There's been a little bit of delay since my last post. We had some of the hottest weather this year which caused me to avoid slaving over a soldering iron, and I had some other personal commitments going on otherwise which ate up some time.
It consists of a handful of 7 and 16 segment LED displays to provide the visual representation of the state of the counters within the clock.
As I mentioned in my previous post, I thought I had bought common cathode displays, but it turns out I bought common anode instead. The difference is as follows.
In a common cathode configuration, all of the "negative" pins of the LEDs are commoned together such that you supply positive voltage to the anode ("positive") pin of the LED to make it light up.
In a common anode configuration, the "positive" pins are commoned together, and to make an LED light up you ground the "negative" pin.
In either case you still need (ideally) a current limiting resistor per LED. You can in theory use a single resistor for an entire display, but the brightness of the segments will vary depending on how many of them are lit at a time.
How this affects my build means I just need a "few" more transistors on the display board than originally planned. "Few" as in 102 to be precise. The way I use them is as a switch to connect the cathode of each LED segment to ground. This enables current to flow from the positive rail, through the LED segment, then through its current limiting resistor and finally down to ground.
The decoders were originally planned to output a positive signal, that is that the output is high when the LED segment should be on, and originally it was planned to have the 2.2K resistors in the final driving gates of the decoders. But now, I use my standard issue 100K resistor, and the output of the gates in the decoder feeds to an associated transistor on the display board to switch a segment (or two in some cases) on or off.
So most of the circuitry on this board is resistors and transistors used to switch segments on and off.
Other than that, I'm doing something a little different with this post and combining both the design and build log in to one. The display board is reasonably simple, it holds no state like the decoder boards, and construction was quite straight forward and was completed some weeks ago. So ...
And then proceeded to solder everything in to place.
Due to the way I had laid out my bus pins I had to leave a reasonable amount of slack on each of the wires leading up to the top row of displays so that I could shuffle them around in order to get the bottom row of displays etc in. A little bit more thought in that area could have gone a long way to making construction a lot less fiddly, e.g. by placing some of the calendar signals on the top connector instead.
After completing this much work, it was time to try it out and much to my delight, when I plugged it in I got an error free readout. I manually cycled through all of the hours and minutes to ensure they all displayed the correct digits, and only had to fix a couple of segments that had been swapped around due to an inconsistency with how I wired some of the transistors up.
That last photo isn't the greatest, but I was actually able to use it as a clock at that point in time! Yay!!!
And after achieving this I decided to take a break from this project, plus the hot weather we've had around this time of year, plus other personal commitments that I had going on.... So that explains the rather long gap since my last post.
And at this point I noticed that my linear regulator was starting to generate quite a lot of heat due to the increased current from all of the LEDs. So I decided to up rate the heatsink a bit:
It does still get a bit warm, but the new heatsink has much more surface area to dissipate heat from and it seems to be doing a good job so far.
The next step was to place all of the displays for the calendar readout. And after doing so, I ended up with the following mass of wires. And the stack is building up quite nicely, too:
The squiggly bit of wire you see weaving between all of the pins on the rear side of the board is just pulling all of those pins high to turn on the associated segments so I could see what everything was wired and working, and thus would be able to turn on all of the required segments for each character that needs to be displayed for the various readouts. It remains to be seen if they are all in the right order though...
So now there are only 2-3 boards left to build: the calendar module and its decoder board, and having left the clock on overnight it really does add significant amounts of light to my bed room, so I will definitely be building a module which is able to dim the display between certain hours. I'm still toying with ideas of how best to do that.
And of course, below are links to the design files for this board.
Design
What will be the last board in the clock (but not the last in the build) is the display board, which makes all of the internal workings of the clock visible to us simple humans.It consists of a handful of 7 and 16 segment LED displays to provide the visual representation of the state of the counters within the clock.
As I mentioned in my previous post, I thought I had bought common cathode displays, but it turns out I bought common anode instead. The difference is as follows.
In a common cathode configuration, all of the "negative" pins of the LEDs are commoned together such that you supply positive voltage to the anode ("positive") pin of the LED to make it light up.
In a common anode configuration, the "positive" pins are commoned together, and to make an LED light up you ground the "negative" pin.
In either case you still need (ideally) a current limiting resistor per LED. You can in theory use a single resistor for an entire display, but the brightness of the segments will vary depending on how many of them are lit at a time.
How this affects my build means I just need a "few" more transistors on the display board than originally planned. "Few" as in 102 to be precise. The way I use them is as a switch to connect the cathode of each LED segment to ground. This enables current to flow from the positive rail, through the LED segment, then through its current limiting resistor and finally down to ground.
The decoders were originally planned to output a positive signal, that is that the output is high when the LED segment should be on, and originally it was planned to have the 2.2K resistors in the final driving gates of the decoders. But now, I use my standard issue 100K resistor, and the output of the gates in the decoder feeds to an associated transistor on the display board to switch a segment (or two in some cases) on or off.
So most of the circuitry on this board is resistors and transistors used to switch segments on and off.
Other than that, I'm doing something a little different with this post and combining both the design and build log in to one. The display board is reasonably simple, it holds no state like the decoder boards, and construction was quite straight forward and was completed some weeks ago. So ...
Build
I started off by laying out the displays on a board to determine an arrangement and spacing that I liked. I'm not going to worry about any kind of separator between the digit groups in the clock portion, I'm hoping the physical separation of the digit groups will be sufficient to make it easy enough to read. The digits/alpha groups in the calendar portion I think will be obviously different enough to make that readable, so separators shouldn't be necessary there either.
And then proceeded to solder everything in to place.
Due to the way I had laid out my bus pins I had to leave a reasonable amount of slack on each of the wires leading up to the top row of displays so that I could shuffle them around in order to get the bottom row of displays etc in. A little bit more thought in that area could have gone a long way to making construction a lot less fiddly, e.g. by placing some of the calendar signals on the top connector instead.
After completing this much work, it was time to try it out and much to my delight, when I plugged it in I got an error free readout. I manually cycled through all of the hours and minutes to ensure they all displayed the correct digits, and only had to fix a couple of segments that had been swapped around due to an inconsistency with how I wired some of the transistors up.
That last photo isn't the greatest, but I was actually able to use it as a clock at that point in time! Yay!!!
And after achieving this I decided to take a break from this project, plus the hot weather we've had around this time of year, plus other personal commitments that I had going on.... So that explains the rather long gap since my last post.
And at this point I noticed that my linear regulator was starting to generate quite a lot of heat due to the increased current from all of the LEDs. So I decided to up rate the heatsink a bit:
It does still get a bit warm, but the new heatsink has much more surface area to dissipate heat from and it seems to be doing a good job so far.
The next step was to place all of the displays for the calendar readout. And after doing so, I ended up with the following mass of wires. And the stack is building up quite nicely, too:
The squiggly bit of wire you see weaving between all of the pins on the rear side of the board is just pulling all of those pins high to turn on the associated segments so I could see what everything was wired and working, and thus would be able to turn on all of the required segments for each character that needs to be displayed for the various readouts. It remains to be seen if they are all in the right order though...
So now there are only 2-3 boards left to build: the calendar module and its decoder board, and having left the clock on overnight it really does add significant amounts of light to my bed room, so I will definitely be building a module which is able to dim the display between certain hours. I'm still toying with ideas of how best to do that.
And of course, below are links to the design files for this board.
- GitHub schematics repo
- display-board.sch EAGLE schematic file
- display-board.pdf
No comments:
Post a Comment