General > General Technical Chat
Cheapo clock tear up.
<< < (9/13) > >>
MK14:

--- Quote from: emece67 on January 05, 2022, 05:57:55 pm ---No need for any ratcheting mechanism, you can remove the whole wheel train and the rotor will still turn always in the same direction. The asymmetry of the electromagnet poles will do the trick. The friction wheel seen in the photos (that with crescents) introduces the necessary friction to allow manual hand setting the mechanism.

Interestingly, such uni-directional motors can be forced to walk backwards applying, instead of a single pulse, two adjacent pulses of different polarities (and the first one being shorter than the other). The first (and reverse) pulse starts the reverse movement, inertia makes the rotor pass the mid-way point, then the 2nd pulse completes the backward movement.

Regards.

--- End quote ---

If I remember correctly. When the battery, eventually gives up the ghost, but before changing to a fresh AA battery. The clock can go into this funny mode. Whereby it lacks the strength to operate correctly, so the time remains fixed at some point. But the clicking (ticking) and seconds hand, continue, where it flutters, between say, 29 seconds and 30 seconds, then 29, 30, 29, 30 ...
Which I assume, is perhaps related to the mechanism of being able to go backwards, if the pulses were arranged in the way you describe.
It sounds like a fun and cool project, to make a typical mechanical/analogue (battery) clock, run backwards in time. Presumably if you re-arrange the digits on the face, it could read the time correctly, but backwards.

The (typically) black blob IC (chip on board, COB), is also of special interest for me. I sometimes wonder how fast it could be clocked, as it would make a powerful binary counter/divider. If they can be adjusted, to correct for quartz crystal calibration errors at the factory (some people seem to say they can be, but I haven't seen a definitive source, that I remember). Then there must be something more complicated inside, I wonder what ?. Maybe a further binary counter, and (presumably) some kind of mechanism (really cheap in massive bulk quantities), that allows it to retain the calibration setting. Perhaps an array of diodes/zeners  (blown to change the bit, by high voltage pulses, at the factory), like very early PROM devices.
Surely the price point, doesn't allow it to be an OTP MCU, which would be an obvious solution, but a bit too expensive for that kind of application. On the other hand, there are thread(s) here about $0.03 MCUs, so maybe.
MK14:

--- Quote from: tooki on January 05, 2022, 08:16:46 pm ---Posting an interesting original post doesn’t give you the right to be the colossal asshole you’ve been to every single person who replied to you here.

--- End quote ---

In my observation in this thread. If your very first post (or maybe later ones), disagrees at all, with his opinion and/or (apparently) massive list of weird rules and regulations. He immediately adds you to his ignore list, replies to your post, but the post ends with words to the effect, 'I'm ignoring any responses you might make'.

That is NOT the way friendly open/free forums, such as this one works. Especially technical forums.
PlainName:

--- Quote ---and seconds hand, continue, where it flutters, between say, 29 seconds and 30 seconds, then 29, 30, 29, 30
--- End quote ---

We had a clock which, when the battery was low, was OK on the 0-30 side but lost time on the 31-60 side, presumably due to a very small imbalance in the second hand.

We do actually have a clock that I had written off as not worth taking more effort over than clicking a 'buy now' link on Amazon - it seems to work but gradually loses time, and not at a trivial rate. Thanks to this thread maybe I'll dig into it and see if it can be saved...
DrG:

--- Quote from: dunkemhigh on January 05, 2022, 04:05:52 pm ---
--- Quote ---I have, on occasion, thought about getting something that I could interface, e.g.,
--- End quote ---

I got around to doing that once. It kind of worked, but positioning was critical and I wound up having to stagger writes to the oLED display otherwise the signals in the ribbon cable would interfere with the RF decoding. But commercial clocks with similar gubbins also suffer from that (we seem to be in an RF hollow - even TV reception is dodgy), and the clocks typically only sync at night and when most of the toys are turned off.

If I were needing to do it again, I wouldn't really consider it. With easy WiFi you can do faultless ntp on demand, and if you're even lacking that a cheap GPS module would likely be sufficient. Ain't gonna run for 10 years off an AA, though.

--- End quote ---

Agreed. The WiFi route makes sense and I have done that, but at least in my case, the controller logs onto my ISP through the router. At some times in the past that was a chronic problem. In fact, I even built and used a circuit to reset the router every night. Later, I had the router on a remote switch to reset it easily from various locations.

True, my current router and ISP are much better and I almost never reset it for any reason at all, but to not have to rely on the ISP at all is a plus.

Of course, that plus becomes a minus when some of the experiences you report occurs.

I was looking again at this product during morning coffee,  https://universal-solder.ca/product/canaduino-60khz-atomic-clock-receiver-module-wwvb-msf-jjy60/ and it still interests me. I was looking at some of the GitHub code for projects folks have used and it looks quite feasible to me (there is even an Arduino project https://github.com/ahooper/WWVBClock and stand-alone projects as well as stand-alone projects https://github.com/bhall66/WWVB-clock).

But, I have no earthly idea how to evaluate how good or bad that antenna might be. It is not even clear to me if the new V3 version improved that aspect  Can you [or anyone I am not ignoring] get a sense for how good it [is] from the specs they give...


--- Quote ---The advantages over other modules from other manufacturers:
– wide operating voltage range 2* to 15V
– 60 mm tuned high-gain, high-Q antenna
– current consumption <100µA (LEDs off)
– super high sensitivity of 0.4V (RMS)
– power down control pin
– fast start-up
– AGC control on/off

--- End quote ---

Right now I am still waiting for the new clock from Amazon which should come in tonight. If the WWVB works for the clock, I am thinking (rightly or wrongly) that the board would work also.

I mean, it's not like I need another project :) but this is something that I find interesting.
PlainName:

--- Quote ---Can you ... get a sense for how good it from the specs they give...
--- End quote ---

Sorry, it's meaningless to me! And I suspect only meaningful in comparison to something else. For $20 it could be a few evenings worth of fun decoding the signal. And waiting a couple of minutes for a stream to almost complete before detecting corruption.

Navigation
Message Index
Next page
Previous page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod