must be high, to use HSO.2, TX must be idle.
1X33213
the
DJ A output current
2.3.5
Analogue Drive
This section of the COMMS circuit consists of 1/4 of a TL084 op-amp coupled
with an LM6321 buffer amp. The TL084 provides an almost ideal op-amp for the
low current output of the DI A, and is arranged as a
converter . Enclosed in its feedback loop is the LM6321 buffer (IC7). This device
provides +/- 300mA drive and is stable into very heavy capacitive loads. It is
purpose designed for driving long lines. It incorporates current limiting and
thermal shut-down protection . Voltage protection is provided by a bidirectional
12V transzorb, D 17, with its capacitance providing half of the LC RFI rejection
network. The D54 output requires a source impedance of 4R 7, the AMX output,
lOOR. The same amplifier is used for both as the protocols are supported on a
mutually exclusive basis.
The standard output current has to be converted to +5V by the resistor in the
feedback loop of the amplifier, hence this resistor is:
5/.057
An intelligent feedback system is used to set the PWM value, and this
compensates for differences in processor Voh values and de-sensitises the
requirement for the accuracy of this resistor . Even so, 2% components are used in
the whole subsystem so as to limit the possible tolerance build up. An 8k2 resistor
is used.
= 8.77k
virtual earth current to voltage
2.3.6
RS 485 (DMX/SMX)
The RS485 output provides the several mutually exclusive functions on MX :
-B10-
Issue 6
Page 19
1X33213
FUNCTIONAL BLOCK DESCRIPTION
2.3.7
1.
2.
3.
The IC used is a standard 75176. The receive function is permanently enabled as
the receive section of the processor UART may be disabled internally. Protection
is provided against transients with transzorbs across the driven lines to ground.
These, together with inductors, provide RFI protection.
To drive the SMX protocol. This is achieved with the 80C196 UART.
To drive the DMX512 protocol. This is achieved with the 80C 196 UART.
To drive the AMXl 92 sync. pulses. The HSO.2 output of the processor is
used for this function. lOOR resistors are used in series with the outputs as
the AMX spec. requires.
RS 232/RS 485
The RS232 port is implemented with a 75155 single channel transceiver. This is
powered from the
This is sufficient to maintain the
specification . Protection against interference and voltage transients is again
provided by transzorbs and inductors.
+/- 9V rails which produces a signal range of approx. +/- 7V.
+/- 3V swings required by the RS232
2.3.8
It was decided (in the Design Specification) that modem control lines would not
be implemented, as this would only serve to complicate the software, increase the
cost of the hardware and give the user no real added benefit. However, as
as a DCE, the DCE's required output modem control lines (pins 5,6 and 8) are
pulled high (active). Sometimes receiving equipment locks out if these lines are
not valid, this prevents that situation occurring.
This port also has an RS 485 transceiver (7 517 6) to enable direct use of SMX. Its
transmit and receive functions are separately enabled by /ENABLE_RS485_TX
and/ENABLE_RS485_RX, allowing comprehensive loopback tests. The gating
arrangement around IC 35, 21, 23 toggles the RXDB line between the RS232
transceiver and RS485 transceiver. The application software can decide on the
receive source and can
transceiver and associated components are not fitted for
tum the line around during SMX comms. The RS485
/B (standard) build.
MX acts
MIDI
The MIDI interface is implemented using variants of the circuits detailed in the
MIDI specification. The optocoupler chosen is the GE Hl lLl, as it provides the
required sensitivity (l.6mA), is specified to work up to 1 Mbaud, and
incorporates a schmitt trigger output. Using this, rather than a darlington output
device, results in better rise/fall time control, and
factor that governs the durability of the interface because the signal becomes
tum on/off time ratio. It is this
Issue 6
-811-
/
Page 20
FUNCTIONAL BLOCK DESCRIPTION
more asymmetric, the more MIDI devices it passes through . The absolute skew is
irrelevant, but the asymmetry is not. The HllLl specification states:
Ton
This would only result in 2% asymmetry.
The actual figures achieved depend on the matching of the drive current to the
LED. The input resistor has been increased above the MIDI spec. to 820R, to
reduce saturation, giving a good performance at 2.SmA. In the actual application
(SmA worst case), there is still approx. 100% overdrive, this being maintained to
allow for LED degradation over time and component tolerance. The figures
achieved were:
= 900ns, Toff = 2.Sµs, giving a 5% asymmetry
Ton
The through and output signals are driven by a 74AC14, IC3, capable of
supplying 24mA, and a lower level of voltage protection is supplied by BAS 16
clamp diodes. RFI protection is provided by the series inductors. The MIDI
interface was seen to pose less of an interference and transient hazard as it only
involves connection of short cables in temporary installations .
1X33213
= 0.65µs , Toff = l.2µs
2.3.9
2.3.9.1
2.3.9.2
Test Facilities
In common with the other elements of the circuit, the COMMS circuits
incorporate extensive test facilities.
DUART
The DU ART has the internal capability to allow most of its registers to be read
back by the processor. This should give adequate indication of DUART
operation .
DIA, Analogue Drive and Synch. Signals
There is no direct test on the
little difficult to test! The output of the analogue driver is fed back to an analogue
multiplexer for sampling by the processor
DI A and driver block as a single entity. The input is protected by a 4k7 resistor
the
and diode clamps. The 10 bit resolution of the
to be performed .
DI A converter itself, as being current based, it is a
A/D. This allows a functional test of
AID could allow monotonicity tests
The D54 synchronisation signals can be tested by splitting the control of the
HSO.0 and HSO.1 signals. By enabling the
-812-
DIA (/D_A_ENABLE = low) and
Issue 6
Page 21
1X33213
FUNCTIONAL BLOCK DESCRIPTION
setting the sync. signal (D54_SYNC) high, it is possibie to perform an additive
function between the D/ A and the sync. current,
check
The
this
and the
limited by the power supply reference tolerance of +/-2%.
t.lie result.
DI A reference signal, D _A_REF, is also·fcd back to a!! analogue multiplexer,
allows the processor to check that both the PWM and integrator are working,
absoiute value of the reference for setup purposes. The accuracy will be
usingJhe read-back AID input to
2.3.9.3
2.3.9.4
2.3.9.5
RS 485 (DMX/SMX)
As indicated in the RS485 section, both RX and TX are permanently enabled, this
allows the processor to read the actual RS485 data that
use of loopback connectors.
RS 232/RS 485
. The RS232 drivers require loop back connectors to be · used to check their
operation. The DUART has to
reception. This
/ENABLE_RS485::._TX and /ENABLE_RS485_RX. To perform a test on this
port, both the functions
receiver disables .the RS 232 receiver.
arranged so that the line is tri-state during
MIDI
The MIDI interface drivers are tested in the same way as the RS232 drivers.
loop backs being used so that the processor can transmit and receive its
port also has an RS485 transceiver and this has 2 control lines
need to .be enabled. The act of enabling the RS 485
transm.it data out of the B port and check it on
Th~ /ENABLE_RS485_TX 8ignal is
&11d after reset.
it i3 transmitting without
cwn data.
2.4
2.4.1
Issue 6
Audio
I
The Audio input to M:X provides comprehensive automatic level control to assist
the user. This
requirements for a 60dB dynamic range; this is achieved using a
compandor in ALC configuration .
~frcuit is arranged so_ that it meets the Design Specification's
ffillips NE571
·
Input
The front end of the circuit consists of a differential amplifier based ·on a 1L084
with a gain of approximately .0.5 and 6dB/octave roll-off above 20kHz. The input
is AC coupled in both signai and
ground lines and has RFI suppression using
-813-
Page 22
FUNCTIONAL BLOCK DESCRIPTION
appropriate LC networks. This gives the input immunity to noise, DC and 50Hz
common-mode signals. The left and right channels are algebraically added in this
configuration.
