Repairs

Introduction

I have been know to do repairs to machinery and electronic equipment. Usually I get involved when the only other options are to throw it out of spend a fortune replacing large parts.

 

Japax Spark Eroder

This beast is about 25 years old and contains a lot of ttl, cmos and linear circuitry. Unfortunately it's been failing bit by bit over recent years. I was asked to get it going again.

click for larger picture

A main interface board

click for larger picture

Motor control board bench testing

The picture on the left shows one of nine main interface boards containing logic that operates between the main CPU module and the I/O that is the machine itself.

The picture on the right shows one of the main motor control boards being tested on the bench following repair. Bottom left of the main board is a small PIC based pulse generator board configured specifically to test this module. Depending on which input is triggered it outputs a specific number of pulses; up to 320,000. These are traced through the circuitry and counted by the 6-digit counter on the left. The other small board just to the right of the counter is a current limited 5V power powe supply for the main board under test.

The counter was built specially to test these boards. There are 4 motor drives in the machine, of which three were faulty. The counter is special in that it has two count inputs; up and down. It can track the movement of the machine in two directions confirming the return to origin or indicating which part of the circuitry is faulty.

The motor drives have been quite a challenge because we didn't have an circuit diagrams, just a small book with a block diagram and test point information. The boards originally contained Toshiba TC logic which is a cmos version of 7400 series ttl. This has been replaced with 74HC series logic. Unfortunately it'snot that simple. The new chips are much faster than the old ones. Since the board relies on the propogation delays in a couple of places, delays were installed. In other areas the rise and fall times of clocks were too slow for some of the 74HC chips causing multiple pulses to be produced.

 

 

 

next section

.

Notes:

  • I haven't thought of any yet.

PMB home

last updated: 11 September 2004