MAI COMPANY CONFIDENTIAL FIELD INFORMATION BULLETIN SERVICE GROUP(S): MINI MFG: BASIC FOUR NACS TYPE(S): 4108 MFG. MODEL(S): 2000 CATEGORY: CPU DESC: MICRO (EAGLE) *** TABLE OF CONTENTS AS OF 02/14/95 *** FIB 00001.General Information.........................................01/25/95 FIB 00002.Service manual corrections on 4-way controller jumpers......09/13/84 FIB 00003.LAN Controller switch settings..............................09/24/84 FIB 00004.Correction to MCS controller switch settings in Service Manu09/24/84 FIB 00005.Packing material must be removed from inside cpu upon instal09/24/84 FIB 00006.Installation instruction for second Winchester disk.........09/24/84 FIB 00007.Hints for troubleshooting an inoperative system console term12/20/84 FIB 00008.Upgrade instructions for 7.1*27 and installing SSN..........01/25/95 FIB 00009.Unexpected diagnostic program execution.....................02/13/85 FIB 00010.Floating point math software gives different results than 1302/20/85 FIB 00011.Jumpering detail information for old and new style CMB PCBAs10/26/87 FIB 00012.Mounting and handling of the Winchester disk drive..........02/21/85 FIB 00013.Installation and warranty policy............................03/18/85 FIB 00014.Announcement of the 2000 System Reference Card..............03/18/85 FIB 00015.Initial installation problems due to console unit configurat05/22/85 FIB 00016.Second hard disk drive installation information.............05/22/85 FIB 00017.False memory parity errors caused by below LARL CMB PCBA....08/02/85 FIB 00018.Hard disk controller jumpering information..................10/22/85 FIB 00019.ERROR -2 and ERROR -62 during backup to the MCS tape drive..12/05/85 FIB 00020.Various problems when the console terminal is set-up on a fo12/26/85 FIB 00021.Instructions.for formatting the 85MB Maxtor disk............12/26/85 FIB 00022.ERROR 0's on non spooled serial printers....................01/30/86 FIB 00023.Intermittent failures of 53 MB and 20 MB disk drives........02/26/86 FIB 00024.Installation instructions for 1/2" streamer (MTS)...........02/24/89 FIB 00025.Corrupted data on MCS backup................................03/26/86 FIB 00026.Won't load operating system, system hangs...................03/26/86 FIB 00027.Hard disk controller revision required for OS Level 7.2A....05/07/86 FIB 00028.Detailed info concerning hard disk failing sector determinat07/16/86 FIB 00029.Sys 2000 Flaw Maps, DUTIL-Drive Formatting..................08/07/86 FIB 00030.System slowdowns/hangs won't load - caused by unterminated c01/13/87 FIB 00031.Various disk start-up problems on power-up sequence.........03/09/87 FIB 00032.Missing defect maps on MAXTOR drives........................03/31/87 FIB 00033.1488.& 1489 socket locations for I/O ports on CMB & 4-way PC04/06/87 FIB 00034.ERROR 31 when running ORIGIN on OS Rel 7.2A and 7.2B........04/08/87 FIB 00035.New firmware for MTCS controller required for Tandberg MCS d04/22/87 FIB 00036.New Kernel PROM allows 1/2" MTS boot........................04/27/87 FIB 00037.PCBAs below LARL may cause Winchester disk errors...........05/05/87 FIB 00038.File list gets corrupted on O.S. Releases 7.2...............06/05/87 FIB 00039.System 2000/3000 Hard Disk Substitution.....................09/14/87 FIB 00040.Service Report form for 2000 service calls (send to Prod. Su10/07/87 FIB 00041.Unable.to load SSN tape succesfully.........................10/07/87 FIB 00042.Command 'who' shows 'admin' logged onto unused terminal(s)..10/12/87 FIB 00043.Firmware corrections for 4-way & 8-way self test problems...10/19/87 FIB 00044.ESD (Electro-Static Discharge) kit is available.............11/09/87 FIB 00045.ESCAPE key problem while in BASIC mode......................10/30/87 FIB 00046.Memory Diagnostic Test Failure on Release 7.2B..............11/02/87 FIB 00047.Memory parity errors on 7.3 diagnostics.....................11/20/87 FIB 00048.Slave printer attached via DS-5XX workstation...............11/20/87 FIB 00049.Terminal.hang when opening R0 and LP [ WPS 407 ]............12/29/87 FIB 00050.New installation procedures for 2000/3000 [ WPS 422 ]......02/16/88 FIB 00051.PARADYNE MODEM CONFIGURATION................................02/23/88 FIB 00052.SYS Crash When Adding Memory................................08/06/88 CPU-BASIC FOUR-2000---------------------------------Table Of Contents Pg01 of 02 FIB 00053.2000/3000/4000 LARL History [ WPSF 475 R2 ].................03/21/90 FIB 00054.NVRAM errors................................................11/01/88 FIB 00055.New 1/2" Streamer Tape Controller (STC) [ WPSF 487 ].......12/20/88 FIB 00056.Workaround for terminal hangs on OS release 7.5B............01/29/90 CPU-BASIC FOUR-2000---------------------------------Table Of Contents Pg02 of 02 FIB 00001 01/25/95 *** General Information *** This F.I.B. group contains information regarding all hardware in the CPU enclosure EXCEPT I/O devices (i.e. floppy disk drive, 1/4" streamer tape, fixed disk drive). For information on an I/O device, consult the F.I.B. group for that device. Software information will be located in the Software section in F.I.B. group BOSS/IX,BBVII. If a software problem appears to be hardware it will be included in the CPU group. *** NACS MACHINE TYPES AND FEATURE NUMBERS *** MAI Basic Four 2000 Desk Top Computer - Model 4108 - Sorbus Type - 4108 *** AVAILABLE DOCUMENTATION *** Sorbus Environmental Handbook ........................................S-HDBK-027 MAI Basic Four 2000 Service Manual....................................S-HDBK-B71 MAI Basic Four 2000 Technical Reference Manual........................S-HDBK-B73 MAI.Basic Four 2000 Desktop Error Logging Reference Manual............S-HDBK-B70 MAI Basic Four 2000 Eagle Reference Card..............................S-HDBK-B67 5 1/4" Winchester Disk Controller Service Manual (order from B/4).....M8158 4-Way Controller Service Manual (order from Basic Four)...............M8155 LAN Controller Service Manual (order from Basic Four).................M8157 *** AVAILABLE TRAINING *** All SPX models (2000,3000,4000,1800) are included in one formal course available in Tustin. *** SPECIAL TOOLS *** System Serial Number (SSN) software and associated IC PAL is required for OS releases 7.1*27 and later. These tools must be ordered but will only be shipped to those people who have signed a non-disclosure agreement for the 2000. To order SSN tapes for OS levels 7.3 and later (for formally trained personnel) contact Software Distribution. NOTE - See the 2000/3000 Handbook FIBs for systems using OS releases 7.3 & up. MM538160.....Cartridge Tape, SSN (levels prior to 7.3) MM538170.....Floppy Diskette, SSN (levels prior to 7.3) MM538180.....I.C. PAL, SSN........(all OS levels) MM538040.....B/4 907529-001.....Loopback cable, 4-way 1 port MM538050.....B/4 907530-001.....Loopback cable, 4-way 2 port (2 required) MM538060.....B/4 907588-001.....Loopback plug, parallel printer port (CMB) ????????.....B/4 907565-001.....Loopback cable, LAN ????????.....B/4 907641-001.....Loopback plug, CMB serial port *** SPECIAL ADMINISTRATIVE PROCEEDURES *** For lock and key authorization phone numbers and hours of coverage reference FIB 00005 in the General, Basic Four, Administrative Procedures FIB group. ORIGINATOR: N. JONES CPU-BASIC FOUR-2000---------------------------------FIB001 Pg001 FIB 00002 09/13/84 *** Service manual corrections on 4-way controller jumpers *** Service Manual BFISD 8079P, page 5-15, contains incorrect information on 4-Way controller jumper settings. Jumpers are set according to the artwork (PCB detail) used. At the present time 2 different PCB details may be used. Set each port according to its use: Terminal/Printer or Modem. PCB detail 904741-001 (current production) must be set as follows: Terminal Printer: Modem: 1 to 7 1 to 2 2 to 8 3 to 4 3 to 4 7 to 8 9 to 10 9 to 10 13 to 15 13 to 14 14 to 16 15 to 16 PCB detail 904943-001 (future production) must be set as follows: Terminal/Printer: Modem: 1 to 3 1 to 2 2 to 4 3 to 4 9 to 11 9 to 10 10 to 12 11 to 12 13 to 15 13 to 14 14 to 16 15 to 16 ORIGINATOR: J. VAN OORDT CPU-BASIC FOUR-2000---------------------------------FIB002 Pg001 FIB 00003 09/24/84 *** LAN Controller switch settings *** Lan controller switch settings on Service Manual BFISD 8079P, page 5-24, are inaccurate due to recent Engineering Changes. Switches on Lan controller P/N 903405-001, MM531060 should be set as follows: SWITCH 1 - SW1 determines the node address (Station number), bias and termination of the network. Every Lan controller on the network must have a unique node address, from node 01 to node 63. Node 00 is not allowed. Example: set switch SW1, positions 3 to 8, as follows: 1st. system: Set positions 8 to OFF (Node 01) 2nd. system: Set positions 7 to OFF (Node 02) 3rd. system: Set positions 7,8 to OFF (Node 03) Terminator . Station Address s . Bias . . SW1 positions: 8 7 6 5 4 3 2 1 . . . . . . . . . . . . Binary Value: 1 2 4 8 16 32 Note: SW position set to OFF = asserted. Bias(SW1-1): Only 1 controller on the network should have SW1-1 set to OFF. All other controllers should have SW1-1 set to ON. Terminator(SW1-2): Only 2 controllers (one at each end of the network) should have SW1-1 set to ON, all other controllers should have SW1-2 set to OFF. Terminating resistors (100 ohms) in the tap boxes are no longer required. SWITCH 2. - SW2 sets the board address and Bus arbitration on the system bus, 1 board per system is normal, 2 is maximum. Set switch SW2 as follows: 1st. board: Set positions 4,5,7 to ON 2nd. board: Set positions 4,5,6,8 to ON A 2nd. board is normally used only with diagnostics. ORIGINATOR: J. VAN OORDT CPU-BASIC FOUR-2000---------------------------------FIB003 Pg001 FIB 00004 09/24/84 *** Correction to MCS controller switch settings in Service Manual *** Service Manual BFISD 8079P, page 5-20, contains incorrect information on MCS controller switch settings. MCS controller P/N 903406-001, MM531070 should be set as follows: Switch 1 (8 positions). Set positions 5 and 6 to OFF Set all other positions to ON ORIGINATOR: J. VAN OORDT CPU-BASIC FOUR-2000---------------------------------FIB004 Pg001 FIB 00005 09/24/84 *** Packing material must be removed from inside cpu upon installation *** The MAI 2000 Desk Top Computer system (also known as the Eagle system) is being shipped.with packing material inside the plastic cabinet. This packing material holds the modules and disk drive in place. It is most important that this packing material be removed before power is applied, as it could possibly restrict air flow and cause thermal stress and/or unit failures. The floppy disk drive is also shipped with a cardboard retainer installed as media. This must be removed prior to powering up the system. ORIGINATOR: J. SANKA CPU-BASIC FOUR-2000---------------------------------FIB005 Pg001 FIB 00006 09/24/84 *** Installation instruction for second Winchester disk *** In order to install the second Winchester Disk Drive, it will be necessary to remove the cable, P/N 907633-001, that is plugged into the Adaptec Module at Location J 2 and the disk drive. This cable will be replaced by cable, P/N 907644-001, which is a daisy chain cable. Plug proper end on J2 of Adaptec Module and other two plugs on Drive 1, J1 and Drive 2, J1. Plug in cable, P/N 907632-002, to J1 on the Adaptec Module and the other end to Drive 2, J2. The Drive Address Switches (Drive Select) must be set on the second unit. See Figure 1 for Micropolis switch settings. See Fig. 2 for Rodime switch settings. The first drive must also have the Terminator Resistor Pack removed. See Figures 1 and 2. The first drive is shipped with an Operating System installed. The second drive is formatted only. The partitions must be changed to include the capacity of the second disk drive. DRAWING INFORMATION: See Mini Alert 298 hardcopy for figures 1 and 2 drawings. ORIGINATOR: J. SANKA CPU-BASIC FOUR-2000---------------------------------FIB006 Pg001 FIB 00007 12/20/84 *** Hints for troubleshooting an inoperative system console terminal *** SUBJECT: System 2000 (Eagle), inoperative console The console terminal may fail to operate correctly when the NVRAM contains incorrect data. This condition may occur after a CMB board is replaced or after writing.new NVRAM data. Use the following procedure to activate the console and write correct parameters in the NVRAM: 1. Set the terminal to 9600 baud, 7 data bits, 1 stop bit and odd parity. 2. Connect the terminal to CMB board port 0. 3. While observing the LED on the left side of the CMB board, power up the system. o The LED should turn on immediately, then go off 4 times and turn back on. This sequence indicates that the CMB self tests (ROM resident) have passed. o If the LED does not function as indicated above, replace the CMB or first memory board to correct the problem. o After the CMB self tests have passed, the console should print the results of the self tests on the screen. Incorrect printing or no printing at all may indicate a malfunctioning terminal, a defective cable or incorrect data in the NVRAM. 