Видео

Cool enough to win….

CS is defined in the sketch. If you use my sketch unaltered, it defaults to D8. It's up near the top of the main sketch: #define CS D8 Beware that different versions of the 8266 boards confusingly mark the same pin with different labels. If you don't have documentation, or are unsure, it's best to modify a simple 'Blink' sketch (standard example) to blink the pin - then connect up a LED (with 470-ohm or similar resistor) or use a multimeter to check which pin is actually going high and low. The other pins are the standard SPI pins for the 8266 - and should be labelled on your board or in your board's documentation MOSI (master out slave in - connect to Din) and SCLK (serial clock, or maybe just CLK). 8266 numbering of the pins is confusing - there's a GPIO number, and then a D pin number. Sometimes the D number printed on the board is different to the one you use in your Arduino sketch. A blink sketch 'blinking' a selected pin is the surest way to find out. You should be able to rely on the MOSI and SCLK labels printed on a board, or the board's documentation, though. If you have an oscilloscope, you can also monitor the clock and data lines to see that they're changing rapidly.

@@ceptimus Thanks for the reply. I appreciate the help. Your display is great.

If the pictures don't have to change during run-time, then it's best to put them in __code memory. That way, they don't take up any RAM at all (they're stored with the program, in FLASH memory), and your program can still access them. That's the way I stored the tune data, etc. for this program. If the data does have to change at run time, you can still put it in FLASH memory - though of course, writing to that memory is much slower than to RAM, and if you do it too many times you can eventually wear out the FLASH memory, and then the chip is bricked. Some of the STC chips do this by treating part of the FLASH memory as EEPROM. You can partition the FLASH memory into a FLASH partition, which remains read only at run time, and an EEPROM partition which can be written and read. stcgal allows you to change the partition boundary, so you might have 16k FLASH and 1k EEPROM, or split the same 17k flash to, say, 8k FLASH and 9k EEPROM.

@@ceptimus I'll have to look into that but it sounds like I might be out of my depth. I did use the __code memory like you suggest but I've been using the stc-isp software to upload the programs to the STC12C5A60S2. But I still need to get the buttons working and my stretch goal is some basic I2C so I can link up some of these project boards with each other or a RaspberryPi.

Got my buttons working in the main section of the code but clearly need to try that interrupt loop that you used in the hourglass for them to be more useful. I also did some reading about the SDCC. Read through Chapter 3 and found out about the --model-large compile option. Not sure if this what you were talking about but it got rid of my DSEG problem and let me add more images. I also added a delay profile for each image so got rid of all of the hot spots when there is only one or two LEDs lit up in a given row. I got up to 70 images before it started acting weird and drawing weird stuff without warning.

Yes. Will do. Stand by for a fresh video in the next couple of days.

@@ceptimus Thank you, no need to hurry, enjoy fully the end of 2024...

I didn't do anything special, and my Ferris wheel is still on the standard firmware. Just connected the 3.5mm jack to the 'earphone' output on my sound recorder (I guess a phone or other MP3 player would work too), and selected the right mode using the remote control (press 4 or 7). Had to wind the volume up pretty high on the sound recorder to get the LEDs to light up right to the outer ring.

@@ceptimus I'll give it a go and let you know how it works out. BTW If you have the elevator kit, or something similar, I'll happily share the image files I typed up.

There is a test mode that switches all the LEDs on at the same time. Then, they do get out of sync after a second or two. I think the code in the micro that cycles through the patterns has been written so that it never leaves any LED lit for very long (maybe it just briefly switches off a LED every second or so) - that keeps them synced pretty nicely. I'd like to get hold of some of the two-wire RGB LEDs that you can control exactly, by briefly (a few microseconds) removing power from them, and using that method to send them coded messages. But the only way to obtain them seems to be to buy a string of Christmas lights that use them, and harvest them from there - I've not found any online stores that sell packets of those LEDs to build into your own projects. It would be nice to substitute some of those LEDs in a gadget such as this, and reprogram the micro to send the appropriate pulse trains to them. But I don't think I'll do that on this Ferris wheel, even if I can obtain the LEDs - 180 LEDs are too many to solder!

@@ceptimus Yes, that would be quite a challenge and time consuming 👍👍

@@ceptimus I swear youtube is deleting my comments or something. Anyway, you should have access to the files now. Once you get the timing down everything seems to work perfectly every time. I am noticing some differences between my 6 star kits. They don't all run the program at the exact same speed, might need to make some slight adjustments for my final project.

@@paulandrusyszyn9177 Thanks. Yes, I've got all the files now, and I've ordered what I think might be a compatible kit from Ali Express: www.aliexpress.com/item/1005005604090511.html Once the kit arrives, I'll investigate further, and probably post a video.

