Arduino: Eh, just grab some pliers and bend the pins

The Something Awful Forums > Discussion > Hobbies, Crafts, & Houses > Arduino: Eh, just grab some pliers and bend the pins

«‹›89 »

My Rhythmic Crotch: Jan 13, 2011

Conflict for the usage of pin 10? SD and the display need their own CS pins.

# ? Sep 8, 2019 20:03

Adbot: ADBOT LOVES YOU

# ? May 29, 2024 18:54

huhu: Feb 24, 2006

I never use the SD_CS pin in the sketch I posted.

# ? Sep 8, 2019 20:52

My Rhythmic Crotch: Jan 13, 2011

Ah I shoulda looked more closely, looks like you're using i2c to connect to the display, in which case no CS needed.

Uh, grasping at straws, you could try using the other i2c interface (I think the 328p has 2 of them?) If that fails, try another display library, or start digging into the one you're using looking for incompatibilities.

It does not look like you're out of flash or ram... The flash and ram usage reported there at build time is coming from GCC, which I would place a lot more faith in than anything else in the chain. Though I suppose you could mess around with the optimization settings and try to get a smaller build, just as a sanity check.

My Rhythmic Crotch fucked around with this message at 21:48 on Sep 8, 2019

# ? Sep 8, 2019 21:45

babyeatingpsychopath: Oct 28, 2000; Forum Veteran

If commenting out SD.h makes the problem go away, then the two libs are likely using the same resources. See if you get different behavior putting SD.h first.

# ? Sep 8, 2019 21:59

TwystNeko: Dec 25, 2004; *ya~~wn*

I've decided to finally finish my second DeadMau5 head. The basic build is mostly done, but now I'm working on the electronics. Currently, it's a bit overkill - an ESP32 to run the LEDs in the eyes, a Nano to do audio beat detection and signal the ESP32 to "pulse" the eyes. There's also a super-basic circuit to bring in the audio, from a BT audio speaker amp.

example of how it looks: https://imgur.com/a/ZMLBzcR - you can see the blue LED on the right, that's the beat detection.

I'd like to see if I could consolidate the ESP32 and the Nano into one board.

Relevant code:
Nano Beat Detection: https://pastebin.com/a3Q56NWk - uses a LCD display for tuning the pulses, will probably remove when done
ESP32 LED code: https://pastebin.com/Q4Gpk2s0 - just basic stuff, uses an interrupt to catch the pulses from the Nano.

Any suggestions? Or is it better off separated? Supplying power's a bit of a pain, as it's 9V for the audio amp, 5v for the nano, 3.3 for the esp32. I'll probably put a separate battery in for the LED supply, just to aid battery life.

# ? Sep 9, 2019 18:16

Cocoa Crispies: Jul 20, 2001; Vehicular Manslaughter!; Pillbug

babyeatingpsychopath posted:

This is neat, and i approve of the concept. Are you going to need a fairly beefy microcontroller to pulsecount 2+10 LEDs?

Ambrose Burnside posted:

I've got a spare ESP32 I was going to use and I'd be amazed if that isn't fast enough

I haven't had trouble using an ESP32 to drive 256 LEDs, the only stuff I have had issues with it is roaming between multiple wifi APs on the same SSID.

Cocoa Crispies posted:

ESP-32 owns bones. I'm using one to control an LED grid over wifi, with an OLED status display. It's nicer than the ESP-8266 because instead of requiring a shitload of interrupts and fiddly timing to pulse the LED matrix, it has some kind of peripheral that handles that for you.

I got started because I saw a neat display at SHA-2017 that was also world-controllable, and figured I could fix some problems that one had while making it not a whole-wall thing. Only in the last month (after seeing similar toys at another computer festival) did I get off my duff and put it together, and I'm pretty happy I did.

# ? Sep 9, 2019 20:58

Cocoa Crispies: Jul 20, 2001; Vehicular Manslaughter!; Pillbug

TwystNeko posted:

Any suggestions? Or is it better off separated? Supplying power's a bit of a pain, as it's 9V for the audio amp, 5v for the nano, 3.3 for the esp32. I'll probably put a separate battery in for the LED supply, just to aid battery life.

The ESP32 I have is happy with 5V, and also provides a 3.3V pin.

I power it off a USB power bank, the LEDs off the ESP's 5V pin, and the ESP 3.3V pin goes to a level shifter for the benefit of the LEDs (and simplifies the breadboard layout too).

# ? Sep 9, 2019 21:01

mobby_6kl: Aug 9, 2009; by Fluffdaddy

Cocoa Crispies posted:

The ESP32 I have is happy with 5V, and also provides a 3.3V pin.

I power it off a USB power bank, the LEDs off the ESP's 5V pin, and the ESP 3.3V pin goes to a level shifter for the benefit of the LEDs (and simplifies the breadboard layout too).

Doing pretty much the same thing too, at least during development now. Though my board works fine with 3v signals. Also lol people were freaking out how you totally need 3 amps to drive the board but under typical conditions it's drawing like 400ma with the ESP that's receiving on wifi and doing number crunching.

TwystNeko posted:

I've decided to finally finish my second DeadMau5 head. The basic build is mostly done, but now I'm working on the electronics. Currently, it's a bit overkill - an ESP32 to run the LEDs in the eyes, a Nano to do audio beat detection and signal the ESP32 to "pulse" the eyes. There's also a super-basic circuit to bring in the audio, from a BT audio speaker amp.

example of how it looks: https://imgur.com/a/ZMLBzcR - you can see the blue LED on the right, that's the beat detection.

I'd like to see if I could consolidate the ESP32 and the Nano into one board.

Relevant code:
Nano Beat Detection: https://pastebin.com/a3Q56NWk - uses a LCD display for tuning the pulses, will probably remove when done
ESP32 LED code: https://pastebin.com/Q4Gpk2s0 - just basic stuff, uses an interrupt to catch the pulses from the Nano.

Any suggestions? Or is it better off separated? Supplying power's a bit of a pain, as it's 9V for the audio amp, 5v for the nano, 3.3 for the esp32. I'll probably put a separate battery in for the LED supply, just to aid battery life.

I'm not entirely sure what it's supposed to do but it sounds like it should be possible to do everything with just the ESP, receive BT audio, do the beat detection, and flash LEDs to your beats. I think this would really simplify the circuits.

I'm having a pretty strange situation and it seems it's more of an Arduino weirdness than C++, so maybe someone else also experienced it. Basically, things break if I move the definition of a class method from the main .ino file to a separate cpp file. This minimal example is supposed to fill the mentioned led board with white color:

code:

class dummy_test {
 public:
  void clear_display(PxMATRIX *d);
};

  void dummy_test::clear_display(PxMATRIX *d) {
      Serial.printf("%p\n", (void*)d);
      d->fillScreen(0xFFFF);
  }

This works as expected unless the definition is in another source file. Everything else is fine, the line gets printed with the correct pointer address but it just doesn't do anything. I don't have the entire C++ standard memorized, so I can't think of anything to explain this, but i do know Arduino does some weird poo poo when building it so maybe some knows what's up.

# ? Sep 16, 2019 23:53

mod sassinator: Dec 13, 2006; I came here to Kick Ass and Chew Bubblegum,
and I'm All out of Ass

How much flash memory and RAM is your sketch taking? If you're getting close to the limits (like 90% flash used and lots and lots of RAM used at runtime) I've seen compilers do some really nasty and bizarre things, particularly when they try to optimize.

One important difference between defining the function code in a header (what works for you) vs. in a source file (what's failing) is that in the header the compiler is much more likely to just inline the code directly wherever the function is called (and IIRC the standard actually mandates this, although who knows if compilers really follow that as we realized long ago compilers and compiler tech is much better at figuring out what to do for performance vs. what a standard says). This is a very important difference as in one case your code hums along fine and executes a few extra instructions when your class member is called vs. in the other case it has to throw some variables on the stack, go invoke a function, pull those results off the stack, etc. and if your stack is already running very low on memory... boom things can fail.

None of this is helped by Arduino hiding errors and warnings from the compiler unfortunately. Ten years ago C++ compilers sucked (and 10 years before that they _really_ sucked) so Arduino just hides all their complaints. But nowadays compilers are pretty drat good and we've had decades of time to see all the fun ways C++ lets you shoot yourself in the foot. So long story short, see if you can turn on warnings and errors and see what happens. I have a strong suspicion you might find some weird warnings that give you a hint of ways classes and pointers might be being abused in unexpected (and unspecified behavior) ways.

# ? Sep 17, 2019 00:16

Duck and Cover: Apr 6, 2007

I plan on following this but I have some questions. https://www.hackster.io/mrelia100/particle-photon-flip-dot-clock-2d83a7 As I've never done anything Arduino related and believe electricity to be some sort of magic I figure it wise to ask here.

If I wanted to power this using a rechargeable battery what would I need? What size battery would be reasonable (I mean the longer the better for run time but I'd want to keep it from being longer/wider and not too thick)?

The displays manual says the following. For a 14x28 panel (mines 7x28).
Max. current consumption 680mA
Idle current 40mA
Energy to swap one complete panel 160ms x 680mA

https://www.sparkfun.com/products/12708
This will allow me to not need an internet connection for basic clock function right?

# ? Sep 17, 2019 01:45

mod sassinator: Dec 13, 2006; I came here to Kick Ass and Chew Bubblegum,
and I'm All out of Ass

I'd step back and think about what you really want out of the project. Is it just the cool look of a dot display, or do you want the sound and whole experience? I say this because old mechanical tech like that is just a plain bitch to run and maintain. Moving parts stop moving without oil/lubrication over time. Moving parts need a lot of power to... move. If it's just a neat eye catching display you want then a TV screen showing a web page that animates little fake moving dot displays (or whatever else you can imagine) would be a lot easier to manage.

If you're dead set on a portable version of that display it's going to be challenging. It looks like they run off 24 volts of power which these days is a little odd to find easily. It's not a common battery voltage so you're going to have to put some batteries in series or find a switching power supply that can boost a small voltage up to 24 volts at an amp or more of current. My gut tells me a couple car batteries are probably what you'd want depending on how long you want this thing to run. But before you run off and go do that, I'd get a real EE involved or talk to someone who has used those displays off battery power. Batteries are big chemical vats of energy and when everything goes well they are magic that makes voltage, but when things go wrong they are big things filled with caustic acid and a ton of potential energy... dropping a screwdriver across some leads or connecting things the wrong way could very quickly cause a huge problem.

Based on those specs you need almost a full amp of current to switch the display. That's a lot of power, especially if you want the display to constantly be changing. Batteries are measured in amp-hours, that is how many hours of time they can provide the specified current. Some AA batteries are ~1 amp hour, so you could in theory run the display with it changing all the time and get an hour of runtime from AA batteries. Car batteries are about 50 amp hours (they don't actually measure it in amp hours but you can convert), so the display could go wild for maybe 2 days with car batteries. Still not too great, so you probably want some way to keep the batteries trickle charged and tone down the activity of the display to extend its runtime.

Alternatively consider a gas powered generator if this thing really has to be portable and away from any power source. Plug a 24 volt switching power supply in (like the guide shows) and just keep it topped up with fuel. Honda makes great little quiet, low emission generators.

But I have a feeling this is more for a neat idea to just display information. I'd go for a virtual display on a TV screen in that case. There's a reason you don't see these mechanical displays around everywhere anymore...

# ? Sep 17, 2019 02:51

My Rhythmic Crotch: Jan 13, 2011

Duck and Cover posted:

I plan on following this but I have some questions. https://www.hackster.io/mrelia100/particle-photon-flip-dot-clock-2d83a7 As I've never done anything Arduino related and believe electricity to be some sort of magic I figure it wise to ask here.

If I wanted to power this using a rechargeable battery what would I need? What size battery would be reasonable (I mean the longer the better for run time but I'd want to keep it from being longer/wider and not too thick)?

The displays manual says the following. For a 14x28 panel (mines 7x28).
Max. current consumption 680mA
Idle current 40mA
Energy to swap one complete panel 160ms x 680mA

https://www.sparkfun.com/products/12708
This will allow me to not need an internet connection for basic clock function right?

Huh, I had no idea any companies still sold those things. Pretty expensive for your first DIY thing.

Looks like it needs 24 volts, so your battery is going to be pretty chunky to get the required voltage.

Yes, an RTC will keep a pretty decent clock without requiring any active connection to anything.

You know those things are really loud too, right? What are you trying to make or do?

# ? Sep 17, 2019 02:55

Cojawfee: May 31, 2006; I think the US is dumb for not using Celsius

How did you even get one? Nothing I can find even lists a price (aka expensive).

# ? Sep 17, 2019 03:09

Duck and Cover: Apr 6, 2007

My Rhythmic Crotch posted:

Huh, I had no idea any companies still sold those things. Pretty expensive for your first DIY thing.

Looks like it needs 24 volts, so your battery is going to be pretty chunky to get the required voltage.

