Note: This manual may cover
several products or versions of one product.
Other products or versions covered by this manual may still be active.
PC Comscope Bit Error Rate Testing
Reference Manual
| Table of Contents |
|
|
|
1.0 Introduction 2.0 Installing The Device 3.0 Running The BERT Program 4.0 The BER Test 5.0 The BLER Test 6.0 The CHER Test Copyright / Warranty |
|
|
Special Technical Notes
Because the BERT program makes extensive use of the Pcs video BIOS some users
may notice problems with interrupt latency at higher bit rates. This is because
some BIOSs are not 'well behaved' as far as re-enabling interrupts after they
have been entered via a software interrupt. The problem may also show up with
the keyboard BIOS. The video BIOS problem will manifest itself as bit or
character errors that occur whenever you hit a key. The distribution diskette
contains a short TSR program called INT.COM that will solve these problems. This
program will intercept the video and keyboard interrupts, re-enable interrupts,
and then call the BIOS routine. Simply enter INT at the DOS prompt to run this
program. If this does not solve the problem you should also examine your CONFIG.
SYS and AUTOEXEC.BAT files for other TSR programs which may interfere with the
PC Comscope.
1.0 Introduction
The BERT (Bit Error Rate Test) program option enables you to obtain a
statistical measure of the quality of a data line. The BERT option records the
number of bits, characters, or blocks with errors. The program may be used in a
loop-back mode, with another PC Comscope or BERT tester, or derive error rates
from live data.
The BERT program consists of three distinct tests:
The BER (Bit Error Rate) test generates and checks a continuous stream of bits
and measures the number of bits in error. The test can generated an synchronize
with a number of standard bit patterns, including those generated by other
manufacturers' equipment. This test also derives a block error rate figure and
the error percentages specified in G.821. The BER test conforms to CCITT
recommendations V.52 and G.821. However, the BER test does not attempt to
normalize the figures to 64 Kbit/s as specified in Annex D of G.821 - all
figures assume the actual bit rate is 64 Kbit/s.
The BLER (Block Error Rate) test derives a block error rate figure from live
HDLC protocol data. The BLER test reports the number of aborted, short, and long
frames and frames with CRC errors. The BLER test also monitors for Reject and
Selective Reject frame.
The CHER (Character Error Rate) test generates and checks asynchronous data. The
test can generate and synchronize with the same standard bit patterns as the BER
test, although they will of course include start and stop bits for each
character. The test can also synchronize on an arbitrary data stream and report
the number of breaks, parity, and framing errors, allowing you to derive
performance figures from live asynchronous data lines.
2.0 Installing The Device
The BERT program module is named BERT.OM4 which is supplied on a diskette
labeled PC Comscope Supplemental Disk. Installing the program simply consists of
copying this file to the directory where you run the PC Comscope program. If you
purchased the Model 903 PC Comscope or the update software with the PEPI
program, the module was included on the diskette and was installed when you
installed the PEPI software. In that case proceed to Section 3.
If you have purchased the BERT software separately you will have a diskette with
the file BERT.OM4 on it. You must copy this file to your working diskette or to
the directory on your hard disk where the PC Comscope files are kept.
1. Make a working copy of the diskette:
| 'MARK' | is a continuous 1's pattern. |
| 'SPACE' | is a continuous 0's pattern. |
| '1:1' | is an alternating 0's and 1's pattern. |
| '1:3' | is a pattern of a 1 followed by three 0's. |
| '3:1' | is a pattern of a 0 followed by three 1's. |
| '1:7' | is a pattern of a 1 followed by seven 0's. |
| '7:1' | is a pattern of a 0 followed by seven 1's. |
| 'CCITT-511' | is the 511 bit pseudo random pattern of recommendation V.52. |
| '1023' | is a bit pseudo random sequence. |
| 'FOX' | is the standard 'Quick Brown Fox' message. |
The FOX pattern is only defined in the CHER (asynchronous) mode. The FOX
pattern is handled differently by the BERT test. The BERT test will send the
standard 'Quick Brown Fox' message but only checks for breaks, parity, or
framing errors on received data without verifying the actual data contents.
This allows you to test live data lines with a MONITOR adaptor.
3.5 Block Size
This menu allows you to select or enter the number of 8 bit characters in
each block. In the BLER mode received frames that are longer than this will
be considered 'long' and in error. In BER mode the block size will affect
the Block Error Rate figure. Internally, the BERT reads in line data in
blocks of the size selected. Each block with one or more bit errors is
counted as a bad block. The block size also has a small effect on
re-synchronization's. If the BERT test finds 5 characters in a row with one
or more bit errors its considers itself out of sync. It will discard the
rest of the data in the block and re-synchronize with the next block of
data. You will also note that the block size is the minimum resolution of
the counters. Making the block size small will slightly degrade the maximum
speed the BERT test can run at.