The ALC circuit generates a standard Od.Bm signal (0.775 RMS) and the
minimum accepted signal is-43dB from this level ( 43dB is approx. the max. gain
available). This represents a signal of 5.5mV RMS. The maximum level seen by
the compandor will be approx. 11 V peak-peak (3.8V RMS), determined by the
limit of the TL084 output. The gain of the input circuit, has to be set so as to allow
approx. 1
The gain of the input circuit is set to 0.5 so that the compandor just about sees
sufficient level at 10m
possible level of+ 14dBm.
Orn V to 1 0V RMS to be accepted.
V RMS input and works linearly up to the maximum
1X33213
2.4.2
2.4.3
Filter
The filter circuit follows the input circuit and is placed before the ALC circuit so
that mid and high frequency signals do not affect the gain control. This is a simple,
unity gain, maximally flat, 12dB/octave low-pass filter, rolling off above 200Hz.
ALC Circuit
The ALC circuit is built around an NE571 compandor. The design is based on
application notes ANl 74 and ANl 76 for an ALC, with suitable adjustments. The
ALC is constructed around one half of the chip. The adjustments that were made
were:
1. Rdc set to 15k to set the output to +3.6V so that the DC operating level is
satisfactory for an 9V supply.
2. Crect set to l0µF to give a lOOms time constant. The attack time will be
faster than the decay time according to fig. 9 of ANl 7 4 (as required by the
Design Specification). With this configuration lOOms was found to
provide a good subjective response.
The other half of the chip provides an ideal output rectifier from the ALC, as the
precision rectifier element can be accessed independently of the other elements.
The signal is AC coupled into the rectifier, and in order to provide the Design
Specification's required response, a diode is included in the output. This subtracts
a fixed voltage from the (smoothing) capacitor signal, but since this signal is
approximately constant (ALC action), it can be compensated for with extra
output gain - this is not Hi- Fi! The diode does have the added advantage of
giving a small amount of noise immunity. The rise time of the signal on the lµF
-814-
Issue 6
Page 23
1X33213
FUNCTIONAL BLOCK DESCRIPTION
smoothing capacitor is lOmS, determined by the chip's output impedance, its fall
time is set to 50ms by a 47K resistor.
2.4.4
r
Output
The maximum output of the ALC circuit is in the region of (2V - 1 diode drop), ie
approx 1.5V. In order to provide a 5V signal for a schmitt trigger, the signal is
amplified with a gain of approx. 3.5 using 1/4 of the
onto the Audio/MIDI 10k pot.
configuration is used to:
1. Allow the use of only 1 type of pot in the whole product.
2.
3. Allow the processor to check the position of the pot during Audio Test
The configuration is arranged so that MIDI takes precedence. It was not thought
worthwhile providing simultaneous Audio and MIDI control.
A schmitt trigger is used to provide interrupt edges to the processor HSI.2 pin.
This allows software separation of audio level and effects stepping.
Allow MIDI timing signals to be controlled (in software).
procedures.
RV8, together with a MIDI_ENABLE signal. This
11..084. The signal is dioded
2.4.5
2.5
Test Facilities
One of the processor's HSO outputs is divided and AC coupled into the
non-inverting side of the differential amp input stage. This provides a test input to
the audio processing subsystem when the HSO generates suitable frequency
squarewaves. The frequencies are programmed to test the roll-off of the filter
circuit and check for adequate gain.
disconnected during this test.
amp. will behave as a unity gain buffer in this configuration. The resulting signal
will be approx 6m V RMS at the ALC (in the pass band of the filter).
The divide ratio is set to 400: 1 as the differential
Clearly any audio source must be
LED Drive/Keyboard Control
The keys and leds on the control PCB are controlled by an 82C79
keyboard/display driver IC20. This is a clocked synchronous logic block that
continuously scans the keyboard and refreshes the LEDS. It is used in default
display mode (16 character, 8 bit display) and in encoded mode, with the
keyboard in sensor mode.
Issue 6
-815-
Page 24
FUNCTIONAL BLOCK DESCRIPTION
The clock rate is set to 80KHz by an internal software controlled prescaler, which
divides down the 2MHz basic clock provided by division of the processor
CLKOUT.
1X33213
2.5.1
2.5.2
LEDS
The 82C79 addresses a 16 x 8 array arranged as 16 sources decoded from four
scan lines, and eight sinks. In this implementation, the first 8 sources are actually
on the fader PCB 's, the last 8 are on the control PCB (VT2-9). All sinks are
provided by ICl 7. On the control PCB, IC14 decodes the last eight states of scan
lines SL0-3, IC13 decodes the first eight for the fader PCB 's. Inter-character
blanking is provided by the chip (/BD goes low), and this disables all drives
(IC13, 14), so removing the possibility of crosstalk between columns.
Because the multiplex drive is nominally 16:1, which is in reality approx. 20:1
when the inter-character blanking is taken into account, the LED's have to be
driven with a reasonably high current. This is set by the 4R7 resistors attached to
ICl 7, andICl 7's saturation voltage and the saturation voltage ofVT2-9 . Approx.
lO0mA was chosen giving an average of approx. 5mA. The LED's chosen are
moderately high efficiency and are intensity matched for the 3 colours.
Keyboard
3
The scan lines SL0-3 are also decoded by IC15 to provide the drives for the
keyboard array. Each row is selected individually as the scan counter increments,
and depending on the key pressed, a code is read into the sensor memory of the
chip. Diodes are provided on each switch to prevent interactions if multiple keys
are pressed. The scan process is continuous and results in eight bytes (of which
only 5 are used) in the 82C79 containing the image of the keys pressed. An active
key appears as a Obit. This method is chosen, as opposed to the de bounced FIFO
mode, so that the software can detect how long a key is held on, allowing
perform the auto-repeat functionality required in the spec. Software also
debounces the keys by reading the sensor image array at 5mS intervals.
it to
FADER PCB
The fader PCB is designed so that it can be used in any of the four required
positions in the system with just one switch change. It is also designed so that all
four PCB 's can be daisy-chained on a single piece of ribbon cable from the
processor PCB. This cable carries a six bit address allowing 24 faders and 12
switches to be addressed, the LED matrix signals, and power supplies. A separate
supply is provided for the LED matrix to reduce noise problems.
-B16-
Issue 6
Page 25
1X33213
FUNCTIONAL BLOCK DESCRIPTION
The ANL_ GND connections are interleaved with the analogue lines out
(ANL_[0 :3]) to minimise cross-talk. The remaining signals are then graded in
order of least to most hostile away from the analogue lines in the ribbon cable.
]
J
[1
]
1
3.1
3.2
Address Decoding
The PCB address is set by the 2 pole/4 way switch. This switch uses 1 pole to
connect the relevant analogue multiplex signal to the analogue bus and generates
3 control bits with the other pole. This 3 bit address is fed into a pair of 4051 's used
as selectors to connect 2 of the 8 column drive signals CA_0 to CA_ 7 to the drive
transistors for the LED's.
There is no per-PCB addressing of the analogue sections of the fader PCB 's, as
there is a dedicated analogue line for each PCB.
Analogue Multiplexing
The analogue multiplexers are supplied with a parallel 6 bit address (via R7-12 to
protect against ESD damage) . The lower 3 bits of this address are supplied in
parallel to each of the 5 4051 analogue multiplexers. A further 4051 multiplexer
talces the 3 upper address bits and multiplexes the 5 multiplexed signals onto one
CM OS buffer, a CA5160. This is used because it can output 0V without a negative
supply and can output 5V on the +9V analogue supply. A buffer was necessary
because of the intrinsic high impedance of the signal, its increased impedance
having gone through 2 multiplexers and the inevitable noise down the ribbon
cable. N.B. The LED switching noise is not a problem because of the architectural
precautions
talcen (Ref. [2]).
ii
•
i
3.3
3.4
Issue 6
Flash Switches
The flash switches are multiplexed in the same way as the faders and are read as
faders. Because the 4 analogue lines are fed direct to the processor port 0, the
80Cl 96 can read the switches in a digital manner. This is convenient because the
4 switches from the 4 possible fader PCB 's are read simultaneously by reading
port 0. Debounce is done in software.
