The Sinclair QL Forum

Hi,

On the Q68, there are 4 pads to install a reset button. In the user manual, it is stated that the switches should be hooked to the upper to pads however it is not clear what are the upper pads and the lower pads. Are the upper pads the ones closer to the keyboard connector and the lower ones are closer to the power plug ?

Also, are the other two pads connected to anything ? The switch I want to install would short the other two pads. Finally, are the two active pads the same as the RESET and Ground signal on the expension connector ?

Thank you.

Francois

OK, a closer look at the PCB shows that the two pads near the keyboard connector are indeed the ones.

F.

FrancoisLanciault wrote:Hi,

On the Q68, there are 4 pads to install a reset button. In the user manual, it is stated that the switches should be hooked to the upper to pads however it is not clear what are the upper pads and the lower pads. Are the upper pads the ones closer to the keyboard connector and the lower ones are closer to the power plug ?

Also, are the other two pads connected to anything ? The switch I want to install would short the other two pads. Finally, are the two active pads the same as the RESET and Ground signal on the expension connector ?

Thank you.

Francois

Hi Francois,

If you fit a switch, what are you wanting to do?

I fitted a reset switch on the first 9 production Q68 PCBs, but the fitting of a reset switch will not reset the Q68. So removed the switch from the parts list.

Derek_Stewart wrote:Hi Francois,

If you fit a switch, what are you wanting to do?

I fitted a reset switch on the first 9 production Q68 PCBs, but the fitting of a reset switch will not reset the Q68. So removed the switch from the parts list.

Well I installed the switch and it indeed reset the Q68 when I press it. Same effect as a power off-on cycle.

[Edit]
On closer investigation, the switch just restart SMSQ/E, same as Alt Ctrl Shift Tab on the keyboard. So I agree, no added value.

HI,

Thank you for the update.

Probably the best place for a reset switch is at the power input header on the board behind the power input socket.

I never really used a reset button for the Q68 or QL.

The function of the optional Q68 RESET button requires a little understanding of Q68 internals.

The Q68 has no classic ROM at $0, but actually SRAM which is prefilled with bootloader code at power-up.

After the bootloader has loaded the operating system, it is no longer present at $0, but has been overwritten by the OS.

The RESET button triggers an actual hardware reset (not some sort of soft reset) but it keeps the memory at $0 unchanged.
So the operating system is restarted and not the bootloader.

During early development it was helpful to do a hardware reset without overwriting the memory, but for the end user I can indeed see no benefit over a power switch. It even has a little extra risk, because a severe crash might have corrupted memory holding the OS itself. The only way to overcome that is to do a power-up.

Hi Peter,

is the current "ROM" (even if copied to RAM) write protected on the Q68? I'm afraid there are more than two or three programs, which write to ROM, just to see, if there is really ROM (and if not, write the stored value back - if one of the better programmer, he at least switched to supervisor mode before, tested, write back and returned to user mode afterwards ;-).

The bootloader ROM coming up at power-up is write protected, but not the OS that was loaded. Like you, I assumed for a long time that some programs might need a write protection, and the Q40/Q60 therefore had a hardware feature to write-protect RAM at $0. This assumption turned out a fairytale, nobody could ever name such a program. For the Q68 I omitted this kind of mechnism.

Why would a normal program (deliberately) need to do this?

I can understand there might be faulty programs that write to stray pointers (thus, maybe also to ROM) and simply need to be fixed, but see no reason why they should need to check whether a certain area is RAM or ROM.

There are copy-protection schemes used by a single-digit number of programs that use dongles in the ROM area (Prospero Fortran and Metacomco compilers), but that can be worked around using other methods.

Tobias