Відео

A technical term, obviously 😉 More accurately, the Arduino can be overloaded with requests/signals and become unresponsive and sluggish, agreed. But it is a relatively slow processor; the Espressif ESP32 is some 20x faster! The RasPi is not a real replacement as it has an operating system, so you can never be quite sure when it will service your latest (hardware) request, but it is even faster than an ESP32 so can easily handle the load.

@@RalphBaconYou are not wrong in what you say. the raspberry pi seems to work great on my predator setup running 7 servos, 3 sets of lights, and a 100 sound effects with no issues. The key difference being you have to set it to run a script at startup that runs another script. Python and C are very similar now you just have to use different syntax for the respective language to accomplish the same tasks...

Glad you got it working and watch out for splinters 😆

Correct. The CYD is effectively a (touch) screen attached to an ESP32 with all the connections made for you. Doesn't give you (m)any GPIO pins to play with, so treat as a self contained display module!

Declare the integer as you have done in the loop but with the word 'static' in front of it. This will ensure it is removed (by the compiler) from the actual loop but is accessible to the code in there. A sort of pseudo global variable!

@@RalphBacon Thank you Ralph😄

Thank you for the feedback.

Luckily, the ticks are the same as milliseconds here. I didn't want to confuse things!

Um... I'm not clear of the context in which you are applying this. I do millis() now minus oldMillis > delay between iterations then reset oldMillis and do the work.

@RalphBacon millis is timeNow, oldMillis is timeStarted and check if it's greater than delay - setTimePeriod. Reset oldMillis - is the new timeStarted, which is now..

I use VSC/PlatformIO and it's an add-in called something like Rainbow Indents (MS will suggest it)

Hmm, it depends on how much current the buzzer is going to take. More than 20mA seems unlikely but a 100Ω resistor will ensure current is limited to 5/100 = 50mA, still a bit high as 40mA max per pin!

@@RalphBacon Yes, makes sense to go higher on the current limiting resistor, but I noticed that you ommitted the resistor altogether and I was curious to know why you did that.

Because my buzzer only consumed a few milliamps, not enough for any concern. Measure it with a multimeter to determine whether a resistor is really necessary!

@@RalphBacon Good point, better to actually test to see if it is needed rather than just mindlessly following general rule!

You can use IR (infrared) diodes to do this. A modulated signal (like a TV remote control) should work.

Glad you got it working OK! Wow! 16 encoders, a glutton for punishment!

No, not really. ESP-NOW is limited to a 256-byte data packet, not a continuous stream. That said, you might be able to send enough packets per second to transmit audio that is clear.

A TTL-level MOSFET will switch cleanly but the propagation delay can only be found by (a) reading the data sheet (to see the delay in switching on) and (b) by trying it out using a 'scope. Best not to use any kind of gate resistor (slows down the gate capacitor charge time) even though some people seem to think they are necessary (I don't, but what do I know?)

@@RalphBacon Ok, thanks, so, my "very basic" assumptions were right... My perspective was to avoid artificially prolonging the data-sheet-delay by wrong-handling, and not to shorten that delay even further, so, thanks. The "trick" with the no-gateresistor, I´ve learned about it recently, too, after I have posted my question (above) to you. Still, the knowledge about the gate-resistor didn´t mean, that my question didn´t need any answer, so, your answer was bitterly needed, still... So, a big "Thanks" again...!!!

I hope so too!

Whatever you do, don't unsubscribe or you will miss any future videos!

Glad you are finding it useful.

Is that from my sketch or have you modified it? It seems to be complaining that the receive data structure is not what it is expecting. Email me the sketch if you still have issues and I can have a look much more easily than trying to do it here! (Email address is in my about page on the channel, only visible on a PC, not a phone).