LED Drive
The addressing effectively selects 2 sequential columns of the 16 total addressed
by the 82C79 for each fader PCB using ICs 2,6. The source drivers are on this
PCB as it was easier to select the address than to select the high current drives. The
sink drivers are the ones on the control PCB. The LEDS are arranged in 2 rows of
6.
-817-
Page 26
FUNCTIONAL BLOCK DESCRIPTION
The peak pulse current, being set by the row sink drives, is the same as that on the
control PCB. It is set at approx lOOmA, as the mark/space ratio will be approx
20:1 as generated by the 8279, giving an average of approx 5mA. The LED's
chosen are moderately high efficiency and are intensity matched for the 3 colours.
3.5 Test Facilities
There is no specific input on the processor to tell the software which model of
board
software will check the 3 test locations of each PCB at power-up. It should find a
sequence of 5V (-10% ), 2.5V ( +/-10%) and 0V ( + 10% ). The tolerances allow for
resistor tolerances and
a fader PCB
multiplexers.
it is controlling. This is done by the test facilities of the fader PCB. The
AID off sets. This unambiguously indicates the presence of
- hence product identity - and the operation of the addressing and
The Hardware Architecture of MX results in a large number of programmable I/O
pins, these being found on the 80C196 and the DU ART, the 2692. The philosophy
has been to use these wherever possible to eliminate
memory decoding. The sections below indicate the uses of these pins.
1/0 port chips and associated
Processor
1/0 Drive
The processor has a large number of 1/0 pins some of which are configurable to
perform special I/O functions, some of which are general purpose
which are solely inputs or solely outputs . The Intel manuals are do not group all
this information in one place so it is difficult to determine which pin to use in some
cases, especially as there are three variations of electrical properties :
Quasi-bidirectional - 1/0
will pull low to:
0.45V at 2.8mA
1.5Vat7mA
1/0, some of
1.1.2
will pull high to:
as inputs - leakage
Input only (port 0)
leakage = +/- 3µA
Output only
will pull low to: 0.45V at 2.8mA
will pull high to:
V cc-0. 7 at 30µA
Vcc-1.5 at 60µA
= +/- lOµA
l.5Vat 7mA
Vcc-0 .7 at 3.2mA
Vcc-1.5 at 7mA
Allocation
The choice of I/O pins was made by first defining the use of the special pins, using
them in the most effective way for
MX software and the processor's 1/0
Issue 6
-C1 -
Page 29
SIGNAL DESCRIPTIONS
capabilities, then defining the input only and output only pins for uses that could
only ever require inputs or outputs, then defining the rest with general purpose 1/0
pins.
In MX all the 1/0 devices of the processor are used to their full extent thereby
realising the power of the architecture . This fixes the use of:
UART - RX, TX pins
HSO - 4 HSO pins
HSI - 4 HSI pins
AID - All of port 0
PWM - P2.5, 1.3, 1.4
NMI
EXTINT - P2.2
T2CAP-P2 .7
1X33213
=i
•
I
-
I
•
I
•
I
•
I
•
I
•
I
•
·I
•
I
•
I
•
I
•
I
•
I
•
I
•
I
•
TEST INPUTS - 2 input only pins, must be direct to the processor
Two inputs can be either HSI and HSO pins. In this case 1 of each mode is
selected. HSI inputs are particularly useful because they allow a events to be time
stamped, so reducing interrupt latency problems .
HSO.0 and HSO .l were chosen for D54 sync. generation as they can be
programmed to act together for easy software drive under normal conditions, but
can be driven separately for test purposes. The section on COMMS details their
use.
The allocation is detailed below, with the reset state shown for each pin.
is driven by external hardware, the state is shown in brackets:
POWER_FAIL 1 T2CAP - processor has less than 2ms to live
/ENABLE_D_A
AMX_SYNC 0
/AUDIO_TRIG
X(O) RS485 receive
X(O) DUART int signal (+ve edge)
X(1) Test mode input 1
0
0
0
0
0
X(1)
X(O) + edge on 8279 key entry
X(1) - edge on music beat
RS485 transmit
PWM_O - D/A reference voltage
Low enables DIA converter
High adds -5V to
High drives RS485 true (provided TX=high)
Squarewave adds signal to audio input
High drives test LED on
/BO of 8279, low when blanking
DIA output
NMI MEM_CORRUPT
Issue 6
X(O) + edge = attempted illegal memory write
-C3-
Page 31
SIGNAL DESCRIPTIONS
1.2 DUART
1.2.1 1/0 Drive
The 2692 Duart has 7 input only pins and 8 output only pins. The input pins are
CMOS gate equivalents, and IP0-3 have change of state detection capabilities.
The output pins are
control lines and in this mode are open-drain.
1X33213
CMOS gate equivalents also, but OP4-7 may act as modem
•
I
--
1.2.2
Allocation
It was decided that since this device was going to be used (System Architecture),
the output port was best employed to drive the D/ A converter. The output port
drives the D/A (DAC08) directly, although the actual outputs are complements of
the output register. Writing has to be performed in two steps, as there are two
locations - 1 for setting a bit, 1 for resetting it. Therefore the
ensure that this process is indivisible. This may be achieved by writing a word
to the register pair.
The input ports are used for the less important input functions as access to the
processor port pins is much better than to the DU ART pins ( this might require use
of the shared interrupt). One input, IP2, is used to get the clock signal into the
counter-timer for the MIDI channel. This provides with 4MHz derived from
division of the processor's 8MHz CLKOUT signal. The counter-timer must be
set to give the 3 l.25kbaud MIDI frequency.
Four of the input pins are used to detect the status of the memory card (See
Functional Description 2.1.4 ).
The memory map is shown in Fig. 1. The map is divided into code and data spaces
with 64kbytes in each. The map has changed from the original specification (Ref.
[2]) to give more data space, by compressing the peripheral address space to
lkbyte and the card window to 2kbytes. The code space has been expanded with
the bottom 32 kbytes being fixed and the top being paged into 2 equal 32kbyte
blocks.
Memory decoding is achieved with 2 GAL20V8-15 devices and a 74AC138
(IC32). The precise details of the GAL encoding is held in the PLD design file
S4006001.sds (Ref. [24]).
WAIT STATES AND BUSWIDTH
Wait states are controlled with the READY pin. The peripherals and memory card
are run in 8 bit mode (BUSWIDTH
The ROM is run in 16 bit mode, wait states are controlled by the ROM_ WAIT pin.
If ROM_ WAIT is high, (default at power up) 1 wait state is inserted. If the unit is a
48 channel one, then the software switches
completed. ROM then runs with no wait states. This requires lOOnS or faster
parts. 0 wait states are not used on the 12/24 for cost reasons, however, fast
EPROMS will work with 12/24 channel systems.
= 0), with 3 wait states.
ROM_ WAIT to O after BIST has been
J
-_.
•
.,
3
Issue 6
RAM is generally run with 1 wait state. However, in the protected area (see
below), it may be read with
scene and FX information . Advantage is taken of the fact that RAM timings are
faster for reads than writes.
O waits. Protected RAM contains frequently accessed
MEMORY PROTECTION
MX incorporates extensive memory protection. Attempts to write to ROM cause
a non-maskable interrupt (via MEM_CORRUPT) as do attempts to write to the
protected area of RAM or the memory card without first asserting
/ENABLE_RAM. This means that in the unlikely event of a software upset, the
most likely result will be an NMI. The user sees this as a fatal error 1417 "Invalid
The MX control PCB software incorporates extensive Built-In-Self Test (BIST)
functions and the general philosophy of these is outlined in the Design
Specification
These are resident in the standard product's software, which is the same across all
models.
It is required that the control PCB is given an in-circuit- test to detect obvious
manufacturing defects before attempting to run any of the BIST tests. It is
desirable for the fader PCB to be in-circuit-tested, but not essential.
Both the fader and control PCB 's are tested using the control PCB BIST software.
