Shop OBEX P1 Docs P2 Docs Learn Events
$2 Clock Kit - Page 8 — Parallax Forums

$2 Clock Kit

1234568

Comments

  • ercoerco Posts: 20,244
    Another weird thing is that to get anything displayed. I have to short both electrolytic caps before powerup to reset the micro. Residual charge prevents booting. This PCB is poorly designed, I think even a working clock is barely operative and marginal, so it doesn't take much to stop it. Component tolerance, solder flux, or wind direction might be enough.

    In contrast, the smaller rectangular PCB clocks work great, we've built two of those. You just don't know which one the eBay seller will send you.
  • erco,

    If the ebay seller has a good rating then tell them to get rid of all those crappy kits, and you will likely get a replacement. refund, or both.

    By the way when are you publishing your book? :P
  • ercoerco Posts: 20,244
    Thanks Genetix, I do enjoy lively communications with my sellers. :)

    Book? That monthly column for SERVO magazine last year nearly killed me! Actually I do enjoy writing and the monthly deadlines force me to produce. Jon and Gordon warned me going in, and BOY were they ever right. As much as I hate deadlines and time constraints, they keep me moving.

    I wrote my own material for an undergrad course in electromechanical toy design I created several years ago. The weekly handouts evolved over several years into some useful and understandable material. My students kept asking for various reprints for many years afterward. It was nice to see that many of them had saved and used their dog-eared and well-worn copies for several years. An expanded compilation of those might be book-worthy. Hmmm...

  • jmgjmg Posts: 15,140
    erco wrote: »
    Another weird thing is that to get anything displayed. I have to short both electrolytic caps before powerup to reset the micro. Residual charge prevents booting. This PCB is poorly designed, I think even a working clock is barely operative and marginal, so it doesn't take much to stop it. Component tolerance, solder flux, or wind direction might be enough.

    In contrast, the smaller rectangular PCB clocks work great, we've built two of those. You just don't know which one the eBay seller will send you.
    Do those have the same pinouts ? or did you mean 6 digit vs 4 digit ?
    Can you swap the micro from one, into the other, to check the code content vs PCB design ?

  • Erco, I would love to see those class notes! Of course the really modern thing to do is turn it into an online course with videos...
  • Actualy, even a piezo speaker does have polarity, despite, in schematics, it is generaly represented as a non-polarized component.

    To be true, after processing, they must be sorted, to separate the ones that have the positive pole connected to the shiny metal plate, from the ones that have the oposity polarity.

    Some of them does indeed come with a (+), (-) or even a red (positive) or black (negative) dot of ink, to indicate the polarity of that particular device.

    The reasoning behind the marking is the same as for coil-driven speakers: ensure unanimous phase when you need to parallel two units, or more.

    As for the displacement of its "cone", its so subtle (100nm/V, yes, 100nm/V, acording to one of the linked sources), thus you would need to apply 1.8VDc to fit a single P2 basic-geometry node element, sideways turned, underneath (OnSemi ONC18).

    https://electronics.stackexchange.com/questions/231460/how-to-determine-the-polarity-of-a-piezoelectric-buzzer

    https://en.wikipedia.org/wiki/Atomic_force_microscopy
  • David Betz wrote: »
    Cluso99 wrote: »
    J2 is connected across the speaker LS.

    So it's a speaker connection.
    If that's the case it's interesting that it is marked + and -. Do speakers have polarity?
    erco wrote: »
    @David Betz:

    I got two of the same 4-digit clocks you have (square PCB), neither of mine worked either. Nothing displayed at all until I removed the upper silicon diode, which feeds into the battery backup connector. Both have 88:88 displayed now, anyway. Yank your upper 1N4007 diode and see what happens. Something hinky going on here.

    BTW, I Googled and found your post at https://www.eevblog.com/forum/chat/c51-4-bits-electronic-clock-kit/
    Well, at least I'm not the only one. I couldn't find anything wrong with my soldering. I ordered a second kit thinking that the MCU was defective on mine but it hasn't arrived yet.
    I got my second clock kit. It isn't the same as the first one I bought. It has the smaller PCB. I'm not sure whether the program in the MCU is likely to be the same or not so I'm hesitant to try plugging it into my already assembled first kit.

  • jmgjmg Posts: 15,140
    David Betz wrote: »
    I got my second clock kit. It isn't the same as the first one I bought. It has the smaller PCB. I'm not sure whether the program in the MCU is likely to be the same or not so I'm hesitant to try plugging it into my already assembled first kit.
    You could build the second one, then when that is working, play swap the MCU.
    If you post pictures of the PCBs copper, side by side, it should be easy enough to check pin mapping.
    The open-drain design these PCBs use, is fairly forgiving of some pin swaps. The display will just get more scrambled.

