The ‘Lock’ LED, D4, is driven from a port of the FPGA. It is lit when the PT51 status indicates that
horizontal lock has been achieved and the input is a valid standard supported by the PT51.
Five DIP switches provide standard selection for the SM08 (see Chapter 2) and are read by the
FPGA microprocessor.
U8 is a proprietary copy-protection IC. U8 calculates a checksum from a PT13 generated data
stream, and the calculated checksum from U8 is compared with an FPGA internally generated
checksum. If the two do not match the SM08 module is shut down. This means that even if the bit
stream of the FPGA/EEPROM is captured the PT51 IP core will not run without U8 being fitted.
Sheet 6.
Sheet 6 is the analogue front end (AFE). The analogue video inputs, which may be aCVi, NTSC/PAL
or 960H formats, may be either coaxial or twisted pair (build option for the SM08).
If twisted pair, they are terminated in 100Ω (R26) and converted from differential to single ended
outputs by U9 and U10. If coaxial inputs they are treated as a pseudo-differential input, with the
ground screen of the BNC input connector connected to ground via R27 and C37, which affords
some hum rejection for long cable runs.
U13 provides a mid-rail (2.5V) bias voltage with low output impedance as the AFE runs from a
single supply voltage (5V).
U11 is a voltage controlled amplifier. The PT51 decoder provides automatic gain control and the
‘GAIN_CONTROL’ PWM output of the PT51 is used to control the gain of U11. The PWM output is
first clamped to the 2.5V reference voltage to avoid supply variations affecting the gain (D8). R42
and C52 provide low pass filtering of the PWM signal to convert it to an analogue voltage. This is
buffered and amplified by U13 to provide a 0-5V control voltage (although the voltage control
range of U11 is only 2.5V ± 650mV.
The output of U11 is AC coupled into U12A, which is a high input impedance, low input bias current
op-amp. The average DC level of video varies widely and to ensure it can be scaled into the
analogue to digital converter (ADC) properly it must be clamped. The AFE uses a sync tip clamp.
The most negative part of the AC coupled video is clamped to the VCLAMP reference voltage,
which is the most negative voltage reference voltage of the ADC. The ADC requires a 0.5V to 2.5V
(2V pk-pk) input signal and the sync tips (the most negative part of the video signal) are clamped,
using ‘ideal’ diode D7 and U12B, to the 0.5V ADC reference. The black level of the video (back
porch level) is measured and used to restore the video black level by the PT51 decoder.
Sheet 7.
U19 is the ADC, a 10-bit 80MHz component that is clocked at the rate shown in Table 2. The ADC
accepts a differential input, but is used in single-ended mode. The clamped video from U12A is
applied to the VIN+ input, and VIN- input is biased to mid rail (1.5V). U14 provides the ‘clean’ 3.0V
supply for the ADC. The output of the ADC, ADC[9:0] is the straight binary, digital composite video
which is applied directly to the PT51 decoder.
Sheet 8.
The aCVi interface allows bidirectional data to be sent between transmitter and receiver during
dedicated lines of the vertical blanking interval. Data from the transmitter to the receiver is
treated the same as video data, and the ADC ‘data’ line is decoded within the PT51.
Data to be sent from the SM08 to the transmitter is output from the PT51 and FPGA as the
DATA_OUT port. This binary data is converted into a current by U16 and R53 and injected into the
cable input (DATA_INS to AFE input, Sheet 6).
Sheet 9.
The Y, Cb and Cr video and the Hout, Vout and Fout synchronizing signals from the PT51 decoder
are formatted into an HD-SDI/SDI format inside the FPGA. The synchronizing signals are modified
to be compatible with the HD-SDI/SDI Flag requirements and the line count is derived for HD
outputs. The video and the timing signals (TRS) are combined and scrambled according to the HD-
SDI/SDI specifications and cyclic redundancy check (CRC) is added.