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.

Filament Joiner Part 2 (With Display and Knob)

Thanks to the current corona-virus crisis, the parts I ordered for the filament joiner project were taking forever to arrive. But now that they have finally arrived, I can put them to good use. These were the parts ordered: 0.96" OLED display SSD1306 Rotary switch encoder KY-040 Here is the final circuit diagram: The OLED display is connected to the SCK and SDA pins of the Nano (A2 and A3 respectively), and powered by 5V and GND. The rotary switch encoder is connected as follows: VCC => 5V GND = > GND CLK => D9 DT => D8 SW => D2 My prototype board now looks like this: The updated code for driving the knob and display is available in  heater-with-display.ino in the Github repository . We now have a fairly compact (about 7cm x 5cm) and independent filament joiner (no need to connect to PC) that is driven solely by a 12V power supply. Here's how to use it to join printer filaments. More usage details in my previous post .

Adding "Stereo Mixer" to Windows 7 with Conexant sound card

This procedure worked for my laptop (Thinkpad E530) with a Conexant 20671 sound card, but I suspect it will work for other sound cards in the Conexant family. I was playing with CamStudio to do a video capture of a Flash-based cartoon so that I can put it on the WDTV media player and play it on the big screen in the living room for my kids. The video capture worked brilliantly, but to do a sound capture, I needed to do some hacking. Apparently, there was this recording device called "Stereo Mixer" that was pretty standard in the Windows XP days. This allowed you to capture whatever was played to the speaker in all its digital glory. Then under pressure from various organizations on the dark side of the force, Microsoft and soundcard makers starting disabling this wonderful feature from Windows Vista onwards. So after much Googling around, I found out that for most sound cards, the hardware feature is still there, just not enabled on the software side. Unfortunately, to