notes last updated Mar 26 2016 Herb Johnson Working too fast to make Web pages. Just jotting down notes, and editing photos. March 15th: Continued to work on boards that failed from robinson system (green dot, socketed), also boards from nonworking 2012 system (red dot, socketed) plus nonworking spare boards (dots of other colors, solder-in chips). 0) tried Z80 probe - can't run real-time software and also do memory tests and dumps, etc. ith_z80ide 1) memory board from spares yellow dot #2 lubed up logic chips but not RAM chips didn't check for shorts - blew up +/- 18V fuses ith_spr_64k_fuses replaced both caps and fuses, but one was backwards - blew it up and another fuse ith_spr_64k_caps1 replaced cap and fuse, still not working. read resistance from +12 to ground on board = 14.2 ohms? 30 ohms?!! experimented with removing chips - resistance up with more chips out ith_spr_64k_ram 30 ohms?!! 60 ohms? few hundred ohms? pulled chips from an unused S-100 card - very high resistance so I replaced 3/4 of the Ithaca chips, got resistance up to thousands of ohms memory works! ith_spr_64K_testok but later Mar 15th found it fails if flexed or not flexed. 1a) mar 15 - socketed spares #2 64K card (yellow dot). This is the board where I replaced 3/4 of the RAMs. While it works, when flexed it would cause a running program to change activity. ith_rob_64k_flex.jpg. Put this board aside for further examination. I have two good working 64K boards (spare #1 and Robinson boards). 2)tested socketed Robinson floppy controller, got it working, seemed to be problems with IC contacts (ith_rob_fdc_result, ith_rob_fdc_test1) 3) tested 2012 floppy controller red dot ith_2012_fdc tried to work but could not get to screen display, also seemed intermittant and sensitive to movement of board fried working disk - tested on Z-100 track 00 format failed! Eventually discovered that IFORMAT has option to format only system tracks and do system-track transfer. That worked! Also found, contacts in 50-pin cable was pulling away from cable. put connector in a vice to squeeze cable back onto pins. Works. In the course of testing, discovered that Kipp disk w/ STARTUP 3.4 could be booted! It required a few CR's to set the baud rate (as documented) ith_kipp_boot 4) while pulling boards out, noticed backplane was flexing I forgot to put in remaining screws! Fixed that. ith_rob_bus_screws.jpg 5) testing socketed Robinson serial board (red dot) Seems to have a problem, when cold showed the prompt and responded. When warm, displayed odd garbage and then ran a bunch of ???????? lines. ith_rob_ser_inter.jpg replaced RS-232 chip pair, no change Discovered when probing with fingers near the 2nd interrupt chip IC 20 and jumper J13? that the board would stop sending characters. reseated jumper and 2nd interr controller. reading schematic, the 9519 at IC 20 and J13 is the interrupt controller for the serial ports, and pins 17-20 near my finger are part of that circuit. ith_rob_ser_sch, ith_rob_ser_probe (Mar 15th) traced lines around J13 and IC20. This Rev B board has a new jumper not on Rev A schmatic. it connects the INTA gated by A1 lines to either the same address (rev B) or to A1-seperate addressed lines (rev A). ith_rob_ser_revb.jpg Found the U20 pin 13 and 16 were not making good contact with the U20 jumper - no ohmmeter continuity. Cleaned the pins and socket, used cleaner. Board works OK now! Test was to read off diskette sectors while flexing board, look for gaps changes - ith_rob_ser_test1, test2 6) Mar 15 - socketed Robinson CPU board (green dot). Never worked even after previously replacing a number of 74ALS74 TI brand chips with pins so black and so corroded the pins would break off. Used DeOxit on all chips previously, no reasonable operation. (Bus display board showed activity but not normal memory-test and boot operation on bus.) ith_rob_cpu_rep2.jpg photo shows several chips taped to show they were replaced. Persistant problem: When the board was inserted and powered up - it started to boot and display the startup message. Further tests of flexing the card seemed positive sometimes, booting CP/M and running CRC program. Other times loads CP/M or not, or loads but