4. If the problem is isolated to the NVRAM proceed as follows: o Power down the system. o While observing the LED, power up the system and be ready to press ESCAPE several times. Wait until the LED goes off for the first time. ESCAPE should be pressed while the LED is blinking on and off. o Pressing ESCAPE during the self tests forces the firmware to load default parameters, to ignore NVRAM data and to jump to the firmware debugger . o Default parameters loaded for the console are: 9600 baud, 7 data bits, 1 stop bit and odd parity. 5. After is printed on the console screen enter: config ; config commands are displayed on the screen read ; contents of NVRAM is read into RAM display ; parameters are displayed on the screen 6. Enter the correct parameters for the console and the other devices if necessary. Re-write the NVRAM. If step by step instructions to enter parameters and write the NVRAM are needed, refer to the Debug Users Guide in Section 2 of the Diagnostic Manual. 7. Power down and power up again. The console should now be operational. ORIGINATOR: J. VAN OORDT CPU-BASIC FOUR-2000---------------------------------FIB007 Pg001 FIB 00008 01/25/95 *** Upgrade instructions for 7.1*27 and installing SSN *** SUBJECT: MAI 2000 upgrades Basic Four will be shipping to existing installations a package to upgrade the MAI 2000 to 7.1*27 Operating System. The package will contain the following items: 1. New firmware proms for the CMB. Installation instructions are incorporated in this document. 2. New firmware prom for the 4-way. Installation instructions are incorporated in this document. 3. Media containing 7.1*27 Operating System. 4. A new serial number label that must be put in place on the rear of the system. In order to install the new O.S., the CMB and 4-Way Controller boards must be upgraded with the new firmware. A new System Serial Number (SSN) must be installed on the system. The two types of media containing the SSN Software, and only the SSN Software, have the following part numbers. These programs will be issued only to those who were trained and signed the Non-Disclosure agreements for the MAI 2000. MM538160 Cartridge Tape, SSN MM538170 Floppy Diskette, SSN There are 2 versions of the CMB module presently in use.They are interchangeable The old artwork version has the Basic Four part number 903441-001 and contains many cuts and jumpers. On this version, after the installation of new proms, all that is necessary to install the SSN is the SSN Software. The new artwork version has the Basic Four P/N 903441-002 and contains very few cuts and jumpers. On this version the installation of the serial number requires the SSN Software and the installation of the special I.C. PAL, P/N MM538180 Basic Four P/N 911016-004, in location 8Z on the CMB. After installation of the serial number be sure to remove your PAL and reinstall the PAL that was at location 8Z. Failure to do so will show up as a failure during the CMB self- test immediately after power is turned on. Once the system has been upgraded and tested for the 7.1*27 O.S., a MAI-BFISD Software Support person should be called to the site to install the O.S. The Service Call for upgrading the system is not billable, use Service Code 11 on the IR. MAI 2000 UPGRADE INSTRUCTIONS 1. Check the following for the proper Revision Levels: CMB Rev. "S" or later Four-Way Rev. "J" or later MTC Rev. "L" or later LAN Rev. "D" or later Do not proceed any further until you obtain modules with the above listed revision levels. If the system is without any of the above modules, disregard that portion of the list. CPU-BASIC FOUR-2000---------------------------------FIB008 Pg001 2. Remove the IC at location 6S on the CMB and replace it with BFIS P/N 165047-037 FMWR 1. 3. Remove the IC at location 6N on the CMB and replace it with BFIS P/N 165047-036 FMWR 0. 4. Mark CMB to show Revision Level 'AD'. 5. Remove the IC at location 6J on the four-way and replace it with BFIS P/N 165047-042. 6. Mark 4-way to show Revision Level 'P'. 7. Re-install all modules in their proper locations. 8. Turn on the display that you are planning to use for the installation of the serial number. 9. Turn on the MAI 2000 System and continuosly hit the 'Escape' Key on the key board until the lower case appears on the screen. 10. Type in lower case 'config' and CR, displayed on the screen will be the following: Config Menu command description read read Config NVRAM display display config terminal specify terminal type boot specify boot dev console specify system console download specify download port printer specify printer port write write config NVRAM debug enter Debug menu display menu 11. Type in 'display' and CR. - Displayed on the screen will be the following: term type : evdt (or type of your terminal) boot port : wd0 console port : sc0 baud : 9600.......etc. download port : sc1 baud : 9600.......etc. printer port : pit0 12. Type in 'write' CR. - Displayed on the screen will be the same screen as above plus the last line will read: write it (y/n)? type in 'y' done 13. Turn power off the MAI 2000 momentarily and then back on. While the system is doing the normal self-test depress the CTRL Key and the letter 'C' Key simultaneously several times to invoke the alternate load. If successful the last line on the screen will be: Boot device: At this point load the media that contains the SSN program in the appropriate device and type in the device name. e.g. 'fd' for floppy or 'cs' for tape. CPU-BASIC FOUR-2000---------------------------------FIB008 Pg002 The next input the system will request is: System file: Type in lower case 'ssn' CR The system will then boot from the floppy, or the Streamer, whichever was selected above. 14. If the load failed the following message is displayed: Error during load of Press 'RETURN' Key to reboot ('^C'=altload, '^S'=selftest): If the load is successful a Boot Complete message is displayed followed by: NVRAM System Serial Number Update Program Rev. A Current Serial Number : 2xxx-xxxxx (read from the NVRAM) New Serial Number : 2 Enter a properly formatted serial number (only numeric values with a dash '-', in the fourth position are accepted), terminated by a . 15. Refer to GENERAL-BASIC FOUR-ADMINISTRATIVE PROCEDURES FIB 5 for current lock and key information. 16. Lock value : xxxxxxxx (randomly generated) Enter Key : Enter the correct 8 alpha numeric characters, received from an authoriza- tion center, terminated with a . If the Key is incorrect the following message is displayed: FAILURE TO ENTER THE CORRECT KEY!! Press 'RETURN' Key to reboot ('^C'=altload, '^S'=selftest): If the key is correct the following is displayed: Update Serial Number (y/n)? If is entered the System Serial Number is updated in the NVRAM. If it is successful or is entered the following message is displayed: NVRAM System Serial Number Update complete. Press 'RETURN' Key to reboot ('^C'=altload, '^S'=selftest): The program will verify that the serial number is properly written. It will attempt to write the serial number two times. If it fails after the second attempt the following message is displayed: NVRAM System Serial Number Update FAILED!! Press 'RETURN' Key to reboot ('^C'=altload, '^S'=selftest): The program will always terminate with the general reboot message. 17. At this point the OS *27 can be installed by the Basic Four representative 18. Return the proms removed from the CMB and the four-way to Sorbus West. ORIGINATOR: S. SANKA CPU-BASIC FOUR-2000---------------------------------FIB008 Pg003 FIB 00009 02/13/85 *** Unexpected diagnostic program execution *** Sometimes after writing the NVRAM on the CMB board with the system serial number or new configuration parameters, a byte in the NVRAM is set incorrectly. This incorrect byte may cause an unexpected loading and execution of a Logic Test when running diagnostics. Writing F0 in location 680000 should correct the auto load and execution problem after loading the diagnostic executive. The incorrect byte can be re-written with the correct data as follows: Prompt Enter 1. Jump into debug by hitting escape or CTL Z during self tests. 2. Jump into config config 3. Read NVRAM read 4. Jump to debug by pressing ESC 5. Display NVRAM location 680000 db 680000 and change it to F0 680000 Fx F0 (and CR) 6. Jump to config config 7. Write NVRAM after checking all write parameters 8. Jump to debug debug 9. Load diagnostics executive boot fd0 [or cs0] The diagnostic executive should load and wait for commands before loading and executing a Logic Test. ORIGINATOR: J. VAN OORDT CPU-BASIC FOUR-2000---------------------------------FIB009 Pg001 FIB 00010 02/20/85 *** Floating point math software gives different results than 13xx CPU *** The BASIC language on the MPX series and 2000 series CPUs will round all floating point numbers. This is not the case with 13XX systems. On the 13XX, numbers are sometimes rounded and sometimes not. Only certain expressions are rounded. The following are examples: : 13XX : 2000 : 8000 MATH EXPRESSION : RESULT : RESULT : RESULT -------------------:--------------------:--------------------:------------------ 1/N : .083333333333 : .083333333333 : .083333333333 : : : 12/C : 2.0 : 2.0 : 2.0 : : : (1/N)*(12/C) : .166666666666 : .166666666667 : .166666666667 : : : 1/((1/N)*(12/C)) : 6.000000000024 : 5.999999999999 : 5.999999999999 ------------------------------------------------------------------------------- These slight differences can cause large deviations on the same expressions used on both systems. Programs on the 13XX that have expressions that take into account this innaccuracy may not work the same on the 2000 or MPX series. The expression has to be re-written to account for the automatic rounding off by the 2000 and MPX series. ORIGINATOR: BASIC FOUR CPU-BASIC FOUR-2000---------------------------------FIB010 Pg001 FIB 00011 10/26/87 *** Jumpering detail information for old and new style CMB PCBAs *** Currently, there are two types of CMB boards available as spare parts: 1. MM531010, P/N 903441-001, original CMB board. It does not require the I.C. security PAL to write the system serial number. 2. MM531011, P/N 903441-002, new CMB board in current production. It requires a security I.C. PAL to write the system serial number. The new and the old CMB boards are interchangeable, except that the new board requires using the security PAL P/N MM538180/911016-004 to write the system serial number. Parts orders will be filled with either type of CMB board accord- ing to availability. All CMB boards will be shipped with the new firmware level B2.9 contained in prom 165047-036 (located at 6N) and prom 165047-037 (located at 6S). This firmware level is a requirement for OS 7.1A *27 and above. Systems in the field running on OS levels prior to *27 use old firmware level A3.1 contained in proms 165047-15 and 165047-16. Use the old proms with the new CMB until the system is upgraded to *27. The firmware level is printed on the console screen during the power-up self-tests just before the system serial number. To observe screen information use CONTROL C to stop automatic system loading. Refer to MSA No. 334 for system upgrade to OS level *27 and above. On new CMB 903441-002, the I.C. PAL P/N 911016-003, located at 8Z must be replaced by the Security I.C. PAL P/N 911016-004 while writing the system serial number on the NVRAM. The security PAL should be removed after writing the system serial number and the standard PAL should be re-installed. CMB 903441-001 does not require the use of the security PAL. The system serial number is written using the program 'ssn' contained in cartridge streamer tape P/N MM538160 or in floppy diskette P/N MM538170. An authorization key is needed from Basic Four while using the program 'ssn' to write the system serial number. Call Basic Four with the lock number and the system serial number to obtain the proper key. Refer to MSA No. 334 for procedure to write the system serial number. When replacing a CMB board, verify that its jumpers are set correctly. All CMB boards are shipped with the serial ports set for terminal/printer. If a modem is used some jumpers must be changed. CMB SERIAL PORTS JUMPERS SETTINGS: Old CMB P/N 903441-001 (PCB artwork 904862-001) PORT 0. Also known as Port A. Terminal/Printer jumper setting: JMP A: 1-3, 2-4, 5-6, 7-9, 8-10, 11-13, 12-14 Modem jumper settings: JMP A: 1-2, 3-4, 7-8, 9-10, 11-12, 13-14 PORT 1. - RS 232 - (Also known as Port B) Terminal/Printer jumper settings: JMP B: 