[Coco] Did this really happen?

[Arthur Flexser]

See my 2nd correction I sent out earlier, after realizing the ghosting is
to addresses offset by every 128K ($20000), not 64K ($10000) as I
erroneously said initially.  So, looking at your table just every OTHER row
(even numbered rows), you can see that the 128K CoCo 3 is repeating due to
ghosting, while the 512K is maintaining the separate values you poked in.

Congrats on receiving a free and unexpected memory upgrade!

Art

On Mon, Sep 6, 2021 at 12:47 PM M. David Johnson <mdj at bds-soft.com> wrote:

> Thanks Art -
>
> I tried the 62000/72000 last night
> and noted there was no
> differentiation.
>
> But I got the general idea and did
> this, which did work and does seem
> to confirms that I do indeed have
> 512K.
>
> I'm not completely sure what the
> results mean in detail though.
>
> 1000 '*****
> 1010 '*
> 1020 '* MSIZCHEK.BAS
> 1030 '* MDJ 2021/09/05
> 1040 '*
> 1050 '* 512K MEMORY SIZE CHECK
> 1060 '*
> 1070 '* WITH THANKS TO
> 1080 '* ART FLEXER
> 1090 '* FOR POINTING OUT THE
> 1100 '* THE GENERAL CONCEPT
> 1110 '*
> 1120 '*****
> 1130 '
> 1140 'POKING INTO MEMORY
> 1150 LPOKE &H8A3E,31
> 1160 LPOKE &H18A3E,62
> 1170 LPOKE &H28A3E,93
> 1180 LPOKE &H38A3E,124
> 1190 LPOKE &H48A3E,155
> 1200 LPOKE &H58A3E,186
> 1210 LPOKE &H68A3E,217
> 1220 LPOKE &H78A3E,248
> 1230 '
> 1240 'PEEKING AND REPORTING
> 1250 V = LPEEK(&H8A3E)
> 1260 PRINT "&H08A3E ";V
> 1270 V = LPEEK(&H18A3E)
> 1280 PRINT "&H18A3E ";V
> 1290 V = LPEEK(&H28A3E)
> 1300 PRINT "&H28A3E ";V
> 1310 V = LPEEK(&H38A3E)
> 1320 PRINT "&H38A3E ";V
> 1330 V = LPEEK(&H48A3E)
> 1340 PRINT "&H48A3E ";V
> 1350 V = LPEEK(&H58A3E)
> 1360 PRINT "&H58A3E ";V
> 1370 V = LPEEK(&H68A3E)
> 1380 PRINT "&H68A3E ";V
> 1390 V = LPEEK(&H78A3E)
> 1400 PRINT "&H78A3E ";V
> 1410 '
> 1420 'EXIT
> 32767 END
>
> Results:
>                        PEEKS
>                ----------------------
>                    VCC     REAL COCOS
>                ----------  ----------
> ADDRESS  POKE  128K  512K  128K  512K
> -------  ----  ----  ----  ----  ----
> &H08A3E    31   217    31   217    31
> &H18A3E    62   248    62   248    62
> &H28A3E    93   217    93   217    93
> &H38A3E   124   248   124   248   124
> &H48A3E   155   217   155   217   155
> &H58A3E   186   248   186   248   186
> &H68A3E   217   217   217   217   217
> &H78A3E   248   248   248   248   248
>
> Thanks again.
>
> Could you point me to something that
> would explain the details?
>
> --
> M. David Johnson
> mdj at bds-soft.com
>
>
> -----Original Message-----
> From: Coco [mailto:coco-bounces at maltedmedia.com] On Behalf Of Arthur
> Flexser
> Sent: Sunday, September 5, 2021 10:51 PM
> To: CoCoList for Color Computer Enthusiasts
> Subject: Re: [Coco] Did this really happen?
>
> Oops again!  The two addresses should be $72000 and $52000 (not $62000) for
> a proper test--ghosting is to addresses $20000 apart (128K), not $10000
> apart (64K).
>
> Art
>
> On Sun, Sep 5, 2021 at 10:31 PM Arthur Flexser <flexser at fiu.edu> wrote:
>
> > Oops, Basic may mess with address 0, safer if you try it with $72000 and
> > $62000 instead.
> >
> > Art
> >
> > On Sun, Sep 5, 2021 at 10:28 PM Arthur Flexser <flexser at fiu.edu> wrote:
> >
> >> A quick check for 512K is to see if values poked in "ghost" to other
> >> addresses that are $10000 different.  For example,
> >>
> >> 10 LPOKE &H70000, 111
> >> 20 LPOKE &H60000, 222
> >> 30 PRINT LPEEK(&H70000), LPEEK (&H60000)
> >>
> >> If the values printed out by line 30 are both 222, you have 128K;  if
> >> they are 111 and 222, you have 512K.
> >>
> >> (This is off the top of my head, somebody please correct me if I've made
> >> a mistake here.  It's been quite a few years.  Try it on your 128K CoCo
> as
> >> well as the 512K (?) one to make sure this works properly.)
> >>
> >> Art
> >>
> >> On Sun, Sep 5, 2021 at 6:16 PM M. David Johnson <mdj at bds-soft.com>
> wrote:
> >>
> >>> Before I get too excited, I'd like someone, more proficient than I, to
> >>> confirm my apparent unexpected blessing.
> >>>
> >>> I just purchased two 128K CoCo 3's off of ebay.
> >>>
> >>> The first one worked fine. It said "128K Color Computer 3" on the top,
> >>> but
> >>> just to confirm, I took a flashlight and looked through the top slots.
> I
> >>> read MCM41464P on one of the chips (a 64K x 4 RAM I believe, implying a
> >>> total 128K x 8 memory).
> >>>
> >>> When the second one arrived, it also said "128K Color Computer 3" on
> the
> >>> top. But it had something rattling around loose inside the case. When I
> >>> opened it up I discovered that it had empty sockets where I expected
> the
> >>> RAM
> >>> chips to be. And the thing rattling around loose was an L-shaped board
> >>> with
> >>> sixteen MCM6256AP12 chips on it (each a 256K x 1 RAM I believe,
> implying
> >>> a
> >>> total of 512K x 8 memory).
> >>>
> >>> After I said a quick "Thank You, Lord" under my breath, I re-seated the
> >>> board in the headers provided on the motherboard, and closed the CoCo
> >>> back
> >>> up, I plugged it in and it too worked fine.
> >>>
> >>> Have I judged correctly?
> >>>
> >>> And, could the CoCo 3 512K Memory Checker available on the CoCo Archive
> >>> be
> >>> used to double-confirm this?
> >>>
> >>> --
> >>> M. David Johnson
> >>> mdj at bds-soft.com
> >>>
> >>>
> >>> --
> >>> Coco mailing list
> >>> Coco at maltedmedia.com
> >>> https://pairlist5.pair.net/mailman/listinfo/coco
> >>>
> >>
>
> --
> Coco mailing list
> Coco at maltedmedia.com
> https://pairlist5.pair.net/mailman/listinfo/coco
>
> --
> Coco mailing list
> Coco at maltedmedia.com
> https://pairlist5.pair.net/mailman/listinfo/coco
>