Sand Clock - critique

[dsage]

I recently built the sand clock. I guess I was initially taken by the cool nature of it but now that I have it built I find it a bit annoying having to listen to the vibration of the erase cycle every minute (or more).You certainly cannot place it anywhere "typical" of having a clock - like in the living room while watching TV or something. Anyway, it is what it is I guess. Now that I have it, I don't use it for this reason.
But on that topic can anyone enlighten me as to what the program is doing to decide when to plot the time. I would have expected it to watch for a change in the minute data and initiate a new plot. But on occasion my clock plots the time several times per minute. Sometimes twice in a row and it's often the same time plotted. I suspect there is something wrong with the firmware.
This issue makes the clock even more of an annoyance.

Any insight would be helpful.

If anyone comes up with a retrofit to perhaps run a fine brush across the sand or something I would welcome hearing about it.

Sage

[SteveThack]

I've had several issues with this clock and will go over them here.

The first thing to say is that the plastic parts are beautifully cut and look great. The parts all go together really well. It's a good design mechanically.

1/ One area I'm dubious about is the several places where an ordinary machine screw is screwed into the plastic, cutting its own thread on the way. It leaves the plastic under considerable stress and in my experience with acrylic and similar, this invariably causes fine star-like cracks to radiate from the hole after some time - months or a year or so. I think making a machine screw cut its own thread is poor engineering and I hope the design might be reviewed to either tap the holes in advance (one is already tapped) or use nylock nuts.

2/ The holes in the base of the sandbox for the M2.5 machine screws have been made slightly too large. Again, these are supposed to be threaded as you force the machine screws into them, but in mine the screws went in very easily and immediately stripped the thread. In the end I had to drill out the holes to 2.5mm and fit longer M2.5 screws with nuts on the underside.

3/ The wires from the vibrator motors are only just long enough, and are soldered to the PCB. If you need to remove and replace the sandbox the soldered joints are pulled somewhat and may well fail. I would recommend the PCB is modified to use a connector of some sort for the motors, such that the motors can be unplugged, allowing the sand box to be removed completely.

4/ I have found the same thing as dsage: sometimes the software goes mad and it erases and writes the time in the sand several times in rapid succession. I've noticed something interesting: sometimes, after writing the time, the right hand servo isn't returned to its normal home position. Whenever that happens it immediately erases and writes the time again. I've had it do this several times in one minute, until eventually the right hand servo goes back to the home position and normal service is resumed.

It may be that the non-homing of the right hand servo is a red herring. All I can say is that I've noticed that pattern on mine. Perhaps there is more to it.

4a/ While I've been typing this, I've noticed another bug which might be related. Sometimes, after erasing the sand, the servos move the pen to the left, but then stop with the pen still in the air. After a brief pause the pen is homed again and the sand erased once more. Mine has just done three such erases in succession without writing the time.

5/ I can tell from reading the article that Elektor have had a great deal of trouble getting the vibrator motors to erase the sand satisfactorily. Well, let me tell you: they still haven't cracked it. I assembled it exactly as described, and frankly the sand isn't smoothed out at all well after a five second burst of vibration.

I got an immediate improvement by melting a large blob of solder onto each eccentric motor weight, increasing the vibratory force. Of course this might cause more rapid wear of the motors, so I would recommend you keep that in mind if you want to try it yourself. This is another area that needs revising.

6/ dsage is exactly right: this clock is far too noisy to have in your home unless you live in a castle. There needs to be a convenient on-off switch so that you can disable the erasing and writing, switching it on only for display purposes. Yes, you can just unplug it, but this is less convenient and of course it doesn't keep the clock battery topped up like that.

Oh, one last thing: the calibration process doesn't work that well. Despite following the process to the letter with great care, my numerals all lean to the left, and get larger from left to right. I'll be fine-tuning the calibration to sort that out.

In summary, this is a fantastic idea and the kit is really well made and looks great. Forcing machine screws to cut their own thread is a bodge and that aspect needs revisiting. (Probably just pre-threading the holes will be enough.) The software bugs are so obvious I'm astonished they weren't found during testing. Mine malfunctions a few times every hour. Finally, the sand smoothing doesn't work very well at all. This is another area that needs revisiting. This kit needs a Version 2 update as soon as possible.

[sigo]

Dear Sage and Steve,
We will work through it and come back to you asap.

Ralf Schmiedel
Elektor customer Service

[planofencecompany]

Sand clocks and hourglasses combine history. Information about hourglasses, also known as sand clocks.

[Jessica1]

hi everyone,
i often wondered myself the physics behind the sand clock..
now,imagine a 1 hour sand clock or hour glass (the sand in upper part takes an hour to settle down completely).if anyone knows the physics of the sand clock,kindly let me know the following:

1) How much is the effect of gravity on the sand particles inside the clock ?
2) Normally the sand clock is kept vertical to the ground plane.If the clock is tilted to some extent making an angle to the ground,will the sand clock measure the time as usual (1 hour) or does time vary

[tiwarikalpna582]

thanks