1-3, 2-4, 5-6, 7-9, 8-10, 13-15, 14-16 JMP G: 15-16, 17-18, 19-20, 23-24 JMP K: 1-2 CPU-BASIC FOUR-2000---------------------------------FIB011 Pg001 JMP H: none Modem jumper setting: JMP B: 1-2, 3-4, 7-8, 9-10, 13-14, 15-16 JMP G: 15-16, 17-18, 19-20, 23-24 JMP K: 1-2 JMP H: none PORT 1 - RS 422 - (Also known as Port B) JMP G: 1-2, 3-4, 5-6, 7-8, 9-10, 11-12, 13-14, 21-22 JMP B: 11-12 JMP H: 2-3 Be sure that the following jumpers are disconnected: JMP G: 15-16, 17-18, 19-20 JMP K: 1-2 NEW CMB P/N 903441-002 (PCB artwork 905011-001) Serial Ports configuration between a Terminal/Printer or Modem is determined by orientation of the "zero-ohm-packs". PORT 0. Also known as Port A. Terminal/Printer jumper set-up: Pin 9 of zero-ohm-pack is in pin 1 of socket A. Modem Jumper set-up: Pin 1 of zero-ohm-pack is in pin 1 of socket A. PORT 1. - RS 232 - Also known as Port B. Terminal/Printer jumper setting: Pin 9 of zero-ohm-pack is in pin 1 of socket B Pin 1 of zero-ohm-pack is in pin 1 of socket K Nothing in sockets G & H Modem jumper set-up: Pin 1 of zero-ohm-pack is in pin 1 of socket B Pin 1 of zero-ohm-pack is in pin 1 of socket K Nothing in sockets G & H PORT 1. - RS 422 - (Also known as Port B) Pin 1 of the zero-ohm-pack is in pin 1 of socket G Pin 1 of the zero-ohm-pack is in pin 1 of socket H Nothing in sockets B and K CMB MISCELLANEOUS JUMPERS: FOR BOTH P/N 903441-001 AND 903441-002 L: 1-2 C: none (1-2 etched) E: 1-2, 7-8 N: 1-2 P: none (1-2 etched) F: 1-2 M: 1-2 R: none S: none D: 2-3 S1: 1-2 (1-2 to force an interrupt for dumps, 2-3 to produce a hard reset.) CPU-BASIC FOUR-2000---------------------------------FIB011 Pg002 P/N 903441-002 ONLY: V: 1-2 T: none (1-2 etched) on Rev. AN and lower W: none T: 2-3 on Rev. AP and higher X: 1-2 ORIGINATOR: JAVIER VAN OORDT CPU-BASIC FOUR-2000---------------------------------FIB011 Pg003 FIB 00012 02/21/85 *** Mounting and handling of the Winchester disk drive *** SUBJECT: MAI 2000 WINCHESTER DISK DRIVE MOUNTING The mounting plate studs protrude through the shock mounts so that when the drive is placed on a flat surface the studs contact the flat surface and defeat the shock mounting. Under these circumstances it is imperative that all the Winchester Disk Drives be handled with extreme caution, otherwise either the surface(s) or head(s) may be damaged. ORIGINATOR: J. SANKA CPU-BASIC FOUR-2000---------------------------------FIB012 Pg001 FIB 00013 03/18/85 *** Installation and warranty policy *** The MAI Basic Four 2000 will be installed and maintained like all current MAI Basic Four products. Sorbus will perform pre-site surveys, install the systems, install add-on equipment and provide warranty service in accordance with the standard agreement between the two companies. ORIGINATOR: E. UNTEN CPU-BASIC FOUR-2000---------------------------------FIB013 Pg001 FIB 00014 03/18/85 *** Announcement of the 2000 System Reference Card *** The Reference Card for the System 2000 (Eagle) was designed to help to run diagnostic programs. It contains, in addition to the diagnostic programs, hardware and software information needed when running diagnostics. A very condensed format has been used to save space. The user is assumed to be a person already trained on the system and in the use of the diagnostics. The Reference Card contains the following: Diagnostic Section: o List of all programs names and their purpose. o How to boot the executive and use its commands. o How to run the diagnostic programs. o How to run diagnostic utilities. o Firmware routines used with diagnostics. Hardware Section: o LARLs for all PCBAs. o Hard disk drives jumper and switch settings, status bytes and commands. o Streamer status bytes and commands. o Floppy drive jumpers, status bytes and commands. o Floppy Head Alignment information. o Memory boards addresses and switch settings. o CMB board jumpers settings. o Controllers jumpers and switch settings for: - 4-Way - Lan - Streamer - Winchester adapter and Adaptec. o External cables part numbers. Software Section: o How to install and upgrade the Operating System. o How to backup and restore the root partition. o How to install a key in software products. o How to print the system error log. o System Memory map. o Hexadecimal to decimal conversion table. See hardcopy of Mini Alert 353 for the complete reference card. ORIGINATOR: J. VAN OORDT CPU-BASIC FOUR-2000---------------------------------FIB014 Pg001 FIB 00015 05/22/85 *** Initial installation problems due to console unit configuration *** At the present time all Systems 2000 (Eagle) are being preset at the factory with the system console on the first port of the first 4-way controller. This configuration is used regardless of the final hardware configuration. The system console configuration is resident in the NVRAM of the CMB board. The NVRAM is configured prior to shipment as follows: term type: evdt boot port: wd0 console port: fw00 baud:9600 data bits:7 stop bits:1 parity:odd flow:xon The NVRAM should be re-configured to move the console to CMB port 0, in any of the following conditions: o the system does not have a 4-way controller. o the system has a 4-way controller but there is a malfunction and the terminal connected to 4-way port 0 does not respond. o 4-way cable is not available. PROCEDURE TO INSTALL A SYSTEM USING CMB PORT 0 AS THE CONSOLE. 1. Connect the console terminal (EVDT, EDT or 7270) to the first port on the CMB using Eagle cable P/N 907753. (old part number was 907454) If the system was shipped without a CMB to terminal cable, use one of the following cables to connect a terminal to the first port on the CMB: o 907756 Cable Adapter, CMB port to 4-way cable. Market Code: KA016AAZ This cable adapter allows a 4-way cable to be connected to the CMB. o 906694/MM730750 Cable, I/O Panel to Terminal (13XX type). Or any equivalent terminal cable as 906274 etc. These cables are for temporary use only. Customer should only use cables designed for the system 2000. 2. Enter by pressing ESC immediately after power-up or by pressing CTLR Z during the memory portion of the self-tests. 3. After the prompt, type 'config' to enter configuration mode. 4. After the prompt, type 'read' to read contents of the NVRAM. 5. Type 'display' to display contents of NVRAM. 6. To re-configure the system console type 'console'. Enter the following information as prompted by the system: console dev: sc console unit: 0 baud rate: 9600 data bits: 7 stop bits: 1 parity: odd flow control: xon If necessary change the terminal type by typing 'term'. Available types are: 'edt', or 'evdt', or 'vdt7270', or 'other'. 7. Save the new configuration by entering 'write' and responding with a 'y' to CPU-BASIC FOUR-2000---------------------------------FIB015 Pg001 the write prompt. 8. Exit the mode by typing 'debug'. 9. Power down the system, and then power up again. System should load using CMB port 0 as the console. After the Operating System is installed, re-configure the system ports using the utility programs. ORIGINATOR: J. VAN OORDT CPU-BASIC FOUR-2000---------------------------------FIB015 Pg002 FIB 00016 05/22/85 *** Second hard disk drive installation information *** Before installing the 2nd Hard disk in the system cabinet, verify the following: 1. Check the settings of the jumpers/switches on the second drive for the correct address: drive number 1. Refer to the Eagle Ref Card, page 10 (Mini- Alert 353). 2. Check if Dual cable received to install second drive is the correct type: o Systems with single board disk controllers require MM538250/907606-001 o Systems with 2-board disk controllers require MM538140/907644-001 3. Check Write Pre-compensation jumpers on disk controller. Refer to MSA 348. 4. Install 2nd. drive in the system cabinet and boot diagnostics. o Load 'dutil' and using the Read Only Utilities, display the defect map to verify if the drive was formatted correctly. o Re-formatting will be needed if the drive is unable to display the defect map. PROCEDURE TO CREATE ONE LARGE PARTITION ON SECOND DRIVE AFTER INITIAL INSTALLATION. NOTE: THIS PROCEDURE IS FOR TESTING PURPOSES ONLY. ANY PERMANENT CHANGES TO THE DISK PARTITIONS SHOULD BE DONE BY THE CUSTOMER OR THE SOFTWARE VENDOR. 1. Read superblock of second drive (drive 1) to verify its validity: ADMIN> usb /dev/rwd1 2. Write drive 1 parameters to file 'drive1' in directory '/etc/diskdesc': ADMIN> usb /dev/rwd1 desc=drive1 3. Display description file 'drive1': ADMIN> ved /etc/diskdesc/drive1 THIS IS A SAMPLE OF A DESCRIPTION FILE FROM A MAXTOR 140MB: ** volume id BASIC FOUR INFORMATION SYSTEMS DIVISION M A I 2000: Formatted by DUTIL: ** class : hds : sects : bytes/sect : # cyls : blocks : rwc : wpc : defects 0:15:17:512:918:233580:919:919:1: ** partition # : starting block : length in blocks 4. Modify partitions as needed. Only partitions 0 and 1 are required. Delete any other partitions. Use CTRL n to move cursor down to line desired. Use CTRL p to move cursor up to line desired. Use CTRL k to delete line desired. Use CTRL d to delete character desired. 5. Add or change partition 1 as follows: 1:0:-1: Delete any other partitions. THIS IS A SAMPLE OF THE MODIFIED DESCRIPTION FILE: ** volume id BASIC FOUR INFORMATION SYSTEMS DIVISION M A I 2000: Formatted by DUTIL: CPU-BASIC FOUR-2000---------------------------------FIB016 Pg001 ** class : hds : sects : bytes/sect : # cyls : blocks : rwc : wpc : defects 0:15:17:512:918:233580:919:919:1: ** partition # : starting block : length in blocks 1:0:-1: 6. Press 'CTRL c' to end and answer 'Y' to put edits. 7. Re-write superblock using modified parameters from file: '/etc/diskdesc/drive1' ADMIN> usb /dev/rwd1 type=drive1 -save -set 8. Make devices for 2nd. drive on '/dev' directory. ADMIN> makedev /dev/rdr1 u 14 33 ADMIN> makedev /dev/dr1 b 14 33 9. Make filesystem on newly defined devices: ADMIN> makefs /dev/rdr1 10.The file system on the second drive may be accessed by mounting it to the root partition: ADMIN> mount /dev/dr1 /mnt ORIGINATOR: J. VAN OORDT CPU-BASIC FOUR-2000---------------------------------FIB016 Pg002 FIB 00017 08/02/85 *** False memory parity errors caused by below LARL CMB PCBA *** Systems Affected: Systems with new style CMB board, MM531011/903441-002, at revision AN or below. The CMB board revision level should be checked in systems that have unexplained dumps or crashes, or when the Error Log shows Parity Errors in different memory locations. This problem only affects CMB P/N MM531011/903441-002 (the new style CMB). The old style CMB, P/N MM531010/903441-001 is not affected. If the CMB P/N MM531011/903441-002 is revision AN or lower, order a new board from Sorbus West. All affected CMB boards at Sorbus West are being upgraded to revision AP or above. ORIGINATOR: J. VAN OORDT CPU-BASIC FOUR-2000---------------------------------FIB017 Pg001 FIB 00018 10/22/85 *** Hard disk controller jumpering information *** SUBJECT: HARD DISK CONTROLLER JUMPERS (WRITE PRECOMP) The Hard Disk Controller for the System 2000 contains jumpers to set write pre- compensation according to the drives being used. The jumpers allow write precompensation to be set as follows: o always off o always on o same cylinder as reduced write current Disk drives specifications for drives used with System 2000: Drive type: Type of write precompensation recommended: 20MB Rodime always on 50MB Micropolis same cylinder as reduced write current 140MB Maxtor always off Two different types of Hard Disk controllers are available for the System 2000, the original controller is a 2-board controller set, the new controller is a single-board controller. Their settings are indicated below, use the vendor P/N for identification. 2-BOARD HARD DISK CONTROLLER - One BFISD board and one Adaptec board. (Current production). Jumper is located on the Adaptec board, P/N 907649-001/MM531050, near connector J1. This one jumper sets both drives. If the system has only one drive or two drives of the same type, the jumper is set as follows: Drive type: Jumper: Write precompensation: 20MB Rodime R to T always on 50MB Micropolis R to S same cylinder as reduced write current 140MB Maxtor R to U always off Note: Jumper position R-U is equivalent to no jumper installed. If the system has two drives of different type, the setting of the jumper is determined as follows: Rules: Perform one step only, the first that applies. 1. If any of the drives is a 140MB Maxtor, set jumper to R-U, and do not perform steps 2 and 3. 2. If any of the drives is a 50MB Micropolis, set jumper to R-S, and do not perform step 3. 