Windell Oskay

Forum Replies Created

Viewing 15 posts - 1,456 through 1,470 (of 1,518 total)
  • Author
    Posts
  • in reply to: Further help needed debugging #20883
    Windell Oskay
    Keymaster

    It’s very likely to be repairable if you can identify the trouble spot; don’t give up yet! :)

    in reply to: Further help needed debugging #20881
    Windell Oskay
    Keymaster

    If diode D2 is not installed correctly, or if it has a bad connection, that would indeed cause this symptom.  Check to see if there’s a bad solder joint there, or if perhaps a trace on the board has been torn.

    in reply to: Bulbdial Alarm Clock #20879
    Windell Oskay
    Keymaster

    Yes, see some hints about how to modify Arduino to work without a crystal here:  http://arduino.cc/en/Tutorial/ArduinoToBreadboard

    Generally however, I would recommend sticking with the crystal and putting the speaker elsewhere; you only need one output pin to do it.
    in reply to: Peggy 2LE some colums lighting halfway #20856
    Windell Oskay
    Keymaster

    Fantastic– Glad to hear it!

    in reply to: Bulbdial Alarm Clock #20877
    Windell Oskay
    Keymaster

    One thing to consider is how you’re going to reprogram the clock to take advantage of the speaker.  If you’re going to use the USB-serial interface, then repurpose the ISP pins.  If you’re going to use the ISP interface, repurpose the serial pins. 

    The crystal *is* used by the Bulbdial. So, if you want to run without it, you’ll need to reprogram the chip with an ISP interface (to set the fuses to run off of the internal timebase), and likely make other changes to the programming as well.
    The Alpha Clock Five and Meggy Jr RGB kits use a little magnetic speaker/buzzer for their noises and alarms.  It only requires one pin from the MCU (plus a transistor) to drive it.  That type is likely a good choice.  Use the contact form on our web store if you’d like to buy one from us.
    in reply to: Alpha Clock Five observations (presales/feature request) #20872
    Windell Oskay
    Keymaster

    We are currently working on an all-new version of the Alpha Clock Five firmware, which has an updated toolchain and will (hopefully) be a little easier for people to modify.  The new firmware is currently at the stage of adding the menus, options, and other functions. 

    The primary supported functions of Alpha Clock Five are (1) as a serial-controlled data display device and (2) as a bedside-type alarm clock.   We are always interested in hearing new feature requests, but there is a definite tradeoff to consider in terms of feature completeness versus simplicity and usability.    So, while I like the idea of adding features like countdown and interval timers, they may not actually be good things to add to the main branch of the firmware.  We could certainly consider a separate fork of the firmware (i.e., an available download) that would have those features instead. 
    – We could do a boxing timer, that displays the round number and time remaining
    – We could do a lap timer, reset by a button, or so forth.
    – We could do a simpler countdown timer.
    The trickiest part for any of these is defining what the behavior *should* be– which button press does what, and what is displayed an so forth –such that it’s intuitive and easy to use.   If you would care to make such suggestions, we’d be happy to listen.
    BTW, I see that you’ve left this request on our wiki as well:
    (That’s the best place for feature requests– if anyone has others, please add them there as well.)
    in reply to: Octolively – different colour LEDs on a single module? #20859
    Windell Oskay
    Keymaster

    It’s no problem to do so, so long as you use the correct load resistor for the different LEDs. 

     Step 8 of the assembly guide mentions this possibility: “Advanced tip: If you are using multiple LED colors on a single Octolively module, make sure that the resistor by each LED is the correct type: 220 ohm for red/orange/yellow/yellow-green, 82 ohm for blue/green/white.”
    There will be intensity differences due to the relative brightness of the different LED colors.  It is possible (but not particularly trivial) to correct that in firmware.   (Some LEDs are just brighter than others.)  Another thing that you can play with is the load resistor value for each LED, so long as you *increase* rather than decrease the values. 
    in reply to: Do I need to keep my circuit boards in static bags? #20857
    Windell Oskay
    Keymaster

    No, not necessarily.  While you shouldn’t scuff your feet along the carpet and then try to zap the boards, they’re pretty robust with respect to general handling.  

