Author Topic: The (long) story of the relay clock  (Read 9689 times)

0 Members and 3 Guests are viewing this topic.

Offline schmitt trigger

  • Super Contributor
  • ***
  • Posts: 2310
  • Country: mx
Re: The (long) story of the relay clock
« Reply #50 on: December 20, 2019, 04:27:52 pm »
I have said it earlier, but will say it again:

Magnificent project. I wish I had 1/100 the patience you have.
 
The following users thanked this post: Spemo

Offline SpemoTopic starter

  • Regular Contributor
  • *
  • Posts: 53
  • Country: de
Re: The (long) story of the relay clock
« Reply #51 on: December 20, 2019, 05:54:17 pm »
I have said it earlier, but will say it again:

Magnificent project. I wish I had 1/100 the patience you have.

Thank you very much. It sure was a long way, but atleast I feel comfortable about giving these someone else, if anyone wants one (who knows).
 

Offline digsys

  • Supporter
  • ****
  • Posts: 2209
  • Country: au
    • DIGSYS
Re: The (long) story of the relay clock
« Reply #52 on: December 20, 2019, 09:45:11 pm »
Boards came up nice, especially with the black resist. Are you doing a final video with the sound?
Hello <tap> <tap> .. is this thing on?
 
The following users thanked this post: Spemo

Offline Cerebus

  • Super Contributor
  • ***
  • Posts: 10576
  • Country: gb
Re: The (long) story of the relay clock
« Reply #53 on: December 20, 2019, 11:06:26 pm »
Fantastic project!!
PS: do you plan a wrist version of the relay watch?  :D

sure, why not!
3 Kilograms in weight, 50000mAh Batterypack to get through the day, and a massive 35x22cm case on your arm, does that sound ok for you? ;D
Price would be 1599,99$...you know, because of the wrist strap.

At least it'd be a bit lighter than this wristwatch, though perhaps not quite as accurate.

Anybody got a syringe I can use to squeeze the magic smoke back into this?
 
The following users thanked this post: Prehistoricman

Offline SpemoTopic starter

  • Regular Contributor
  • *
  • Posts: 53
  • Country: de
Re: The (long) story of the relay clock
« Reply #54 on: December 21, 2019, 08:54:46 am »
Boards came up nice, especially with the black resist. Are you doing a final video with the sound?

Yes, once I have made a case for it I will make a video with sound.



At least it'd be a bit lighter than this wristwatch, though perhaps not quite as accurate.



Yep  :-+
 

Offline george.r

  • Contributor
  • Posts: 20
  • Country: 00
  • not to be confused with other georges
Re: The (long) story of the relay clock
« Reply #55 on: December 23, 2019, 06:23:16 am »
Brilliant project, and commendable dedication.  :-+

Great to see the learning curve here
Copper scribbler, garbage hoarder
 
The following users thanked this post: Spemo

Offline SpemoTopic starter

  • Regular Contributor
  • *
  • Posts: 53
  • Country: de
Re: The (long) story of the relay clock
« Reply #56 on: January 04, 2020, 08:34:10 pm »
Hey everyone,

I had some time making a short video of the latest version.

Check it out:




Have a nice day

Daniel
 

Offline DaJMasta

  • Super Contributor
  • ***
  • Posts: 2344
  • Country: us
    • medpants.com
Re: The (long) story of the relay clock
« Reply #57 on: January 04, 2020, 09:16:40 pm »
Have you run into any reliability issues in all your versions?  Had any relays wear out?


At a cycle per second, I suppose it would take a while, but I'd assume in the long term, the relay(s) switching the flashing dot would be the first to go, since they're switching 60 times for every minute hand switch.  If you end up doing another version some time, you could use two relays to switch each flashing dot, SPDT or similar where one input for each relay is high and one is low.  Then if you had a pair of LEDs in opposite polarities between the two relays, you could drive each relay half as many times per minute, so long as they are offset in phase by 90 degrees, and get the dot to light whenever one relay is connected to power and one to ground.

Anyways, with the lifetime of a relay and a one cycle per second switching, it may be entirely unnecessary.  Neat project with a very satisfying sound!
 
The following users thanked this post: Spemo

Offline SpemoTopic starter

  • Regular Contributor
  • *
  • Posts: 53
  • Country: de
Re: The (long) story of the relay clock
« Reply #58 on: January 04, 2020, 09:46:33 pm »
Have you run into any reliability issues in all your versions?  Had any relays wear out?