3. If the drives are both 20MB Rodime, set jumper to R-T. Note: Jumper position R-U is equivalent to no jumper installed. The Adaptec board also contains jumper block J5. Verify that all jumpering positions are open. 1-BOARD HARD DISK CONTROLLER - No separate Adaptec board. (Future production). P/N 903496-001, MM531150. Jumpers are located next to connector J3. Each drive is set independently by a separate jumper. Jumper B is for drive 0 Jumper C is for drive 1 CPU-BASIC FOUR-2000---------------------------------FIB018 Pg001 Set Jumper B or C according to drive type. Drive type: Jumper: Write Precompensation: 140MB Maxtor 1A to 2A always off 20MB Rodime 1B to 2B always on 50MB Micropolis 1C to 2C same cylinder as reduced write current The single board controller contains another jumper (Jumper A) to set the board address. Verify that jumper A is set at 1-2. ORIGINATOR: J. VAN OORDT CPU-BASIC FOUR-2000---------------------------------FIB018 Pg002 FIB 00019 12/05/85 *** ERROR -2 and ERROR -62 during backup to the MCS tape drive *** PROBLEM: Customer gets ERROR -2 and ERROR -62 while backing-up the hard disk to the Magnetic Cartridge Streamer. SYMPTOMS: o Problem is highly intermittent. o Usually fails in long back-ups of more than 400 files. o Problem persists after replacing the MCS and the MCS controller. SOLUTION: Two work-arounds are suggested by Basic Four to circumvent this problem until a permanent solution is found: 1. Do a 'filesystem' backup instead of a file-by-file backup. Filesystem backups are faster than file-by-file backups but individual files can not be restored from the backup tape. When a restore is required, the whole 'filesystem' must be restored from tape. Example of how to do a 'filesystem' backup of the root partition: o Root to the boot partition o use filesystem option on 'mcssave' and 'mcscompare' as follows: ADMIN> mcssave name=backup -f -v /dev/root ADMIN> mcscompare filesystem=/dev/root backup 2. Do multiple small file-by-file backups of less than 400 files each. When using multiple small backups, the customer should be able to restore individual files from the backup tape but the customer has to know the saveset where the file was saved. Multiple small backups may be done by writing a command file that executes the backups automatically cancelling the need to enter a command for each backup. Multiple small backups should not use any more tape than a single backup if the '-append' option is used. Example of how to do multiple small backups: o create command file 'mcs_backup' using 'ved': ADMIN> ved mcs_backup o enter commands in command file 'mcs_backup' mcssave name=backup1 -v -r /bin mcssave name=backup2 -v -r /etc -a mcssave name=backup3 -v -r /sys -a mcscompare backup1 backup2 backup3 o execute command file 'mcs_backup': ADMIN> command mcs_backup ORIGINATOR: J. VAN OORDT CPU-BASIC FOUR-2000---------------------------------FIB019 Pg001 FIB 00020 12/26/85 *** Various problems when the console terminal is set-up on a four-way *** Problem: When the console terminal is on the first port of the four way . controller (fw00), intermittently, there is a problem in doing CTL C, CTL S, CTL Z, or 'CR' to re-boot. Solution: Basic Four knows about the problem, an interm solution is to put the console terminal on sc0, the first port on the CMB. All of the functions will work then. ORIGINATOR: M. PEPPER CPU-BASIC FOUR-2000---------------------------------FIB020 Pg001 FIB 00021 12/26/85 *** Instructions for formatting the 85MB Maxtor disk *** Problem: The current level of the diagnostic media (DIA-26.2) shipped with the System 2000 does not contain an entry in the formatter program to format the 85MB Maxtor hard disk drive. Solution: To format the 85MB Maxtor, select the 'Micropolis 85 Megabyte' entry on the formatter menu. The 85MB Maxtor and the 85MB Micropolis have the same number of cylinders and heads, therefore the formatter program uses the same parameters to format either drive. ORIGINATOR: J. VAN OORDT CPU-BASIC FOUR-2000---------------------------------FIB021 Pg001 FIB 00022 01/30/86 *** ERROR 0's on non spooled serial printers *** Problem: The task running a serial printer reports ERROR=0 during a print job without apparent reason. Symptoms: The problem occurs in multi-user mode when a terminal is running a print job and another terminal signs-off. The terminal running the print job gets an ERROR=0. If the directive 'ERROR' is typed at this point, the system returns File System Error Code: -62. The problem only occurs when the printer is configured as a non- spooled printer. Solution: The following work-around is suggested until Basic Four corrects the problem on the next Operating System release. o Use the Utilities to configure the printer with spooling. o Use the option OPTS="-off" in the OPEN statement on print jobs when spooling is not desired. Example: 10 OPEN (1,OPTS="-off")"LP" ORIGINATOR: J. VAN OORDT CPU-BASIC FOUR-2000---------------------------------FIB022 Pg001 FIB 00023 02/26/86 *** Intermittent failures of 53 MB and 20 MB disk drives *** PROBLEM: Intermittent read failures Usually fails when the drive is cold in either of the following manners: 1. When booting the operating system from hard disk it fails to read the loader and the system jumps to DEBUG. 2. After reading the loader, the system partially loads the operating system and then hangs. SOLUTION: The Rodime drives should be formatted during initial system installation and when a drive is installed as an add on or replacement. The Rodime drives, due to the absence of a servo track are more susceptable to changes on head positioning may be affected by temperature variations and jarring during transportation. ORIGINATOR: J VAN OORDT CPU-BASIC FOUR-2000---------------------------------FIB023 Pg001 FIB 00024 02/24/89 *** Installation instructions for 1/2" streamer (MTS) *** Basic Four is selling a 1/2" Streamer Tape Option for the System 2000. This option enables the System 2000 to use the Model 4402 Cipher 1/2" Streamer Tape Drive, the same drive used with other Basic Four systems, to do backups and restores of the hard disk. The tape drive may also be used to transfer data files to other Basic Four systems. The following items should be available at installation time: o Formatter Controller box, MBF Model 4602. It contains the formatter and the power supply. o SCSI Bus Adapter Board P/N 903528-001 to be installed inside the 2000 system. o Interface cable P/N 906608-017, from 2000 Bus Adapter to Controller Box. o Cipher 1/2" Streamer Tape Drive, MBF Model 4402. o Operating System Level 7.2A*27 or above. o Diagnostics Media Level DIA 7204 or above. HARDWARE INSTALLATION: 1. Open the controller box by loosenig 1 screw on each side and sliding the tray towards the front. Inspect formatter controller board (NCR ASSY 348-0014807A) to verify switches and jumpers settings: o Dip SW. 1 at location 7W: all positions should be set to CLOSED. o Jumpers W22, W23, W24 at location 6J must be OPEN. o Jumper J16 at location 4G should be OPEN. o All other jumpers on the board should be ON. Jumpers are numbered from 1 to 33, position 12 is not used. o Terminator Packs in place at locations 6H and 7B. Test the operation of the controller by running the self-tests: Controller self-tests are started when the power switch is turn ON in the controller box. Self-tests may be re-started by pressing the Reset Sw. on the formatter board. There are 8 Leds marked from '7' (MSB) thru '0'. The upper half should always display a '9' (bits 7 and 4). The lower half changes with the test being run. The LEDs should display: '9F' while doing Data Buffer Test. '90' at the end if everything worked OK. 2. Position the controller box within 6 feet of the System 2000. Allow 4 inches. clearance in the back for proper air ventilation. 3. Place the Streamer Tape Drive in top of controller box, 2 persons are needed. The 4 rubber stand-offs should fit in the 4 holes in the controller box. 4. Before attaching the 2 ribbon cables from the controller box to the tape drive check the position and orientation of the cables in relation to the tape drive. The cables may have to be twisted, on the tape drive end, in order to connect the cables as follows: J1 on the formatter board should connect to P1 on the tape drive. J2 on the formatter board should connect to P2 on the tape drive. On both cables, make sure that pin 1 in one end is connected to pin 1 in CPU-BASIC FOUR-2000---------------------------------FIB024 Pg001 the other end. NOTE: WHEN PROPERLY CONNECTED THE CABLES WILL BE CRISSCROSSED AND TWISTED, IF THE STREAMER DOES NOT WORK AFTER INSTALLATION RECHECK FORMATTER TO TAPE DRIVE CABLING! 5. Install SCSI Bus Adapter, P/N 903528-001 in the system 2000 controller stack. Verify that on dip switch SW. 1, all positions are set to ON. 6. Connect cable P/N 906608-017 from the 2000 Bus Adapter to the controller box. The end marked "PA" should be attached to the Bus Adapter. Viewing both units from the rear, pin 1 is located as follows: On the 2000 Bus Adapter board, pin 1 is on the right hand side. On the Controller Box, pin 1 is on the left hand side. 7. Set dip switch at location 5W or 8W on the drive formatter board inside the streamer drive (there is only one dip switch on the board but may be in either location) positions 1, 2, 4, 5 closed , the rest open (drive 0). This will parity check the data into the streamer drive. TESTING THE OPERATION: 1..Turn power on to the controller box, the tape drive and the system 2000. The system 2000 console will not print a test result for the Bus Adapter during the self-tests because the Bus Adapter board itself does not perform self- tests. 2. Load a reel of tape in the tape drive with the write enable ring in place. After the tape is successfully loaded, the LOAD and the WRTEN lights should be on. Press the ON LINE switch to place the drive On Line. 3. Using Diagnostics to test tape drive operation: Boot diagnostic media level DIA 7204 or above. Load 'mts'. Enter 'quick' after the prompt to run the tests in quick mode. The 'mts' tests in quick mode takes about 8 minutes to run. It should run with no errors. Tests 1 to 11 check the Bus Adapter board by itself, the rest of the tests require a ready tape drive and controller box. 4. Using the command language to test the tape drive with the Operating System: The following commands are included in 7.2A*27 release: mtslabel, mtssave, mtsrestore, mtscompare. These commands operate in a similar manner to the ones used with the 1/4" mcs. Use 'mtslabel' and 'mtssave' to write to tape, and then use 'mtscompare' to compare the tape to the disk. PARTS LIST: The following parts are available as spares: MM532010 903528-001 PCBA, SCSI BUS ADAPTER MM532020 400591-001 PCBA, 1/2 STREAMER FORMATTER MM532030 906608-017 CABLE, ADAPTER TO FORMATTER MM550140 400446-002 POWER SUPPLY, FORMATTER MM532070 907846-001 CABLE, FORMATTER TO MTS, RH MM532080 907846-002 CABLE, FORMATTER TO MTS, LH MM560010 907174-030 CABLE, FORMATTER TO PANEL MM532130 345019-001 FAN, FORMATTER CONTROLLER BOX ORIGINATOR: JAVIER VAN OORDT CPU-BASIC FOUR-2000---------------------------------FIB024 Pg002 Modifed by N. Jones CPU-BASIC FOUR-2000---------------------------------FIB024 Pg003 FIB 00025 03/26/86 *** Corrupted data on MCS backup *** Customer MCS backups must be done in Single User Mode, or the backup may contain corrupted data files. Files that have been OPENed and had information in them updated, may be saved to tape as corrupted files. This is because the Header information of a file is not usually updated until that file has been CLOSEd. Single User mode is recommended, so as to prevent users from logging on and processing data while the backup is in progress. 1. To go into Single User Mode from Multi-User Mode, type in the following under ADMIN: ADMIN>shutdown -- -1 -s 2. The customer can now follow his standard Backup Procedure. If his procedure requires him to be in Multi-User Mode, that is to enter an Account Name to proceed with the backup, the procedure will have to be altered to run in Single User Mode. 3. To go back to Multi-User Mode, depress CTRL-D under ADMIN and enter 'multi' at the prompt: single, multi, or shutdown? multi Reference: CTS No. 8613 ORIGINATOR: J. VAN OORDT CPU-BASIC FOUR-2000---------------------------------FIB025 Pg001 FIB 00026 03/26/86 *** Won't load operating system, system hangs *** Problem: Operating System fails to load from hard disk. Symptoms: Usually fails when the disk drive is cold. Solution: Perform the following steps to decrease the occurrence of boot problems: 1. Install new PROMs in the CMB board as described below. 2. Install new PROMs in the Hard Disk Controller as described below. 3. Reformat all Rodime disk drives during initial installation. 