Problems with P&E's simulator

[peg]

Hi all,

I believe I have found 2 problems with P&E Micro's ICS08QT/QY ICS V1.16

The first is that the OSC2EN bit of PTAPUE register is implemented in reverse.

The second is that level sensitive KB interrupts are implemented for both levels, i.e. interrupt triggers constantly.

Here is the minimal code that exposes both problems:

$Include 'qtqy_registers.inc'
 ORG $F800
* compile this then start simulator on this
* issue var porta
* issue var ptapue
* issue inputa ff
* now step through code
Start NOP    ;at this point porta in variables window shows %00111111
*     ;this is correct as only lower 6 bits are implemented
 lda #$00   ;keeps simulator quiet when interrupt occurs
       ldhx    #$0000   ;ditto
 CLR PORTA   ;not really required here
 mov #%00000000,DDRA  ;Port A all ins
 mov #%00111111,PTAPUE ;pullup all
 NOP    ;at this point porta in variables window shows %00101111
*     ;this is wrong as bit 4 should work as an input
 mov #%10111111,PTAPUE ;pullup all and enable OSC2 on PTA4
 NOP    ;at this point porta in variables window shows %00111111
*     ;this is wrong as bit 4 should be outputting OSC
*     ;now continue to expose next problem
 bset 1,KBSCR   ;mask keyboard interrupts
 bset 0,KBSCR   ;set to edge and low level sensitive
*     ;if you comment out the above line edge sensitive KBI works as advertised
 bset 2,KBIER   ;enable ints on A2

* blah,blah

 bset 2,KBSCR   ;ack int just in case
 bclr 1,KBSCR   ;unmask kbd ints
 CLI    ;as soon as you get to here you will be locked in KBI int no matter what
*     ;level is on A2
 bra *

kbi_int
 BSET 2,KBSCR  ;ack int
 NOP
 rti

 org $FFE0

 dw kbi_int

 org $FFFE

 dw Start  ; Reset

This is the current version (i believe) and seems to be a couple of years old.

Has anybody else experienced this?

Or am I missing something?

BR Peg

[bigmac]

Hello Peg,

Are you operating with "simulation only", or are you driving some hardware?  I am reminded that, for the MMEVS with the M68EML08QTQY emulation module, two of the vector locations have been moved, including the KBI vector (to $FFE4).  I wonder if part of your problem might be of this nature.

Regards,
Mac

[peg]

Hi Mac,

Simulation only!

The KBI problem I found several months ago and  know my code works on real hardware. In full simulation you only get to simulate your KBI interrupt code over and over once you set level sensitive.

The OSC2EN problem I only discovered the other day after I sent every device I had out for field testing. They reported an issue so I was doing a full simulation and found 1 input didn't work in the simulation.

BR,

Peg

[peg]

Hi all,

These two issues have now been acknowledged by Mark Cukier of P&E and have been noted for correction in the next release.

BR Peg