Custom pin support.

[MokaSoda]

I want to use this firmware because of critical bugs on stock clock kit. Since this kit is different from banggood varients, I want to modify pin numbers. However I cannot change A1~A4 pin numbers of this source code because I don't know the way this code works. Could you tell me the way to change these pins? (In this schematic, A1-A4 is P3.4-P3.7)

[zerog2k]

your schematic has quite a few differences ntc, ldr, ds1302 all on different pins... can you scan a better copy of that schematics as pdf or sharper/cleaner jpg? we can add to the docs section of the repo to go along with this "new" model can you tell me where you got this, or did it have some name? (I'm trying to think of how to distinguish between the different models... As none of them really seem to have a name, but sometimes use same chip with different layouts, I'm tempted to just call them something unimaginative, like HW_MODEL_A, HW_MODEL_B, etc)

[zerog2k]

I created a new branch for testing which tries to consolidate and expose all of the various board dependent settings into one file, hwconfig.h: https://github.com/zerog2k/stc_diyclock/blob/hwmodels/src/hwconfig.h

then, looking at your schematic, I added your board as HW_MODEL_C (because I'm aware of at least two others prior ;)

I compiled a firmware from this branch with the following options: make CFLAGS="-D WITH_ALT_LED9 -D WITHOUT_LEDTABLE_RELOC -D HW_MODEL_C" you can try this out and see if it works for you: main-HW_MODEL_C.zip

[MokaSoda]

I bought this from Taobao(Origin of Aliexpress). Its name is DY-020 Here's more crisp schematics (sorry for upside down images..) After, I upload compiled firmware from you, I notice that LED is also Okay(All of them works normally). However, it shows "1F:7F." and the clock minute doesn't count up after a minute. (F is defined in this sourcecode so it should be F).

[zerog2k]

Can you try setting the clock? Maybe the DS1302 has initial junk values (happens if no power and backup battery is dead)? If this is the case, the process of setting the values should clear this out. If not it may be something else. I just double-checked the hwconfig.h for your model (what we will call "HW_MODEL_C" for now) and it looks correct.

[vargham]

I have another pinout version with STC15F204EA. I figured out the connections, and I will share it. This version has one 4 digit display. I could also change pin definitions in your code. This one from Aliexpress

But I have the same problem as above. When I turn on, it displays 1F 7F and it is never changes. I can change mode with button 2. Next mode is temperature, and it is works correctly. I can change C display to F with button 1. Next mode is unknown for me. It displays an empty digit and then -8- I tried long and short press combinations with both buttons, but I could not set up the clock.

[zerog2k]

@vargham I see this has a slightly different 4 digit led module (uses single one rather than several separate led digits) which results in a different scheme required to display properly. Also other stuff (like ds1302 chip) may have different connections.

Do you have a schematic diagram for this? Likely we need to create another hardware model variant to handle this one. Please open up a new issue so we can track adding support for it.

[vargham]

may have different connections

Everything have different connection. It is not a problem, I figured out all the connections. Clock, buttons, led segments, everything. The display is OK, I wrote a little test loop inside your main() and it iterates over all symbols. And your program correctly displays the temperature. Only the clock setting is wrong. I will draw the schematic and post in a new issue. I also could fork your repo, commit changes, and you can check them.

[zerog2k]

Yes realized that display is working properly if you can see the temperature... Does the kit include schematics? Can you take a good picture or scan and include here? I can abstract any differences in hwconfig.h

[zerog2k]

@vargham try this: update these pin mappings to match your ds1302 connections to stc mcu: https://github.com/zerog2k/stc_diyclock/blob/master/src/hwconfig.h#L84-L89 (note DS_CE is RST pin on ds1302)

Looking at another schematic on aliexpress, it might be something like:

 #define DS_CE    P1_2
 #define DS_IO    P1_1
 #define DS_SCLK  P1_0
 // needed for asm optimizations
 #define _DS_IO   _P1_1
 #define _DS_SCLK _P1_0

[vargham]

No, there is no schematic. I have to draw.

I can abstract any differences in hwconfig.h

I made changes in hwconfig.h with new HW_MODEL_D. Is this occupied? I can change if needed.

[zerog2k]

yeah - i dont think we have really good names for them as there are so many and they are super generic... HW_MODEL_D is fine ;)

[vargham]

OK, here is the pinout:

01  P2.6    Button2
02  P2.7    Button1
03  P1.0    DS1302 PIN_7 (SCLK)
04  P1.1    DS1302 PIN_5 (DATA I/O)
05  P1.2    DS1302 PIN_6 (CE)
06  P1.3
07  P1.4
08  P1.5
09  P1.6    LDR
10  P1.7    TEMP
11  P0.0
12  VCC
13  P0.1    BUZZER Negative
14  GND

15  P3.0    7SEG_a  RX
16  P3.1    7SEG_b  TX
17  P3.2    7SEG_c
18  P3.3    7SEG_d
19  P3.4    7SEG_e
20  P3.5    7SEG_f
21  P3.6    7SEG_g
22  P3.7    7SEG_dp
23  P2.0    7SEG_4
24  P2.1    7SEG_3
25  P2.2    7SEG_2
26  P2.3    7SEG_1
27  P2.4
28  P2.5

[vargham]

I will fork, commit. Discuss the error, and pull request if solved. Is It OK?

[zerog2k]

sure... please put some comments in the hwconfig.h about your new HW_MODEL_D (e.g. link from aliexpress, etc), and I'll take a PR if you can test it out ;)

[vargham]

Here is the issue in the forked repo: https://github.com/vargham/stc_diyclock/issues/1

[aFewBits]

I probably have schematics and code for that model. Please post a picture of the component side of the circuit board as there are at least three different models. Further differences are that the LED's can be either common cathode or common anode and the display LED is all direct drive (with no current limiting resistors) from the I/O ports so you have to diddle the port STC port configurations to increase the current capability of the I/O. The end result is that you end up with somewhat different digit intensities, depending upon how many segments are illuminated. It's a really down and dirty design but they work and they are more compact in size.