The purpose of this entry is to describe the assembly, programming and testing of the alternate Clock Display Module – whereby instead of a dedicated display driver chip, the module will be controlled by its own micro controller. This was created under a separate GitHub branch as UCtl_small. After the relative ease of getting the MAX7221 version of the display module working, it is easy to think that this version to go just as smoothly. This article will describe more about the approach that will be taken and the planning for the final API. Subsequent posts will document the execution and results.
In one of the latest modules of Contextual Electronics, the services of OSH-Stensils was introduced with the purpose of trying out reflow in an electric frying pan. I don’t consider myself at the stage to try it out (my work bench is just not big enough for another appliance), but the idea of using the stencil had me intrigued. Not long after submitting the board for fabrication, I also ordered the associated stencils, along with a jig to hold the work in place. The assembly went very smoothly! The stencils are a great help.
They also helped to demonstrate that you don’t need a lot of paste at all to get a good soldering job. The trickiest was the transistors. They would have to be the smallest parts I have soldered to date. I am very grateful to have been introduced to the stencil idea before attempting this board.
The board is now assembled and I am very satisfied with the result, now comes the testing. I think the is the most exciting oart of the development. All the previous work of sifting through data sheets and trawling through vendor sites looking for parts, creating the schematic and working through the layout puzzle all comes down to this monent! Will it work? Or will it sit there is stoney silence and do nothing? Now that I am at this point, I realise that this board, although functionally the same as the MAX7221 based display module, is significantly different in the way it is to be “brought up”. The MAX7221 based board has given me confidence with SPI interfacing. However, this board is still dumb. The processor has not yet been programmed. So there is my first challenge – onboard programming. In the design, I included a pin header for the programmer connector.
One Step at a Time
I have come to realise that I will need some type of systematic approach to the problem. In a similar vein with the MAX7221 based board, starting with a “Heartbeat” and building on that, I will need to start small and gain confidence that a section is working before proceeding with the next. I figure, I can tackle this with the following areas:
- Programming interface
- SPI Slave
- Display Module API
In this section, I need to prove I have everything connected correctly in oder to achieve processor start up plus inboard programming. To confirm this, as much as it seems extraneous work, I will bread-board a processor with the pin header and a LED just to conform the “Hello World” scenario. If this does not work on the bread-board with the same configuration of the display module, then I know I have something fundamentally wrong. “Hey, wait a minute”, I hear you say. “Shouldn’t this have already been done?”. Well, yes, it was but a while ago and I to just try it again to double checkand then I have a model that I can pull around, if required. Once I have the confidence with respect to the set up, then I can transfer the “Hello World” excercise to the display module board. It does not matter that it will not be displaying any digits, the main thins is that it blinks a LED or segment – any or perhaps a sequence. The idea here is that this will prove that the basic infrastructure is working. The programming interface and the IO lines to the LEDs.
The next step after asserting the basic infrastructure is to start bullding the firmware for the display module proper. This means setting up the display module as a Slave. The first response should be at least to react to a specific register-value pair. This could be considered the test mode for the module (see the API specification below) and could even reuse the test code in the previous test.
Display Module API
Once I have successfully reached this stage, everything is in place to start defining the actual API for the module. An overview of the possible commands is shown in the following table. Some of these features are nice-to-have (*) and will be subject to further development once the basic features are implemented.
Basically it will work using eight bit register-value pairs. The resister will specify an operation or feature and the value will configure that operation or feature. Some of these operations and features will be mutually exclusive, so a priority will need to be fleshed out. Further refinement of the descriptions will be subject of further posts as the features mature.
|| The 10s digit for the hour part of the time
BCD value from 0-2
|| The units digit for the hour part of the time
BCD value from 0-9
|| The 10s digit for the minutes part of the time
BCD value from 0-5
|| The units digit for the minutes part of the time.
BCD value from 0-9
|| The colon (between the hours and minutes)
0x00 – Colon Disabled
0x01 – Colon Enabled
|| The indicator lamps that are featured on the display
0x00 – All indicators OFF
0x01 – Top indicator ON
0x02 – Bottom indicator ON
0x03 – Top and Bottom indicator ON
|| Controls the intensity of the display
|| Enables the display unit
0x00 – Display unit OFF
0x01 – Display unit ON
|| Defines which digit or display element should blink. The data value of the register is a set of bit-wise flags for each element. OR-ing the various bit enables the digits to be selectively enabled for blocking, or not.
0x00 – No blink
0x02 – Digit 1
0x04 – Digit 2
0x08 – Digit 3
0x10 – Digit 4
0x11 – Colon
0x12 – Indicators
0x1F – All
|| Controls the blink rate of the digit
0x00 – Lamp test off
0x01 – All segments on
0x02 – Sequence run*
Serial Data Format
D15 D14 D13 D12 D11 D10 D9 D8 D7 D6 D5 D4 D3 D2 D1 D0
MSB ADDRESS LSB MSB DATA LSB
The next steps are already described in the body of this post. Needless to say the board needs to be powered up and programmed with the basic “Hello World” example to prove the connections and basic infrastructure.