behaves oddly. Examination of the board immediately found a 74LS00 with a pin outside the socket. tests better but not good. Another mis-inserted chip was spotted with a pin bent underneath. So...I lifted every chip to find any more bent or unseated pins - found none. Still some signs of corrosion on IC pins. tests were better, attempts to boot. Then I lifted every jumper and wriggled them, in some cases removing the jumper to examine pins. Found one area aroud U14 where there was poor contact, touching that area in operation caused change to bus status lights on Bus Probe. ith_rob_cpu_sxtn.jpg ith_rob_cpu_u14.jpg, S-100 bus 2 & 4MHz clock. removed chip TI brand 74LS124 and found brittle pin - replaced with another TI chip, which I solder-tinned by hand to insure contact. No change after testing. Also: soldered every open through-hole. No change in operation. ith_rob_cpu_vias.jpg Changing S-100 socket on bus did not change CPU board behavior. Conclusion: board has IC sockets with poor contact producing intermittant problems that permit CP/M boot but damaged results. Considering replacing all sockets or similar measures. Mar 16 - spare CPU board. yellow dot. Has a problem with displaying HLT and MWRT when reset, should not do that. May be a problem with jumpers or connections between HLT line on Z80 pin 18, pins 1 and 2 on LS14 at U6, pin 3 on U38 which provides buss with HLT signal. I think there's a broken connection from U6-2 to U38-3. Similar problem, MWRT thru jumper J0, U35? pin 15 LS244, and U32 LS02 pin 10 out, pin 8/9 in. ith_spr_cpu_MWRT.jpg Put board aside for later work. Mar 17 - tested a few memory cards to get more confidence in results. One issue is that the memory cards don't "like" running from the extender card. It's not unusual for S-100 boards to see different signals on an extender card. The card introduces cross coupling between lines. I found better results on memory boards when run in the chassis and not on the extender. Also possible it's some flexure of the board. Mar 17 - work on 2012 system CPU (red dot). PUlled from front-panel chassis, changed a few jumpers to match the non-frontpanel setup. Seems to work OK! CPU board seems stable enough. Here's the tests I run. I've noticed that the complete system seems to need minutes to "warm up", otherwise even if I can boot CP/M, the programs don't run correctly for the first number of minutes (several). ith_2012_cpu_test, First of course it must boot CP/M. After that I run "CRC CRC.COM" to check for memory and CPU function, it has to run correctly to get the correct checksum. ThenI run CRC *.* to check all file checksums against list. Then I run MEMRS over a segment of memory as I wriggle CPU, memory to see if they are affected. ith_2012_cpu_crc Mar 17 - work on Robinson CPU (green dot) This board seemed to run intermittantly in previous tests. Close examination today showed 1) one of the chips was a TI 74ALS74 which has black corrosion problems - I replaced it. 2) corrosion on the jumper pins. ith_rob_cpu_rust.jpg. also _rust2.jpg I cleaned all the jumper pins by scraping with screwdriver, then oil with DeOxit. Board performed more stable afterwards. mar 17 - work on 2012 system FDC (red dot). ith_2012_fdc2.jpg. doesn't work consistently. Replaced a number of TI TTL chips with black corrosion. ith_2012_tirust.jpg. Results seem better, CRC's run better. mar 21 - establish a set of four reliable cards, mark as white dot "#1" set. They are: Robinson I/O card (green dot), spares floppy controller (blue dot), spares #1 64K card (yellow dot), spares CPU card (blue dot). These will be my primary system cards. ith_pri_cards.jpg Mar 21-22 now testing out a set of secondary cards, which worked more or less reliably. Put four of them in chassis to test. Problem: boots CP/M but it fails to load and run programs like CRC and MEMRS. Problems isolated to 64K memory board, yellow dot #2. Replacing that card with white-dot #1 produced more stable system. When #2 yellow-dot 64K flexed it would fail or load intermittantly or constantly attempt to load CP/M. One issue: a number of chips still the TI black-rusted ones. Replaced the remaining TI 74ALS74's and TI 74LS90 and it performed