The resident tests may be run in a number of ways, depending on the required
procedure, each mode is optimised for that test procedure. The control PCB
incorporates 2 test inputs on the test connector (PL5)-these are called Ml and
M2. The four possible combinations of these pins give three test modes and the
normal user mode. These pins are connected directly to the processor, so require a
minimum of hardware to perform.
/RESET (available at the test connector) must be asserted briefly when
changing modes.
(IX 33154) and the System Architecture Document (lX 33180).
These pins are only sampled at RESET, so
!!I
~
!!
1.1
2
2.1
Additionally, the tests may be run from the menu system on the LCD, the modes
are split into SERVICE tests, which require the test jig to be attached, and USER
tests, which require no jig. The tests may be run cyclically, to help find spurious
failures, during soak test for example.
Definitions
DMM - Digital Multimeter.
OVERVIEW OF TEST FACILITIES
Test LED
The control PCB incorporates a test LED, which appears on the back panel of a
complete unit. The normal sequence of operation for this LED is:
Issue 6
-E1-
Page 37
-1
I
TEST SPECIFICATION
1. A brief flash ON after RESET to indicate that the LED is operational.
Failure should be investigated immediately as it means that either the
processor is not running, or the LED itself is not working.
2.
3. A short period ON (approx. 1/2 - 1 sees.) while the LCD is tested.
Having completed the LCD test, the LCD will indicate all further test
information.
A short period
is non-destructively tested.
is faulty.
LED goes off during this period, the LCD is faulty. Having passed the
LCD test, the LED should then remain on as long as the unit is powered
and not RESET. The system automatically resets on finding RAM or LCD
(fatal) errors and therefore keeps re-trying the test sequences.
1X33213
OFF ( approx. 1/2 - 1 sees.) while a small part of the RAM
If the LED does not re-illuminate, the RAM
If the
E: '
l
I
t_ !
t ~
t
2.2 Test Modes
The two test connector pins Ml and M2 provide the following tests:
1. Ml=0, M2=0-Mode 0
Birth:
This is the functional test for the complete control PCB before assembly with the
control panel. This mode destructively initialises all the data areas in the RAM
and does all possible tests in sequence. It is required that all elements of this test
are passed before the control PCB is passed to be assembled mechanically with
the control panel.
The test jig must be connected to run this test sequence correctly. A part of this jig
is also a Gold Brick set of components-fourfader PCB 's must be connected and a
memory card inserted in the card socket. Fig. 1 illustrates the required test jig.
NOTE: The memory card will be formatted during the test procedure, - it
must contain no valuable data!
When the test sequence completes, the board goes into normal operating mode.
2. Ml=0, M2=1 - Mode 1
Cycle:
In this mode, all tests that require no operator input are run cyclically. The control
PCB must have been born and requires the same test jig as Birth above. This mode
-E2-
Issue 6
Page 38
1X33213
TEST JIG
To Ref 1952 MIDI in
connector (SK2)
[
INTERCONNECTION DETAILS
DMM
Vol~0
+ -
TEST SPECIFICATION
RS 232 Loopback Connections
...----"2'"' To Ref 1952
.__ __ ___,. connector [SK3]
MIDI Loopback Connections
4 4
PL2
5 5 PL3
-------PLB connector [SK5)
PL~ 25 way '0'
:
3
:, To Ref 1952
J MIDI out
SW2
Mode Selection
Switch
Test Connector Details
1. ITEMS
SK1 - 14 way MOLEX socket [O.
Diodes are all type BAT 85
SW1 - 1 pole 6 way rotary switch
SW2 - 1 pole 4 way rotary switch
SW3 - SPST switch
ModeO
Mode 1
Mode2
o..-M2,de3
Closed = 'Reset'
Open = 'Run'
SW3
SPST
SW1 +9V
Voltage Selection J .----9"'-V'------"'--4
Switch
BAT85
BAT85
BAT85
BAT85
CONSTRUCTION DETAILS
Loopback Connector Details
1. PL 1 is 25 way '0' type connector [free)
11
2. PL2, 3 are 5 pin DIN connectors
l
~
+5V anal
/Reset
VRef
M1
M2
+15V
7
[free)
To Ref 1952
Test Connector
[PLS)
2. Mount all switches on suitable panel
3. Wire all components as per interconnection details
4. Check all connections with continuity tester
5. Label switches as shown
Figure 1 - Test Jig Arrangement
Issue 6
-E3-
Page 39
TEST SPECIFICATION
is designed to allow extensive testing in environmental chambers and may also be
run while cycling the power to the unit to check for memory corruptions etc.
3. Ml=l, M2=0-Mode 2
Fader Test:
In this mode, the unit will only test one or more connected fader PCB's. This
requires considerable user input as all switches and faders have to be exercised to
pass the test. This mode is provided for use on a gold brick system when just
testing fader PCB 's.
4. Ml=l, M2=1-Mode 3
Normal:
1X33213
In this mode, the unit will perform a power-up self test sequence which will be a
subset of the BIST functions, but will be invisible to the user unless a fault is
found. The fault will then be reported as in the tests above. A single message will
normally be displayed telling the user that self tests are in progress. The total time
for these tests will depend on whether it is a cold or warm start.
A cold start is performed when the board was last powered down deliberately
using the on/off switch. A fairly extensive set of tests are performed and then the
board is used as normal.
A warm start is performed when the board was last powered down by a mains
failure. A short set of tests are performed to check the integrity of the Scene, FX,
Patch and Setup memory and then the board is used as normal. Any running
effects or fades continue from their position at power failure.
2.3 Test Procedures
These are split into the three: those for the control PCB, those for the fader PCB,
and those for the finished product.
2.3.1
2.3.1.1
Control PCB
Equipment Needed
1. 4 off tested captive fader PCB's (REF 1953) with inter- module ribbon
cable.
-E4-
Issue 6
Page 40
1X33213
TEST SPECIFICATION
2.3.1.2
2.
3.
Procedure
1. After in-circuit-test which should detect opens/shorts, missing, wrong, or
If a 48 way system is used in NORMAL mode with slow EPROMs, the system
will crash after the
the EPROM wait states for a 48 way system at this point. A system crash may
corrupt the memory, so causing memory corruptions to be registered at the
next power up. A
Test jig (fig. 3):
DMM.
reversed components, the current version of MX EPROMS should
(low order) in socket for IC36, odd (high order) in IC39.
If it is known that the control PCB is to be part of a 48 way system, the
EPROM type should have lOOnS or less access time (part no 159/B435).
If the system is to be a 12 or 24 way, the EPROMs may have up to 200nS
access time (159/B412). Mark the PCB accordingly on the rear.
The software allows either type of EPROM set to be used for all
modes other than NORMAL.
CAUTION!
BIST sequence. This is because the software switches out
BIRTH sequence should be performed if this occurs
be installed, even
2. Install the battery link, LKl if not already installed.
3. Connect the test jig (Fig. 1) including transformer power supply.
4. Set the test mode to 0, set RESET to RESET.
5. Switch on the power using the PCB mounted switch.
6. Check the following voltages with respect to pin 14 of PL5:
Nominal value Min. Max.
+15V
+5.1V(nominal 5V} 4.95V 5.25V
+SV_ANL
+9V
-9V
VREF +5.1V
14V 24V
nominal 5V - 0.2 nominal5V
BV 10V
-10V
4.95V 5.25V
-av
'
I
Issue 6
-ES-
Page 41
TEST SPECIFICATION
7. Switch RESET to RUN
8. The BIST will now start to run. Firstly the test LED should light in the
cycle described above, then, as the tests are performed, a banner will
appear in the LCD to indicate which test is under way.
RETRY, if the test does not pass, turn off the unit and find the fault.
Re-do the test procedure until all tests are passed.
NOTE:
a) Some tests are so fast that the banner will not necessarily be seen.
b) Appendix 1 describes each test. The possible errors produced are
1X33213
If a test fails, first
Faults will be presented (as an ERROR) with accompanying
diagnostic information .
listed by number.
c) Appendix 2 describes each error, its diagnostic information and
an indication as to the possible cause of the fault, and suggested
corrective action.
d) The test software is not capable of testing the LCD backlight.
the display does not light up when the RESET is switched to RUN,
this fault should be investigated before proceeding.