4. On Maxtor and Micropolis disk drives perform a disk drive read test to decide if reformatting is required during initial installation. Use the Read Only Utility Option 2 of 'dutil' in the Diagnostics. If any read error occurs, reformat the drive. NOTE: When formatting a hard disk always enter all media defects from the manufacturer's list attached to the drive and the defects read from the drive's defect map. 5. If the loading problem persists, ask the local Basic Four Branch to install the 7.2 loader on the 7.1 Operating System. They may have to contact Marketing Support in Tustin to obtain the loader. CMB PROM UPDATE _______________ New Firmware PROMs have been released for the CMB board. The new firmware contains several fixes and improvements. It will be installed on all new systems manufactured by Basic Four after March 17, 1986. The new firmware contains corrections to fix the 2 following problems: 1. Failure to boot the Operating System from hard disk. This is an intermittent problem that usually happens when the drive is cold. This problem only occurs with some drives, it is more prevalent with the Rodime drives. The new firmware fixes only the booting problem that causes a drop to 'debug' immediately after printing "booting from wd0". Other problems to load, after reading sector 0, are not fixed by this change. 2. Unable to reload the Operating System when the console terminal is attached to the 4-Way Port. The system does not respond when pressing 'RETURN' to reboot. CHANGE IS NOT MANDATORY - INSTALL AS REQUIRED ONLY. This change is not mandatory in the field. The PROMs should be ordered and installed ONLY in systems that exhibit any of the 2 problems indicated. The new firmware will be installed on all the boards at Sorbus Materials West and the Repair Center. HOW TO DETERMINE THE REVISION LEVEL OF THE PROMS ON THE CMB. CPU-BASIC FOUR-2000---------------------------------FIB026 Pg001 The revision level of the CMB PROMs can be determined without removing the covers by running the self-tests and examining the 3rd. line printed on the screen. Look at the area just before the SSN: If B2.9 is printed, old PROMs are installed on the CMB. If B3.1 is printed, the new PROMs are already installed on the CMB. To perform a visual check, remove the system top cover and examine the PROMs at location 6N and 6S: Old PROMs New PROMs PROM at location 6N 165047-036 165047-051 PROM at location 6S 165047-037 165047-052 PARTS ORDERING INFORMATION: Preferred solution: Order the following 2 PROMs and install them on CMB board. YY017500 165047-051 IC PROM, 3.1 CMB 6N YY017501 165047-052 IC PROM, 3.1 CMB 6S Alternate solution: Order new CMB board P/N MM531010 at rev. AU HARD DISK CONTROLLER FIRMWARE UPDATE ____________________________________ Inspect PCBA PROM at location indicated below. If PROM is vendor 400108-x or 400510-x (any revision), or 165053-001, remove and replace with PROM YY017502, . MBF #165053-002, Vendor #410041-00 Revision A or 2A. ICN MBF Part Number Vendor Part Number PROM Loc MM531050 907649-001 ACB4000 U12 MM531050 907649-001 ACB4000A 9D MM531150 903496-001 5P NOTE: Return removed PROM to Materials West for reprogramming. ORIGINATOR: J. RIPPL CPU-BASIC FOUR-2000---------------------------------FIB026 Pg002 FIB 00027 05/07/86 *** Hard disk controller revision required for OS Level 7.2A *** The new 2000 Operating System level 7.2A requires an upgrade on the Hard Disk Controller. This upgrade is necessary to provide errorless Overlap Seeks on systems with Dual Drives. The change consists of replacing IC PAL at location 3M on the Single Board Hard Disk Controller. Overlap Seeks enable the system to execute seeks on both disk drives at the same time. As a result of this, disk performance is improved. The change only affects the Single Board Hard Disk Controller. The Dual Board Hard Disk Controller, used with systems shipped before May 1985, is NOT affected . The Dual Board Hard Disk Controller is unable to perform Overlap Seeks. If the customer desires to upgrade to the Single Board Hard Disk Controller, he should contact the local Basic Four Branch to order Market Code DN012AA according to Marketing Announcement No. 360. CHANGE IS MANDATORY IN THE FIELD - INSTALL ON ALL SYSTEMS 2000 USING SINGLE BOARD HARD DISK CONTROLLER. DETERMINING IF THE CHANGE IS REQUIRED: Part affected: P/N MM531150/903496-001 PCBA, Single Board WD Controller Change is already installed if the IC PAL at location 3M on the Hard Disk Controller is Basic Four P/N 911017-004 (or Sorbus P/N YY017543). The change is needed if the IC PAL is P/N 911017-001 or 911017-002. If change is required, order parts as follows: YY017543 911017-004 IC, PAL WDC-Location 3M The change will be installed on all the boards at Sorbus Materials West and at . the Repair Center. Upgraded boards will be marked with Revision J. ORIGINATOR: J. VAN OORDT CPU-BASIC FOUR-2000---------------------------------FIB027 Pg001 FIB 00028 07/16/86 *** Detailed info concerning hard disk failing sector determination *** Read errors caused by defective media, that occur during customer operation, are difficult to isolate to a specific sector because the sector number causing the error may not be printed along with the error. Read errors may be caused by marginal sectors that were not detected when the drive was formatted or by marginal sectors that became more critical after customer use. When a read error occurs while running the operating system, the system prints ERROR -2. If the error occurred while running a Basic application program, Basic prints ERROR= 3. In either case the number of the block (sector) causing the error is not printed on the screen. The Error Logging Facility in the System 2000 records all errors in the Error Log file if the system was operating in Multi-User mode. The System Error Log should be examined to determine the sector (block) causing the error. To display the Error Log, type the following after the ADMIN prompt: ADMIN> /sys/errlog /etc/error.log _ p In a Winchester disk error, the following fields are significant to find the sector number: o Field 'Minor' indicates the partition number where the error occurred, and the partition number points to the disk drive: Drive 0 contains partitions 0 to 31. Drive 1 contains partitions 32 to 63. o Field 'A' contains the number (in decimal) of the starting block of the data to be transferred. o Field 'B' contains the number of blocks to be transferred. Field B must contain a 1 to isolate the defective sector. If Field B contains a value greater than 1, a read test must be performed on the suspected area to find the defective sector. Refer to Error Logging Manual BFISD 8279P (Sorbus S-HDBK-B68) to interpret the other fields of the error log. READ TEST USING THE OPERATING SYSTEM. If the error entry in the Error Log was for a multi-sector transfer, proceed as follows to perform a read test forcing the system to make an entry on the Error Log while reading one sector at a time. NOTE: The system must be in multi-user to log errors in the error file. ADMIN>debe if=/dev/wd0 of=/dev/null bs=1b -noerror -pg skip=xxxxb count=yyyy Note: Replace xxxx with the contents of Field A. Replace yyyy with the contents of Field B. o the above command will read the number of sectors specified by Field B on wd0 . (1st.drive) starting with the sector indicated by Field A. If an error occurs CPU-BASIC FOUR-2000---------------------------------FIB028 Pg001 an entry will be made on the error log. To check the second drive use '/dev/wd1' Errors encountered during the above read test should cause entries in the Error Log. These new error entries should have Field B equal to 1, isolating the error to one block. Sectors causing repetitive read errors, found thru the Error Log, should be added to the defective map when formatting the drive. STEPS REQUIRED TO TRANSLATE THE LOGICAL SECTOR NUMBER TO PHYSICAL LOCATION. To convert the logical block number to physical location (Cylinder, Head and Sector) two steps have to be taken: STEP 1. CONVERT THE DECIMAL NUMBER OBTAINED FROM THE ERROR LOG TO HEXADECIMAL. 1. Boot the diagnostics and load 'dutil' 2. Obtain block number from the error log (Decimal number). 3. To convert block number from decimal to hex. Enter after the prompt: dec xxxxxx ;where xxxxxx equals the block number in decimal. The program will print the block number in hex. Write down this number. STEP 2. TRANSLATE THE LOGICAL BLOCK NUMBER (IN HEX) TO PHYSICAL SECTOR: 1. Enter after the prompt: unit X ;to select unit desired, X = 0 or 1 block yyyyyy ;where yyyyyy equals the block number in hex 2 The program will read the block and print it contents on the screen. The first two lines contain the logical sector number and the physical location in hex, as follows: SECT: xxxxxxxx EXP CYL: xx HEAD: xx B.F.I. : xxxx Write down this information which will be used when formatting the drive. If there is an error in the ID field of the targeted sector, the program will not print the physical location. It will be necessary to read the blocks before and after the targeted sector to determine the location of the bad sector. Use the attached worksheet to determine the physical location. READ TEST USING THE DIAGNOSTICS. An alternate method of determining bad sectors is to use the 'read no compare' option of the Read Only Utilities of 'dutil'. o boot diagnostics and load 'dutil'. o select 'Read Only Utilities' and then 'Read No Compare' option. o select disk drive Type and drive number. The utility will read all the sectors printing the errors as it goes along. At the end it will print all errors encountered in a Bad Sector Table which MUST be translated since it is in Hex and may contain inaccurate sector information. To translate the Bad Sector Table proceed as follows: o Press ESC after the Bad Sector Table is printed. o after 'dutil' prompt type: dutil> convert The program will convert the Bad Sector Table to decimal. This new table should CPU-BASIC FOUR-2000---------------------------------FIB028 Pg002 contain accurate CYL, HEAD and BFI information. MAPPING DEFECTIVE SECTORS ON THE HARD DISK The following sequence is recommended at formatting time to map all marginal sectors: IMPORTANT: Always print the existing defect map before re-formatting to keep a record of defective sectors. Use the 'Display the defect map' option on the Read Only Utilities of 'dutil'. Select either of the two following formatting sequences according to time available, the first sequence performs a surface analysis in addition to formatting. 1. Long formatting time (example: Maxtor 140MB takes 2.5 hours). o Use Format, Surface Analysis and Defect Mapping option. o Enter all manufacturers media defects from the label attached to the drive o Add all sectors from the existing defect map read from the drive. . o Add the new defects encountered during customer operation. The new defects data may be in hex. To enter hex data precede entries with a 'h'. 2. Short formatting time (example: Maxtor 140MB takes 10 minutes). o Use 'Format, No Surface Analysis, Defect Mapping' option. o Enter all manufacturers media defects from the label attached to the drive . o Add the new defects encountered during customer operation. The new defects data may be in hex. To enter hex data precede entries with a 'h'. WORKSHEET TO DETERMINE THE PHYSICAL LOCATION OF A DEFECTIVE SECTOR. 1. Enter all defective blocks from the Error Log under Block (Dec) in Table 2. 2. Convert blocks numbers from decimal to Hex. 3. Read defective block using BLOCK command enteringH4y1lock number in HEX. 4. If the program returns: EXP CYL: xx HEAD: xx B.F.I.: xxxx , enter these values in Table 2. 5. If the program does not return the EXP CYL, HEAD & B.F.I. while reading the . targeted sector, read the previous sector, and then the next sector. 6. Using the B.F.I. values reported while reading the previous and next block enter HD and CYL on the proper lines in TABLE 1. B.F.I. determines in which of the 17 sectors of the track is the block located. 7. The defective sector is located between the two sectors read successfully. 