Date : Tue, 06 Dec 2005 15:05:20 +0000
From : Richard Gellman <splodge@...>
Subject: Re: Test Sideways Ram
It suddenly occurs to me that the program won't work on anything lower
than a Master 128, in that case you will need to do the following:
change 60-70 to:
60 ldx#0:lda #&AA:sta (&70),X:lda (&70),X:sta &81
70 lda #&55:sta (&70),X:lda (&70),X:sta &81
check that ?&F4 returns the ROM of BASIC (should be 15). If not, change
40 to
40 lda &fe30: pha
I believe on the B ROMSEL is actually readable (ISTR its not on a
Master, hence interrogation of &F4, the MOS's copy).
Line 120 specified which sockets to check. On the master these are 4 to
7. Change to your sockets as required.
-- Richard
Richard Gellman wrote:
> Paul J wrote:
>
>> Hi,
>>
>> My friends machine has a solidisk sideways ram with solidisk dfs. I
>> am trying to test the sideways ram, but can not load the utilities
>> disk as the disk is shot. The last time I got the menu program to
>> load and run, it just locked up.
>>
>> Is there a short basic/assembly prog I can type in to test sideways
>> ram, write a few bytes to a rom socket and read them back ?
>> Found the call to read a byte from ROM, but not write a byte ?
>>
>> There are a couple of roms in this machine that do not seem to
>> display any header on boot or respond to *HELP command. Is there a
>> short basic/assembly prog I can type in to interrogate a rom at a
>> specific socket ?
>>
>> Regards,
>> Paul
>
>
> RUN this:
>
> 10 P%=&900
> 20 [OPT 2
> 30 sei
> 40 lda &f4:pha
> 50 lda &80:sta &fe30
> 60 lda #&AA:sta (&70):lda (&70):sta &81
> 70 lda #&55:sta (&70):lda (&70):sta &82
> 80 pla
> 90 sta &fe30
> 100 cli
> 110 rts:]
> 120 FOR b=4 to 7
> 130 t = FALSE
> 140 PRINT "Testing Bank 0...."
> 150 ?&80 = b
> 160 FOR a=&8000 to &bfff
> 170 !&70 = a
> 280 CALL &900
> 290 IF ?&81 <> &AA OR ?&82 <> &55 THEN PRINT "Fault@... ";b;"
> address &";~a:t=TRUE
> 300 NEXT
> 310 IF t THEN PRINT "Bank ";b;" FAILED"' ELSE PRINT "Bank ";b;" PASSED"'
> 320 NEXT
>
> It'll tell you which banks are faulty, and at which address.
>
> -- Richard
>
>