3.6 Speed
This menu allows you to select or enter the bit rate of the data. It is only
relevant in CHER (asynchronous) mode or with a DCE adaptor where the PC
Comscope provides the data clocking. The speed will default to the speed in
effect when the BERT test was selected, so you either enter the speed here
or in the main PC Comscope program.
3.7 BERT Function Keys
3.7.1
F2 - Return to PC Comscope menus
F10 - Exit from tests
These two function keys actually have identical effect - they return you to
the PC Comscope main program.
3.7.2 F6 - toggle data display: ON/OFF
You can have the line data displayed in the upper half of the screen or
leave it off. Displaying the data will limit the maximum line speed you can
run at but may be helpful when trying to get your configuration right.
Hitting function key F6 will toggle the data display ON or OFF as indicated
on this line. You can only turn data display on or off before starting the
tests.
3.7.3 F9 - Start testing
Begin test.
4.0 The BER Test
The BER test generates and analyzes a continuous stream of synchronous data
bits. The patterns supported are such that the test can synchronize with any
point in the stream. The BER uses the first 8-10 bits of the data stream to
synchronize and then verifies all further data bits. If the test receives 5
consecutive 8-bit characters containing bit errors the test will consider
itself out of sync and re-synchronize.
The BER test also reports a Block Error Rate figure. The blocks are somewhat
arbitrary since the BER test does not actually generate data in blocks but a
continuous stream. However, since many protocols generate data in blocks, so
that any one bit error requires the whole blocks to be scrapped, this figure
may give a more useful measure of line quality. You should select a block
size near the average block size of the protocol to get the most meaning
from this figure.
4.1 Elapsed Time: hh:mm:ss
This is the hours, minutes, and seconds since the test began or the test was
last restarted with F6.
4.2 Pattern
This is the pattern selected in 3.4 above.
4.3 Bits Received
This value does not include the characters remaining in a block when an
out-of-sync condition was detected.
4.4 Sent
This is the number of characters transmitted by the test. This figure is not
exact - it is calculated from the elapsed time multiplied by the bit rate.
If there are large differences between this value and the Bits received it
suggests either long out-of-sync times due to gross data errors or loss of
data clocking, probably due to loss of carrier on the modem.
4.5 Errored Bits
This is the number of bits found to be in error. Note that this value does
not include errored bits during an out-of-sync condition. Thus you must
check both the calculated error rate AND the number of re-syncs to evaluate
the condition of a line.
4.6 BER
This value is the number of bits in error divided by the number of bits
received. The value is given in exponential notation, where 1.00E-006 is
equivalent to 1 times 10 to the minus 6th, or 1/1,000,000.
4.7 BLKS Received
This is the number of blocks received. The number of bits in each block is
displayed in 4.17 below. The number of characters in each block (which is
1/8 the number of bits in each block) is selected in 3.5 above. The BER test
reads and processes line data in blocks of the size selected.
4.8 Sent
This is the number of blocks transmitted by the BER test. Like 4.4 above
this value is calculated from the elapsed time. The BER test does not
actually generate data in blocks but in a continuous stream, so this figure
is simply the number of bits transmitted divided by the block size.
4.9 Errored BLKS
This is the number of blocks with one or more bit errors. Blocks during
which an out-of-sync condition were detected are included in this figure.
4.10 BLER
This is the number of errored blocks divided by the number of blocks
received. The value is given in exponential notation, where 1.00E-006 is
equivalent to 1 times 10 to the minus 6th, or 1/1,000,000.
4.11 Errored Seconds
This is the percentage of seconds that had one or more bit errors. Seconds
when the circuit was deemed 'unavailable' (see below) are not included in
this calculation.
4.12 Severely
This is the percentage of seconds with bit error rates of 1/1000 or worse.
Seconds during which an out-of-sync condition was detected are also
considered severely errored. Seconds when the circuit was deemed
'unavailable' (see below) are not included in this calculation.
4.13 Degraded
This is the percentage of minutes with bit error rates of 1/1,000,000 or
worse. Seconds when the circuit was deemed 'unavailable' (see below) are not
included in this calculation.
4.14 UNAVAIL: or UNAVAIL!
This is the number of seconds the circuit is deemed 'unavailable.' The
circuit is considered unavailable when the bit error rate is 1/1000 or worse
for 10 consecutive seconds. The circuit is considered available again at the
beginning of 10 consecutive seconds with bit error rates better than 1/1000.
When the circuit is unavailable this line will read 'Unavail:.' Note that
there is a 10 second delay in recognizing then beginning an end of the
unavailable condition.
4.15 SYNC: 'Out!' or 'In'
This will display whether the BER test is currently 'In' sync with the data
stream or 'Out!' of sync.