At a cycle per second, I suppose it would take a while, but I'd assume in the long term, the relay(s) switching the flashing dot would be the first to go, since they're switching 60 times for every minute hand switch.  If you end up doing another version some time, you could use two relays to switch each flashing dot, SPDT or similar where one input for each relay is high and one is low.  Then if you had a pair of LEDs in opposite polarities between the two relays, you could drive each relay half as many times per minute, so long as they are offset in phase by 90 degrees, and get the dot to light whenever one relay is connected to power and one to ground.

Anyways, with the lifetime of a relay and a one cycle per second switching, it may be entirely unnecessary.  Neat project with a very satisfying sound!
Yes, I have run into reliability issues with the hand wired designs, and everytime I had the relays in regular IC sockets on the hand wired ones. They often lost contact because the perfboards were quite flexible.

Also, if you look at earlier posts in this thread, you see how much wireing is needed to make one of these work - something has to fail at that point.


with the relays, I purchased used relays from ebay at 0,20€/piece, so I don't know how long they will last - some have failed already, some came unuseable right away. You pay cheap you get cheap.


as for getting the 1Hz signal and the blinking dot:

to get these clocks somewhat accurate, I have to use a crystal and some ICs, they put out a 1Hz signal with 50% duty cycle, the blinking dot is connected to that output through a transistor(so no problems there). But, the counter modules require a duty cycle of 15 to 20% (on time), so there is a circuit with 2 relays converting that.
It works like this:
The first relay is turned on by the output of the ICs, shortly after that, the second relay turns on and breaks the connection between the first relay and the IC. With the other set of contacts of the first relay and a power connection the short pulses for the counter modules are created.

 So every seconds 2 relays turn on, which would obviously fail first.

I am not worried about contact failures inside the relays, since they only have to switch about 100- 200mA, which is 5 to 10% of their maximum rating. So either the coils go weak or the mechanical parts fail.


If I find a way to replace this duty cycle converting circuit with something, I might do it - but that would also mean it loses some of the ticking noise. It's what you hear most of the time.


 

Offline Prehistoricman

  • Regular Contributor
  • *
  • Posts: 216
  • Country: gb
Re: The (long) story of the relay clock
« Reply #59 on: January 04, 2020, 10:13:14 pm »
You are nuts! I thought it was crazy when I made an EEPROM programmer on perfboard that has 30-40 wires.

I watched the video and you have the worse case of European number 1s I've ever seen. It's like an M got cut in half.
Anyway they're very cool. The night dimming function is great. I like the orange LED one.
Does it make much more sound when the hours turn over?

Why do you use so many relays? 4 7-segment displays and 1 segment for the seconds surely means you only need 29?

Offline SpemoTopic starter

  • Regular Contributor
  • *
  • Posts: 53
  • Country: de
Re: The (long) story of the relay clock
« Reply #60 on: January 04, 2020, 11:22:31 pm »
You are nuts! I thought it was crazy when I made an EEPROM programmer on perfboard that has 30-40 wires.

I watched the video and you have the worse case of European number 1s I've ever seen. It's like an M got cut in half.
Anyway they're very cool. The night dimming function is great. I like the orange LED one.
Does it make much more sound when the hours turn over?

Why do you use so many relays? 4 7-segment displays and 1 segment for the seconds surely means you only need 29?


Hey, that thing with the number 1 : look closely, number one is just a simple / , the other thing is a 7 ;)

I use this many relays because this clock uses decimal counters - so:

10 digits for 10 seconds (=20 relays + 2 to carry over to the next module)
6 digits for the tens of seconds (=12 + 2 relays)
10 digits for 10 minutes (20 + 2)
6 digits for minutes (12+2)
10 digits for 10 hours (20+2)
and 3 for hours (=6 relays)

In this version I just have no 7 segment displays for the seconds - the circuit still has to be there though. The relays are doing the counting.

Startup: 2 relays
1Hz signal: 2 relays
24 Hours reset: 3 relays
auto dimming at 20:00 : 2 relays

Total: 109 relays

The more it has to turn over at once the louder it gets, so turnover from 19:59 to 20:00 is where you get the most noise. it has to turn over all the modules plus the dimming circuit.

Have a nice day

Daniel
 

Offline Prehistoricman

  • Regular Contributor
  • *
  • Posts: 216
  • Country: gb
Re: The (long) story of the relay clock
« Reply #61 on: January 04, 2020, 11:40:33 pm »
Hey, that thing with the number 1 : look closely, number one is just a simple / , the other thing is a 7 ;)

No, on your handwriting on the underside of the orange LED clock.

I didn't understand that you used relays for counting rather than only switching the LEDs. That makes a lot of sense actually.


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf