> > > It appears that there's a DIP shunt block on the CPU board where each shunt > > corespond to a row of DRAMs. You should check all 4 positions are shorted. > > According to the technical manual: > > www.bitsavers.org/pdf/sage/sageandstride/Technical_Manual-1983.pdf > > they should all be open.
Yes, you're right. A jumper is fitted for a non-populated bank. That's quite clever in that the sort of shunt originally fitted is the type where you break the shorting bar with a screwdriver. So it comes with some shorted, if you upgrade the RAM you can break more (it is assumed you never downgrade the RAM :-) > > I think that the power-on test checks the first 128K RAM, then sees if the > > top > > location (? first location) of the next 128K works if so, it tests that > > RAM, and > > repeats 128K at a time to work out how much RAM it has. So you may have > > defective RAMs in the second 128K. > > Also, the manual write it would tell me, if the test fails with the > location of the bad Address. But it say 128K, and just boots ... Ah yes... But if the location it uses to check if the bank exists at all fails (perhaps there's one totally bad chip in the bank) then the first test will fail and the machine will assume the bank does not exist. It wll not give an error. -tony