4.16 RE-SYNCS
This is the number of times the BER test has gone out of sync with the data
stream. As explained above, the BER will attempt to re-synchronize when it
receives 5 consecutive 8 bit characters with bit errors.
4.17 BITS/BLK
This is the number of bits in each block received by the BER test. It is 8
times the block size selected in 3.5 above.
4.18 BER Function Keys
4.18.1 F2 - Freeze/Resume Display
F2 allows you to temporarily halt updating of the display, to allow you to
check or record the results to date. When you hit F2 to freeze the display
this line will change to 'Resume display.' Hitting in F2 again will cause
the numbers to be updated again. The test continues normally even when the
display is frozen.
4.18.2 F4 - Display Code
If you have data display enable (3.8 above) F4 will change the display to
HEXadecimal. Hitting F4 again will change it back to the selected display
code.
4.18.3 F5 - Inject Error
Hitting F5 will invert the state of one bit in the stream generated by the
PC Comscope, emulating the effect of a single bit error.
4.18.4 F6 - Restart Test
F6 resets the bits and blocks in error counters.
4.18.5 F7 - Clear Errors
F7 only resets the bits and blocks in error counters.
4.18.6 F10 - Exit Test
F10 stops testing and returns you to the BERT menus.
5.0 The BLER Test
The BLER testis used to derive the block error rate from live HDLC protocol
streams. The BLER test will work with most protocols that use the HDLC
format, including X.25 and SDLC. The BLER test will not work in loop-back or
end-to-end configurations since it does not generate any data. The BLER test
will thus be used only with MONITOR electrical adapters. The BLER test
monitors for good, aborted, and long frames, and frames with CRC errors. It
also monitors for Reject or Selective Reject frames which indicate an error
took place in the opposite direction. The BLER test monitors both send and
receive data streams but calculates the BLER for each differently. One of
the data streams is assumed to be coming from data equipment adjacent to the
PC Comscope and will thus not contain any errors. The BLER for this data is
figured as the ratio of Rejects counted on the other data stream to the
total number of blocks sent. The BLER test assumes that the remote end of
the circuit will issue one Reject for each frame that gets corrupted in
transit. This is not necessarily true, so the actual BLER may be somewhat
higher than what this test reports. The other data stream is assumed to be
from a remote location and may thus contain errors. The BLER for this data
stream is figured as the ratio of the sum of aborted, long, and frames with
CRC errors to the total number of frames received. If you select DTE (3.3
above), receive data will be assumed to be from the remote and transmit data
will be assumed to be generated locally. If you select DCE, transmit data is
from the remote and receive data is local. Typical setups will use the DTE
setting. The following are the items reported by the BLER test:
5.1 Elapsed
This is the amount of time since the test started or was restarted with F6
(5.17 below.)
5.2 Block Size
This is the block size that was selected in 3.5 above. If you see very many
'long' blocks listed (5.6 below) you may have this value too small.
5.3 RxGOOD
This is the number of valid received data frames counted.
5.4 RxCRC
This is the number of receive data frames with CRC errors.
5.5 RxABORT
This is the number of aborted receive data frames.
5.6 RxLONG
This is the number of receive data frames that contained more than the
selected block size characters. Long frames may occur because the protocol
actually sent a frame that was long or the closing flag of a frame may get a
bit error and two consecutive frames appear to be concatenated. The BLER
test does not distinguish between long frames with or without valid CRC.
5.7 RxREJ
This is the number of receive data rejects counted. Note that the protocol
will send a receive data reject because it detected an out-of-sequence
condition on the transmit data side (normally due to a bit error.)
5.8 RxBLER
If DCE (3.3) was selected this figure will be the ratio of TxREJ to total
receive data frames. If DTE was selected this figure will be the ratio of
the sum of aborted, long, and receive data frames with CRC errors to the
total number of receive data frames. The value is given in exponential
notation, where 1.00E-006 is equivalent to 1 times 10 to the minus 6th, or
1/1,000,000.
5.9 TxGOOD
This is the number of valid transmit data frames counted.
5.10 TxCRC
This is the number of transmit data frames with CRC errors.
5.11 TxABORT
This is the number of aborted transmit data frames.
5.12 TxLONG
This is the number of transmit data frames that contained more than the
selected block size characters. Long frames may occur because the protocol
actually sent a frame that was long or the closing flag of a frame may get a
bit error and two consecutive frames appear to be concatenated. The BLER
test does not distinguish between long frames with or without valid CRC.
5.13 TxREJ
This is the number of transmit data Rejects counted. Note that the protocol
will send a transmit data Reject because it has detected an out-of-sequence
condition on the receive data side (normally due to a bit error.)
5.14 TxBLER
If DCE (3.3) was selected this figure will be the ratio of RxREJ to total
transmit data frames. If DTE was selected this figure will be the ratio of
the sum of aborted, long, and transmit data frames with CRC errors to the
total number of transmit data frames. The value is given in exponential
notation, where 1.0E-006 is equivalent to 1 times 10 to the minus 6th, or
1/1,000,000.
