Thanks to Dan for sharing his clock on show and tell! Read more about the clock here.
I’ve seen a few meter clocks in my travels of the web, and I love the idea. A few days ago, I decided that I must have one of my own. Such began the “How to do it” pondering cycle. I had seen builds where the face plate of the meter is replaced. This works, but I wanted to try and find a way to do it without modifying the meter, if possible. After some more ponderation, I came up with what I think is a serviceable idea.
I came across this style of milliamp meter on Amazon. They’re not quite 0-60 mA, but the 0-100 mA (a 0-20mA meter for the hours) is close enough. And they were cheap. So yay.
Part of my requirements were that the clock run off of an Arduino Pro Mini I had lying around, and with minimal additional parts. In order to drive the meters with some degree of precision, I would use the PWM pins to vary the effective voltage across a resistor in series with the meter. This would, by the grace of Ohm’s Law, induce a current that, based on the PWM duty cycle, would be scaled in such a way as to move the needle on the meter to the corresponding hour, minute, or second.
One minor issue came up in the form of the max current the GPIO pins on the ATMega328 chip can source/sink. The pins can source/sink a maximum of 40mA, a bit far from the 60mA needed for the minutes and seconds meters. Enter the transistor.
Using a simple NPN transistor switch circuit, I was able to provide the current for the minute and second meters from the 5V supply. The PWM signals switch the respective transistors on and off, effectively varying the voltage across the resistors in series with the meters.
The resistor between 5V and the meter is actually 2 1/4 watt 100 Ohm resistors in parallel for an effective resistance of 50 Ohms. The two in parallel was necessary as 5V x 0.06A = 0.3W (more than 0.25 that a single 1/4W resistor can handle safely).