Duh; sorry, I saw the M100 references and didn't notice that we're talking 
about a 102.

My 102 is similar to the 24K M100, with three soldered chips (3 x Sony 
CXK5864PN-15L) and one Mitsubishi M5M5165P-12L (M6) in a socket.

Don't know what the history of yours is, but I'd actually suspect M6 (or the 
decode circuit); I think a bad M9 would keep the computer from working since 
part of it is used for system variables and memory is actually counted from the 
top down.

Good luck!

The stylized F is Fujitsu's logo:

https://www.westfloridacomponents.com/manufacturer-logos.html

m

  ----- Original Message ----- 
  From: Josh Malone 
  To: Model 100 Discussion 
  Sent: Monday, July 04, 2016 8:03 PM
  Subject: Re: [M100] 32k M102 suddenly showing 24k


  ​
   20160704_192827.jpg​​
   20160704_193204.jpg​​
   20160703_204651.jpg​Well - on closer inspection, they are *all* soldered, 
but I'm not sure that they were soldered at the factory. The left-most chip was 
clearly done by hand -- although I suppose it might still have come from the 
store that way.


  The left-most chip (M9) is different from the rest - it's an NEC D4464 vs the 
other 8464 ICs.


  Model is 26-3803, Serial 809001955. I'm guessing it's a later "revision" 
where they added the final 8k once RAM prices dropped enough to sell it 
fully-loaded. The uncleaned flux on the top of the board clearly looks 
hand-soldered to me. Do you figure this is the most likely place to start 
looking for a bad chip? Any reason to suspect the NEC over the other 3 chips. I 
don't recognize the logo (serif capital F with horizontal bars above and below).


  Hope the large pictures go through. :-/


  -Josh



  On Mon, Jul 4, 2016 at 6:44 PM, Mike Stein <mhs.st...@gmail.com> wrote:

    *All* of them? Are you sure? Even the left-most one?

    I always believed that there were only two models (in NA):

    26-3801, an 8K base memory version with three expansion sockets, and
    26-3802, a 24K base memory version with only one expansion socket.

    What's your model (and serial) number?

    m
      ----- Original Message ----- 
      From: Josh Malone 
      To: Model 100 Discussion 
      Sent: Monday, July 04, 2016 5:38 PM
      Subject: Re: [M100] 32k M102 suddenly showing 24k


      No, they're all soldered. I was hoping a reseat would be warranted, but 
alas.



      On Jul 4, 2016 3:39 PM, "John R. Hogerhuis" <jho...@pobox.com> wrote:

        If you have socketed ram chips you may need to reseat then.  


        -- John. 

        On Monday, July 4, 2016, Josh Malone <josh.mal...@gmail.com> wrote:

          Thanks. For some reason, finding the 100 cold-start procedure via 
Google wasn't working for me.


          However, the free memory hasn't changed. It's still stuck at 24k. I 
guess it's time to get out the scope/meter/etc. and see what's going on. Not 
looking forward to this and not exactly sure what to try next.


          BTW - shouldn't your sig take the Model 100 into account? :)


          -Josh



          On Mon, Jul 4, 2016 at 8:04 AM, Ron Lauzon <rlau...@gmail.com> wrote:

            Hold the reset button on the back, then press Control-Break.



            On Mon, Jul 4, 2016 at 7:07 AM, Josh Malone <josh.mal...@gmail.com> 
wrote:
            > What procedure should I use for a cold start? I figured the 
memory power
            > should do it.
            >
            >
            > On Jul 3, 2016 10:52 PM, "John R. Hogerhuis" <jho...@pobox.com> 
wrote:
            >>
            >> Sounds like it is crashed.  Try cold starting it and see if 
memory is back
            >> to normal. The typical ways this happens is running an ML 
program with a bug
            >> or attempting to load a untokenized basic program with a BA 
extension.
            >>
            >> Logical operators can be used for masking and whatnot. But to 
shift in
            >> BASIC you need multiplication and division by powers of two.
            >>
            >> -- John.
            >>
            >> On Sunday, July 3, 2016, Josh Malone <josh.mal...@gmail.com> 
wrote:
            >>>
            >>> So,
            >>>
            >>> I was sitting down to hack on some basic code
            >>> 
(https://www.reddit.com/r/retrobattlestations/comments/4qs0f3/july_is_basic_month_the_challenge_turtle_graphics/)
            >>> and typed in a quick basic statement to verify that Tandy 102 
basic has
            >>> bitwise operators (it seems to). The command was something like 
"print c and
            >>> 2" or something stupid like that. To my surprise, my 102 reset 
itself. All
            >>> my files were gone (backed up, so no real loss) but now the 
menu is showing
            >>> 21446 bytes free. Some gremlin has stolen 8k of my RAM!
            >>>
            >>> I'm so perplexed by this, I just opened up my 102 to visually 
verify is
            >>> has 32k (indeed, all 4 RAM chips installed - although 1 is a 
different chip
            >>> than the other 3). I've let the system sit for >1 hour with 
memory switch
            >>> off and power removed. No change.
            >>>
            >>> What on earth could have happened?
            >>>
            >>> Last clue I have is that I re-loaded FILSIZ.BA and when I run 
it, I get
            >>> OV error on line 9.
            >>>
            >>> Thanks for any help anyone can provide.
            >>>
            >>> -Josh




            --
            Ron Lauzon - rlauzon at acm dot org
               Homepage: http://webpages.charter.net/rlauzon/
               Weblog: http://ronsapartment.blogspot.com/

               DNRC: Lord of All Things That Are Fattening

               "To be sure, conservative radio talk show hosts have a built-in
               audience unavailable to liberals: People driving cars to some
               sort of job." - Ann Coulter

            Microsoft Free since July 06, 2001
            Running Ubuntu 16.04




Reply via email to