LytBrite wrote:Interesting how many ways there are to do the same thing.

Oh geez, now you got me thinking...

instead of a timer module, why not PID the grinder?!
You'd need a scale that outputs a voltage, which you can run to a general purpose PID input, then grind directly onto the scale. The PID algorithm should compensate for the time it takes grinds to leave the burrs and hit the scale (needs to be doserless, of course). Simple process - 0g to <target weight>.

I've got a spare PID and some SSRs lying around...now all I need is a load cell.

=8-)

Caveat - I was being mildly faceitious, though I'm actually considering playing with this grinder idea just for fun. I'm also not a PID expert by any stretch of the imagination, but my office mate used to work on industrial control systems, so I've discussed these systems a bit.

PID is an algorithm, also used to describe a piece of hardware that implements said algorithm.
PID controllers can be used to control a system, with feedback from the system, and used to smooth some aspect of the system to damp inertia, for example.

So a PID will damp thermal inertia in the boiler of an espresso machine, allowing it to be held at a constant temperature, as opposed to a simple thermostat, which will have hysteresis. You set the "process value" - the temperature to hold, and feed back using a thermocouple. You can also find programmable PID controllers, which have a time-based component allowing you to accurately ramp the temperature over time. This is the kind of PID controller I use in my coffee roaster.

But PID algorithms are also used in motor controllers for things like elevators. A properly tuned elevator will start smoothly, and stop smoothly, lined up perfectly with the floor. A poorly tuned PID algorithm will lead to the elevator overshooting the floor, and oscillating a little bit. In this system, the PID doesn't measure temperature - it gets feedback from the elevator on its position, and aims for a target position. The PID primarily deals with starting and stopping the elevator - once it is at speed, it's pretty constant.

Using a scale with a setpoint, let's say 16g, where the scale cut the grinder off upon reaching the set point, the inertia of the motor would grind a little bit more once it was shut off, and there would be grinds that hadn't yet hit the scale, so you'd overshoot. One could compensate by having the motor shut off a little early, and if you never changed the beans you were grinding, this might work; the extra amount over overshoot would be a constant that could be accounted for, but this amount might change with a different blend.

Using a PID might allow the controller to compensate for these factors and grind the perfect dose every time, with no overshoot, based on weight, not time.

Practical considerations may totally render this approach invalid. I may very well be crazy.