128MB SIMM. Stupid. Bought 'new' from MemoryMasters a couple of years ago. They gave me a partial refund (because it was going to cost a lot to send it back to the US anyway). But it seems a shame to just toss it in the can, so I'd like to see if could maybe be dropped back to, say, 64MB. I've a true 64MB SIMM which looks physically identical to this 128MB, just without the ICs on the 2nd side, but before I go mad and pull all 8 ICs off the back, I'd like to try and figure out where the failure is if possible, as there's a 50/50 chance it's in the lower 64MB and I'd then also need to swap out a single RAM on the front.
So... RISC OS 3.6 and an Arm610 in a RiscPC to get POST:
We see the two 64MB banks (&40) giving a total of 128MB (&80). There's no detail in the RiscPC TRM about this output and my brain turns to goo when trying to read the POST code... but to start with -- any ideas what the "9" could be all about? The 128MB of RAM runs from 10000000 to 18000000 so '9' doesn't make sense to me here. Ideas? It looks like top bit set on the memory addressing, but I don't know what it means.
EDIT: After "some faffing about" it's changed to beWhich I read as being still a chip fail in the &800 range. Hm. Doesn't seem like a simple stuck bit.
So... RISC OS 3.6 and an Arm610 in a RiscPC to get POST:
Code:
SELFTEST R2.05ROM :VRAM :02 MByteData :Addrs :Byte :DRAM 0:40 MByteData :Addrs :Byte :DRAM 1:40 MByteData :Addrs :Byte :DRAM 2:00 MByteData :Data-F FFFFFFFFDRAM 3:00 MByteData :Data-F FFFFFFFFDRAM80 MByteIOMD :IOMD :D4E7 V.2SRAM :IOinit:Speed :Speed 3E80.1.0RAM :RAM : 02000000RAM : 10000000RAM bad 9000080CWR-0 RD 02000000VIDC :Mid0 0ARM ID:skippedFAIL : 0004008AEDIT: After "some faffing about" it's changed to be
Code:
RAM : 10000000RAM bad 90000824WR-1 RD FFFFFFEFStatistics: Posted by IanJeffray — Tue Dec 02, 2025 4:05 pm