
Evaluation Kit User Manual
EV9000
2001 Consumer Microcircuits Limited
40
UM9000/2
This test measures the bit error rates of various message structures, depending upon which chip
type is selected. A normal receive acquisition bit and level sync. sequence is used for each
transferred message. Device B is used as the transmitter, device A as the receiver.
4.6.6.2 Description
Static BER and DQ test 'B' to 'A' using formatted messages. The number of messages is
determined by the 'BER Test bits' buttons. CMX909B uses TSB tasks followed by TDB to
transmit, SFS followed by RDB to receive. CMX969, FX919B and FX929B use T4S and T24S
tasks followed by THB, TIB, TLB or TIB + TLB to Tx, SFS followed by RHB or RILB to Rx.
AQLEV and AQBC/AQSC are automatically added to SFS, providing these buttons have been
selected. (Also see Appendix B for further details on the CMX969 test).
This test transmits messages containing pseudorandom data from channel B to channel A and
calculates the bit error rate based on a comparison of sent and received data. Various message
structures are each used to transfer the amount of data configured in the BER Test parameter
(see Figure 22). The pseudorandom sequence length is configured via the Rand seq parameter
(see Figure 23).
Receive PLL (clock) and level acquisition sequences are initiated with each message, providing
the AQSC and AQLEV buttons have been selected.
Where applicable, the following are reported for each message structure tested:
1.
The type of transmit command used.
2.
The number of messages transferred.
3.
The number of blocks in each message.
4.
The number of data bits transferred.
5.
The number of data bits in error.
6.
The number of blocks having CRC errors.
7.
The number of Data Quality register readings taken.
8.
The average Data Quality reading.
9.
The number of frame sync detect aliases.
10. For the latter blocks in the message:
•
The number of data bits transferred.
•
The bit error rate.
Some results are calculated for the latter blocks in the message to quickly indicate whether errors
are uniformly distributed among all blocks in a message.
An alert box provides the option to save results to a file (default name ES9000.XLT or similar)
upon test completion. When saved, the log file includes totals for data bit errors and CRC errors,
grouped according to the block number in which those errors occurred. These totals are
accumulated over all messages of a given type and can be helpful in detecting whether a block’s
cumulative errors are related to the block’s position within a message e.g. the start, middle, or
end. Table 18 illustrates the block result accumulation method for a hypothetical test having four
messages, each message consisting of six blocks.