8. Use the information recorded in Table 2 when formatting the drive. TABLE 1 TRACK n TRACK n + 1 (Next Head) BLOCK CYL HD BFI* BFI** SEC BLOCK CYL HD BFI* BFI** SEC HEX HEX HEX HEX HEX HEX HEX HEX HEX HEX HEX HEX _____ ___ ___ 0096 (0096) 00 _____ ___ ___ 0096 (0096) 00 _____ ___ ___ 02D6 (O2D8) 01 _____ ___ ___ 02D6 (O2D8) 01 _____ ___ ___ 0516 (051A) 02 _____ ___ ___ 0516 (051A) 02 _____ ___ ___ 0756 (075C) 03 _____ ___ ___ 0756 (075C) 03 _____ ___ ___ 0996 (099E) 04 _____ ___ ___ 0996 (099E) 04 _____ ___ ___ 0BD6 (0BE0) 05 _____ ___ ___ 0BD6 (0BE0) 05 _____ ___ ___ 0E16 (0E22) 06 _____ ___ ___ 0E16 (0E22) 06 _____ ___ ___ 1056 (1064) 07 _____ ___ ___ 1056 (1064) 07 _____ ___ ___ 1296 (12A6) 08 _____ ___ ___ 1296 (12A6) 08 _____ ___ ___ 14D6 (14E8) 09 _____ ___ ___ 14D6 (14E8) 09 _____ ___ ___ 1716 (172A) 0A _____ ___ ___ 1716 (172A) 0A _____ ___ ___ 1956 (196C) 0B _____ ___ ___ 1956 (196C) 0B _____ ___ ___ 1B96 (1BAE) 0C _____ ___ ___ 1B96 (1BAE) 0C _____ ___ ___ 1DD6 (1DF0) 0D _____ ___ ___ 1DD6 (1DF0) 0D CPU-BASIC FOUR-2000---------------------------------FIB028 Pg003 _____ ___ ___ 2016 (2032) 0E _____ ___ ___ 2016 (2032) 0E _____ ___ ___ 2256 (2274) 0F _____ ___ ___ 2256 (2274) 0F _____ ___ ___ 2496 (24B6) 10 _____ ___ ___ 2496 (24B6) 10 * B.F.I. when using Single-board WD Controller. * B.F.I. when using 2-board WD Controller >@$HH TABLE 2 BLOCK (DEC) BLOCK (HEX) CYL (HEX) HD (HEX) B.F.I (HEX) __________ __________ _________ ________ _________ __________ __________ _________ ________ _________ __________ __________ _________ ________ _________ __________ __________ _________ ________ _________ __________ __________ _________ ________ _________ __________ __________ _________ ________ _________ ORIGINATOR: J. VAN OORDT CPU-BASIC FOUR-2000---------------------------------FIB028 Pg004 FIB 00029 08/07/86 *** Sys 2000 Flaw Maps, DUTIL-Drive Formatting *** The necessity for keeping and maintaining an up-to-date flaw map for the WDD can not be overemphasized. This holds especially true with possible adaptec firmware problems, where blown formats may necessitate reformatting of the disk unit(s). Some systems may have been shipped without accompanying flaw maps for the disk drives (prior to 3/86). If you have such a system and your unit does not now have a current flaw map, make one at your earliest convience. It is up to you to make sure it can be used as a viable tool. A good format and surface analysis is absolutely essential for reliable operation of the disk unit. The following is suggested - - every disk should have a flaw map. It should include the MFG. list as well as the flaw map printed by the dutil program. - never format a drive unless necessary. You can determine this by running the disk diagnostics, concentrating on the seek tests and using the read- only option. Run the read-only verify option with ecc off to detect possible soft errors. (The error log should also be looked at to determine weak areas of the disk). - If the disk's flaw map is inadvertantly lost, a printout of the defect list can still be obtained by selecting printer command option sp or option lp while running under dutil, and then displaying the defect list. - If the drive has been formatted before and the service rep wishes to format, always use option #5 (update defect map). use option #$4 (format, no surface analysis, defect mapping) only if update defect map fails. Below are listed the various dutil options available; recommended procedures for using them and a concise explanation of what they do. OPTION: 1. Format, surface analysis, defect mapping - use this option to first format a disk unit as if there were no defects, then perform a surface analysis with hardware ecc and retries disabled. After surface analysis, the MFG defects are entered by the user. The surface analysis results are merged with the MFG defects to form a more complete list. (Use this option when no defects are known to exist, or when the defect list is lost). NOTE: Option 1 is the only formatting option which will do 2 formats on the disk; the first without defects, the second with the defect list just created. 2. Format only, no defect mapping - this is the quick format. use this option when no defects are desired, or none are to be mapped out. 3. Surface analysis, report results - surface analysis specifically for use after formatting. This analysis will destroy data on the disk. Use this option after the last formatting option is complete to ensure no further soft errors exist. If defects are found by this option, rerun option #5 to add new defects to flaw map. 4. Format, no surface analysis, defect mapping - this format option is intended to permit the user to enter known defects. This option does not use a previously saved defect list, it is intended primarily to manually CPU-BASIC FOUR-2000---------------------------------FIB029 Pg001 enter defects. Option #3 should normally be run after this to ensure that no additional defects exist. 5. Updating defect map - this option is used to add a new defect to the existing defect list created by a previous format. Option #3 should have been run prior to this option. Error log information saved by the o/s is also an excellent source of intermittant bad spots. This option is the only one that uses the saved defect list. Use this option to reformat the disk unit with the old (original) defect list. To do this, simply carriage return when prompted for defects. If no saved defect list exists and no defects are entered, this option informs the user and no formatting will take place. 6. Change disk label - this option only changes the disk label in the superblock. The re-write operation will either take place immediately or during the next formatting process, at the user's option. If the choice is immediate, dutil will create a default label; if the choice is during formatting, the label can be user-defined. 7. Read maxtor defects from disk - only for maxtor disk units which have never been formatted under dutil. Defect list written on the drive by the mfg may be optionally re-written for later dutil use by the user. This eliminates the necessity to manually enter all defects. Do not select this option if the drive has been previously formatted by dutil. To check, use the read-only utilities menu and read defect list. To determine what to do and when, use the following parameters - CONDITION ACTION Maxtor drive never before formatted. option 7 Non-maxtor drive never before formatted and option 4, then option 3 defects are known. Non-maxtor drive never before formatted and option 1 no defects are expected or list is lost. Drive has been formatted before. option 5, then option 3 When defect list on the disk is corrupted. option 4 (enter all defects), then option 3, then option 5 if new defects are found Add a new defect list to an existing list. option 5 Reformat using existing list as a normal option 5 (enter C/R to maintenance procedure. "defect entry" prompt in order to use old list), then option 3, then option 5 if new defects are located To test format and data reliability option 1, read-only utilities The suggested number of surface analysis patterns for use is - as many as time will allow... the more patterns selected, the more thorough the analysis, and the more reliably the unit will operate. ORIGINATOR: P. KOEHLER CPU-BASIC FOUR-2000---------------------------------FIB029 Pg002 FIB 00030 01/13/87 *** System slowdowns/hangs won't load - caused by unterminated cables *** SYMPTOM: System slowdowns/hangs, failure to load, some diagnostics run slow. PROBLEM DETERMINATION: Check for any RS-232 cables connected to the CPU which are not connected to a serial device. FIX: If at all possible disconnect all unterminated cables. See F.I.B. in the Basic Four Cables section titled "RS-232 PERIPHERAL CABLE TERMINATOR" for details on proper cable termination. ORIGINATOR: N. JONES CPU-BASIC FOUR-2000---------------------------------FIB030 Pg001 FIB 00031 03/09/87 *** Various disk start-up problems on power-up sequence *** SYMPTOM: Disk drives failing to come ready on power-up sequence. PROBLEM DETERMINATION: Check revision level of single board WD disk controller PCBA (MM531150,903496-001), must be REV. R or above. FIX: Replace WD disk controller PCBA with a revision level R or above ,or perform one cut and one jumper as follows: CUT: 5J-5 on solder side JUMPER: 5H-8 to 3N-2 on component side ****NOTE**** If re-work is to be done, PCBA should at least be at revision level "N". revision "N", requires new firmware PROM (MM531051, 165053-005) at IC location 5-P. ORIGINATOR: J.O'BRIEN CPU-BASIC FOUR-2000---------------------------------FIB031 Pg001 FIB 00032 03/31/87 *** Missing defect maps on MAXTOR drives *** Surface Defects detected by the manufacturer must be entered at formatting time to prevent read errors during customer operation. Defect Maps on Maxtor Hard Disk drives are supplied in two ways: on a label attached permanently to the drive, or in a sheet of paper shipped with the drive. This sheet is sometimes lost during shipment, installation, etc. To obtain the manufacturer defect list when the sheet of paper containing the manufacturers defect list is missing, proceed as follows: 1. Take note of the Maxtor Model number, and serial number on the Maxtor ID label located at the rear of the drive. 2. Call Basic Four Customer Satisfaction, and ask for WDD Flaw Map for the drive, supply them with the model and serial number. Telephone numbers: Outside California: (800) 848-0600 Inside California: (714) 730-3387 ORIGINATOR: J. VAN OORDT CPU-BASIC FOUR-2000---------------------------------FIB032 Pg001 FIB 00033 04/06/87 *** 1488 & 1489 socket locations for I/O ports on CMB & 4-way PCBA'S *** On the Basic Four 2000 System all 1488 and 1489 (line driver & receiver) IC's have been socketed. The purpose of this change was for chip replacement, as opposed to PCBA replacement. The IC location for each particular I/O port on the CMB and 4-WAY PCBA is as follows: CMB PCBA - MM531011 (903441): GRID LOCATION GRID LOCATION 1488 1489 SERIAL PORT A (0) 1B 1C 1D SERIAL PORT B (1) 3B 3C 3D 4-WAY PCBA - MM531040 (903390): GRID LOCATION GRID LOCATION 1488 1489 SERIAL PORT A (0) 4C 2B 3C SERIAL PORT B (1) 4C 3B 5C 3C SERIAL PORT C (2) 5C 5B 6C 3C SERIAL PORT D (3) 6C 6B 3C NACS PART NUMBERS FOR 1488 & 1489 ARE: MM160171 - 1488 MM160172 - 1489 ORIGINATOR: J.O'BRIEN CPU-BASIC FOUR-2000---------------------------------FIB033 Pg001 FIB 00034 04/08/87 *** ERROR 31 when running ORIGIN on OS Rel 7.2A and 7.2B *** An error 31 can occur in ORIGIN 2.0A *21 on the 2000 levels 7.2A and 7.2B in program "READFILE". This is caused by a problem in BASIC with the CALL routine on OS levels 7.2A and 7.2B. The data area used by the variables that are passed in a CALL is not cleared when the variables are modified. The following program changes will alleviate the error 31's in the ORIGIN program 'READFILE'. LOAD "READFILE" o Modify statement 30: OLD 30 ENTER TYPE$, RETRN$ NEW 30 ENTER TYPE$, RTRN$ o Append to statement 40: OLD 40 RETRN$="" NEW 40 RETRN$="", RTRN$="" o Add Statement 9015: 9015 LET RTRN$=RETRN$ SAVE the proram This workaround will be implemented in the next release of ORIGIN. ORIGINATOR: T. TIANGCO CPU-BASIC FOUR-2000---------------------------------FIB034 Pg001 FIB 00035 04/22/87 *** New firmware for MTCS controller required for Tandberg MCS drive *** A new firmware for the MTCS controller is now available. This new firmware is needed to support the new MT 4406 tape drive. The rework on the board is done as follows: o Remove prom MBF part number 165047-044 location 4-N. o Replace with prom MBF part number 165047-086. o Change part number on the controller to: 903406-002. RESTRICTIONS: o DO NOT USE THE NEW FIRMWARE ON CONTROLLER 903406-002 IF USING O.S. LEVEL 7.2 C. o NEW FIRMWARE AND CONTROLLER 903406-002 WILL WORK WITH THE; MT 4403 & MT 4406. ORIGINATOR: C. FIGUEROA CPU-BASIC FOUR-2000---------------------------------FIB035 Pg001 FIB 00036 04/27/87 *** New Kernel PROM allows 1/2" MTS boot *** A new release of the boot prom (B4.2) for the MAI 2000 is now available. This new boot prom will add support to allow you to load from MTS on the 7.3A O.S. release. The notable changes are as follow: 1. The mnemonic for MTS boot is "ts". A. When doing an MTS boot, specify "ts" at the alt load prompt: "boot device" 2. After "Test Unit Ready" is complete the tape is rewound to BOT. 3. After tape rewind is complete the tape is spaced forward one file in-order to bypass the tape label. 4. When a sense command is issued as a result of a "Check Condition" the status for both tape and disk is now displayed in the same format. A. The error message has the following format: tscs-dd-ee-cc-ssssssss o Where "tscs" is the MTS check status,("wdcs" for the disk). o "dd" is the board number involved, 00 or 01. o "ee" is the ending status from the controller, (normally 02). o "cc" is the command code that received the error. o "ssssssss" is the four byte of sense information. 5. Corrected a bug where entering "Ctl-C" 19 times hung the system. 6. The memory self-test had a bug in the testing of memory that is now corrected. 7. Corrected "NVRP" error that incorrectly stated presence or absence of the security PAL. 8. Corrected a problem with the handling of DTR and RTS when an intelligent synchronous modem on a multi-drop network is present. 