0%
0%

# doodle clock #2

robotic writing clock

Similar projects worth following
11.6k views
i built the first doodle clock http://hackaday.com/2012/03/14/robotic-doodle-clock/ 3 years ago as a joke. But to watch the clock work was so mesmerizing that i wanted it to be a practical desk clock. The biggest problem with the first clock was the drying up of the markers after just 30 minutes of working. So the solution to that i found was the magnetic writing boards http://www.walmart.com/ip/Fisher-Price-Doodle-Pro-Basic-Purple/26012794 made for children. I used 2mm cylindrical magnets inside a solenoid to write and erase the text. Next i replaced the servos with small geared steppers to make the clock very silent and smoother. And finally all parts were designed and 3d printed . I also got the final pcb manufactured .

The clock uses two 1:100 geared 15mm stepper motors to move the arms configured in a scara fashion. The motors are located at the base to keep the weight very low on the arms. The tip of the arms contains two solenoids with cylindrical magnets inside them. I initially tried electromagnets but to get the text to be as dark as what is written by the magnet which comes along with the board needed a lot of power. The entire clock is run by a atmega 644p with a arduino bootloader. The motors are run by the standard stepstick and the coils are run by a l293dd dual h-bridge.

The kinematics for the arm was solved by a user on the reprap forum.

the only problem that remains is that the board gets scratched only after a day of use. i am looking into covering it with a scratch resistant film or plexiglas.

kinematics

We have x,y and we need two angles: one represents the height and the other the angle
first we need distance from x,y to axis :h= sqrt(x^2+y^2)
now angle between arms with law of cosines: angA= (L1^2+L2^2-h^2)/(2*L1*L2)
the angles of the parallelogram always add up to 360: angB=180-angA
calculate real height of arm: hA=sen(angB )*L2
height of arm's joint: hB=y-hA
ang2=asen(hB/L1)
ang1=angB+ang2

angB=180-ACOS((L1^2+L2^2-x^2-y^2)/(2*L1*L2))
ang2=90-(ACOS((L1^2-L2^2+x^2+y^2)/(2*L1*RAIZ(x^2+y^2)))+(ATAN(x/y))
ang1=ang2+angB

• 1 × atmega 644p
• 2 × stepstick
• 1 × l293dd Interface and IO ICs / Peripheral Drivers and Actuators
• 2 × 1:100 15mm stepper motors
• 2 × 12mm x 18mm solenoids
• ### process

ekaggrat singh kalsi10/04/2015 at 10:21 2 comments

prototype 1:

the first prototype used one motor on the base and one motor mounted on the base of the arm which drove the upper arm by a thin cable . The backlash of both the motors added up making the test not so precise .

prototype 2 :

the second prototype I replaced the cable with a connecting rod as the cable kept slipping but the issue of the backlash remained.

prototype 3:

Share

## Discussions

ekaggrat singh kalsi wrote 10/11/2023 at 06:49 point

magnets are from aliexpress. I wound the solenoids on a brass core using some magnet wire.

Are you sure? yes | no

Vandini wrote 10/10/2023 at 09:05 point

Hi, could you tell me where you bought the spolenoids and the magnets? Thanks in advance

Are you sure? yes | no

Steve wrote 02/10/2019 at 17:55 point

This clock is so incredibly cool.  I would probably sit and watch it for hours.

Are you sure? yes | no

ekaggrat singh kalsi wrote 02/11/2019 at 07:57 point

thanks a lot for liking it.. already working on v3 lets see where this goes!

Are you sure? yes | no

Normanclanger wrote 04/05/2016 at 08:36 point

Hi - nice project.  How do you know the start position for your stepper motors?  Do you have the concept of a home position - can't see any micro switches to help you find home.  Thanks, Tim

Are you sure? yes | no

ekaggrat singh kalsi wrote 04/05/2016 at 09:20 point

there are extremely small ( 5mm x 5mm ) opto endstops hidden in the base of the arm.. but if i had to do it agiain i would use hall switches much less of a headache

Are you sure? yes | no

Thomas Sjobom wrote 10/08/2015 at 07:40 point

want one

Are you sure? yes | no

derek.bever wrote 10/07/2015 at 13:44 point

Are you sure? yes | no

ekaggrat singh kalsi wrote 10/08/2015 at 03:56 point

pposted on thingiverse

Are you sure? yes | no

bienonline wrote 10/07/2015 at 07:22 point

Vvery impressive, indeed. I want to reproduce your clock. Would you be so kind and publish a link to the description of the motors and offer the files for 3D-printing?

Are you sure? yes | no

ekaggrat singh kalsi wrote 10/08/2015 at 03:57 point

okay will share soon on github or youimagine ..

Are you sure? yes | no

DeepSOIC wrote 10/04/2015 at 19:37 point

Lovely! I want one =)

Are you sure? yes | no

Alex Rich wrote 10/04/2015 at 11:43 point
this is sweet, keep in mind the "description" of your project crops without telling you, that appears to have happened to your description

Are you sure? yes | no

ekaggrat singh kalsi wrote 10/04/2015 at 12:32 point

thanks . fixed it !!

Are you sure? yes | no

## Similar Projects

Project Owner Contributor

### HDD Clock

Bjorn Bosell

Project Owner Contributor

### Tired Clock

Michael Möller

Project Owner Contributor

### Edgytokei

ekaggrat singh kalsi

Project Owner Contributor

AberDerBart

# Does this project spark your interest?

Become a member to follow this project and never miss any updates