Hi all,

I've successfully breadboarded a very basic motor controller using an AT89S51. The circuit has a any ordinary 3-4 MHz crystal. I've used the SDCC compiler to build the hex file and it works as desired.

Since I want the final PCB to be small, I try using a 89C2051 which I programmed (successfully) via my PC's parallel port (see schematic). I've taken care to use only ports 1 and 3 so I could use the same hex file in both prototypes.

However, when breadboarding the 2051 circuit (see schematic), strange things started to happen. No matter what I do, I can't get the 2051's internal oscillator to work. I tried a range of different crystals and ceramic resonators (some of them grounded), several different 20-40nF ceramic capacitors, no capacitors, larger power supply electrolytics, pull-ups in P1.0 and P1.1 pins (?), pushed the reset button a thousand times... nothing worked. I've also tested two different chips (89C4051's) with the same results.

I also wired a quick-and-dirty external oscillator -- built from an 74HCT14 -- to pin 5 and it worked at times (that's how I know the program works with the 2051 chip). However it was quite erratic and stopped too frequently.

Another curious fact is that if I grab pin 5 in my hand the oscillator DOES start working (induced 60Hz noise, I suppose), but that's not a solution, of course...

Any clues? Isn't it possible to breadboard a 89C2051/89C4051 at all? Is there any layout tricks I should know about?

Thanks in advance for any hints!