Forum Replies Created
-
Posts
-
The chips are, fundamentally 16-channel devices; only 24 LEDs are actually on at any given moment, and doing so should not present any problems. Switching quickly (multiplexing) between 8 rows makes it look like all 192 LEDs are on at the same time.
And, sorry for the slow reply!
-WindellI see that I didn’t answer the question about the rear projection clock face.
The standard clock face (included with the Bulbdial Clock kit) is made of opaque (non-see through) acrylic plastic, and has a white face with printed numbers on it. The gnomon (spike) protrudes up the same side that the numbers are printed on, and that face is installed facing towards the inside of the clock. When viewing the clock, you look through the three rings of LEDs to see the clock face and shadow hands cast by the LEDs and gnomon.By contrast, the optional Rear Projection Clock Face is made of translucent and textured light-diffusing acrylic plastic. One side of the clock face is smooth, with printed numbers on it. Unlike with the regular clock face, the gnomon (spike) is attached on the “back” side of the clock face, protruding from the side without printed numbers. The clock face is installed with the gnomon pointing towards the inside of the clock (towards the LEDs), and the numbers on the outside of the clock. When viewing the clock, you look at the side with the clock that has the printed clock face, and see the shadow hands– in rear projection –cast by the LEDs and gnomon upon that surface.
Originally, we had planned to use wide viewing angle (40-degree) LEDs in the kit, but we were unable to get the right ones in time for the initial launch of the Bulbdial, so we used standard narrow-angle (~20-degree) instead. A year or so later, when we were finally able to get the wide-angle LEDs in the right quality, we began shipping those in the kits. These are pretty neat because they light up nearly the whole face, making neat cyan, magenta, and yellow shadows with a white background. However, even though it was a better set of LEDs in certain ways, most people turned out to prefer the look of the narrower LEDs, and (for better or worse) we were somewhat forced to switch back to the original (narrow-angle) LEDs.
My expectation, since you got your kit recently, is that you *do not* have the wide angle LEDs that you were reading about in the forums– we haven’t shipped any kits like that in a very long time, and we’ve exchanged most of those for narrow LED sets. It is true that there is a bit more variation and width in the blue and green LEDs than we might like, but we’ve generally found it to be within what we expect from them. However, thanks to this feedback, we’re starting to look at a few other LED suppliers that might (hopefully) be able to give us a little better consistency in that regard. (BTW, the red LEDs in current-generation Bulbdial kits are actually Cree C503B-RANs, so you might want to skip swapping those, wbp.)I’m not sure of the best way to ensure that the seconds/minutes hands are always distinguishable and readable. (It’s sometimes even a challenge with analog clocks!) You might try playing with the relative intensities of the LEDs; that can help to bring up the contrast. You might also consider aiming them a little bit differently. If the blue (seconds) are aimed further out, you’ll get a different character of shadow that doesn’t overlap as much.Putting in the Chronodot too early should not cause any problems, and the orientation of the crystal does not matter.
What stage of construction are you at, presently?I have not seen anything like this happen before. If the chip is in backwards, or shorted in certain ways, I could imagine this happening. We’d be happy to give you a replacement chip, but it’s critical to identify what caused the problem. If you cannot, then the same thing will most certainly happen again after you replace the chip.
It’s possible (but unlikely) that there’s an issue with the LED matrix itself. If so, it would be possible to test by swapping it out onto another Meggy Jr, to see if it works there. You can bring your unit by our shop during business hours, but it might be helpful to make an appointment (off-forum, please) to ensure that we’re expecting your visit.If you would like to send it back to us, please use the contact form: http://shop.evilmadscientist.com/contact
(Edit: To set up the RMA! You can’t fit a circuit board in the contact form.)
The bootloader from the UNO is not strictly compatible with the one on the Duemilanove (and Peggy 2 default), although I *believe* that it can work if you select UNO as your board type. Not entirely sure.
This is starting to get a little more complicated; Do you want to send your Peggy 2LE back to us for diagnosis?Right now, we need to figure out what’s preventing programming, because this is a much more fundamental problem than the one with the LEDs.
Again, I would suggest that you actually remove U2 and U3 while debugging this– and do not put them back in –until you are sure that you can get the microcontroller to work normally.Using both sets of jumpers like you did could potentially damage the microcontroller, but that shouldn’t be an issue since you’ve got the extra chip to try. Are you sure that the extra has the same Arduino bootloader on it?By, “the same result,” do you mean that it still can’t be programmed?
The programming error would suggest that the board is resetting, possibly due to an overcurrent event when it’s lighting up the LEDs like that.
Previously, you said that you were able to program it. If it can’t be programmed, the “CPU” is probably also not in a state where it can correctly light the LED display. You should reload the peggy2_minimal sketch to have a known-good state before going further. If necessary, gently remove the LED driver chips (U2, U3) so that it will not drive the LED display, so that you can reprogram it.It looks like you might have wire jumpers in both the P2 and SER slots– is that correct? If so, please fix that first and see how the rest is behaving.
First off, are you getting the same pattern exactly, regardless of the program that is being run?
And, can you please describe a little more clearly what you are seeing? Q7 and Q23 are the names of two transistors, which control rows 7 and 23, respectively, starting with row 0 on the top. Is it these two rows that are showing the issue, or are you actually seeing a problem with two of the columns? Is the rest of the board lighting, or are these the only rows/columns that are showing any light?Normally, if two back-to-back LEDs light at the same time, that indicates that one of the LEDs is backwards. You’ve already checked for that, with the flat sides. Just in case, double check that the “leadframes” — the metal shapes within the plastic lens –are also oriented in the same direction as each other.
The second thing that could be happening is if there is an accidental connection (e.g., a solder bridge) between signal lines LED1 and LED4. This could be at D14, D41, or elsewhere on the board.Yes, it sounds like this is a reasonable setup, and should work work pretty much as you describe it, all without any programming.
You do not necessarily have to use 12 V DC pumps, but I would recommend it. 12 V DC water pumps are plentiful, quiet, and inexpensive– many are used for applications like desktop PC CPU coolers.Also, I wonder if there might be good alternatives to requiring a button press to reset the demo. For example, you could have a small hole in the pipe that drains out the water over the course of a minute or so, even if they don’t press the button to reset the exhibit. The kids would still see the water standing there, unless they were quite patient to wait for it to drain all the way. With programming, you could also change the “reset pump” controller to be triggered by the two main buttons, but not to run the pump until after a delay of (say) 30 seconds.Yikes! The instructions are correct, so it seems like they must have mis-wired it at the factory. I’m glad to hear that it’s working, though!