9.
10.
11. Switch the unit back on and check that it performs a cold start -ie the
12. Repeat above but this time pull the power plug out, leaving the unit
Switch off the unit using the ON/OFF switch and disconnect the two last
fader PCB's-so making the system a 24 way.
Set the test mode back to
banner appears with the words
display also shows
does a warm start, this is indicative of a fault in the power fail circuitry
(R51-53 or IClO or the pf signal itself).
switched on. Re-power the unit and check that a warm start is performed.
If a cold start is performed, the /POWER_ ON signal is suspect . Switch the
unit off using the ON/OFF switch.
NORMAL.
'Running self test ... '. Check that the
'MX 24 Version ... ' If this does not occur, ie the unit
If
•
I
•
I
•
I
•
I
•
I
•
I
13. When all tests have been completed, and passed, remove the test jig, fit
the test passed label
EPROMS are fitted,
panel.
2.3.3 below.
Note: EPROM change may be required for final product - see
and indicate on the label whether fast or slow
and pass the PCB on for assembly with the control
-E6-Issue 6
Page 42
1X33213 TEST SPECIFICATION
14. There are no adjustments or other tests to perform.
2.3.2
2.3.2.1
2.3.2.2
Fader PCB
Equipment Needed
1. Captive MX control PCB (REF 1952) set to run fader test, Ml=l M2=0.
2. DMM.
Procedure To Test 1-4 Panels Simultaneously
This test procedure can be used for 1-4 fader panels. Panels not connected will
be included in the test cycles. Ideally four panels should be tested together, so
not
that all four positions for SWl 3 on all PCB 's can
1. Check boards for manufacturing defects.
2. Put all faders at zero, set SW13 on the test panels :
Panel one to position 1
Panel two to position 2
be tested.
i
I
Panel three to position 3
Panel four to position 4.
3. Connect fader panels to captive
4. Check the power rails with voltmeter w.r.t ANL_GND at PLl pin 33:
Nominal Value
+5V_ANL nominal 5V - 0.2
+9V
+5V_LAMPS
nominal 5V - 0.2 nominal 5V
min. max.
av 10V
5. Take control PCB out of reset. LEDs will now be in a short flash sequence:
1.
2.
All LEDs flash on - check for uniformity of brightness.
Single LED on, chasing from LEDl panel 1 to LED12 panel 4check for no shorts
MX and power up in RESET.
nominal 5V
check at IC9 pin 16
check at IC1 pin 7
check at emitter of VT2
Check
!
I
I
Issue 6 -E7-
Page 43
-1
I
TEST SPECIFICATION
6. LEDs will all go on for a short time, then off.
The LCD will now indicate the flash button to press, (from SWl panel 1 to
SW12 panel 4 ). On pressing each button, the respective channel LED will
indicate acceptance and the LCD will prompt for the next one. A stuck
key will be indicated by the respective LED being on and an error
message on the LCD. There is an opportunity to re-try the test at this
point.
After SW12 position 4 the next test will start.
7. Take each fader in turn (from VRl panel 1 to VR24panel 4) and set to full
and back to zero. The LEDs will light for any column with faders not at
zero. Check that each fader gives a smooth transition of percentage
values, with reaching 1 % or less at the bottom, 99% or greater at the top.
8. In order to check that the addressing works for on all PCB 's in all
positions, do the following three tests. Set SW13 on the test panels :
1X33213
I
I
I
I
I
I
I
I
:J
I
-i
Panel one to position 4
Panel two to position 1
Panel three to position 2
Panel four to position 3.
Set the board to RESET momentarily, and then to RUN.
Perform test 6 again.
9. Set SW13 on the test panels:
Panel one to position 3
Panel two to position 4
Panel three to position 1
Panel four to position 2.
Perform test 6 again.
Set the board to RESET momentarily, and then to RUN.
10. Set SW13 on the test panels:
Panel one to position 2
Panel two to position 3
Panel three to position 4
Panel four to position 1.
-ES-
Issue 6
Page 44
i
j
i
1X33213
TEST SPECIFICATION
~
~
~
~
~
~
~
i
~
~
2.3.3
Set the board to RESET momentarily, and then to RUN.
Perform test 6 again.
11.
If passed test fit tested label to all PCB 's.
Finished Product
1. Check that EPROM set is of the correct type. Use 159/B412 type
EPROMS for 12/24 way systems, 159/B435 for 48 channel.
2. Connect the power supply and switch the unit on.
3.
4. Connect an EC90 dimmer (set for auto MUX protocol mode) to the D54,
Check that the unit performs a cold start. If not
again. Verify that no faults appear during power up tests. Verify that all
LEDs illuminate on
DMX and AMX outputs in turn. Using the menu selection SETUP->
-> DIMMER, verify that each output works satisfactorily. Note that
EC90 software versions prior to Al do not support SMX.
all fader panels.
turn the unit off and on
10
2.3.4
5.
6.
7.
Using the Menu system, access the ERROR LOG. Check that there are no
entries in the log, if there are, investigate the cause.
When power up tests pass, leaving no trace in the error
CLEAR menu, and choose the WHOLE SYSTEM option,
by YES. This step ensures that the system has default settings for the customer.
Verify that a memory card fits into the unit and reads / writes correctly.
log, access the SETUP
and press CLEAR, followed
Service
MX provides access to most of the tests via the menu system. The operator is
provided with both USER and SERVICE type tests. The distinguishing
charactaristic is the fact that the jig (fig. 3) has to be used to get error-free results
on a good system when executing service tests. All USER tests may be run
without the jig. The tests are detailed in Appendix 1.
Removal of Small Surface Mount Components. . . . . . . . . . . . F20
Removal of Integrated Circuits Using Cutters . . . . . . . . . . . . . F21
Issue 6 -i-
Page 46
Appendices
1
1.1
APPENDIX 1 - MX SELF TESTS
Introduction
All tests are performed on the premise that if successful, they will return control to
the user without any message. It is therefore assumed that if the given test does not
produce an ERROR it is successful. There is no positive indication of "PASS".
All test failures are indicated by an ERROR in the format described in this
appendix. This appendix lists the tests with the possible ERROR numbers that
may be generated.
The following tests are performed on BIRTH, in the order listed:
A TOD REF TEST
FADER CONNECTIONS TEST
RAM TEST
ROM TEST
NMITEST
CHECK DATA AREAS
A TOD REF TEST
LCD CONTRAST TEST
LCD TEST
D TO A REF TEST
BATTERY TEST
DTOA TEST
D54 SYNC. TEST
RS232TEST
MIDI TEST
DMX/SMX TEST
AUDIO INPUT TEST
MEMORY CARD TEST
FADER MODULE TEST
CONTROL MODULE TEST
Issue 6
-F1 -
Page 47
APPENDICES
1X33213
The following tests are performed on COLD START, in the order listed:
A TOD REF TEST
FADER CONNECTIONS TEST
RAM TEST
ROM TEST
NMITEST
CHECK DATA AREAS
A TOD REF TEST
D TO A REF TEST
BATTERY TEST
DTOA TEST
D54 SYNC. TEST
RS232 TEST (internal only)
MIDI TEST (internal only)
DMX/SMX TEST
FADER MODULE TEST (LEDs and stuck keys only)
CONTROL MODULE TEST (LEDs and stuck keys only)
The following tests are performed on WARM START, in the order listed:
A TOD REF TEST
FADER CONNECTIONS TEST
CHECK DATA AREAS
The following tests are performed on FADER TEST, in the order listed:
A TOD REF TEST
FADER CONNECTIONS TEST
FADER MODULE TEST
<repeat>
The following tests are performed on CYCLE TEST, in the order listed:
FADER CONNECTIONS TEST
RAMTEST
ROM TEST
NMITEST
CHECK DATA AREAS
A TOD REF TEST
LCD CONTRAST TEST
LCD TEST
D TO A REF TEST
BATTERY TEST
DTOA TEST
D54 SYNC. TEST
RS232TEST
MIDI TEST
DMX/SMX TEST
AUDIO INPUT TEST
-F2-
Issue 6
Page 48
1X33213
MEMORY CARD TEST
FADER MODULE TEST
CONTROL MODULE TEST
<repeat>
However no parts of the tests requiring user input (faders, keys etc) are executed.