9. Added 1/2" tape boot support. The new boot proms are required for 7.3A O.S. release at sites which requires MTS boot. All other sites need this release for better bootstrap error information. The part number and location for these proms are: 1. B/4 PART NUMBER 165047-084 loc. 6N on CMB checksum 03BF 2. B/4 PART NUMBER 165047-085 loc. 6S on CMB checksum 631E CPU-BASIC FOUR-2000---------------------------------FIB036 Pg001 ORIGINATOR: C. FIGUEROA CPU-BASIC FOUR-2000---------------------------------FIB036 Pg002 FIB 00037 05/05/87 *** PCBAs below LARL may cause Winchester disk errors *** At revision 'AM' an ECN was installed on the 2000 system CMB PCBA which added a capacitor to the PFD controls. Basic Four engineering has made tests and proven that systems without this ECN, when subjected to AC electrical disturbances, will destroy data on the Winchester disk to the point that reformatting of the drive is required. There have been several ECNs applied to the 2000 system componants which greatly effect system reliability and performance. It is therefore highly recommended that all 2000 systems be brought up to LARL as soon as feasible in order to reduce downtime and maintenance costs. ORIGINATOR: H. MITCHELL CPU-BASIC FOUR-2000---------------------------------FIB037 Pg001 FIB 00038 06/05/87 *** File list gets corrupted on O.S. Releases 7.2 *** Problem: An existing filelist will become corrupt if it is modified without first being displayed. The problem exists in all levels of 7.2 O.S. but has been corrected in 7.3A. Workaround: After choosing the modify option and entering the filelist name, choose the display option. After the filelist is displayed, the add option can be used to add filenames and/or filelist name to the existing filelist without filelist corruption. ORIGINATOR: D. Wenzel CPU-BASIC FOUR-2000---------------------------------FIB038 Pg001 FIB 00039 09/14/87 *** System 2000/3000 Hard Disk Substitution *** Defective hard disks must be replaced with the same type of hard disk to avoid customer problems. 53MB Rodime and 50MB Micropolis hard disk drives are not interchangeable due to the differences in Formatted Capacity and Average Seek Time. * Do not replace 53MB Rodime (44MB formatted) with 50MB Micropolis (43MB formatted) due to the 1MB difference in capacity. Error -2 may occur when restoring a filesystem (saved from the Rodime) to the Micropolis. * Do not replace 50MB Micropolis with 53MB Rodime because of the Average Seek Time, the Rodime is slower than the Micropolis. The customer may complain of slower disk operation. * SPECIFICATIONS OF HARD DISKS USED WITH THE SYSTEM 2000/3000 MFG. FORMATTED NACS # OF # OF AVERAGE MANUFACTURER MODEL CAPACITY I.C.N. HEADS CYLs SEEK TIME ------------ ------ --------- ------ ----- ---- --------- 20MB Rodime RO 204 22MB TT524100 8 320 90 msec 40MB Rodime RO 203E 33MB MM534055 6 640 55 msec 53MB Rodime RO 204E 44MB MM534060 8 640 55 msec 50MB Micropolis 1304 43MB MM534020 6 830 30 msec 85MB Maxtor XT 1085 71MB MM534070 8 1024 28 msec 105MB Maxtor XT 1105 88MB MM534030 11 918 27 msec 140MB Maxtor XT 1140 120MB MM534040 15 918 27 msec 190MB Maxtor XT 2190 160MB MM534080 15 1224 30 msec ORIGINATOR: J. Van Oordt CPU-BASIC FOUR-2000---------------------------------FIB039 Pg001 FIB 00040 10/07/87 *** Service Report form for 2000 service calls (send to Prod. Support) *** 2000 SERVICE CALL REPORT FORM Date __________ Call no.__________ 1. S/N of the System _____________________________________________________ 2. Current board revision levels (LARL) DESCRIPTION LARL (installed) CMB (old style) MM531010 (903441-001) _____ CMB (new style) MM531011 (903441-002) _____ 256K memory MM531030 (903368-001) 1__ 2__ 3__ 4__ 5__ 6__ 4-Way conroller MM531040 (903390-001) 1___ 2___ 3___ Adaptec contr MM531050 (907649-001) _____ WDC Adapter MM531100 (903439-001) _____ Single bd. cnt. MM531150 (903496-001) _____ MCS controller MM531070 (903406-001) _____ Power Supply MM533000 (907472-001) _____ Lan controller MM531060 (903405-001) _____ 1/2" Bus Adapt MM532010 (903528-001) _____ 1/2" Stream Format MM532020 (400591-001) _____ 3. Is user installing and removing cables with power on system (Y/N) ___ 4. Description of problem reported by customer ___________________________ _______________________________________________________________________ _______________________________________________________________________ 5. What action taken to fix problem & is it fixed ________________________ _______________________________________________________________________ _______________________________________________________________________ 6. System on Table or Floor ______________________________________________ 7. Disk drive types ______________________________________________________ 8. How many users on system ______________________________________________ 9. Does customer power system on and leave on, or is system powered on & off throughout the day ________________________________________________ 10. What type periphal equipment __________________________________________ _______________________________________________________________________ 11. Enviromental condtion (Temp,Static,Clean,Carpetting)___________________ _______________________________________________________________________ 12. Humidity ? ____________________________________________________________ ORIGINATOR: Allan Mollyhorn CPU-BASIC FOUR-2000---------------------------------FIB040 Pg001 FIB 00041 10/07/87 *** Unable to load SSN tape succesfully *** SYMPTOM/TEXT: Upon installing the SSN PAL chip and selecting alt load from 'cs' and system file 'ssn', the system goes through the sequence of booting from the SSN tape but never requests the new serial number. The problem began with the revision upgrade of CMB PROMs at location 6N to 165047-051 from 165047-036 and location 6S to 165047-051 from 165047-037. As of this date, No permanant fix has been implemented. FIX: The problem appears to be some type of timing problem due to the new PROMs. By removing any disk controllers and disconnecting the DC power plugs from the. Winchester disk drives the ssn tape will work properly. ORIGINATOR: Allan Mollyhorn CPU-BASIC FOUR-2000---------------------------------FIB041 Pg001 FIB 00042 10/12/87 *** Command 'who' shows 'admin' logged onto unused terminal(s) *** A condition has been reported from the field, that just after system boot the 'who' command and 'ps' command shows 'admin' logged in at another terminal or terminals. These terminals are usually connected to a modem. This should not be viewed as a problem since system security has not been violated; to log on to the system will still require an Operator ID and Password (if required). The reason that 'admin' shows logged in is that when an or ^C is entered at the Escape prompt, admin is called to execute the '/bin/login' command. This is the command that prompts for the operator ID and password. In some cases, a modem will produce oneof these characters which in turn calls admin to run '/bin/login'. If this is a problem for your customer, the work around is to turn off the modem during system boot. ORIGINATOR: W. Moore CPU-BASIC FOUR-2000---------------------------------FIB042 Pg001 FIB 00043 10/19/87 *** Firmware corrections for 4-way & 8-way self test problems *** A new firmware is available to correct a specific problem on the 4-way and 8-way controllers. PROBLEM: At power on, as part of each controller self test, data output from Serial Communication Controller (SCC) ICs were internally looped back. The SCC data out-puts were disabled from going to the serial line by hardware. The disabling action was causing a NULL character inadvertently sent out to the line which, on DIABLO interface type printers would cause a garbage character (`?') printed. RESOLUTION: The new firmware resolve this problem by not performing SCC data loop back test as part of the self test. In addition, the 8-way controller fixes minor problems related to abnormal conditions. The ECN number for the 4-way controller is 12276. The ECN number for the 8-way controller is 12266. DRAWING INFORMATION: See WPS hardware bulletin #374 for detail. ORIGINATOR: Carlos Figueroa CPU-BASIC FOUR-2000---------------------------------FIB043 Pg001 FIB 00044 11/09/87 *** ESD (Electro-Static Discharge) kit is available *** A new ESD kit MBF part number 916175-001, MM538260, has been released for the MAI 2000. This kit, should be installed on a system that is having intermittent hardware problems, such as disk errors and system hangs, only after the following have been checked to ensure they are correct: 1. All boards in the system are at the Lowest Acceptable Revision Level. 2. All cables meet MBF cable standards and are properly grounded. 3. The power to the system and the peripherals is clean, that is the lines are free of noise. ESD fixes have already been installed in all systems shipped with a system serial number of EG109644 or greater, so this kit should not be installed in these systems. ORIGINATOR: W. Moore CPU-BASIC FOUR-2000---------------------------------FIB044 Pg001 FIB 00045 10/30/87 *** ESCAPE key problem while in BASIC mode *** A condition has been found where the ESCAPE key on the MAI 2000/3000 is lost in BASIC. This condition is generated when an error occurs and the error handling logic causes a tight program loop. Since the loop is so tight, the ESCAPE is not seen and the condition gives the appearance that the terminal is locked. An example of this would be: 00100 READRECORD(1,KEY="ABC",ERR=00100)A$ Where the KEY is missing. Another example would be: 00100 SETERR 00100 200 X=I/B Where B=0 (which sets an invalid divide by 0 condition) Presently there is no work around for this condition. If your program gets in this condition use the 'ps -a' command to find the task executing and then use the 'kill' command to kill the BASIC task. This will bring your terminal back to the Log-on screen. It's recommended that you use a SETTRACE to troubleshoot terminal hang problems. This condition will be corrected in 7.4A. ORIGINATOR: W. Moore CPU-BASIC FOUR-2000---------------------------------FIB045 Pg001 FIB 00046 11/02/87 *** Memory Diagnostic Test Failure on Release 7.2B *** SYMPTOM: The memory diagnostic finds "a system parity error" on the second pass. Follow the usual procedure to load and run the memory test. Execute the test by entering 'quick', the first pass is uneventfull. But, during the second pass, when the test shifts to test the second block of memory, a memory error is registered. I have checked four system 2000's with 7.2B diags. All exhibited this failure. At the MAI B/4 sales office in Norwalk, CT. I used two different 7.2B diag tapes which also failed. FIX: Right now th fix appears to be upgrading the level of the diagnostic tape. ORIGINATOR: Robert Kent CPU-BASIC FOUR-2000---------------------------------FIB046 Pg001 FIB 00047 11/20/87 *** Memory parity errors on 7.3 diagnostics *** SYMPTOM: Memory parity errors on diagnostics, level 7.3 . System parity error occurs after test is shifted to second test location. Logout after memory tests shows no errors. FIX: No workaround available at this time. ORIGINATOR: Jon Fuller CPU-BASIC FOUR-2000---------------------------------FIB047 Pg001 FIB 00048 11/20/87 *** Slave printer attached via DS-5XX workstation *** SYMPTOM: Slave printer (non-B/4) attached to a DS-5XX workstation would not function as a slave. Print screen function worked okay. When trying to print to it as a slave the following message displayed: . - SLAVE PRINTER CONFIGURATION ERROR. FIX: Set up printer as follows under 2000 utilities port config: - TERM TYPE............... VDT 7270 - SLAVE PRINTER ATTACHED.. YES * - BASIC FOUR INTERFACE.... YES - PRINTER TYPE............ WHISPER - READ STATUS TIME OUT.... 7 WRITE TIMEOUT.... 