So, I built this a while back and I finally decided to show it to the world. It’s one of the few custom tools I built and that actually was and is in use for an extended amount of time.
It is a DIY PCB etching machine.
The core components are a heat-resistant glass bowl (kitchenware for making casseroles) and a heating foil. The foil is rated to be operated at 12V (1A I think). But that was not enough heating power to heat up some 330 mL of etching solution to the desired 45°C in reasonable time. So I took a sizable 24V transformer which I had lying around to power the foil.
The temperature of the liquid vessel is measured with an ordinary rectifier diode (1n4148). The diode is glued to the bottom of the glass bowl with superglue. The thermometer circuit is super simple: Just a voltage divider consisting of an 1k resistor and the (forward biased) diode:
The voltage drop on the diode is then measured with the ADC of a microcontroller (ATTiny45). The hotter the diode, the lower the voltage dropped over the diode (because semiconductor physics). I directly switch on and off the mains power for the heating foil transformer with a small solid state relay . The heating system reacts quite slow, so I need no fancy control loop. I simply switch on heating power if the measured voltage is smaller than the value corresponding to 45°C. And I switch it off if the reading is above.
I recycled the motor and gear system which used to move the tray in an old CDRom drive. I attached a metal bar to the drive gear and put some neodymium magnets on top. Everything is installed in a record card box. Right above this magnetic rotor I installed a “false floor” on top of which I put the glass vessel with the heating foil. I chose cards of PCB raw material for that because it is nice and thin, stable, heat resistant and most important: It does not block the magnetic field from the rotor.
The magnetic “fish” which stirs the liquid is made of a small stack of neodymium magnets in the center and the “bodies” of two small electrolytic capacitors on the sides to make it longer. The magnets and the capacitors are then sealed in shrinking tube. If you heat up the tube and then press it firmly together at the ends with pliers it seems to become sufficiently waterproof. Not only waterproof: Shrinking tube also seems to survive indefinitely in sodium persulfate etchant.
The rotor speed is set with a potentiometer which is also read out with the microcontroller’s ADC. The motor is driven with a single standard n-channel MOSFET which is connected to a PWM output of the microcontroller. It’s straightforward: As you might have guessed, the potentiometer sets the duty cycle of the PWM.
I hope this was inspiring.