    For long term storage, preventing moisture should generally be the chief concern– store vented in a cool dry place, or sealed with a desiccant. 
    Antistatic bags will work fine, as will polyethylene bags (those made of the same type as ziploc bags).  Avoid other types of plastic, especially styrene, acrylic, and other types that seem particularly “staticy.”
    in reply to: Peggy 2LE some colums lighting halfway #20854
    Windell Oskay
    Keymaster

    Hi Relaxing, 

     Every LED in each vertical column is connected together by a printed copper trace, on the back side of the Peggy 2LE circuit board, that runs along each row and connects the “round” hole of each LED together.
    (The “square” holes connect every LED in a horizontal row, on the front side of the circuit board.) 
     So, it appears that you have some broken traces on the circuit board. Either the circuit board was scratched in those three places where the column turns off, or the trace was somehow otherwise damaged.

    The good news is that it isn’t so hard to fix. 

    Identify, on the back side, the three exact places where you have breaks– in those three columns between the working LEDs and non-lit LEDs. In each of those three places, solder a short wire (e.g., a piece of clipped resistor lead or LED lead) between the pin of the lower (working) LED– the pin in the round hole! — to the pin of the upper (non-lit) LED– again, the pin with the round hole.

    Fix one at a time; make sure that the fix works on the first of these locations before going on to the next. 

     Good luck, and please let us know how it goes. :)
    in reply to: ATmega644A and STK500? #20852
    Windell Oskay
    Keymaster

    While I have not done this myself (I do not have an STK500), it should definitely be possible. 

    Let me ask a couple of things to point you in the right direction. 

    – Was this a preprogrammed ATmega644A (e.g., for the Alpha Clock Five)? If so, the fuse bits are set to require an external crystal oscillator. Make sure that you’ve provided one, or are programming it on a board that provides one. 

    – Are you using the latest version of Atmel Studio? The ‘644A is a relatively new device, not supported by all older versions. 

     – Is it possible that your STK500 needs a firmware update? (I’m not sure that it does, however, we have needed to run firmware updates on our Atmel AVRISPmkII devices.)
    in reply to: Help needed debugging interactive LED display #20850
    Windell Oskay
    Keymaster

    Ah– bad timing for my reply!

    On the next panel, see if you can identify any similar types of soldering issues– they’re responsible for a good 90% of things that ever go wrong. :)
    in reply to: Help needed debugging interactive LED display #20849
    Windell Oskay
    Keymaster

    > The other LED’s in the quadrant of the panel I mentioned are off. 


    Then, there are (at least) two independent problems here.  

    Problem 1:  The LEDs are not responsive, stuck in one “direction”
    Problem 2:  The LEDs in the other half of the quadrant are all off.

    There is no single place on the circuit that a single problem could cause both of these to happen.  First, then, try to figure out why the other quadrant is fully off.  It’s likely a soldering problem.
    in reply to: Help needed debugging interactive LED display #20845
    Windell Oskay
    Keymaster

    > Two of my six panels have an arrangement of 5 lights (trapezoidal pattern) that stay lit up and do not interact in any way. I haven’t been able to find the problem. Any ideas?

    One panel at a time: Are those 5 LEDs on full bright, or only a little bit? And, how are the other 10 LEDs in that quadrant acting?

    First thing to double check is that each LED in that half-quadrant is oriented correctly (check the LED “collars” to make sure that the flats all point the same way) and that each LED in the quadrant is correctly soldered. Check each solder joint to make sure that they look shiny and clean.

    > Is there an easy way to add music to the device so that when the lights are going, music is automatically triggered?

    There are a few different ways if you know some electronics, but I’m not sure what you regard as “easy” — you could sense and divide the output signal from each node, but then you also need to figure out what kind of device you plan to drive from that output.

    in reply to: Peggy 2 RGBW Blue LEDs not lit #20838
    Windell Oskay
    Keymaster

    If you are indeed getting some LED light in every row (or at least, almost every row), then the AVR microcontroller is indeed running and scanning between the rows.  If so, it should be possible to reprogram it, and that’s a separate issue.

    If pushing down on the socket improves things, then either the chip isn’t fully seated in the socket (and check the others, too!) or it’s not soldered correctly.  At this point, a soldering issue is the most likely type of thing that’s going wrong.
    in reply to: Open pins on Alpha Clock Five #20771
    Windell Oskay
    Keymaster

    Revised firmware is one of our active projects right now; likely will be complete within the next few weeks. 

    We do not currently have a plan for a new case design (except for a few minor updates).  For daisy-chained use, if you don’t need the buttons, we’d suggest mounting the displays with angle brackets to a horizontal surface– you can fit an arbitrarily large number next to each other.
Viewing 15 posts - 1,456 through 1,470 (of 1,518 total)