5.15 BLER Function Keys
F2 allows you to temporarily halt updating of the display, to allow you to
check or record the results to date. When you hit F2 to freeze the display
this line will change to 'Resume display.' Hitting F2 again will cause the
numbers to be updated again. The test continues normally even when the
display is frozen.
5.15.2 F4 - Display Code
If you have data display enabled (3.8 above) F4 will change the display to
HEXadecimal. Hitting F4 again will change is back to the selected display
code.
5.15.3 F6 - Restart Test
F6 resets all counters to zero. It is the equivalent of exiting and then
re-starting the test.
5.15.4 F7 - Clear Errors
F7 only resets the FxCRC, RxAbort, RxLong, RxREJ, TxCRC, TxAbort, TxLong,
and TxREJ counters.
6.0 The CHER Test
The Character Error Rate test generates and checks asynchronous data
characters and reports the ratio of characters with errors to total
characters received. The CHER test can run in loop-back, end-to-end, or live
data configurations. The CHER test supports the same data patterns as the
BER test as well as the standard "Quick Brown Fox" message. The CHER test
does not verify the data characters in the FOX pattern but only counts the
number of breaks, parity, and framing errors. This allows you to derive a
CHER rate from live data streams. If you are using the FOX message you must
select the number of bits/character and the parity before starting the BERT
program. For all other patterns the CHER test will force the PC Comscope to
8 bits/character, 1 stop bit, and no parity. The following are the items
reported by the CHER test:
6.1 Elapsed Time
The time since the test was started or restarted with function key F6 (6.14
below.)
6.2 Pattern
Displays the currently selected pattern being generated and checked for. If
FOX is selected the test does not verify the content of received data
characters.
6.3 CHARS Received
This is the total number of characters received, including errored
characters. Breaks are not included in this figure.
6.4 SENT
This is the number of characters transmitted by the PC Comscope.
6.5 GOOD CHARS
The number of characters with no errors.
6.6 BREAKS
The number of breaks received. A break is a space (zero) condition for 10 or
more bit times. Breaks are not counted as errored characters and are not
figured into the BLER calculation.
6.7 FRAMING ERRORS
The number of framing errors detected. A framing error means that the
character did not include a valid stop bit.
6.8 PARITY
The number of parity errors detected. This error will only show up when
using the FOX pattern and parity enabled.
6.9 BAD DATA
The number of characters with incorrect data. This error will not show up
with the FOX pattern.
6.10 CHER
This is the ratio of the sum of characters with parity or framing errors or
bad data to the total number of characters. The value is given in
exponential notation, where 1.00E-006 is equivalent to 1 times 10 to the
minus 6th,or 1/1,000,000.
6.11 CHER Function Keys
6.11.1 F2 - Freeze/Resume Display
F2 allows you to temporarily halt updating of the display, to allow you to
check or record the results to date. When you hit F2 freeze the display this
line will change to 'Resume display.' Hitting F2 again will cause the
numbers to be updated again. The test continues normally even when the
display is frozen.
6.11.2 F4 - Display Code
If you have data display enabled (3.8 above) F4 will change the display to
HEXadecimal. Hitting F4 again will change it back to the selected display
code.
6.11.3 F5 - Inject Error
Hitting F5 will alter the data in a single character that it generates,
producing a single 'Bad data' error.
6.11.4 F6 - Restart Test
F6 resets all counters to zero. It is the equivalent of exiting and then
re-starting the test.
6.11.5 F7 - Clear Errors
F7 only resets the breaks, parity, and framing error counters.
6.11.6 F10 - Exit Test
F10 stops testing and returns you to the BERT menus.
Copyright
Copyright (c) 1991 by Telebyte Technology, Inc., Greenlawn, New York 11740.
PC Comscope and PC Comscope II are registered trademarks of Telebyte
Technology, Inc.
This manual and the BERT software are copyright by Telebyte Technology, Inc.
Telebyte grants to the original purchaser of the software the right to make
several copies of the software for backup purposes only. You are also
authorized to make such copies of this manual as are helpful to the use of
the software so long as the manual is copied in its entirety with this
notice as a permanent part of the copy.
Disclaimer of Warranty
The BERT program option is sold without warranty of any kind. Telebyte makes
no claims or guarantees concerning the accuracy, performance, or suitability
of this product. If you like what it does, keep it. If not, send it back.
You are given a period of thirty days from the purchase date to evaluate the
suitability of this product for your purposes. If you determine that the
product is not suitable you may return the software to the address below for
a full refund. Any copies of the software must also be destroyed at that
time. Please call before returning the software for a Return Authorization
Number, RMA.