12 * - Spooler must be turned off for the slave printer ("printer parameters" option in the spooler utilities). ORIGINATOR: Barry Matthews CPU-BASIC FOUR-2000---------------------------------FIB048 Pg001 FIB 00049 12/29/87 *** Terminal hang when opening R0 and LP [ WPS 407 ] *** When RO and LP are opened in this order in a basic program it will cause the terminal to hang. The workaround is to reverse the order that the devices are opened. Examples: OPEN (1) "RO" --- will cause a terminal hang OPEN (2) "LP" OPEN (2) "LP" --- will NOT cause a terminal hang OPEN (1) "RO" ORIGINATOR: Pamla Thatcher CPU-BASIC FOUR-2000---------------------------------FIB049 Pg001 FIB 00050 02/16/88 *** New installation procedures for 2000/3000 [ WPS 422 ] *** MAI Basic Four is now configuring and formatting all MAI 2000 systems to the sales order, which is the method by which the MAI 3000 has always been shipped. The packaging has been modified to ensure that the PCBA's will not come lose during shipment. This action should result in minimizing field failures caused by exposing the PCBA's to excessive handling and ESD related damage. The following is the recommended procedure for installation of the MAI 2000/3000 at a customer site: o Unpackage the CCA cabinet from shipping cartons. o Open CCA cabinet and remove the packing material from inside the cabinet. (Follow instructions inside CCA cabinet) o Make sure that all PCBA's are seated correctly. (DO NOT change the switch settings on any of the controllers) o Close CCA cabinet. o Plug the unit into dedicated line receptacle. o Power-on the system and perform a standard load. o Do an alt.load and run the SIT diagnostics. o Install diagnostics into hard disk. (Optional) If the system does not power up correctly then perform the normal troubleshooting procedures. Verify that all diagnostics run successfully before you continue with your normal installation procedures. ORIGINATOR: Carlos Fiqueroa CPU-BASIC FOUR-2000---------------------------------FIB050 Pg001 FIB 00051 02/23/88 *** PARADYNE MODEM CONFIGURATION *** PARADYNE MODEMS CONNECTED TO THE MAI 2/3000 SYSTEMS AT 7.3 OR HIGHER MUST BE SET UP UNDER THE" HOST MODEM CONFIGURATION" CALLED OUT IN HARDWARE ANNOUNCEMENT NUMBER 16 SECTION 3.4.1.1 . FAILURE TO DO THIS WILL RESULT IN BUFFER OVERRUNS, OR TERMINAL HANGS ,OR TERMINAL UNABLE TO TRANSMIT TO THE CPU. SEE FIB #30 UNDER COMMUNICATIONS BASIC FOUR. ORIGINATOR: J WHALEN CPU-BASIC FOUR-2000---------------------------------FIB051 Pg001 FIB 00052 08/06/88 *** SYS Crash When Adding Memory *** SYMPTOM: After adding a 256k memory card the system would crash with a code of 0 or 21 sometimes saying the configuration was bad. Re-installing config or OS tape did not help. FIX: Changing the disk controller fixed the problem. ORIGINATOR: Doug Hampson CPU-BASIC FOUR-2000---------------------------------FIB052 Pg001 FIB 00053 03/21/90 *** 2000/3000/4000 LARL History [ WPSF 475 R2 ] *** The following controllers should be checked for the LARL (Lowest Acceptable Revision Level). A brief comment on the ECN history is also included to avoid unnecessary board replacement. NOTE: The revisions that are skipped means that there were no functionality changes, only cosmetic and documentation changes. (*) = LARL (**) = ECN Implementation Code 6A Controller Part # Rev. Comment ------------ ------ -------------- -------------------------- WDC 903496 Y 10/10/88 Improve ESD immunity R ** 2/19/86 Multiple Errors, Incorrect Power- Up Sequence P >8MB addressing range. N Firmware change- Data corruption fix, Drive unit fix. ............................................................................. CMB 2nd Artwork 903593 Y 10/18/88 Improve ESD immunity V 01/18/88 EPROM Diagnostic update U ** 10/5/87 New MDS Firmware Bug fix, Sys. hang at BOOT when MODEM/MUX attached to 8-Way port 0. T MDS Firmware Change. N Timing problem causing system to fail MEM Test at boot. ............................................................................. CMB 3rd Artwork 903631 AE 6/12/89 Parts changed because of possible cache parity errors. Note: SCSI disk drives require a Update firmware to add 903631-004 on the 3/4000 and SCSI to 2500. 903631-005 on the 2500. (903631-005) AB * 5/17/89 Add SCSI firmware for 7.5B. (903631-004) AA 2/7/89 Change PAL firmware to support 7.5B func. Remove 32 IC sockets from PCBA. W 1/31/89 Remove resistor at R46. CPU-BASIC FOUR-2000---------------------------------FIB053 Pg001 Controller Part # Rev. Comment ------------ ------ -------------- -------------------------- CMB 3rd Artwork 90363 V 8/14/88 Use common CPU and (continued) Boot PROM for 2500/ and 3000/4000. U 8/24/88 Eliminate occasional ERROR message "BAD INTERRUPT THROUGH VECTOR XX" one IC vendor change. T 8/18/88 LED component change. S 7/27/88 Improve ESD Immunity. R 5/19/88 Release Boot PROM for the 2500. L * New MDS Firmware to support the MAI 4000 system. K ** 10/05/87 MDS Firmware change- Bug Fix, Sys.hang at BOOT when MODEM/MUX attached to 8-way port 0. H MDS Firmware change- Bug fixes. ............................................................................. 8-Way 903577 L * 3/28/89 Component reliability improvement. K 9/14/88 Remove sockets. J 8/18/88 LED component change. H 7/27/88 Improve ESD immunity. G 6/6/88 Added new firmware for 7.4E. F ** 8/10/87 Update Firmware & Hardware for Bug fix & avoid bad character output to printer. C SCC Timing problem. ............................................................................. 4-Way 903390 Y 10/10/88 Improve ESD immunity W ** 8/17/87 Firmware change; Bypass SCC loopback selftest. P Firmware change; Required for 7.1A. ............................................................................. MCS 903406-001 AE 10/10/88 Improve ESD immunity AD 03/14/88 Timing change Y * Data miscompare fix. T Pal & Firmware change to correct timing problem. ............................................................................. MCS 903406-002 AB * Firmware change to work with 60/120MB streamer tape drive. ............................................................................. 1 MB Memory 903533 D ** 8/3/87 Leads short on control panel connection pins. ............................................................................. CPU-BASIC FOUR-2000---------------------------------FIB053 Pg002 Controller Part # Rev. Comment ------------ ------ -------------- -------------------------- 4 MB Memory 903575 D ** 8/3/87 Leads short on control panel connection pins. B Connect missing ground. ............................................................................. LAN 903405 M 10/10/88 Improve ESD immunity D ** 08/13/84 Add terminator with switch. C DMA timing/disable ............................................................................. 1/2" Tape Ctlr. 903639 F * 2/7/89 Firmware update to support STC CD05. E 10/24/88 Update firmware to support GCR. C 5/17/88 Improve System ESD immunity. A 2/5/88 Product release. ............................................................................. MTS 903528 A * Production release ............................................................................. 128/256K Memory 903368 H 6/26/87 PCB Catch. G 7/24/86 Improve release on PCBA Latch/Ejector catch. F 5/10/84 New artwork 905060. A 12/13/83 Production Release. ............................................................................. WDC Adapter Board 903439 B * Change spacers ............................................................................. Adaptec Board 907649 E * Firmware change-Data corruption fix. D Substitute Rev.A Adaptec firmware. ............................................................................. CMB 16 BIT 903441-001 AU * Boot Proms Rev.B3.1 ............................................................................. CMB 16 BIT 903441-002 BE * 1/2" Streamer tape boot ............................................................................. EIA 903645 E * 8/18/88 LED component change. C 4/5/88 System hangs if more than 12MB memory and external control config. A 9/14/87 Product release. ............................................................................. EIB 903643 C * 8/18/88 LED component change. A 9/14/87 Product release. ............................................................................. Dual SCSI Ctlr 903667 E * 3/9/89 Product release. DRAWING INFORMATION: See WPS Field Bulletin 475 Revision 2 for a complete ECN history. This FIB is an extraction of important ECNs. ORIGINATOR: Dan Arteritano/Norm Jones CPU-BASIC FOUR-2000---------------------------------FIB053 Pg003 FIB 00054 11/01/88 *** NVRAM errors. *** SYMPTOM/TEXT: When trying to boot a 2000 system NVRAM errors were encountered. This occured immediately after abnormal termination of the SSN program while trying to change the system serial number. When booting the system, " NVRAM ERROR " was displayed and the system jumped into Debug. The SSN program would not run. FIX: When this error is encountered, the NVRAM information must be re-written. On 2000 systems, enter config mode by typing "config" at at the debug prompt. Then use the "read", "display" and "write" commands to re-write the NVRAM information. On 3000 systems, the NVRAM can be set to default values by setting the CMB switch, position #3 to on. In both cases, be sure to record the current settings if possible before re-configuring NVRAM. ORIGINATOR: Barry Matthews LOGON: 05DPS4 CPU-BASIC FOUR-2000---------------------------------FIB054 Pg001 FIB 00055 12/20/88 *** New 1/2" Streamer Tape Controller (STC) [ WPSF 487R ] *** A new 1/2 inch Streamer Tape Controller (STC) is now available for all BOSS/IX systems. This new controller is a replacement for the current SCSI Bus Adapter and NCR ADP-52 controller subsystem . It supports 1/2 inch tape drives using the Pertec (1600 bpi) format. The STC can be used with the MAI 2000/2500/3000/4000 systems. o There are no O.S. restrictions for the STC controller. WARNING: For all diagnostic releases the "MTS LOGIC TEST" will fail on test #7; the error message will read "Bus Error While Bus Master". This error will prevent other tests from running. *** This will be fixed on a future diagnostic release *** MBF P/N NACS P/N Description 903639-001 MM002679 1/2 inch Mag Tape Streamer controller PCBA 907065-005 MM002680 25 Foot Cable (2 ea. required) or 907065-006 MM002681 10 Foot Cable (2 ea. required) Cable Connection _____________________________________________ | |---| | | | J | | | 1/2" Tape Controller | 4 |<---------------* | (top view) |---| | | | | | | |---| | | | | J | | | | | 3 |<-----------* | | |---| | | | | _________________ __________________ | | | | | J2 | | J1 | | | | | |_________________| |__________________| | | | |_____________________________________________| | | to CMB | | | | | | ______________________________________________ | | | | | | | Cipher Tape drive | | | | (rear view) | | | | | | | | ________ ________ _______ ________ | | | | | P1 | | P2 | | P1 | | P2 | | | | | |________| |________| |_______| |________| | | | |_____________________________^__________^_____| | | | | | | | *---------|---* *--------------------* CPU-BASIC FOUR-2000---------------------------------FIB055 Pg001 NOTE: There are no switches to be set on the STC PCBA, but there are 10 jumpers which should be set as follows: JP 1 --- out JP 6 --- 2-3 JP 2 --- 1-2 JP 7 --- 1-2 JP 3 --- 1-2 JP 8 --- 1-2 JP 4 --- 1-2 JP 9 --- 1-2 JP 5 --- 1-2 JP 10 -- 1-2 ORIGINATOR: C. Figueroa/N. Jones CPU-BASIC FOUR-2000---------------------------------FIB055 Pg002 FIB 00056 01/29/90 *** Workaround for terminal hangs on OS release 7.5B *** Below is a BASIC program and the different ways to run it as a temporary workaround to free terminals which become locked at the BOSS/IX screen. The command given in a previous message may not have worked in all cases because it did not clear the 8-way buffer. OTHER TIPS: o Turn off the key-repeat on terminals. o Make sure that the cable shield is grounded to pin 1 on the terminal end. WAYS TO RUN THE PROGRAM: The program can be run in 4 different ways. 1. Run the program when terminals are locked. 2. Run the program continuously on a terminal which is normally not used. 3. Run the program as a ghost. 4. Run the program in background mode. This is preferred to running as a ghost. HOW TO RUN "TRMFRE" o To run the program when terminals are locked enter: RUN "TRMFRE" The freed terminals will be displayed then hit ESCAPE to stop the program. o The program could be run on a terminal which is used infrequently and will display on the screen continuously; enter: RUN "TRMFRE" The advantage of running the program in this is manner is that if a terminal receives a large amount of data, it will be displayed on the screen and can be viewed. o To run the program as a ghost enter the following: START 64,"/fix/TRMFRE","G0" Make sure that there is a ghost configured in the vconf. There will be no display. o To run the program in background, a command can be created to start the program. Enter the following: ved /fix/TRMFRE exec -d /bin/basic s=64 -nr pgm=/fix/TRMFRE.pgm CTRL-C and Y filemodes +x +x /fix/TRMFRE (This will make the file executible) After the system has booted, the operator can go into CPU-BASIC FOUR-2000---------------------------------FIB056 Pg001 admin and enter /fix/TRMFRE and the program will run in background. PROGRAM LISTING: 5 REM" TRMFRE - Unlocks terminals on BOSS/IX 7.5B 10 BEGIN 20 CLEAR 30 SETESC 700 40 LET STATE=0 50 DIM A$(1),B$(512) 60 LET TTYINFO$=DEVINFO 70 LET I=1 80 LET TTY$=TTYINFO$(I*10+1,5) 90 IF TTY$(1,1)<>"T" THEN GOTO 20 100 LET P=POS(" "=TTY$) 110 OPEN(1,ERR=320)TTY$(1,P-1) 120 LET COUNT=0 130 PRINT TTY$(1,P-1),":",DAY,":",STR(TIM:"00.00"),":" 140 FOR J=1 TO 512 150 PRINT (1)'BI', 160 INPUT RECORD(1,SIZ=1,TIM=2,ERR=210)A$ 170 LET COUNT=COUNT+1 180 LET B$(COUNT,1)=A$(1,1) 190 PRINT (1)'EI' 200 NEXT J 300 PRINT (1,ERR=310)'EI', 310 CLOSE(1) 320 WAIT 2 330 LET I=I+1 340 IF STATE=0 THEN GOTO 080 350 END 700 REM" ESCAPE PROCESSING 710 LET STATE=1 720 GOTO 300 730 END ORIGINATOR: Harvey Mitchell CPU-BASIC FOUR-2000---------------------------------FIB056 Pg002