Jump to content


  • Content Count

  • Joined

  • Last visited

Everything posted by techknight

  1. techknight

    68K (no Mac) Designs OK?

    No, not really. GCC would be ideal if it works the way I am imagining it would. Just not familiar with GCC. when I was looking up Retro68K compiler, it was all mac-based so I didnt know if it could compile straight binary that isnt "mac". I almost need a "hello world" example or something. Maybe there is one out there already.
  2. techknight

    68K (no Mac) Designs OK?

    I have to write code that receives data from a data card, I have to write code to speak to another 68K on the bus to render graphics, and then I have to write code that runs on the other 68K CPU that actually renders the graphics, and enters into DMA transfers across the VMEBus to grab the graphics to be displayed from shared asset RAM. The graphics card has its own CPU basically acting as a GPU, but again it downloads its program from the main CPU card when its up and running. What I am saying is, there is no OS. so the application has to do everything. Including building and setup of the Vector table in RAM, as well as the Trap handlers for all I/O operations. Including that of reading the RTC IC. (so I have to write a driver for that). As well as a Vertical Blanking interrupt handler for doing things like on-screen Roll, and Crawl, as well as keeping software timers. So no, its not so simple. But my main question was is there a modernish C/C++ compiler that will build stand-alone drivers/vectors/applications for a 68K that isnt attached to the macintosh specific ROM, etc... Also there is no "system calls" or anything of that nature. its basically lights and clock-work with no code. So I have to write the code to make it run again. Basically, Bare-Metal programming. There is no ROM code (well there is, but no API/Documentation on how to use the built-in kernel, and the guy who wrote it is dead, so it must be replaced) Basically, write a program, function, subroutine, etc. that compiles into an organized BIN file that I can either load and execute into RAM from a bootloader, or Burn into PROM and the 68K will boot and run it.
  3. techknight

    68K (no Mac) Designs OK?

    I am curious about this thread, because I would like to know how to compile C/C++ programs for a 68K single board computer that is NOT a Mac. I was looking at Retro68K, but I think its setup to target the Mac arch, no? I was looking at BMOW uCLinux option, but he made a ton of modifications in the kernel/BSP and not really any good comments on why or what they do. (Not familiar with a linux kernel, or 68K ASM). I have been buried head deep in a project since December that is baseed on the MC68010, it has Two of them, its a VMEBus system with multiple cards. I have been in the process of reverse engineering the system back into schematics for trying to re-create the software to allow this unit to function again. Since the original software was downloaded over satellite and is long long gone. it does have a Ready Systems RTOS Kernel in ROM, but again since I am not that fluent in 68K, or even worse, C compiled 68K that abuses the stack, I cant make heads or tails of it. So rebuilding something equivalent is necessary.
  4. I figured I would share this, Uniserver found the bad connection, and I looked at the couple CC boards I have here and one of them had the exact same issue. Ill take pictures of the actual issue, and the fix a little bit later. But attached is a picture that Uniserver has drawn up. Basically, one of the traces to the CUDA chip crystal oscillator rots out, and breaks the crystal connection to one of the load capacitors on the bottom of the board, and sometimes the CUDA chip itself. So, if the CUDA chip doesnt have a clock pulse, Well, its not going to power on or keep time/ADB. In the diagram, youll see a dashed yellow line. Thats the trace that breaks. Well, one of them.
  5. techknight

    This Does Not Compute SE/30

    a good ultrasonic clean may help it, there could be conductive electrolyte stuck in the data bus somewhere. Its hard to say without seeing the board. You could even go as far as to wash the board in a dishwasher, dip it in IPA to displace the water, and bake it. I have had to do this numerous times with Mac Portable boards that had weird untraceable grimlins. Also when testing ROM connections, you need to go further back. you need to go back into the CPU area and pickup the connection testing from there. or even the PDS slot itself. the vias right next to the battery share the same tie points as the muxes, so the muxes can be broken along with the ROM SIMM socket, away from the PDS/CPU. I recall running into a few boards where I had to run patch wires from the ROM points back to the PDS. 90% of the time though, I find the connections broken at the SCC/SCSI/ASC/SWIM area. But I havent found it to stop the booting process, unless the IRQ line from the SCSI IC is broken which will cause a floating condition. and can stop the machine from booting by holding a vector.
  6. techknight

    This Does Not Compute SE/30

    The low impedance of the 5V bus is normal. Your measuring the effective load of all the chips including the PALs. Do you have the pattern with no boot chime? if so your address/data bus is broken between the ROM and the system. So somewhere, you have an open trace. Guarenteed. I cant count how many of these boards I used to fix with various issues. What your going to have to do is grab the schematics, and start buzzing out ROM back to the rest of the system bus, to see where issues may reside. There is a break somewhere. If you get slow chimes instead of no sound at all, the ROM is booting but there is an issue with the system's ability to see and control the RAM. If this is the case, the 74F253 muxes could be the culprit here, or the traces near the first SIMM socket where that leaky cap resides. Sometimes traces broken at the ASC/SWIM/SCSI/SCC chip could mimic these symptoms. Edit: Just watched your video, As far as I can tell, the CPU cant see the ROM properly. there is a capacitor right next to the battery compartment that gets replaced, well its "goo" likes to break the address/data traces right there usually at the Via, or right at the ROM SIMM socket itself. So test each pin of the ROM socket back to the rest of the system on a matching point, that is where I would start if I were working on the board.
  7. Got a couple of these power supplies in. Both dead. one of them had 5V standby, the other one does not. Both have good fuses. This post is about ONE, as I only got the parts for one so far. Basically, if your fuse is good, but your missing your 5V standby, IC1 is likely shorted, and R4 blows open when IC1 shorts. R4 is the fusible resistor to IC1, and IC1 is a self-contained chopper mosfet, and SMPS controller designed specifically for standby power supply output. These ICs are discontinued, so you gotta scour ebay for NOS supplies. After changing out those components, the power supply fired right up. no problems.
  8. techknight

    Notice to Powerbook Duo Users!