  • jmg wrote: »
    David Betz wrote: »
    I got my second clock kit. It isn't the same as the first one I bought. It has the smaller PCB. I'm not sure whether the program in the MCU is likely to be the same or not so I'm hesitant to try plugging it into my already assembled first kit.
    You could build the second one, then when that is working, play swap the MCU.
    If you post pictures of the PCBs copper, side by side, it should be easy enough to check pin mapping.
    The open-drain design these PCBs use, is fairly forgiving of some pin swaps. The display will just get more scrambled.
    I'll try building it over the weekend. I also got a second 6-digit clock kit. It's the one with only a single button.

  • erco, I'm checking every day on P2 developments and erco's eBay finds.

    Spring is coming, and I need more of your eBay 'deals' for purposes as yet unknown simply because they are so cheap.
  • ercoerco Posts: 20,244
    MCUs are not the same, I swapped, no display. No smoke, still works.
  • erco wrote: »
    MCUs are not the same, I swapped, no display. No smoke, still works.
    Were you ever able to get the clock with the PCB like mine to work?

  • ercoerco Posts: 20,244
    No, back burner for now. Sucks!
  • David Betz wrote: »
    jmg wrote: »
    David Betz wrote: »
    I got my second clock kit. It isn't the same as the first one I bought. It has the smaller PCB. I'm not sure whether the program in the MCU is likely to be the same or not so I'm hesitant to try plugging it into my already assembled first kit.
    You could build the second one, then when that is working, play swap the MCU.
    If you post pictures of the PCBs copper, side by side, it should be easy enough to check pin mapping.
    The open-drain design these PCBs use, is fairly forgiving of some pin swaps. The display will just get more scrambled.
    I'll try building it over the weekend. I also got a second 6-digit clock kit. It's the one with only a single button.
    I built the second clock kit. This is the one with the smaller PCB. It works fine. I tried swapping the MCUs and the new clock didn't work with the MCU from the original one. I did buy a programmer for the MCU but I'm not sure when I'll have time to try to reflash the MCU on the non-working clock.
  • jmgjmg Posts: 15,140
    David Betz wrote: »
    I built the second clock kit. This is the one with the smaller PCB. It works fine. I tried swapping the MCUs and the new clock didn't work with the MCU from the original one. I did buy a programmer for the MCU but I'm not sure when I'll have time to try to reflash the MCU on the non-working clock.

    If the programmer has a read mode, you could try that. If the part is not secure you will get useful hex (but chances are it is secure....)
    Do you have spare 89C2051 ? - the non working one may be faulty ?
    The code I posted above, should at least let you program a stop watch onto the PCB.
  • jmg wrote: »
    David Betz wrote: »
    I built the second clock kit. This is the one with the smaller PCB. It works fine. I tried swapping the MCUs and the new clock didn't work with the MCU from the original one. I did buy a programmer for the MCU but I'm not sure when I'll have time to try to reflash the MCU on the non-working clock.

    If the programmer has a read mode, you could try that. If the part is not secure you will get useful hex (but chances are it is secure....)
    Do you have spare 89C2051 ? - the non working one may be faulty ?
    The code I posted above, should at least let you program a stop watch onto the PCB.
    I bought 10 blank MCU chips to play with but first I have to find the software that goes with the programmer. To make things worse, I'd like to find software that will run on the Mac. However, if necessary I can use a Windows or Linux machine.
  • I found that my programmer could read the hex file from one of the cheap 4-digit clock kits and copy that to a new blank chip. Of course, it's better if you can get hold of source assembly or C code with English comments, variable and function names - that makes changing the program much easier.

    The most interesting thing to do is to write your own code from scratch - but obviously that takes longer...
  • jmgjmg Posts: 15,140
    ceptimus wrote: »
    I found that my programmer could read the hex file from one of the cheap 4-digit clock kits and copy that to a new blank chip. Of course, it's better if you can get hold of source assembly or C code with English comments, variable and function names - that makes changing the program much easier.

    The most interesting thing to do is to write your own code from scratch - but obviously that takes longer...

    Do you have the hex file from that, and the PCB variant / Pin map it 'belongs' to ?

  • ceptimus wrote: »
    I found that my programmer could read the hex file from one of the cheap 4-digit clock kits and copy that to a new blank chip. Of course, it's better if you can get hold of source assembly or C code with English comments, variable and function names - that makes changing the program much easier.

