re: SPAM-LOW: Tempest Ram Trobleshooting with Fluke 9010A

The error being reported means that when it reads/writes to 0x408 and 0x40C it appears that it's addressing the same memory space. Try manually writing the values of 0 to both 408 and 40C and confirm that you can read 0 back. Then write 0xFF to 408. Read both 408 and 40C back. You'll probably get 0xFF back for both. (Sometimes the fault can be the other way so also try initializing them both to 0xFF and then changing one value to 0). IF instead you get back 1111 1011 then I'm mixed up, it's data related and your initial assessment was correct that it's the data lines. Either way, this is typically an internal RAM issue. Sometimes the addressing logic is wrong and sometimes chip select logic. But remember if your ROM sig checks work then it's likely NOT going to be any buffer chip (anything that might latch or buffer the address lines) because then a lot more would be wrong. The error message is telling you something to the effect of "when you read/write to the memory space of 0x408, it appears to make no difference what bit 2 is of the address line". Again, this is from memory and I might be confusing it with another type of error. Manually using the read/write keys to probe memory locations can really help. JB http://tech.quarterarcade.com/ (FIDE still available there) ---------------------------------------- From: "Colin Davies" <colin.w.davies_at_btopenworld.com> Sent: Thursday, September 18, 2008 7:19 PM To: vectorlist_at_vectorlist.org Subject: SPAM-LOW: VECTOR: Tempest Ram Trobleshooting with Fluke 9010A Hi Guys, Sorry if I've posted to the wrong list ( Vector instead of Techtools) - Let me know if its better suited in there.... Having trouble with getting tempest running diagnostics reliably, so I decided to get my new fluke 9010a and 6502 pod out.... Ram runs in two blocks of 1k using a pair of 2114's ...giving 2K x 8 bits alltogether... 1st pair gives ram 0000-03FF - Low Nyble for one ram and High Nyble for the other ( Lower and higher 4 bits in the byte). 2nd pair gives ram 0400-07FF - Bit 2 of the ram.... (Now I was thinking this was Bit 2 of the data bus - but please bear with me) Now I'm having ram failure at 0400>07FF... Now I'm new to the fluke, and didn't quite understand the message... but bear with me.... I've replaced the decoder for Ram0 and Ram1 (P7) and another at B3 (which feeds P7) I replaced H2 and F2 which sit on the data bus also H3 for good measure (which hooks up to the vector generator) Oh... I also replaced the ram R4 in question,one of the pair..... (which I am hoping is the right one - I think that I might have just answered my question here actually !!) I've also removed E2 ( the buffer from the eproms back to the data bus)..... By doing all this, this has allowed me to pretty well isolate the ram from the rest of the circuitry ( If I remove the chips from the sockets etc) - I thought that I had some kind of bus contention problem.... Anyway... the Full message from the Fluke is.....RAM BCD err @ 0408 Bit 2.... Reading the manual in more detail, its saying that the fluke is having problems with bit 2 of the address bus and cant quite identify the difference).... Any ideas ?? - I'm thinking I've actually worked out the problem in that it might be the 2nd ram in question P4.... but I'm not sure quite yet..... Next plan is to replace the 244's on the cpu bus.... From what the maual is saying, the ram passes the tests ok, but there is this BCD error ??? I'll get there, but if anyone is familiar with this kind of fault, please could you suggest a fix... Regards, Colin

---------------------------------------------------------------------------
** Unsubscribe, subscribe, or view the archives at http://www.vectorlist.org
** Please direct other questions, comments, or problems to chris@westnet.com
Received on Thu Sep 18 19:40:43 2008