    DC-DC converters tick if there is no load, tripping it into overvoltage protection, or a short causing it to trip into overcurrent protection. Youll need to go chasing to find out where it is, and what condition it is in. Easiest way is to start resistence checks on the output side of the converters.
  9. techknight

    Notice to Powerbook Duo Users!

    I just stripped down a powerbook duo 270c which has a busted screen. I have noticed that the few electrolytic SMD caps this board has, are starting to leak, leave signs. The large 100uf capacitors in the voltage regulator DC-DC circuits are the ones I am talking about. This is dangerous. Blow your DC-DC circuits, and its Game Over... Permanently.... Granted most of them are tantalum capacitors, There are still 7 aluminum electrolytics, 6 of those being in the Heart. Board has a heart-attack, it dies.... TIME TO RECAP
  10. techknight

    SE/30 Schematics and the missing "Page 9"

    Actually, I wish I had page 9. but ive read schematics for years and years, and by studying the SE/30 schematics as long as I have, I know what page 9 was for.
  11. techknight

    SE/30 CMD-CTRL-POWER Reset

    The electrical reason for all of this, the power key on the keyboard is wired to its own pin the ADB connector. Guess what? its not wired up to anything on the SE/30 board. Thats why it doesnt, and would never work. But, what I Dont know is if the pin on the ADB connector is actually physically connected to the switch, or if its part of the matrix and that is simply an output from the keyboard's microcontroller. However if this was the case, it would need standby voltage to the keyboard at all times which I dont think it has outside of the power key wire.
  12. techknight

    Rescuing my dad's Classic II, last working in 1996

    Awesome father son/daughter time! Cherish it, let me tell ya.
  13. techknight

    Ethernet Hub Repair

    Other thing you need to look out for are the isolation transformers, make sure they are still good and not open circuit from port to port. They love to go pop especially with surges/lightning. I have fixed alot of dead NICs over the years that way.
  14. techknight

    restoring Li Ion batteries

    Some batteries have an EEPROM and you can "revirginize" a smart battery pack. ONCE you replace the cells, that is. Newer batteries are I2C/SPI SMBus and special software can be used to reset the microcontroller. Older batteries, need to be more invasive.
  15. techknight

    Floppy Disk Drive Read Head Repair

    Yea, even if you tried to repair this, youve lost all possiblity of obtaining cylinder/track alignment between sides.
  16. techknight

    SE/30 Schematics and the missing "Page 9"

    Page 9 is meaningless for the most part. it defines all the connectors and test points for a flying probe tester/scanner. Or bed of nails.
  17. techknight

    PowerBook 165 Grid Effect on LCD

    Umm, you broke the glass on the LCD! that thing is done bro. Look in the corner. You have to be super careful when redoing LCD displays, gotta pay attention to detail and not manhandle it. Scrap it, get another LCD and try again.
  18. techknight

    Portable and 12 volts !!!

    the PMGR chips are the same.
  19. techknight

    Portable and 12 volts !!!

    Hybrid is good, you have 5.18V of power regulation. you are gonna need to find a donor parts board and replace the PMGR IC.
  20. techknight

    Need some quick opinions on this 160 LCD.

    you could try getting a hot bar tip for your soldering iron, and reheating the ribbon back down. They are welded with a material to the glass. this can deteriorate or peel away if it wasnt a good "weld" to begin with. a hot bar will take care of that. But, be careful! In the factory a bead gets laid down by a machine, and then it gets hit with an ultrasonic transducer press and it welds everything together.
  21. techknight

    Portable and 12 volts !!!

    The hybrid is likely fine. The problem is most likely the PMU IC on the other end of the board, its probably toast. But to be sure, with the battery plugged in, press NMI and RESET buttons at the same time. This will reset the PMU. Measure the voltage at the transistor directly in the middle next to your 4-pin molex connector. its output should be 5.2V if you have that, the Hybrid is working. Then what I need you to do is attempt to power it up and feel around the CPU area with your fingers, mainly the 74 series chips near it if they are getting excessively hot. They fail often.
  22. techknight

    Fun with colour on my SE/30

    You could also spray the front CRT Glass with a rear projection paint, and use a micro LED projector with a mirror in the back to extend the throw-distance. It would give it the curved look. now THAT would be cool...
  23. techknight

    Fun with colour on my SE/30

    I have a few of those NEC panels, but they were all CCFL, and I had an 8.4" long ago, only ones I have now are around 10" They are all 90s TFT panels. I am shocked that a generic LCD controller would actually run that display, most of those older displays were a very early parallel standard with not enough RGB color bits to do anything.
  24. I have 3 prodrives now that exhibit the spin up, and spin down without any head actuation. I noticed on all 3 drives when paying close attention, What appears to be tantalum capacitors near teh power jack are actually aluminum electrolytics, and Guess what... They are leaking. everywhere. So, I know the drives work if you give the actuator head arm a little push on the platter, but they wont do it under its own power. I may have just found out why.....
  25. techknight

    VRAM SE/30

    You would have to create an oversampling, something that cant be done with the CRT without cranking up the frequency.