Easy water clock (+ battery clock in replies)

[Altaric]

First post around, so I apologize for my mistakes! I searched around and believe I've found a new and simplier water clock.

Automation only works on binary values, counters are possible but very big.

Liquids on the other hand allow for simple math : addition, and divide by 2.

 

This is how it works :

* Pump on the bottom left is only for initializing the system, can be removed after.

* First valve sets the size of your packets, and the speed of your clock.

* The flow is split in two, half goes forward, half feeds itself. This is so that packets are spaced by an empty pipe. the valve on top synchronizes the looping packets with the incomming packets so that the output is constant.

* The shutoff at the botom is the reset. It is triggered when the two pipes before it are not empty ; that is when you have stacked a full 10k packet there.

Clock period is 10k/valve flow.

The bit of automation below is just confirm the accuracy : edge detector when the shutoff flushes, and a buffer of the desired length => the output has 100% uptime.

 

Hope you like it!

 

[Gurgel]

Nice! I think this is clearer and simpler than other designs. And with tank in between the two sensors, you should be able to scale this to very long times as well.

[Saturnus]

Nice. A little more complicated than needs to be but otherwise clever use of a water loop to remove the miniscule power requirement of a normal water clock. For other variants and use cases you can check my old thread on the subject.

The traditional set up still has tremendous advantages in most cases even though it can use up to 2-3W in extreme cases.

 

[goatt]

I like it, but I also will like it to shrink by a lot. I also like “transitional” water lock, which I think can be shrunken and opened in my preference.

[AMBLCO]

There is only one thing bugging me with that desing and it's the buffer gate. When you reload it resets the timer so it isn't accurate anymore. Otherwise great design!

[Lilalaunekuh]

11 hours ago, Saturnus said:

up to 2-3W in extreme cases.

All most water clocks have a small power consumption and the consumption is greater than the daily drain of a smart battery.

=> I just started to abuse power transformer and smart batteries as timers

(0,66W at less <1s (0,75s worst case) deviation for up to 60 cycle timers)

(A power transformer and a smart battery can be set to any time between 10 and 60 cycles with very high precision.

You can use this method for timers below 10 cycles but than your less power efficient and you need more additional gates.)

[Altaric]

4 hours ago, AMBLCO said:

There is only one thing bugging me with that desing and it's the buffer gate. When you reload it resets the timer so it isn't accurate anymore. Otherwise great design!

The buffer is only used to test the precision of the clock. Real world applications don't need it ; only the AND gate above it.

2 hours ago, Lilalaunekuh said:

All water clocks have a small power consumption and the consumption is greater than the daily drain of a smart battery.

=> I just started to abuse power transformer and smart batteries as clocks

(0,66W at less <1s (0,75s worst case) deviation for up to 60 cycle timers)

(A power transformer and a smart battery can be set to any time between 10 and 60 cycles with very high precision.

You can use this method for timers below 10 cycles but than your less power efficient and you need more additional gates.)

That sounds like a compact clock too, and doesn't require water input to initialize it!

Theorecially, my clock can go up to 10000/0.1/600 = 166.6 cycles, but more realistically to around 10 cycles, then you loose precision in the period setting - which can be fixed with filter gate.

[Lilalaunekuh]

Here are 2 versions of my timer (it isn´t a clock, but you could make one out of it):

- (normal) power transformer:

60   cycles max runtime; 0,75s worst case precision; 0,66w consumption; 6s to reset

 

- small power transformer:

52,5 cycles max runtime; 0,75s worst case precison; 0,66w consumption; 21s to reset

 

=> If the size isn´t critical I would always build the first one.

 

 

[Guest]

3 hours ago, Lilalaunekuh said:

All water clocks have a small power consumption and the consumption is greater than the daily drain of a smart battery.

You can make a pumpless water clock. No power consumption.

This is 2 pumpless water clocks that control a sweeper for a multi-cycle enable/disable loop. Shutoff enables and resets the clock when 120Kg of water has passed the element sensor. Set the valve to control clock interval as per.

[Lilalaunekuh]

1 minute ago, nonoxyl said:

3 hours ago, Lilalaunekuh said:

All water clocks have a small power consumption and the consumption is greater than the daily drain of a smart battery.

You can make a pumpless water clock. No power consumption.

Messed with this idea, but I don´t use it because if you want precision the size can become a problem

Still thank you for pointing out my mistake

[Guest]

14 minutes ago, Lilalaunekuh said:

Messed with this idea, but I don´t use it because if you want precision the size can become a problem

both valid points. the size can make it a pain to find room for it and it lacks a high degree of precision.

[Saturnus]

3 hours ago, Altaric said:

Theorecially, my clock can go up to 10000/0.1/600 = 166.6 cycles, but more realistically to around 10 cycles, then you loose precision in the period setting - which can be fixed with filter gate.

If used as a cycle counter the traditional water clock can go to 9999 cycles with absolute precision though. Only if you want more than 10000 cycles run will you start having a problem. The same limitation of 9999 goes with any interval so if it's seconds then it's 16.665 cycles maximum with absolute precision.

[Altaric]

9 minutes ago, Saturnus said:

If used as a cycle counter the traditional water clock can go to 9999 cycles with absolute precision though. Only if you want more than 10000 cycles run will you start having a problem. The same limitation of 9999 goes with any interval so if it's seconds then it's 16.665 cycles maximum with absolute precision.

I mean precision in how precisely you can set the period time ; it will always run using the same period, but since the period is 10k/valve flow, you're stuck to the 0.1 increments of the valve. Additionally you can make the charging pipe longer, or even use a liquid tank as said @Gurgel.

 

@Lilalaunekuh Updated the post title so that people can find your idea.

[Saturnus]

2 minutes ago, Altaric said:

I mean precision in how precisely you can set the period time ; it will always run using the same period, but since the period is 10k/valve flow, you're stuck to the 0.1 increments of the valve. Additionally you can make the charging pipe longer, or even use a liquid tank as said @Gurgel.

The limitation on a traditional water clock is that hydro sensor can only reliably detect 100g interval when you exceed 100kg. Up to 999.9kg it's still absolutely accurate. And if you go above 1000kg then you start having multiple tiles of water and it should be noted that the bottom tile will increase above 1000kg so there goes your absolute precision out the window. Then again. I really can't imagine anyone needing to count more than 9999 cycles as that would break the current longest game record.

[Lilalaunekuh]

9 minutes ago, Altaric said:

@Lilalaunekuh Updated the post title so that people can find your idea.

Nice one (Thought about starting a thread for it but wasn´t sure^^)

 

If you want to turn my timer into a clock just subtract the reset time from your desired cycle time.

But deviation of the clock (<=0,75s) will add up each cycle.

=> If you need more precision you could go for a clock sensor that will enable the recharge.