Okay a little rant....

We have a design exercise to do in uni - it is a PIC based counting circuit, which counts the number of objects passing between a light source and an LDR on a conveyor belt. We were told that there should be a circuit built from op amps to convert the analogue signal from the LDR to a digital one, so the PIC could count without using any ADC kind of stuff.

Anyway.... The group I was in started off well. We sat down, decided on the number of input/output pins we would need and decided on a PIC16F, as it had exactly the right number of pins for the task. I designed the hardware, setting out a schematic, PCB and finally rendering a 3D image of the PCB.

We came around to programming the PIC, when we encountered a problem which we couldn't identify. We asked for help, and when help came we were told that we shouldn't be using a PIC16F. "You should be using a PIC18F4520, the instruction set is completely different in a 16F and the pins are all analogue." WHAT. We researched the PIC16F and were fairly confident we had a working program - but no second look was made.

"Where does it say in the specification that we have to use a 18f4520?"

"Oh it should be in there."

Hmpf. Well. Okay. Redesign the circuit to use the 18F4520. Now we have a clunky, wasteful, inelegant circuit. But at least it conforms to the imaginary part of the specification. Then we come to build the comparator I designed using op amps.

"Sir, can we have an op amp?"


We get back, find a datasheet for it and.... It's a comparator... okay...

"Sir, we are needing an op amp, not a comparator. The specification says that we should be using op amps."

"Yeah use that."

"But this is a comparator."

"Yes, use that."

"But the specification says to only use op amps."

"Use that."

Hmpf okay.... So I walk off... I manage to scavenge an op amp... Stuff using a comparator....

Then we find that the wires connecting the legs of the LDR to pins for breadboarding have a break in them. Again go up...

"There appears to be a break in the wires here - we tested it and the multimeter is measuring infinite resistance when we measure across the pins, but has a sensible reading when measured across the LDR itself."

"Hmmm. It should be fine. You can still use it."

"Er - if there is an open circuit then the changing resistance will have no effect on the output of the potential divider! The circuit cannot work!"

"Hmmm... It should be fine. Go and try again."

"What - we just tested it and there is 100% a break in it!!!"

"Try again."

Okay... So at this point we are getting frustrated... Our attempts to make an elegant, efficient, WORKING circuit are being put down in favour of a clunky, safe, but easy to mark generic trash.

Don't even get me started on the pains of asking for a 20K variable resistor/pot.

It's just the 'make the rules up' approach that annoys me. It should be the case that a specification is set with the design, and if a project can be shown to meet those criteria, then it should be deemed a success. I doubt my boss would be happy if I wasted company money on using a 40 pin PIC when a PIC with half the pins could be used instead with no loss of functionality. It's like any attempt to actually engineer something to be proud of is shot down. :mad:


Blog entry information

Last update

More entries in General

More entries from Sparky49

Share this entry