OK, got your email and understand that downgrading the ESP framework (eg from 3.0.0 to 2.0.14) solved it for you. This is an ongoing problem with the new Espressif Arduino framework using the newer IDF (Espressif framework) APIs. OLD ESP-NOW callback function: void OnDataRecv(const uint8_t *mac, const uint8_t *incomingData, int len) {... your code...} NEW ESP-NOW (Framework 3.0) callback function: void onDataRecv(const esp_now_recv_info_t *info, const uint8_t *incomingData, int len) {... your code...} If you used to access the (old) first parameter of the sender's MAC address, you can get that now from the 'info' parameter pointer thus: info->src_addr although you will have to convert the octets into printable bytes as we always have had to.

@@RalphBacon Thank you both!!! I just started working with ESP-NOW and I was very frustrated with this invalid conversion error. Have a great day!

Flashing LEDs are all well and good for a visual demo, but most users like yourself will want to do something more useful. Start small, and expand. It's too easy to not spend enough time on drawing out all the possible states and start coding before you're ready! Don't ask how I know this. 🤦‍♂️

Not done, exactly but certainly taking a break. Getting my life back! I've several topics in mind for future videos and I'm sure I'll eventually get the time to do the videos.

By "Factory Reset" what do you mean? When it comes out off the production belt it normally has the Blink sketch loaded (just to give a visual indication it's all working) but that's almost irrelevant. Can you still upload sketches to the board?

Thanks for the info, sounds pretty easy!

So you want a Mars RTC? I've never heard of anything that keeps anything other than Earth time, but you could program a simple Arduino (UNO/Nano) to keep Mars time instead. The downside of using any kind of electronics is that you will be hard pushed to run it on a battery like the DS3231 (pretty accurate). You _could_ put it to sleep for a second, wake up, do the necessary calculations and put it back to sleep again to enable running off a battery but an Arduino will still take several milliamps for the short period it is running. You mention using NTP to set the time (initially). That requires Internet access. If you don't need to run it on a battery then one of the small, newer ESP32 devices do this all automatically (for Earth Time) but you could easily run a simple routine to convert to Mars Time. The ESP32 gets the time from NTP periodically to ensure there is no drift as every microcontroller I've ever tried is not accurate enough for a clock. I'm wondering if a DS3231 that has been correctly set initially could be used with an ATTINY processor of some sort so it can run on a battery for months if not years...? Finally, if I were a Martian I would not say that my day is 24h and 37min (ie based on Earth time). It would just be 24 hours (or whatever the Martians agreed) and each "second" would be a longer-than-Earth second and so on. Now that could be programmed just by counting 'ticks' in a microcontroller (milliseconds) and then translating that to Mars seconds, minutes and hours. Just a thought.

@RalphBacon I want to thank you for responding to my crazy question. The NASA Mars Rover team runs on Mars time, so they know when the window to transmit to the Rover is open.. (Rover being on the Earth side of Mars) So in my scenario, if the Mars Rover had a spotting scope pointed straight up. Every time Earth passed the center field of view, would be noon.. (I know real Mars noon is based on the sun..not Earth) but I want a clock that would tell me when the Mars Rover was on the Earth facing side of Mars.. (6AM to 6PM Mars time) If I had to look it up every day, then there would be no point to using a RTC at all... My RTC is the Ds1307.. It has an error of 7 seconds a day, which means I would have to NTP 7 times a day to keep it accurate to the second.. I was hoping there was a RTC that I could tweak the tick counter to compensate for tolerance errors in the crystal... If it is fast by 7 seconds a day. How many ticks do I need to adjust by so it is accurate to 1 millisecond a day??? Do you know of any RTCs that have a tick count register to adjust how many ticks equals one second? Then I make a 24 hour Mars day = (24h 37m 22.5s Earth time) The only problem this would not fix is a Mars year is over 600 days but I would at least have the Mars 24 hour day worked out... I hope you could do a video on making a Mars clock to solve this problem.. One that you only need to set one time a year or better...

