Since my design was accepted in terms of moving ahead on a prototype, I’ve been working first on the board design. I have decided to abandon the PIC18F2620 in favour of the 18F4685. The reason for this change over is due to the fact that the 18f2620 doesn’t have enough I/O to handle the addition of two analog channels and four I/O for RS-485 communication.
This post is more for my own edification and to help me sort out my thoughts on this issue. I suppose for the sake of following my train of thought while sitting here, I’ll outline the specifications, as I think of them.
Overall feature set:
PWM output for 6-24V DC motor.
RS-485 Out – Rec enabled
16 key keypad in
2 temperature sensors
LED indicators for power/error
Serial out for RS-232
Input for external reflection sensor
2 inputs for timing sensors
So, thats 1+4+6+8+2+2+1+1+2 = 27 inputs
I found some nice Molex headers that are single row, .100 pitch and is latched. Typically I use the friction based header and housings but it needs to be secure inside the housing and thus I’m trying out the new set. Also I’m going to use vertical out terminal blocks in order to save space inside the unit but not necessarily on the board.
I’m also considering adding an RS-232 port along side the board.
Well, this has a been a long few days of work. I have created the ideal rotary viscometer in both terms of price and repeatability. Now, this idea is nothing new but I’m simply happy that it works. I guess it’d be prudent to go in to how it works…
How it works:
The concept is simple. you have a shaft that is separated by a spring, in this case two plastic cups connected by a ball bearing with rods sticking out one side. This shaft is driven by a small gear motor and a set of paddles is attached to the other end. While this is rotating, the difference in the driven and the resisted side is measured with some form of instrument be it a hall effect sensor or a slot type optical transducer.
What I used:
Well, as far as materials are concerned, for the prototype I used blocks of polyethylene to support both the motor and the shafts. The material has pretty low friction properties at low loads and thus I used it as the bearing on the bottom. The actual difference mechanism is made from a high density plastic. I would probably use this material again as it’s light, rugged and easy to machine. For the real unit, I may still use this plastic as it is more than strong enough and it’s easy to machine.
As for the drive I used a design similar to the older designs with PWM motor output, serial output, keyboard input, LCD out and 2 channels of input. This is all controlled with a PIC18F2620 which is more than enough for it’s needs.
The motor itself is a Hsiang Neng gearmotor running at 12 volts. It’s a piece of shit but that’s not important at this stage of the game.
So, finally, a desired feature list:
1. RS-485 Out
2. KU and Cp out
3. Multi-fluid calibration, this feature allows the user to select fluids of any KU value and calibrate the unit by entering them in.
4. Easy to use menus. Too often have I seen automation stuff that’s unintuitive. This hould be easy for the operator to understand and easy for the people to use.
5. Speed selection, so that you can use under-powered motors.
Anyways, here’s a gallery of some of the pictures of what could be a DIY stormer viscometer.