Tests do not report errors on screen, these may
1.2 Test Descriptions
ROM TEST
This test checks the data part of the EPROM set. Due to the 80Cl 96 architecture,
it is not possible to check the code part of the ROM. However, any EPROM
problems will probably
be highlighted by this test.
APPENDICES
be viewed in the error log.
Possible errors: 1390
RAM TEST
This test non-destructively tests the RAM, ie writes patterns to the RAM, testing
each bit in
tum.
Possible errors: 1389
NMITEST
This test checks that a Non-Maskable interrupt is seen when an attempt is made to
write to an area of protected memory. There are 3 areas - the data areas, the
memory card and the EPROM.
Possible errors: 1391
CHECK DATA AREAS
This test checks the integrity of the protected data areas, by verifying their
checksums.
Issue 6
Possible errors: 1384, 1385, 1386, 1387
FADER CONNECTIONS TEST (at every power-up)
This test checks that a valid set offadermodules are connected to the control PCB.
Possible errors: 1388
-F3-
Page 49
APPENDICES
1X33213
LCD TEST
This test invokes the LCD CONTRAST TEST. Having done this it puts all
segments of the LCD on for the user to check visually.
Possible errors: 1393
LCD CONTRAST TEST
This test checks that the LCD contrast can be controlled by the processor. It
changes the contrast between max. and min, checking the LCD reference voltage.
Possible errors: 1393
A TOD REF TEST (at every power up)
This test checks that the AID reference (set by two equal resistors R95, 96) is
reading 2.6V
AID (+I-10%).
Possible errors: 1392
BATTERY TEST
This test checks that the back-up battery is within tolerance. Also tests measuring
circuit, giving indication of possible
AID problems.
Possible errors: 1395, 1396
D TO A REF TEST
This test checks the DI A reference generator by setting it to a known value and
checking against the
AID.
Possible errors: 1394
DTOATEST
Having performed a test to ascertain full-scale on the D/ A, this test checks the
DIA by setting each bit in turn and checking it against the AID.
Possible errors: 1397
D54 SYNC. TEST
This test checks that the processor has full control of the DI A and associated
circuitry. It generates a voltage with the
can be correctly superimposed. It also checks that the
DIA and checks that the D54 sync. pulse
DI A can be enabled I
disabled.
Possible errors: 1398
-F4-
Issue 6
Page 50
1X33213
APPENDICES
RS232 TEST(internal)
This test checks that RS232 part of the DU ART (IC25) is working correctly by
sending / receiving a stream of data internally.
Possible errors: 1399
RS232 TEST ( external)
This test checks that RS232 part of the DU ART (IC25) is working correctly by
sending/ receiving a stream of data external! y using the interface I Cs. It therefore
requires a physical loopback on the 25 pin D type between pins 2 & 3
Possible errors: 1400
MIDI TEST (internal)
This test checks that MIDI part of the DU ART (IC25) is working correctly by
sending / receiving a stream of data internally.
Possible errors: 1401
MIDI TEST ( external)
This test checks that MIDI part of the DU ART (IC25) is working correctly by
sending/ receiving a stream of data externally using the interface I Cs. It therefore
requires a physical loopback between the MIDI IN and MIDI OUT
Possible errors: 1402
DMX/SMX TEST
This test checks that RS 485 port is working correctly by sending / receiving a
stream of data externally using the interface I Cs. No loopback is required, but it is
necessary to ensure that there is no significant load on the port.
Possible errors: 1403
AUDIO INPUT TEST
This is a complex test. It requires the AUDIO/MIDI fader to be set to full to
operate successfully. This is tested before the actual audio input is tested.
An
incorrect setting will give error 404. A series of tone bursts, at increasing
frequency and low level, are then generated by the processor. These are fed into
the audio processor, which should generate appropriate interrupts.
Issue 6
Possible errors: 404, 1405
-F5-
Page 51
APPENDICES 1X33213
MEMORY CA_RD TEST
BEWARE: This test requires a memory card and will FORMAT it, so
destroying any valuable data held.
This test does two things. First, it attempts to create a file on the card called
TEST.BIN (The card is MS-DOS formatted). Having created this file, it checks it
for integrity. In the second part of the test
pages of the card (64 kbyte) to check that the card page register is working
correctly.
Possible errors: 320,322,324,325,326,336,337,338, 1321, 1323, 1406, 1407
FADER MODULE TEST
This test has. a number of sub-tests. The first is a sequence through the flash
LEDs, followed by all on. Next, the LCD prompts the user to press each flash key
in
turn. The associated LED will light in each case. Finally, the LCD prompts the
user to cycle each fader in turn through its full travel. A full excursion is required
in each case. The percentage level is displayed on the LCD. The user should
check for acceptable performance. The faders must all be at
test
it writes a pattern of bytes into the 32
O before starting this
Possible errors: 1408, 1409
CONTROL MODULE TEST
This test has a number of sub-tests. The first is a 'build' through the control PCB
LEDs. Next, the LCD prompts the user to press each named key in
the LCD prompts the user to cycle each fader
in turn through its full travel. A full
turn. Finally,
excursion is required in each case. The percentage level is displayed on the LCD.
The user should check for acceptable performance. The faders must all be at their
0 positions before starting this test.
(Faders set away from front: TIME A, TIME B, B MASTER, D MASTER, FX
FADE TIME, FX STEP TIME. All others set nearest front)
Possible errors: 1410, 1411, 1412
-F6-
Issue 6
Page 52
1X33213
APPENDICES
2
2.1
APPENDIX 2 - ERROR MESSAGES
This Appendix describes all possible errors that MX can generate .
lntrodu ctio n
All ERROR 's are reported in a uniform way using MX's central error handler and
the LCD display.
The format is always:
-ERROR####Ident. string
Diagnostic info.
PROMPT
Each error is allocated a unique number and identification string as a brief
description. Depending on the type of error, some diagnostic information may be
presented . This varies in type depending on the error, and is described below.
The user is prompted with the soft key labels at the bottom of the screen, or if the
error is Fatal (ie the system cannot continue with such a serious fault), the user is
warned that the system is about to RESET itself with the string: "Resetting
system!"
(#### is the error number)
Error numbers appear in two types:
Those with numbers less than 1000 are generally of a less serious nature and are
not logged.
Errors with numbers greater than 1 OOO are regarded as serious and are logged in
the internal ERROR LOG. The log is for both user and service purposes, and
allows the last 100 errors to be viewed.
The errors are not consecutively numbered because each number comprises :
The software module where the error was generated.
The error number within that module.
Some errors, particularly those that are likely to be seen by the user (for example
when a memory card transfer is attempted with no card in place), are translated
into the 3 languages supported. Test and system type errors are only presented in
English.
There are 3 types of general diagnostic data:
Issue 6
-F7-
Page 53
APPENDICES
1X33213
1) Voltages:
These are presented as a result of internal tests in the form:
X.XX Y. YY Z.ZZ, where X is the voltage read, Y is the highest voltage and Z is
the lowest voltage allowed.
2) Addresses/Data:
These are presented in the form
3) Fader levels:
These are presented as
the fader in question.
Other forms of data may be used - these are described in individual cases.
AAA, where A is the percentage of full travel as read from
IIlIHH, where H is a hexadecimal number.
-
.-I
-
I
-
I
•
I
•
I
•
I
•
I
•
-I
•
I
•
2.2
Error Descriptions
ERROR
Identification:
Diagnostic Information:
Type:
Description:
Probable cause:
Remedy:
ERROR
Identification:
Diagnostic Information:
Type:
Description:
Probable cause:
Remedy:
12
"Record Inhibited!"
