My long term goal is to build a large multi faced clock outside in a cupola (old school with hands to tell time). I need to start somewhere so I want to build a small clock for my desk ... a scaled down model if you will. I can even build just one side of the tower but I need to know what ever I start with will be expandable to the 4 sides (or maybe I build 4 completely separate clocks ????).

I have been reading and just too many ideas. Do I use a stepper motor? Do I use a servo (just a geared motor) with an encoder feeding back a pulse for location? This would be a large clock ... say 24" long hands (but balanced so minimal weight, just rotational mass).

Then there is timing. I see many of these controllers don't keep time well and I might need to think about a GPS time signal to "correct" the clock every 12 hours.

Well I need to start somewhere and since I have never touched an Arduino or Raspberry, I need some advice.

I am not even sure which is better or why use one over the other.

I don't even know if there is enough I/O in either. Lets say I build 4 separate clocks (4 sided tower), each with 3 motor (hour, min, sec). Plus inputs for "zero" switches on each of the hands. Or if I go with encodes then 3 encoders per side.

As I said, I have been reading up on this for a few weeks (actually had the idea for about a year now). I just don't have enough knowledge to make an informed decision.

Any advice would be greatly appreciated!

Thanks ... Mike

 

You don't need any of them for a digital clock. It can be made without an MCU.

There are also specialized chips for a real time clock, the Arduino or any MCU reads the time and does whatever you want. They are with an iterface so you can spare the pins. Can you post a circuit?

 

Focus on the mechanics first- that's going to be your biggest challenge.

Three coaxial shafts, home sensors, and a geared or timing belt drive, that cannot slip.
A 3:1 ratio is good, then you get a ratio that is a factor of 60, so the hands line up perfectly.

You could use 1 computer and some stepper drives that can be controlled via a serial port:

https://www.trinamic.com/products/modules/details/tmcm-1110/

 

Your best bet would be to install four individual AC powered mechanical clock movements.
You cannot get better time keeping accuracy than the AC mains frequency (until you get a power interruption).

 

Thanks ...

So, I want an old fashion analog clock (not digital).

I did look at buying movements ... wow ... I was looking at $500 per side for just the movement and then I am not sure how well they would work here up north ... lots of warnings about not holding hands (we get snow and ice here in VT all the time).

I am a machinist by trade and have a lathe and mill here at home. The movement is "relatively" easy for me to build.

This is where I started questioning a step motor vs a motor (ac or dc) with say an encoder feedback (we do this on CNC machines) or ????

When I started reading, the 200 steps did (and still does) give me a "stop and think" moment.

In my mind (and people on other forums have said I am wrong but I just don't understand) that there is no way even with micro stepping to get 60 steps out of a 200 step motor!

Thus I was thinking of a timing belt ... 3:1 ... gives me 600 steps. Then I would move 10 steps per second.

I have the same issue with an encoder ... not easy to find ones that are divisible by 60 (and affordable).

A true AC clock works because there is limited slip so no error. If my clock is fairly big, I might have some weight to it (plus as I said the snow and ice load) ... I think it would never keep accurate time ... plus, building a simple 3:1 belt drive is one thing, building a complete gear system is way over my abilities (well unless I want to make a full time hobby of it).

Just FYI ... I was thinking about mechanical slip clutches in the shafts just in case it gets covered in ice so I don't burn out the motors.

Thanks again .... Mike

 

You don't need encoders, total overkill.

If a step motor stalls from mechanical overload (iced up) no damage will occur, it will just be out of whack until you home it, which should be done every 12 hours.

 

Awesome! I was hoping that was the case. I tried to read up on step motor stall ... I saw where it could get warm ...didn't know if that meant it would be damaged.

So maybe back to my original question ... assuming step motors .... Arduino or Raspberry?

I looks to me like either will drive step motor controllers.

I see a GPS shield for the Arduino so I can get a time stamp off a satellite ... I don't know if the same thing is available for the Raspberry (I don't want to depend on the internet ... unless there is a good reason to).

Mike

 

You are over-complicating a decision thats simpler. To get 60 seconds from a step motor with 200 steps, you just run the motor continuously and count 1 step for 1 second, there is no need to count all steps. The second time you will move the arrow with another 60 steps and so on.

An analog clock means no MCU. Any MCU can drive a motor IC, but not any MCU has a ready GPS shield for it.
The synchronization is easy, you put a real time clock IC and the arduino makes the clock wait a minute or 2 at 1 after midnight.

EDIT:
There should also be a LED to indicate if the clock is out of sync.

This is to get you started.
Step motor driver. Check if it suits your needs.

 

Awesome!

Looks like I will put some parts on order.

It will give me something to play with for the next few months.

I poured the workshop pad last fall but decided to wait till spring to start building the new workshop ... good thing, we have had two good blizzards here in VT in January ... one left 18" of snow and the other 23"

This will be a great learning experience.

Thanks again for all the advice !

Mike