Water Clock (fully Mechanical, So No Circuits, No Electrics)

When I saw the challenge from Instructables about clocks, I thought: "that is not for me". I do not know much about circuits and I do not own the precision tools that are needed to make a mechanical clock. But it made me thinking about how people in previous centuries used clocks and I learned that the water clock has been in use for thousands of years in many variations, with gears, siphons, dropping balls, bells and so on. There are endless possibilities, so I decided to use my creativity to build something that shows the passage of time using only water and gravity to power the clock.

In the next steps I will first explain how a part works and then how I made it.

The top part of the clock (shown in the red circle) consists of a jar filled with water and a floating siphon. The purpose is to get water dripping down at a stable rate.

The water droplets from the floating siphon are collected in the second siphon. That siphon is not floating but glued to the stainless steel cup. When the water level in the stainless steel cup reaches the top of the siphon, the cup empties quickly. So this part of the clock slowly increases in weight and then the weight drops quicky.

Since the stainless steel cup with the fixed siphon is placed on one side of the balance, the other side of the balance moves in the opposite direction. The left side of the balance has a mechanism that can rotate the wooden disk with the numbers 1-12 written on that disk. This disk can rotate in one direction only.

The water that flows out of the second siphon comes in the jar at the bottom of the clock. That jar is there to collect the water until I drain the jar and poor the water again in the jar at the top.

If I would have used just a jar filled with water and with a small hole at the bottom, I would have got a lot of drops per minute when the jar was full and just a few drops per minute when the jar was almost empty. I tested this and found that a full jar gives about 100 drops per minute and an almost empty jar gives about 4 drops per minute. This is because the water pressure on the hole is higher when the jar is full.

I wanted a constant dripping rate, so I used a floating siphon. As the height difference between the water level in the jar and the end of the siphon outside of the jar is fixed, the water pressure driving the water through the siphon is constant and therefore the amount of drops per minute is also constant.

The first photo shows the result of my tests of the variation in dripping rate when I did not use a siphon.

The second photo shows my floating siphon when I was testing it. The float is a piece of Styrofoam, the siphon is made from a copper tube. The washers that hang at the left are used to counter balance the fact that the siphon is only supported at one side, so it tends to tilt. The wooden pieces are used to keep the siphon in the middle of the jar and to restrict the movement of the siphon to an up-and-down movement. The fork is used to prevent the siphon to rotate left and right.

To start the siphon I have to suck the air out once. Air will only get back in when both ends of the siphon are in contact with air and that does not happen when the float lands on the bottom of the jar. Then the water stops dripping, but as soon as I add water to the jar again, the siphon immediately starts to drip again.

I used a piece of Styrofoam and cut some circles with a heated wire cutter. The diameter is chosen as large as possible while still fitting in the jar. The thickness should be enough to get it buoyant even with the weight of the siphon.

The actual siphon is made of a copper brake line of 5 mm outer diameter. Do not use a brake line that has been used so it is not contaminated with brake fluid. The advantage of using a brake line is that it is easy to bend by hand without any tools.

I placed a piece of plastic hose over the brake line and used a plastic washer, so the siphon will stay on top of the Styrofoam float. During testing I kept the plastic hose in place with a hose clamp, but that started to rust, so in the final version I fixed the small piece of hose with hot glue to the siphon.

The position of this piece of hose is important as it determines the difference between the water level and the end of the siphon and therefore how many drops per minute come out of the siphon. (A higher flow rate works better, but drains the jar more quickly.)

The copper side piece serves as an indicator and it also holds the washers to prevent the siphon tilting to much. It was soldered to the main piece. Alternatively it could also be glued.

The wooden pieces were just scrap pieces of wood, so there are more holes than needed. Just a central hole is needed for the copper siphon. To reduce friction, I used a small aluminum tube in the wooden hole so the copper siphon can move up and down more easily as the copper only touches the aluminum and not the wood.

Keeping the friction really low is important to prevent that the siphon does not drop down but gets stuck.

Since I had a side piece on the floating siphon, I could as well use it to indicate the level in the top jar. I cut a piece of copper tube with a diameter of 12 mm and used the pipe cutter to make some lines on the copper. The distance between the marks are equal to the amount the level drops in one hour.

As you can see my jar can only hold enough water for a period of 4 hours, so after 4 hours I have to refill the top jar. I will use it only during a working day, so in the morning I work from 8 till 12 o'clock and in the afternoon I will work from 1 PM to 5 PM. When the top jar is empty, it is time for lunch or time to close for the day.