None
System/Warning
Record lock is activated, disallowing record/modify actions
Turn RECORD/LOCK
288
"MIDI framing error"
None
MIDI interface/Warning
A byte of MIDI
Faulty MIDI connections or MIDI cable
Find fault or shorten cable
337
"Card file closed"
Filename accessed
Memory Card/Warning
The software has detected a fault in the MX card file read
The card file
QUIT
has been edited in some way
card reading
338
"Card file missing"
Filename accessed
Memory Card/Warning
The card file requested was not found on the card
The card does not contain the requested
data
QUIT card reading
339
"Partial Transfer"
Filename accessed
Memory Card/Warning
The card file requested was only partially saved or read
There was a fault in the
data, and the transfer was QUITIED
Try again, use CONTINUE to finish reading card data
ERROR 404
Identification: "Audio fader too low"
Diagnostic Information: % level read, valid max / min % levels
Type:
Description: The audio fader is set too low for the audio input test
Probable cause: Fader set incorrectly
Remedy: Set audio fader to full (10)
ERROR
Identification: "No such handle!"
Diagnostic Information: None
Type:
! Description:
Probable cause: Program fault
Remedy:
data was received with incorrect parity
data SETUP was different to MX,
faulty connections present, or cable
Change EITHER MX OR remote device SETUP accordingly,
find bad connection or shorten cable
Test/Warning
1001
System/Fatal
The system has attempted
System reset
to access a non-existent "device".
too long
ERROR 1002
q
C
r Probable cause: Program fault
~
L
Identification: "No such read!"
Diagnostic Information:
Type:
Description: The system has attempted
Remedy:
ERROR 1003
Identification: "No such write!"
Diagnostic Information: None
Type:
Description:
Probable cause: Program fault
Remedy:
1009
"Duplicate string!"
None
System/Fatal
A duplicate string was found whilst initializing the hash table
EPROMfault
Replace EPROM set
1010
"No such open!"
None
System/Fatal
The system has attempted
to access a non-existent "device"
Not used
-F12-
Issue 6
Page 58
1X33213
APPENDICES
ERROR
Identification:
Diagnostic Information:
Type:
Description :
Probable cause:
Remedy:
ERROR
Identification:
Diagnostic Information:
Type:
Description:
Probable cause:
Remedy:
ERROR
Identification:
Diagnostic Information:
Type:
Description:
Probable cause:
Remedy:
1011
"Power fail stuck!"
None
System/Fatal
The power fail signal is stuck in its active state or power on signal
absent
Inter-PCB cable faulty, PSU chip failure or switch faulty.
Find and correct fault
1192
"LED None
System/Fatal
The LED controller
2MHz clock failure or failure of IC20
Check clock and IC20
1/0 fault"
is not responding to commands
1224
"LCD 1/0 fault"
None
System/Fatal
The LCD is not responding to commands.
LCD cable or LCD controller failure .
Find and correct fault
ERROR
Identification:
Diagnostic Information:
Type:
Description:
Probable cause:
Remedy:
ERROR
Identification:
Diagnostic Information:
Type:
Description:
Probable cause:
Remedy:
ERROR
Identification:
Diagnostic Information:
Type:
Description:
Probable cause:
Remedy:
1225
"LCD - invalid ansi"
None
System/Fatal
An attempt was made to write a non-ansi symbol to the LCD
Program fault
System reset
1226
"LCD - invalid pos"
None
System/Fatal
An attempt was made to write to an invalid position on the LCD
Not used
1321
"Card R/W error"
File written
Memory
The data written to the card was not verified correctly
Faulty memory card or card interface circuitry
Try again,
1323
"Card battery low"
File name accessed
Memory Card/Warning
The card battery voltage is too low for reliable operation
The card battery has fallen below 2.6V and is exhausted
Replace Memory card battery - see repair handbook for details
1384
"Scene corrupt!"
Scene in fonnat Pg.Scene
Test/Warning
A scene checksum has failed
Internal battery discharged, or mem. card battery exhausted,
or power fail detection circuitry faulty
Press CONTINUE
transfer, press CONTINUE
to clear the scene, or, if perfonning memory card
to carry on after ignoring scene or QUIT
to end the transfer.
1385
"Effect corrupt!"
Effect in fonnat Pg.Effect
Test/Warning
An Effect checksum has failed
Internal battery discharged, or mem.
card battery exhausted
or power fail detection circuitry faulty
As error 1384
Test/Warning
The setup checksum has failed
Internal battery discharged, or mem.
or power fail detection circuitry faulty
CONTINUE or
QUIT - cause setup to be reset to factory
defaults
1387
"Patch corrupt!"
Patch number {l or 2)
Test/Warning
A dimmer patch checksum has failed
Internal battery discharged, or mem.
or power fail detection circuitry faulty
"Fader module fault"
Module code
Test/Fatal
A fault has been detected in the fader module connections
Fader module cable fault or invalid id switch combination
Find fault and check switch settings
1389
"RAM test failed"
Address, byte
Test/Fatal
A RAM test pattern has read back incorrectly
Backup battery low or RAM faulty
Find fault or allow battery to charge
1390
"ROM checksum failed"
Checksum read, correct checksum
Test/Fatal
The ROM checksum does not agree with the programmed value
Voltage read, max., min. valid voltages
Test/Warning
The D/A converter reference is out of spec.
Inter-PCB cable faulty or D_A_SET signal not present
Find fault
1395
"Battery volts fault"
Voltage read, max., min. valid voltages
Test/Warning
The battery voltage is out of spec.
Battery discharged or faulty, Inter-PCB cable faulty
Find fault, or charge battery for several hours,
if error persists replace battery.
1396
"Battery ref fault"
Voltage difference read, max., min. valid voltages
Test/Warning
The battery reference voltage is out of spec.
R60 faulty, IC18 faulty,
Find fault
1397
"D
to A fault"
EITHER:Voltage read, max., min. valid voltages
OR:PWM setting, valid max., min settings
Test/Warning
D/A converter is out of spec.
IC6, IC4, IC7 or associated components faulty,
Inter-PCB cable faulty, or IC25 faulty.
Find fault
Identification:
Diagnostic Infonnation:
Type:
Description: The external loopback for the RS232 port
Probable cause:
Remedy:
ERROR
Identification:
Diagnostic lnfonnation:
Type: Test/Warning
Description: The internal loopback in IC25 for the MIDI port is faulty
Probable cause:
Remedy: Find fault or disconnect external MIDI device
1399
"RS232 int-loop fault"
read, byte written
Byte
The internal loopback in IC25 for the RS232 port
Memory decoding faulty (IC28, 32) or IC25 faulty, or data being
transmitted to MX during test
Find fault, or disconnect RS232 device
1400
"RS232 ext-loop fault"
Byte
read, byte written
Test/Warning
is faulty
No loop back connector installed?, ICS faulty
Fit loopback or find fault
1401
"MIDI int-loop fault"
Byte
read, byte written
4MHz clock failure, Memory decoding faulty (IC28, 32),
IC25 faulty or MIDI data being transmitted
is faulty
to MX during test
~
5
5
~
a
'
~
ERROR 1402
Identification: "MIDI ext-loop fault"
Diagnostic Infonnation: Byte
Type: Test/Warning
Description:
Probable cause: No loopback connector installed?, ICS faulty
Remedy: Fit loopback or find fault
ERROR 1403
Identification: "SMX/DMX port fault"
Diagnostic lnfonnation:
Type:
Description:
Probable cause: Short on RS485 lines, IC9, IC3 faulty
Remedy: Find fault
read, byte written
The external loopback for the MIDI port is faulty
Byte
read, byte written
Test/Warning
The DMX/SMX port loopback is not responding.
1405
"Audio test fault"
Frequency at which test failed
Test/Warning
The audio threshold detector has not seen signals at the
required test frequencies.
