Design Flub Fixed

 

Pinout diagram of Zilog Z80 microprocessor

Back around 1989, I received a request to fix an infusion pump, a device that allows medical staff to pump very small, and precise, amounts of fluids and medicines into patients.

I was told it would turn on but wouldn't start pumping. The display showed rubbish and pushing buttons didn't produce any response.

I opened the thing and quickly discovered a Z80 microprocessor, a computer chip that I had been studying for some time with the intention of building my own computer. I was very familiar with it.

From the description of the problem, it seemed as if the unit was not starting up properly. Opening the unit and checking pin 26 (the RESET pin), I found...nothing, the pin was floating (unconnected to anything).

The chip needs the RESET pin to be held at 0 volts for about a split-second after power-on to start the program at the beginning. If the chip is not reset, the program may start at a random line and the unit will likely hang or run erroneously.

A resistor, diode, and capacitor usually sufficed to do the job of resetting the chip but, in this case, there was nothing. I told the owner that I needed to buy the parts and would return the next day.

At home, I checked the Z80 user manual. While it explained what the reset pin was for and how to use it, the manual didn't say it could be left unconnected. Unfortunately, it also didn't say that a reset circuit was required. Anyway, I already had a course of action and I'd know after I put in the parts whether the unit would function or not.

The next day, I brought the parts, a soldering iron, and some soldering lead to the clients 's office and soldered in the parts. These was no provision on the board for them but there was enough space for me to put them in without difficulty. Once finished, I crossed my fingers and turned on the power.

Yes! The infusion pump's display gave an initial display and waited for an input. Someone set it up, inputted some parameters and pressed start. It worked! A few more tests and we declared the unit fit for use.

While basking in the thrill of a successful repair, I was concerned about the seeming design flaw in a commercial medical product. I told the client that they should inform the manufacturer about what we had found so they can recall the units already sold and have them updated. I went home fully expecting to soon receive a call from some engineer wanting more details. Unfortunately, there was none. They had either understood what the problem was and implemented the fix or they had discovered the issue themselves and were already putting out a recall order.

Anyway, this was just one of those repair jobs that sticks in my memory due to its unusual nature.