It can do other stuff at the same time. The video output is all driven by interrupts, so it just works away in the background, without the main 'loop' program having to worry about it - and it doesn't go out of sync. However, the interrupts do occupy a lot of the processor's time, so your main program, if it's doing lots of work, will run slower than it would otherwise.

Happy to help. One thing that might be stopping it compiling is that the later versions of SDCC require you to put parentheses () around the interrupt and using parameters - the earlier version would compile with just the naked numbers.

@@ceptimus I think it's something wrong with the code I got from the QR code that came with the kit instructions. I created my batch file as you showed: set PATH="C:\SDCCFILES\50RGBLEDStarMin\bin";%PATH% sdcc -mmcs51 --code-size 4088 --opt-code-size --verbose 50RGBLEDStar.c -o 50RGBLEDStar.ihx packihx 50RGBLEDStar.ihx > 50RGBLEDStar.hex pause but the errors are as follows: sdcc: Generating code... STC15F2K.h:21: warning 278: return type of function omitted, assuming int STC15F2K.h:21: error 1: Syntax error, declaration ignored at 'sfr' STC15F2K.h:21: syntax error: token -> 'P0' ; column 6 50RGBLEDStar.c:9: fatal error: when writing output to : Broken pipe I tried swapping out the STC15F2K.h with some other .h files unsuccessfully.

@@ceptimus I couldn't get the header file that matched with the code I had so I started over with 8051.h and used the old .c file as a template. I worked out all of the sbits that correspond to all of the diode pairs except for the D29 and D30 pair. Which on a stc15w408as looks like it is pin 13 or P5.5 based on the schematics I've seen. Your thoughts on how to find the missing sbit?

@@paulandrusyszyn9177 I don't have that kit, and I've not got the source files. Can you send me a download link, or email the files to me, and I'll see if I can get them to compile. ceptimus@ceptimus.co.uk

@@paulandrusyszyn9177 __sfr __at (0xC8) P5; __sfr __at (0xC9) P5M0; __sfr __at (0xCA) P5M1; You can manipulate individual bits in the P5 register with the normal |= and &= to set or reset a bit or ^= to toggle. To set P5.5 it would be P5 |= (1 << 5); to clear P5.5 it would be P5 &= ~(1 << 5); I've not tried it, because I don't have the chip, but you may be able to get at P5.5 as an sbit, using: __sbit __at (0xCD); // note adding the bit (5) to the base address (C8) - should work with SFR registers that are divisible by 8

The program shown here takes 2 mins 39 secs for the 'standard' hourglass to cycle. The faster options take respectively, 1 min 17 secs 38 secs 19 secs 9 secs 4.5 secs 2.5 secs 1.25 secs - You can see the time roughly halves with each button press. It would be easy to alter the code to cycle slower or faster - just editing a few numbers and recompiling. You've given me an idea to make a true egg timer version. It will start up flashing five rows, indicating 'five minutes'. Button presses will cycle the setting: 6, 7, 1, 2, 3, 4, 5, ... minutes. A long button press will start it showing the 'trickling hourglass' display until the set time has elapsed, when it will revert to the flashing rows time setting indication. I'll upload a fresh video demonstrating this alternative version - should be up soon (November 2024). The chip doesn't have a crystal controlled clock, so timings are not super-accurate, but should be within about 1%.

@@ceptimus sounds eggcellent, I’ll keep a track of your progress. Also I’ll be getting the usb converter and start some flashing of my own. You have any Idea where in the code I can edit the cycle time.

@@rogierkraan Easiest 'hack' change is to edit the value '804' in the line in the delayMillis() function - bigger values will make the display slower, smaller ones faster. You can put any value up to 65535 in place of the 804. If the number is bigger than 32767, then it's correct to put a U after it: like 40000U, though it will likely work without. For the new version, I'll probably change the delay to work more accurately, by having an interrupt-driven millisecond counter.

@ thanks, you are a great help!

No tengo pensado venderlo. Ahora mismo está en casa de mi hermana y sus hijos, nietos y otras personas lo ponen cuando la visitan.

@@ceptimus vendemelo

You can't do it with just an Arduino, because 12 volts is needed on one of the pins during programming. I did design a programmer based on an Arduino Mega, but it requires extra circuitry. You can find a video about it on my channel.

It is possible to clone an AT89C4051 chip, providing that when the original chip was programmed, the program lock bits, LB1 and LB2 weren't both set. If the lock bits are set, then even though the chip will still run its program, you can't read that memory to make a copy - the chip appears to be erased (full of FF bytes) when you read it. The only way to find out if your chip has its lock bits set is to attempt to read it. This won't harm the chip, and it will still contain its program and run afterwards. If you can read the chip then you can clone it, but if the chip appears empty when read, then there's no way I know of to recover the code from it.