better. Decided to replace dark-legged 74LS667, with chip used on other 64K boards, the AM25LS2521. Continued to perform MEMRS and CRC correctly. This board was previously repopulated with 3 banks of Motorola 4116's. Noticed that some chips on 64K board are 74S-series. ith_spr_64K_oddchips.jpg Did a check across schematics, board layout, and several 64K cards for non-LS chips. Here's the results. IC name four 64K boards parts list schematic U3 74S74 S LS S U4 74S10 S S S U5 74S10 S S LS U9 74S00 S LS S U10 74S74 S LS S U14 74S00 S S S U24 7403 - - LS Also: there's no ALS designations in schematics or parts list. Many boards had 74ALS74's. Mar 23: cleaned edge connector on spare 64K card , DeOxite, got some dark stuff off. ith_spr_64K_edge.jpg Reseated all chips on card. Card performed better afterwards. Stillis not 100% reliable on first power, needs minutes to load and run programs reliably. SEcond set of cards labled #2 white sticker are: I/O card spare yellow floppy controller Robinson green 64k spare #2 card yellow CPU board 2012 sys red Mar 23: since 2nd set of cards seem stable, time to make a proper Ithaca 4.cSTD disk with a set of "Robinson" programs that will work with CP/M as loaded. Since I borrowed the booot track from the SD disk but ISTARTUP from the DD disk, it' slikely CP/M is not "synchronized" (same serial number). Also Robinson disks are mix of 4.c and 3.d BIOS and software. Used CRC.COM to compare files on two disks which were derived from Robinson disks. Found in general two sets of CRC's for files, and some differences between versions of Ithaca programs. ith_rob_crcs.jpg Obtained working DDT. IASM's are same versions but slightly different CRC's. two versions of IFORMAT, different dates of assembly. A version of SET (adjusts parameters in ISTARTUP.COM file) but did not verify it works with 4.c STD I'm using. Two MOVCPM's which almost certainly have 3.d CP/M in them. Two SYSGENs. And no source for the 4.c BIOS in my ISTARTUP (but probably have it elsewhere). Two versions of FORMAT. The 3.d versions of tiny BIOS and the source for the PROM are probably correct, but not checked against the BIOS in the boot tracks. Ultimately I'd have to check the buried serial numbers in the booted CP/M versus whatever's in the SYSGENs or MOVCPMs. Finally, I ignored any 3.d software like XBIOS. March 25th - Pulled out DPS-1 front-panel chassis and FP. Powered from variac, slowly raised voltage. No problems occurred. ith_fp_power.jpg Added 2nd set of DPS-1 cards. Had problems with booting, isolated them to two effects. ith_fp_cards.jpg 1) S-100 connectors on DPS-1 chassis probably dry and slightly corroded. Cleaned indirectly by coating four boards with DeOxit and reinserting a few times. Helped memory board in flexure to run better. 2) the #2 I/O card S-100 edge connector is damaged, has gold missing. Ith_fp_io_edge.jpg but many of those pins not connected. I have my 2nd Ithaca chassis with front panel running now, with the 2nd set of DPS-1 boards. It needs a front-panel reset repair, probably some jumper setting or toggle. But it boots and runs CP/M memory and drive diagnostics. It helped to add DeOxit to the S-100 edge connectors, to lubricate and clean the S-100 slots. [today's photos] Front panel reset did not work. Removed front panel board and confirmed problem was bad clear/reset switch. Happened to have correct DPDT center-off switch execept it's not momentary on both ends, only one end. Doesn't matter. replaced and worked OK. March 26th - replace reset/clear switch on front panel. System boots and reboots OK except on first boot, the startup message has several garbelled characters, as if the UART was not well initialized??? NOt a problem on my #1 set of cards on my no-front-panel system. Tested more software programs. I have two ISYSGEN programs, they copy some memory image into system tracks. I have two FORMAT programs, both work, one also copies system tracks. Need to do more testing of Ithaca utility programs, and to determine how or if they need to have modified CP/M images to match my booting CP/M image and ISTARTUP for 4.C BIOS. Notes on 2nd chassis with front panel will continue on new file possibly new Web page.