Using Arduino To Drive A Quartz Clock / Lavet Stepping Motor - Part 2

[@samjlevy]

Nice work Martin. No matter how niche it is-- it's still fascinating the amount of effort that goes into precisely turning a seconds hand. As well as how much you can delve into from such a "simple" project: oscillator circuitry, effects of probing (hell, effects of gravity!), snubbing transients, arduino coding. There's no end to the learning. Your introduction to oscilloscopes video got me into electronics 4 years ago and now I have a lab that keeps growing. Thanks for the hobby.

[@michaelc.tiberio5761]

Hi @mjlorton. Thanks for your videos. I noticed a couple of things with this one that I thought I'd mention just to see what your take is. 1) I don't remember you talking about sourcing and sinking the coil drive current. You are using a Due output pin to source current into the coil and another pin to sink current from the coil. According to the page linked* below, pin 40 and pin 41 can each source 15 mA but only sink 9 mA. Assuming a low coil resistance, which may not be true, you need to keep the resistor in your loop above 3.3V / 0.009A = 367R. Other than the ability to drive the clock mechanism, what other factors did you consider when choosing the value of the current limiting resistor? 2) When the coil is discharging you are driving a voltage directly into an output pin that is trying to drive a logic low state. The Due pin is connected directly to a processor digital IO pin (see link*) and with that processor pin trying to drive a logic low. I believe that when configured as an input pin it is only tolerant of voltage levels up to 3.3V. I may be mistaken, but I don't think you got down that low with your back-to-back diode protection. I did a quick parametric search on Digikey and had a difficult time finding a diode with low enough reverse bias voltage plus forward bias voltage (both diodes drop some voltage) to keep the IO pin below 3.3V. There were some available, but either sold in bulk (minimum order in the thousands of parts) or very expensive (one was listed at more than 70 USD). What Zeners are you using? What is the forward bias voltage on them? Are you concerned about over-volting the input pin on your Due? (Edit: Just occurred to me that driving a voltage into an output pin is probably not a great idea, even if the current has an alternative path. :) ) * Due currents and pin mappings: www.arduino.cc/en/Hacking/PinMappingSAM3X

[@BlankBrain]

I might be replacing the circuit for a quartz-locked direct drive turntable that uses pulses from a magnetic head under the platter to sense the speed. Your videos are very useful.

[@pfeerick]

Nice follow-on from the part 1 video... and I like how you incorporated being able to rapidly turn the seconds hand in order to be able to 'calibrate' the clock... It was 'minor' details such as this that have put me off trying this sort of thing, that, plus being able to tell the digital clock what time the analog clock is set to, fixes the other issue I was unclear on, which was how to get the clock to effectively set itself... Whilst it would be slow, your implementation would be able to do it... Although, without the use of a bi-directional stepper, I prefer the idea of the user setting the hour / minute nearer to the correct time, and then letting the clock do the final 'tuning/syncing' itself. Again, thanks for the great video, and inspiration! :)

[@MicsigOscilloscope]

Hi @mjlorton. Thanks for your videos. We just see you to use Micsig oscilloscope here. appreciated!

[@andreyvlasenko389]

It is possible to ask to share you Arduino code to convert count of cycles to time? Maybe there is already a library, just could not find..

[@TheRintincola]

Good idea!!

[@srmeister1]

These protection diodes never come into action here. This is because, as opposed to the video, both output pins are never set to "high impedance" state. Instead they are just set to another voltage (GND and 3.3V) and are still low impedance. Also, the "500k" pot is probably set to a very low resistance because i really doubt the coil resistance is in the order of a few 100k Ohms. On the other hand, the set resistance is still high enough in value to limit the current to the I/O pins of the Arduino so it wont die. Long story short, the diode protection is useless in this example and the 500k pot is completely overkill, i guess 10k would be enough to set the voltage. Also, setting the 500k pot without the coil in the circuit creates a voltage divider against both zener diodes...

[@kuro68000]

Thanks, been looking at this myself lately. I want to make something battery powered but with GPS to discipline the clock. Should avoid using the word "cheapo" though, it's racist. I appreciate it's in the original code, not something you wrote.

[@sundhararajanganesan9396]

???

[@kuro68000]

@@sundhararajanganesan9396 Sorry it was in part 1, in the Arduino source code that Martin used.