From: al@petrofsky.org   
      
   > On Friday, July 18, 2014 9:00:32 AM UTC-7, Thomas Lake wrote:   
      
   > > I have a Model 4 that will boot just fine into Model III mode   
   > > (both Cassette BASIC and TRSDOS 1.3) but when I try to boot into   
   > > Model 4 mode, it looks like the horizontal hold is messed   
   > > up. After it's been on for a while (15 min or more) it will boot   
   > > TRSDOS 6.02 OK and work perfectly as long as I don't turn it off   
   > > and back on. Can someone give me pointers as to what might be   
   > > wrong? Also is there a horizontal hold adjustment inside the   
   > > computer?   
      
   Short answer: if you're lucky, you have a late PAL motherboard and   
   you just need to fiddle with the knob of C210 until things work.   
      
      
   Long answer:   
      
   The video signal coming out of the motherboard is supposed to be   
   pretty much identical in the 64-column Model III and 80-column Model 4   
   video modes (60 hz vertical, 15840 hz horizontal) (assuming a   
   U.S. configuration).  Therefore, the fact that there's no problem with   
   64-column mode indicates that there's nothing wrong with the   
   horizontal hold nor with anything else on the video board.   
      
   The problem is probably in the part of the motherboard that generates   
   the 12.672 Mhz signal that is used only by 80-column mode.   
      
   Which type of motherboard do you have?  There are three relevant   
   categories of motherboards:   
      
       1. 8709296 REV A and earlier (early PAL)   
       2. 8709296 REV B and later (late PAL)   
       3. 8709523, all revisions (gate-array)   
      
   George Phillips can give you a Model III BASIC program that will tell   
   you which board you have.  You can also open the case and read the   
   number off the board.  You can distinguish gate-array boards from the   
   outside of the case by (among other things) the presence of an RS-232   
   port pointing out the back, rather than down.  Early and late PAL   
   boards cannot be physically distinguished from outside the case (but   
   they can be distinguished by software).   
      
   1. On early PAL boards, the 12 Mhz signal is generated by a dedicated   
   crystal (Y2) and components C9, C10, C24, C111, R5, R6, R7, R23, R53,   
   U17, and Q2.  If this is your board, I'm guessing one of the   
   capacitors needs to be replaced.  The specs given in the service   
   manual are:   
      
      C9, 47pfd 50V C. Disk NPO   
      C10, 100 pfd 50V C. Disk   
      C24, 56 pfd 50V C. Disk   
      C111, 10 ufd 10V Tant.   
      
   2. On late PAL boards, the 12 Mhz signal is generated from the main   
   crystal by a phase-locked-loop consisting of an SE 564 (U203) and a   
   score of supporting components.  The problem is likely C210.  If C210   
   is a variable capacitor, you may just need to turn the knob to adjust   
   it.  If it's a simple fixed capacitor (as found in the earliest of the   
   late PAL boards), replace it.  It's 6 to 50 pF 2% NPO (if variable) or   
   33 pF 2% NPO if fixed.   
      
   3. On gate-array boards, the 12 Mhz signal is generated by a special   
   phase-looked-loop chip, U8, which only needs three supporting   
   components: C3, C19, and FB1.  I would try replacing the capacitors:   
      
      C3, 10uF 20V Tantalum Elec. Ax. (says 20V on parts list, but 10V on   
         schematic)   
      
      C19, .1uF 50V Mono   
      
      
   George Phillips  writes:   
      
   > Details on the problem and how to fix it can be found here:   
   >   
   > http://www.tim-mann.org/trs80faq.html#[18]   
      
   That's has some helpful info (especially if you happen to have a late   
   PAL board with a variable C210), but be warned that the author was   
   unaware of there being two types of PAL boards, and he confuses   
   gate-array boards with late PAL boards.   
      
   -al   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)