@@ceptimus Thank you. that's really helpful. Is this Arduino set up you videoed able to read a 4051? Cheers

@@landroveraddict2457 Yes, you need to very slightly edit the code. The 4051 chip is pretty much the same as the 2051 except that it has twice as much Flash memory. The signature bytes for the 4051 are 1E 41 instead of 1E 21 for the 2051. I think I have some 4051 chips somewhere, so I'll set one up tomorrow and upload a new video showing it reading.

@landroveraddict2457 See modified sketch here, which will read both 2051 and 4051 chips: ruclips.net/video/tQbi8RlU6DM/видео.html

@@ceptimus You Sir are a god send, thank you

Also, make sure you have a common gnd connection to the nano, FET, and receiver.

@@ceptimus Thank you for the response. I checked all the connections with a multi-meter and it all comes out good. I have a common row for GND on my breadboard and everything is plugged into it (servo, arduino, receiver) as well as power. I have even tried to put the pull up resistor that you said was needed for other transistors with no luck on that as well.

@@fdm225 What model of FrSky receiver are you using? I may have one, and if so, I can try some tests myself.

@@ceptimus Archer RSPlus Mini, the interesting thing is I found a different library where using a soft serial port (and a serial adapter) I can print out the values and I got that to work; so I confirmed that my setup is correct from a hardware perspective. I am looking at the difference now between the two programs as the other project doesn't put the output to control servos and is only printing the values to the SoftSerial port.

@@fdm225 Ah, I don't have any Archer receivers, nor a transmitter, so I can't test that. It does sound like a bug in my code, sensitive to some slight difference, perhaps in the timing, of the SBUS signal from the Archer, is stopping my sketch from working. I hope you have success in comparing the SBUS decode routine in the working project with mine, and fixing the faulty code in my project.

Just search on Ali Express for "diy more pro mini strong" or similar. I just did that and found this link: www.aliexpress.com/item/32842932841.html The links usually expire fairly quickly, but the search will usually turn up several sellers - and you can choose the one with the best price+shipping to your location.

I've got three bare PCBs, with green solder mask like the one in my video - the PCBs cost me about £6 each delivered. P&P to Plymouth is about £3 so you could have one for £9, or two for £15. At the other extreme, if you want a complete picocomputer RP6502 all built up and tested, complete with the Pi Picos, it owes me £72, so say £75 including P&P to Plymouth. I'm sure you could get the parts cheaper though, if you shopped around: I bought them all from Mouser UK, and their prices and delivery charges are pretty steep (though very convenient).

PWM to SBUS: ceptimus.co.uk/PWM2SBUS.zip PWM monitor (use with the Arduino IDE Serial Plotter tool (Ctrl-Shift-L) to get a real-time graph of all 18 PWM inputs (115200 baud): ceptimus.co.uk/PWM18channelMo... SBUS to PWM: ceptimus.co.uk/SBUS2PWM.zip

I've a solution. You can change port 5 -> port 18. Change on main page #define PIN_SUPPRESS 5 to #define PIN_SUPPRESS 18. change to file outputChar.s #define Pin5 &PORTD, _BV(5) to #define Pin5 &PORTD, _BV(3) and #define UNSUPPRESS CBI DDRD, 5 #define SUPPRESS SBI DDRD, 5 to #define UNSUPPRESS CBI DDRD, 3 #define SUPPRESS SBI DDRD, 3 and the white line desepear

Merhaba pwm sinyallerini sbus sinyallerine çeviren alıcı ve verici kodunu bulabildin mi? Haftalardır arıyorum ama bulamadım. Uçağım da bulduğum kodlar çalışmıyor!

You could, providing you don't need a particularly accurate reading from the joystick. You'd run the two signals from the sensor to two analog channels, read them both, subtract one from the other to get the position. Without the, roughly, 50x amplification provided by the op-amp, you'd only get about ten or twenty counts when moving the sensor over the full range. That would probably be fine if you're just using the joystick as a game controller, but if you want to use it control a model aeroplane, then it's worth adding the op-amp and rail splitter to get a more stable signal with better resolution.

@@ceptimus thank you for that informative reply. I was under the assumption the output signals was from gnd to vcc. But I haven't received the PS3 joysticks yet which i ordered, so couldn't test :) The project I am working on is retropie DIY joystick with an ADC to i2c IC. Will be interesting to try.

Obrigado!

The sketch doesn't currently implement an escape code for clearing the screen, though it would be pretty easy to add - It would be Esc[2J The codes for cursor movement are standard VT100 escape sequences, sent by terminal emulators such as PuTTY. Search online for "VT100 escape codes" to find a list of them. I think the ones the sketch currently implements are: Home: Esc[1~ or Esc[H Del: Esc[3~ End: Esc[4~ or Esc[F Up: Esc[A Down: Esc[B Right: Esc[C Left: Esc[D Backspace: Ctrl-H or Ctrl-? Return: Ctrl-M Line Feed: Ctrl-J

@@ceptimus ok...great...thanks.