Wrong components around IC2, IC4, or audio source connected
during test
Re-try, find fault., or remove audio source
1406
"Memory
Byte read, byte written
Test/Warning
The memory card has not read back the test values written
Connector problem, card interface circuitry problem
Find fault, try another card
1407
"Mcard paging fault"
Page number accessed
Test/Warning
The memory card has not read back the test values written
Wrong size
Find fault, try another card
1408
"Flash key stuck"
Corresponding channel number
Test/Warning
A flash key is stuck in the "on" state
Key damaged, corresponding pull-up resistor disconnected, or
something resting on key during power-up sequence
Find fault
1409
"Channel fader stuck"
Corresponding channel number
Test/Warning
"Control key swck"
Key number
Test/Warning
A control module key is stuck in the "on" state
Key damaged, corresponding pull-up resistor disconnected, or
something resting on key during power-up sequence
Find fault
1411
"Control key bounced"
Key name
Test/Warning
A control module key has "bounced"
Key damaged
Find fault or replace key
1412
"Control fader stuck"
Fader number, fader expected,
Test/Warning
A control fader is set at the wrong level
Faders not reset prior to fader
Reset fader!
"Watchdog timed out!"
Module number
System/Fatal
A watchdog timer has expired.
Program fault or system overloaded in some way
System reset
1417
"Invalid interrupt!"
Interrupt number
System/Fatal
An unexpected interrupt was received.
Program fault, probably due
System reset
(HEX)
to noisy environment
Issue 6
-F19-
Page 65
APPENDICES
3
3.1
3.1.1
1X33213
APPENDIX 3
This Appendix describes the procedure for changing surface mount components
Introduction
On MX, since the Control PCB REF 1952 is being in-circuit- tested, there should
be no need to re-work components. However, just in case ...
The removal and replacement of surface mount components is a specialist skill
which takes some practice to acquire.
The following gives an outline of the correct procedure, which should be
practiced on an old/unused PCB.
Suitable Tools and Equipment
Surface Mount Cutters kit STC Components Stock No. 031676R
or
Surface Mount heat re-flow station - Circuit Plating Ltd System 1000
Reflow Station.
-
I
-
I
-
I
•
I
•
I
•
I
•
I
•
-I
•
I
•
I
•
I
•
I
•
I
•
I
•
I
•
I
•
I
3.1.2
Surface mount pick and place machine - Fritsch LM900.500
air jet
Hot
Tempilaq Temperature indicating fluid.
Removal of Small Surface Mount Components.
Apply heat to all contacts of the component using a Hot air jet. When solder is
molten remove component with tweezers.
Apply a small amount of Surface Mount Re-work flux and remove excess solder
with solder braid. Be careful not to apply too much heat or force so as to not lift or
damage any of the pads on the PCB.
Note: Most of the integrated circuits used on the card are sensitive to ESD
(Electrostatic discharge). ESD can permanently damage, or reduce the working
life of these components.
Therefore they should only be handled and fitted in an ESD controlled
environment. Earthed wrist straps, earthed soldering iron tips and storage of static
-F20-
Issue 6
Page 66
1X33213
APPENDICES
sensitive devices in conductive sleeves provides the minimum ESD controlled
environment required.
3.1.3
3.1.4
Removal of Integrated Circuits Using Cutters
Probably the easiest way to remove faulty integrated circuits is to snip the legs of
the IC and then remove the legs left on the pcb with a small tipped soldering iron.
Select a pair of cutters from the kit with small cutting edges that will fit around the
top of the IC leg. Snip the legs near the top, to avoid risk of damaging the PCB
pads or tracks. When all legs have been cut then use a soldering iron to remove the
remaining legs. This is best done by gently wiping the soldering iron tip across the
remaining legs. These will then tend to collect on the tip where they can be
removed .
Apply a small amount of Surface Mount Re-work flux and remove excess solder
with solder braid. Be careful not to apply too much heat or force so as to lift or
damage any of the pads on the PCB.
Removal of Integrated Circuits Using Heat Re-flow Station
Switch on Heat reflow station. Fit a component mask to the both top and bottom
heat jets appropriate to the shape of the IC being removed . This ensures heat
applied is confined to this area.
!
Measure/ Adjust temperature of reflow station to above the melting point of the
solder paste (about 250 degrees C).
Apply Tempilaq Temperature Indicator fluid ( degrees C) or similar to each edge
of the chip, (remember to shake the bottle well!)
Place PCB into pcb guides (adjust if necessary) and centre on component to be
removed.
Move both top and bottom hot air flows into position. They should be
immediately above and below component to be removed.
When the temperature indicator fluid has changed colour, quickly move top and
bottom jets back into the rest position and remove component with a pair of
tweezers.
Parts of the reflow station and the PCB will be very hot - BE CAREFUL!
Apply a small amount of Surface Mount Re-work flux and remove excess solder
with solder braid. Be careful not to apply too much heat or force so as to lift or
damage any of the pads on the PCB.
Issue 6
-F21-
Page 67
APPENDICES
1X33213
Replacement of Surface mount components.
Surface mount components may be replaced by using a reflow station or by using
a fine soldering iron and fine solder. Each method is described separately.
3.1.5
Replacing Components Using Heat Re-flow Station
Remove Solder paste from cold storage and store at room temperature for several
hours before work commences.
Set PICK and PLACE machine regulator pressure to about 35 PSI.
Adjust solder paste dispense time to approx 300ms (030 on the thumbwheel
switches).
Fit PCB into upper slots of the XY table, adjusting movable gate against PCB and
tighten the gate retaining screw.
Fit Solder paste onto XY head. Move solder dispenser into dispense position by
pressing black button on XY head.
Move the XY position control to centre the solder dispenser above component
pad. Gently pull down XY control until solder is automatically dispensed onto
component pad. Release XY control and move to next PAD, repeat until all
components pad have solder paste applied.
Note: Movement in either the X or Y direction can be locked by pressing either
the X or Y button to assist application of the paste.
3.1.6
Move XY control to position the vacuum pick-up nozzle above component to be
fitted. Gently pull down the XY control placing theXY nozzle on top of the centre
of the component. The machine will automatically use a vacuum to pick up the
component.
Note: It is a good idea at this stage to turn on the XY hold function. This will lock
movement of the component when it has been placed in position, to prevent the
operator dragging it through the solder paste when releasing the IC.
Move the component into position carefully lining up the component with the
PCB pads. Gently pull down on the XY control until the component is placed into
position. The machine will automatically release the vacuum.
Re-flow Soldering
Switch on Heat reflow station. Fit a component mask to the both top and bottom
heat jets appropriate to the shape of the IC being removed. This ensures heat
applied is confined to this area.
-F22-
Issue 6
Page 68
~
[~
1X33213
APPENDICES
t~
[!!
[~
3.1.7
Measure/ Adjust temperature of reflow station to above the melting point of the
solder paste (about 250 degrees C).
Apply Tempilaq Temperature Indicator fluid ( degrees C) or similar to each edge
of the chip, (remember to shake the bottle well!)
Place PCB into pcb guides (adjust if necessary) and centre on component to be
removed.
Move both top and bottom hot air flows into position. They should be
immediately above and below component to be removed.
When the temperature indicator fluid has changed colour, check that the solder
has flowed well by inspecting the soldered pins/pads.
and bottom jets back into the rest position and remove the PCB .
Parts of the reflow station and the PCB will be
If all is well then move top
very hot - BE CAREFUL!
Component Replacement Using Soldering Iron
Surface mount components may be fitted to the PCB using a fine tipped soldering
iron and fine solder. The following items are suitable for this method;
Weller EC2002D soldering iron fitted with fine tip.
Farnell Stock No. EC2002D
26 SWG solder eg RS 561-101 or Farnell SMART26 solder.
Carefully apply soldering iron to heat both the PCB pad and the component leg.
Apply solder to soldering iron tip,
component leg and PCB pad. Inspect each solder joint paying particular attention
to
dry solder joints and fine solder bridges. Rework any joints as necessary.