Troubleshooting; Stone Dead; Garbaged Sign-On Message; Disk Drive Won't Work - David Griffith P112 Assembly And Operation Manual

Revision 1.1

Table of Contents

Troubleshooting

This section oﬀers some tips and tricks that I have learned in developing this

board. The tips are in order; starting with a plain dead board, then to more

bits of it working. On the production boards, every fault I have yet encountered

came down to a missed solder joint! Do remember that I cannot electrically test

a part-built board, so don't take my soldering of the SMD parts utterly for

granted. I inspect those joints visually and mechanically, but something may

get through. Experience has shown the area to check ﬁrst is the two "short"

sides of the Z182 CPU chip; almost all the bad joints I have had, have been

there!

6.1

You did of course, ﬁrst use a current-limited power supply, just in case of shorts

:-) At this point, there's no substitute for an oscilloscope. Start by verifying

that the CPU clock oscillator is running (at the right frequency), and the reset

pin is behaving. You can fake a Z80 (or its successor, the Z182) by plugging in

an erased EPROM at U4: the CPU will execute FF opcodes continually. This

will show a regular signature on the CRO. Be aware that the Z80182 drives

the /M1 line diﬀerently to the Z-80: it is only active on interrupt acknowledge

cycles.

6.2

That message is the result of a lot of power-on self testing, so errors in it can

tell you a lot. If it identiﬁes the SMC chip, then the basic interface to that chip

is OK (this has not tested interrupts or DMA). The clock-speed is measured by

setting a spare serial port (on the SMC chip) to 1200bps, and counting CPU

cycles while that port sends one byte. This again, proves out the clocks, and

the SMC interface. The memory layout message veriﬁes the memory mapping

logic, and proves your PCB jumpers are correct!

6.3

First oﬀ, make sure you didn't hook up the drives incorrectly. The P112 was

designed so that you could screw it to the underside of a ﬂoppy drive. This

means that the socket on a ﬂoppy ribbon cable that used to be for "Drive A"

is now for "Drive B" and vice versa. See Section 7.8.

Don't forget to wind out your power-supply current limit before trying the drive!

This problem is most likely to manifest when you try to boot the system. The

boot process does not give much in the way of error messages: you can learn

Table of Contents