    The most interesting thing to do is to write your own code from scratch - but obviously that takes longer...
    Are you using the SP200S-2016 programmer? What software do you use with it?

  • I don't think my SP200S-2016 programmer is working. The Willar software can't detect it and the MCU chip gets pretty warm when I plug it in.
  • ceptimusceptimus Posts: 135
    edited 2018-05-02 07:50
    jmg wrote: »
    ceptimus wrote: »
    I found that my programmer could read the hex file from one of the cheap 4-digit clock kits and copy that to a new blank chip. Of course, it's better if you can get hold of source assembly or C code with English comments, variable and function names - that makes changing the program much easier.

    The most interesting thing to do is to write your own code from scratch - but obviously that takes longer...

    Do you have the hex file from that, and the PCB variant / Pin map it 'belongs' to ?
    The PCB is labelled (on the back in the copper) DIY YSZ-4 860925. I'll see if I can dig out the hex file. It's been a while since I worked on this.
  • ceptimusceptimus Posts: 135
    edited 2018-05-02 07:56
    David Betz wrote: »
    ceptimus wrote: »
    I found that my programmer could read the hex file from one of the cheap 4-digit clock kits and copy that to a new blank chip. Of course, it's better if you can get hold of source assembly or C code with English comments, variable and function names - that makes changing the program much easier.

    The most interesting thing to do is to write your own code from scratch - but obviously that takes longer...
    Are you using the SP200S-2016 programmer? What software do you use with it?
    I'm using a (very simple) programmer I designed and built myself that piggy-backs on an Arduino Mega 2560 - and my own program. You can build your own with a few diodes, capacitors, and transistors. All the circuits, details, and download links for the software are on my blog, here are the links:

    Part 1 (description and hardware): http://ceptimus.co.uk/?p=216
    Part 2 (code): http://ceptimus.co.uk/?p=231
  • David BetzDavid Betz Posts: 14,511
    edited 2018-05-14 18:28
    I just got a replacement MCU from the people who sold me the clock kit. It doesn't work either. Has anyone purchased this kit and gotten it to work?

    This is the PCB:

    https://forums.parallax.com/discussion/download/122478/assembled-front.jpg
  • jmgjmg Posts: 15,140
    David Betz wrote: »
    I just got a replacement MCU from the people who sold me the clock kit. It doesn't work either. Has anyone purchased this kit and gotten it to work?
    Did you check the Xtal was oscillating ok ? You can do that with a meter and a 1N914 or similar diode - usually there is close to full Vcc swing on each pin, averages 50% Vcc
    Pin 1 should also start hi to reset and idle low. ( a leaky cap might hold in reset ?)

  • jmg wrote: »
    David Betz wrote: »
    I just got a replacement MCU from the people who sold me the clock kit. It doesn't work either. Has anyone purchased this kit and gotten it to work?
    Did you check the Xtal was oscillating ok ? You can do that with a meter and a 1N914 or similar diode - usually there is close to full Vcc swing on each pin, averages 50% Vcc
    Pin 1 should also start hi to reset and idle low. ( a leaky cap might hold in reset ?)
    Yes, the oscillator is working fine. I checked that based on a suggestion here quite a while ago.

  • I have to say that the people who sell this $2 clock kit are very responsive to support requests. As I mentioned earlier, they sent me the source code for the clock and they sent me a new MCU chip when my clock didn't work. After the new MCU chip didn't help the problem they offered to send me a new kit or refund my money. I'm not sure what went wrong with my kit assembly but they are certainly trying to resolve my problem. That's pretty amazing for a $2 kit!
  • ercoerco Posts: 20,244
    edited 2018-05-16 17:23
    Good to hear. I have two of that square version and neither worked! The small non-square 4-digit types in my first post all work though.
  • ercoerco Posts: 20,244
    edited 2018-05-16 16:49
    NOW...


    Real men build their own $2 FM receivers! :):):)
    https://www.ebay.com/itm/192119861885

    s-l1600.jpg
  • yetiyeti Posts: 818
    edited 2018-05-16 17:41
    <wearing tin foil hat> $THEY will kill FM here. </wearing tin foil hat>
    I'm angry as [#BEEP#] because of this. In the future we will need to run around with small computers to receive DAB(+) instead of using a few transistors to get FM and that will not be energy efficient at all!

    That's perfectly misimplementing progress!

    Instead of running a small receiver for weeks on a coin cell, we'll then get 8 hours with one charge...
Sign In or Register to comment.