I'm afraid I don't know of a variable/programmable RTC, but the DS1307 is very poor at keeping time, which is why I mentioned the much more accurate DS3231. A daily NTP update is all you need, at most, perhaps even once a week is good enough. If you don't need to run on a battery (like the RTC does) then you can easily program your own Mars clock, ideally with a huge LED display so you don't miss a critical time (like the ones NASA uses when a meteor is about to impact Earth). Using an ESP32, it auto-updates from an NTP. Although you mention "Mars Time", you know how many seconds in a Mars Hour, Day etc so by using a standard RTC DS3231, you can just count the Earth seconds and create your own Mars clock.

@@RalphBacon My thought was to proportionally stretch the 24 hour RTC day, so it takes 24h 37m 22.5s to complete (or whatever the actual calculated Mars day is).... The problem with using (Earth time RTC) is knowing how many days expired since it was set.... Is it possible to edit the RTC library to stretch a 24 hour day? I would be glad to use another RTC... I was looking at the data sheet for the RV-3028-C7 (RV-3028-C7_App-Manual.pdf) and it looks like it has (calibration registers) to adjust Aging compensation with user programmable EEPROM Offset value (Factory Calibrated value may be changed by the user) but I don't know if I could add as much as 37m to a day.. My thought on making the RTC keep the time, is if I set it, I could put it away for a month or two and when I get it out, it is still on Mars time.. Anyway, thank you so much for your ideas.. I would love to see if such a thing is possible..

I was thinking more in terms of counting the seconds from a DS3231 (probably using an interrupt) via the 1Hz output pin, and then in the Arduino program incrementing the user-controlled secs/mins/hours/days etc according to Mars rules. That way, you have full control over all aspects of Mars time.

More likely your antenna is not correctly attached, or you live up a mountain, as I got a number of stations in a town (as the video demonstrated, I believe).

It _might_ work with its weak output - or else wire it to a transistor to boost the signal. Make sure you test your design on a breadboard first!

@@RalphBacon Thank you for your advice. I have more questions. What is Zener Diode model number you put in your circuit? How many voltage and how many watt? and what value of that capacitor?

Are you talking about the fuse resetter circuit, or something else? No Zener in the former, hence my query.

@@RalphBacon At 21:17, there is one component next to the capacitor. It sits between 12v rail and resistor that goes into the collector of the BC547 transistor. It looks like Zener Diode. That's why I'm confused.

Thanks for the feedback!

I'm delighted you did this! Congrats! Start small and then you can increase the complexity of your library (class). 👍

Well, it's been a long time since I learned how to do SM programming, but from what I remember the most important thing it so list (in bubbles) all the possible states that your device could find itself in, then link the states (with an arrow) that defines the action (or change of state) to bring it to that state. Your best bet is to search UA-cam for a simple, beginners intro into this topic - I bet there are loads!

My pleasure!

The receiver module that you use might already have this functionality - or you might have to build one yourself with a simple BT receiver module. You phone might have an app available to do this too, I use one for Wi-Fi so maybe someone has built one for BT too?

Glad it was helpful!

Yes, you at the back, what's your question? Switch/Case? It's just a nice "IF" style construct. Instead of multiple, nested (and unmaintainable) if else... if else... statements we just use ONE switch case to inspect a (numeric) variable, then have a CASE statement for each matching condition. log_v is an ESP32 logging macro that only logs if the build debug level is set to VERBOSE (v). There's a version for d (Debug), i (Info) w (warning) and e (error) too. Try it out, very useful.

@@RalphBacon Thanks for giving me answer. The "case" way I now get, I now use it to se esp wake up reason.

Yes, I've tried this too, but once re-solidified the flux can be too solid for a syringe to work well. You might have to add some extra "flux thinner" aka IPA.

I don't believe there are any issues with the information in this video; it's a barebones ATMega328p chip running in a minimal configuration! What's the problem you're experiencing?

I've had a similar situation; I just got a pair of earphones and plugged them into the device to shut it up, and listen to better quality via the BT method. You could also just use a spare jack plug, not a complete set of headphones to do this. Simple and fairly unsophisticated, but it does the job!

Glad you liked it!

Thank you! Cheers!