I used a marker to write the time on the copper tube. At one side it counts from 8-12 and at the other side it counts from 13-17 h. The copper pipe is placed over an all-thread, so it can easily be rotated. I screwed the level indicator to a piece of wood to support it.

Water from the floating siphon drips at a constant rate and lands in the stainless steel cup. The copper pipe of the second siphon starts in the cup at the blue line. The level in the cup slowly rises until the red line. Then the water level has filled the siphon to the top and the water starts to flow out and the level in the cup drops quickly. When the water level reaches the blue line again, air can enter into the siphon and the water no longer flows out.

Ideally it takes exactly 5 minutes to fill and drain this cup, so measure how much water drips out of the floating siphon in 5 minutes and make that the difference in level between the red and blue line. If you did not nail that exactly, you can either adjust the flow rate of the floating siphon by repositioning the position of the piece of hose or you can put something in the stainless steel cup to reduce the volume of water that the cup can hold between the red and blue line.

I drilled a hole in the bottom of the stainless steel cup with a diameter of 5 mm. The siphon is made of the same copper brake line as the floating siphon, so it also has a diameter of 5 mm. I glued the copper to the cup using an epoxy glue.

Then I cut a piece of wood and made a slot in it, so the cup can stand on the wood without putting any weight on the copper pipe.

The stainless steel cup of the second siphon is placed on one side of a balance. On the two photos above you can see how the balance moves when I increase or decrease the weight. For testing purposes I used two aluminum bars on the location where the cup with the second siphon is placed. Two small aluminum bars have approximately the same weight as the cup just before the second siphon is draining the water. One aluminum bar has approximately the same weight as the cup just after the siphon has drained the water.

The vertical slat that is clamped in the vise on the photo is the part that does not move.

At the left side I screwed a slat at an angle and placed some nuts as weight on that slat. I used nuts so I could easily add or remove a nut to adjust the weight and find a good balance. The slat is screwed at an angle so there is more variation in distance between the nuts and the central vertical slat that is fixed in the vise, so the torque force changes.

I built the balance from 4 slats and I drilled two holes in each slat. The distance between the holes is the same in each slat and is approximately 5". The actual distance is not very important, as long as the distance between the holes is the same in all 4 slats.

At first I used a small hole with a nail as pivot, so I could test how the balance was moving. The test went well so I replaced the nails with bolts and nuts. It is important to keep friction really low, so I used two nuts on each bolt and tightened the nuts against each other so the slats could still move easily.

I did not want to paint the wood because then I would have to wait for the paint to dry. A quick alternative is to slightly burn the wood so I used that method and I like how that looks.

The left side of the balance is moving up and down, but I wanted to use that movement to rotate a dial to indicate the time. The brown locking pin at the top makes sure the dial can only rotate in one direction, which is clockwise. The three photos show how the dial is moving every time that the left side of the balance is moving down. When the left side of the balance is moving up, the dial is not moving because the brown locking pin blocks the dial.

So every time the second siphon is being drained, the left side of the balance will move down and it will rotate the dial one increment. Therefore the contents of the second siphon should be filled and emptied once per 5 minutes.

I first made a paper template for the dial. It consists of two circles and I divided the circles in 12 equal sections of 30 degrees. Then I drew lines from the outer circle to the inner circle exactly on the 12 sections. The size of the circles is not that important, but it is very important to make the 12 sections equal in size.

I used the paper template to make a cardboard version of the dial so I could test if it worked. I also added a small milk can to hold the nuts as weight on the balance.

Testing went well, so I then made a wooden dial with a piece of all-thread as central axis and with several washers as spacers.

The dial is acting as the minute hand of a normal clock. Since a normal clock uses the numbers 1-12, I also used those numbers. For example: if number 3 is on top, it means that it is a quarter past the hour. Alternatively, I could have written numbers 5-10-15...55 instead of 1-2-3...12.

At first I wrote the numbers 1-12 on the dial in the same way as on a regular clock, but that was not correct. My dial is moving clockwise, so at first my clock was counting backwards.

I sanded the dial a bit to remove the incorrect numbers and then wrote the correct numbers on the dial.

On the photos of this step you can see some more details of my contraption, including a photo of the back side of the balance.

I put all elements together on a base plate. All the wood is just connected by screws or by glue. The black support for the top jar is made of steel and is actually a coat rack that I bought at a thrift store.

Now my clock is ready.

I hope you like it.