• Please review our updated Terms and Rules here

Sage & Stride Users

Those Stride 440 mainboard schematics are handdrawn by the well known British Classiccmp enthusiast Tony Duell.
Stride 420,440 and 460 share the same motherboard. The CPU board itself has either a 68000, or a MMU board with 68010 ( req'd for UNIX) , or a 68020 with local fast SRAM.
For the MMU I have entered the schematic for the MMU board into Kicad.
No original schematics for the Stride series are known, but the Sage IV is the predecessor and very closely related.

Start with checking out all TTL ic in the neighbourhood of the battery for valid TTL levels.
Also note that the MC68K itself is orientated opposite to all other IC's in the MB.
 
Do I need to have all 6 lines of the serial port hooked up or can I get by with just 3? xmit/recv/gnd?
Since the Rifa popped on my power supply I've decided to buy a new 130W +5/+12V/-12V power supply for bench testing the board.
If I'm only powering up the main cpu it shouldn't need too much power on each of the lines. The original power supply has good 5V but the 12V is measuring 13.3V and the -12 is measuring -11.6
I'm hoping to get serial port output without anything except power to the main cpu board and a connection to serial port 0.
Thje battery V measures 2.6V with nothing plugged in and I don't see any leakage or damage around the battery.
The lm317 voltage regulator is only putting out 1.6V. The schematic says +2.4V or maybe 2.6V, the notes say this is termination power which I suspect is fopr the VME bus, but the cpu board is not plugged into any vme bus.
Thanks for any additional info you can provide.
dale
 
You should be OK with just RX/TX/GND. 13.3V and -11.6V would not worry me, I would have used the original supply after removing the RIFA's.
The PCB will boot without a floppydrive attached. I would still remove the battery, forcing default for all parameters. Or remove RF14 /33ohm, that also removes the NiCd output from the ciruitry.

Note that the Stride does autobaud sensing. You need the enter a few <CR> after powerup so it can detect the baudrate in use.

However a lm317 output of 1.6 V is definitly wrong. Have no clue as to where that could be coming from.
 
Thanks so much for the extra info The original PS PCB had a resister that was getting really hot, so I decided to purchase a newer power supply. $35 and the screw connectors were the correct size. Voltages are more within range and its about 1/2 the size and enclosed.
I removed the battery thinking that maybe somthing in the cmos ram was messing up. I wanted to force it to reset the serial ports to default. which appears to be 38.4k and no rts or cts.
According to the manual it is looking for spaces, so not just any character would work. Probing while booting up I'm seeing ROM CS and a lot wiggling address and data lines.
I'm not seeing any activity going to the duarts though.
I replaced the 68000 with another 10mhz 68000. no change.
I reseated the ROMs and accidentally bent one pin out, on power on the led stayed red. I reseated the rom correctly and led goes back to green now. So it did have an affect.
I'm going to see if I have an LM317 I can replace into the board.
 
I think I found where the 2.6V from the LM317 goes, on page 9 of the schematics I see a resister pac for terminating ramaddr. there is 2.6V going to there. also sheet 10, goes to a bunch of pacs there.
 
So the 2.6V is OK. Then the only other thing I can suggest is to pull both SRAM. There the parameters are stored, forcing the system to use defaults.
 
id already pulled the battery. do i leave the sram out? Is there src code or disassembly of bootrom? my next step is to clean up my work area and get out more tools. Which is probably what I should be doing anyway.
 
That src listing helps alot. I was able to trace getting through the cmosram and into the start to the duart setup. Unfortunately the roms aren't exactly the same and the addresses get shifted a few words after cmos stuff. Also the schematics show two als138 for i/o decode and my board has a big 74ls154.
I'm using my trusty hp 1630D
 
time for update. I'd found that the mRD signal was corrupted. mRD goes all over so it took awhile to isolate the src of the corruption. After replacing a 74S74, mRD is now cleaned up. I have a reliable serial output. It now gets stuck waiting for the nonexistant HD because nBERR is not getting to 68k.
Thanks again for the rom src code. Although it is not exactly the same as my roms it is close enough. After 40 years I guess we can't expect the chips to last forever.
 
The machine should come up to a prompt, even without disks or floppy. Did you using the MB in standalone mode ?
WRT to harddisk : any MFM disk, or David Gesswein's MFM emulator should work.
Somewhere in my floppy-images is a tool to initialize a harddisk, I did so on a 160MB disk.
Also : your 3.04 ROM indiciates a first version of the MB : this is not compatible with Unistride.
These A-versions had to be returned to the factory to be upgraded in case Unistride was desired.
 
In 1984 we decided not to use the 460 after a few months because Commodore Amiga went to a Sun2 based development environment. It was working fine at the time it was mothballed.

The machine won't come up to a prompt if BERR isn't working as BERR is needed to detect missing HD interface. That's where mine was hung. I fixed the BERR problem so it tries to jump thru the BERR interrupt vector which brings me to the next problem, and a bug in the ROM. The ROMS depend on working RAM before even testing it. The BERR vector is stored in location 000008.. But if RAM is not working it jumps into hyperspace.
It should be at least checking to see if the contents of 000008 are being kept before relying on it.
I would have a put a loop into the rom code to keep writing 8 until it remembers the correct value. This would put it into a loop that could be used for probing the board.

At this time I've determined that writes to the RAM are actually working but the buffer enabling the data back to the CPU is not getting enabled when a read happens. The correct inverted data appears on the input to the buffers, but is not getting gaited to the outputs. So I'm running that signal down.
The chips on these boards weren't designed to last forever, so when powering back up, repeatedly, I expect some mortality. Just wish they had not put so many bypass caps right next to the sides of the chips preventing dip clips from easily being used. Maybe I'll take the time to move the bypass caps to the underside of the board.
 
I'm continuing to document my progress here so the next person that tries to repair one of these boards can follow.
The serial port will spit out the copyright message even if the RAM isn;t working, or even the CMOSRAM because it uses address registers to store return addresses when using branching to subroutines.
 
Stride debugger comes up. 2.5M ram.
I had a tiny short with a line going underneath the chip I replaced.
While I have it running decided to use the sdt command AD to disassemble memory.
So here is a disassembly of the 3.04 rom
 

Attachments

You must log in or register to reply here.
Back
Top