Skip to main content

Sweeping Analog Clock Analysis

The subject of this analysis is the $2 Ikea Stomma wall clock. Unlike the Rusch wall clock that had been the object of my experimentation all along, this clock uses a sweeping clock mechanism, which means the second hand appears to move continuously instead of every second. This also means it is quieter and does not produce the familiar tick-tock sound.

Note: Both clocks appear to have been discontinued at the time of writing. There is only one $2 clock on sale at their website now, which is the Tromma clock. I don't particularly like this clock, because the second hand is missing, which makes it difficult to tell if the clock is working at any instant.

The hacking process of the Stomma clock is quite similar to that of the Rusch clock, so no suprises there. 

Remove front plastic cover

Remove hour, minute and second hands

Remove clock mechanism

Open up clock mechanism. As before, keep the spindle side up helps keep the gear formation in one piece.

Solder 2 wires to the solder points where the thin coil wires is connected to the circuit board. I am using a 2-pin JST connector here.

Reassemble the clock mechanism. As before, I use the soldering iron to create a small hole at the side so that the wires can exit.

I inserted a AA battery into the clock mechanism and hooked up the wires that were just installed to a logic analyzer to check out the signal produced.

So basically, for each second, 16 pulses (each 32ms long) are produced, 8 pulses for each tick pin. They alternate at 62.5ms intervals (62.5ms x 16 = 1 second). So I think there is an extra 1/16 gear added to the clock mechanism, and each pulse moves the second hand by 1/16 of a second. However, I suspect the trade-off is higher power draw, because the pulse width is similar in the 2 mechanisms, so the power draw should be 16x for the sweeping mechanism.

To test this out, I connected the clock to the ESPCLOCK3 circuitry, and wrote a simple ATtiny85 sketch to drive the clock. The observations are as follows:
  • In my test, the power draw is about 10x when driving the sweeping clock mechanism. 
  • The pulse width cannot be lower than 32ms if the clock is to be driven reliably.
  • However, the rest time between pulses can be as low as 8ms (versus the current 62.5ms - 32ms = 30.5ms). This means the maximum rate at which I can fastforward the clock is about 150% (62.5ms / 40ms). In contrast, the maximum rate for a non-sweeping clock mechanism is about a whooping 800% (1sec / 120ms).
  • The stay-in-place ticking trick (by pulsing the same tickpin repeatedly) does not work with the sweeping mechanism.
So it would not be practical to drive a sweeping clock using ESPCLOCK3 primarily because the fastforward rate is too low. If we are a few hours off, it will take forever to catch up. Also, due to the higher 10x power draw, the clock will last only weeks instead of months.

Comments

  1. It would be interesting if swapping the Tick pins would result in the hand moving backwards.
    If so, then this type of movement would be a great choice for Vetinari clock.

    ReplyDelete
    Replies
    1. I don't think that's possible. The order of the tick pins are irrelevant. What matters are the timing of the pulses. Swapping the pins will have no effect on the direction of travel.

      Delete

Post a Comment

Popular posts from this blog

Update: Line adapter for Ozito Blade Trimmer

Update (Dec 2021): If you access to a 3D printer, I would now recommend this solution , which makes it super easy to replace the trimmer line. I have been using it for a few months now with zero issue.

Attiny85 timer programming using Timer1

This Arduino sketch uses Timer1 to drive the LED blinker: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 /* * Program ATTiny85 to blink LED connected to PB1 at 1s interval. * Assumes ATTiny85 is running at 1MHz internal clock speed. */ #include <avr/io.h> #include <avr/wdt.h> #include <avr/sleep.h> #include <avr/interrupt.h> bool timer1 = false , led = true ; // Interrupt service routine for timer1 ISR(TIMER1_COMPA_vect) { timer1 = true ; } void setup() { // Setup output pins pinMode( 1 , OUTPUT); digitalWrite( 1 , led); set_sleep_mode(SLEEP_MODE_IDLE); // Setup timer1 to interrupt every second TCCR1 = 0 ; // Stop timer TCNT1 = 0 ; // Zero timer GTCCR = _BV(PSR1); // Reset prescaler OCR1A = 243 ; // T = prescaler / 1MHz = 0.004096s; OCR1A = (1s/T) - 1 = 243 OCR1C = 243 ; // Set to same value to reset timer1 to

Line adapter for Ozito Blade Trimmer

This is an adapter for Ozito 18V battery trimmer (and possibly some Bosch trimmers as well) that uses a plastic blade for cutting. It lets you insert a 2.4mm trimmer line (about 8cm long) and use that for cutting. Simply cut a length of trimmer line and briefly heat up one end with a lighter so that a little bulb is formed. Then insert the trimmer line into the adapter and slot that into the trimmer as per normal. Make sure the trimmer line is not so long that it touches the safety guard. If that is the case, simply trim off any excess with a cutter or scissors. This part is best printed using PETG, which is a tougher and more flexible material. PLA is more rigid and breaks more easily. However, even with PETG, it will still break when it hits something really hard. Since this takes only 0.5m of material and 15 minutes to print, I will usually print a batch of nine at a time at very little cost. The blades that they sell do not break when it hits a hard object, but