Yes, an RTC will keep a pretty decent clock without requiring any active connection to anything.

You know those things are really loud too, right? What are you trying to make or do?

Clock. Maybe temperature/humidity. I'm more interested in a finish product then I am in the DIY aspect but the programming (hey other people have done that for me in theory I can totally use other peoples code/edit it for my purposes) is simple enough and the other stuff doesn't look too complicated. If it is I have brothers who I can probably get to help me (totally code for them doing it for me).

mod sassinator posted:

If you're dead set on a portable version of that display it's going to be challenging. It looks like they run off 24 volts of power which these days is a little odd to find easily. It's not a common battery voltage so you're going to have to put some batteries in series or find a switching power supply that can boost a small voltage up to 24 volts at an amp or more of current. My gut tells me a couple car batteries are probably what you'd want depending on how long you want this thing to run. But before you run off and go do that, I'd get a real EE involved or talk to someone who has used those displays off battery power. Batteries are big chemical vats of energy and when everything goes well they are magic that makes voltage, but when things go wrong they are big things filled with caustic acid and a ton of potential energy... dropping a screwdriver across some leads or connecting things the wrong way could very quickly cause a huge problem.

Based on those specs you need almost a full amp of current to switch the display. That's a lot of power, especially if you want the display to constantly be changing. Batteries are measured in amp-hours, that is how many hours of time they can provide the specified current. Some AA batteries are ~1 amp hour, so you could in theory run the display with it changing all the time and get an hour of runtime from AA batteries. Car batteries are about 50 amp hours (they don't actually measure it in amp hours but you can convert), so the display could go wild for maybe 2 days with car batteries. Still not too great, so you probably want some way to keep the batteries trickle charged and tone down the activity of the display to extend its runtime.

Alternatively consider a gas powered generator if this thing really has to be portable and away from any power source. Plug a 24 volt switching power supply in (like the guide shows) and just keep it topped up with fuel. Honda makes great little quiet, low emission generators.

It's going to be a clock, which I guess I wasn't clear about. I thought it might be too demanding but was hoping as it's not flipping dots often, and not many when I do that I'd be able to power it with a reasonably small battery set up.

mod sassinator posted:

I'd step back and think about what you really want out of the project. Is it just the cool look of a dot display, or do you want the sound and whole experience? I say this because old mechanical tech like that is just a plain bitch to run and maintain. Moving parts stop moving without oil/lubrication over time. Moving parts need a lot of power to... move. If it's just a neat eye catching display you want then a TV screen showing a web page that animates little fake moving dot displays (or whatever else you can imagine) would be a lot easier to manage.

But I have a feeling this is more for a neat idea to just display information. I'd go for a virtual display on a TV screen in that case. There's a reason you don't see these mechanical displays around everywhere anymore...

Any attempt to emulate a flip dot/split flap/pinball machine on a screen isn't going to look right. It's also just cool technology.

Cojawfee posted:

How did you even get one? Nothing I can find even lists a price (aka expensive).

Money and yeah you have to ask for the price list because businesses are dumb. Shipping was like 90 euros. You can also try ebay but those old bus signs aren't going to be as user friendly.

Duck and Cover fucked around with this message at 06:20 on Sep 17, 2019

# ? Sep 17, 2019 06:04

mobby_6kl: Aug 9, 2009; by Fluffdaddy

mod sassinator posted:

How much flash memory and RAM is your sketch taking? If you're getting close to the limits (like 90% flash used and lots and lots of RAM used at runtime) I've seen compilers do some really nasty and bizarre things, particularly when they try to optimize.

One important difference between defining the function code in a header (what works for you) vs. in a source file (what's failing) is that in the header the compiler is much more likely to just inline the code directly wherever the function is called (and IIRC the standard actually mandates this, although who knows if compilers really follow that as we realized long ago compilers and compiler tech is much better at figuring out what to do for performance vs. what a standard says). This is a very important difference as in one case your code hums along fine and executes a few extra instructions when your class member is called vs. in the other case it has to throw some variables on the stack, go invoke a function, pull those results off the stack, etc. and if your stack is already running very low on memory... boom things can fail.

None of this is helped by Arduino hiding errors and warnings from the compiler unfortunately. Ten years ago C++ compilers sucked (and 10 years before that they _really_ sucked) so Arduino just hides all their complaints. But nowadays compilers are pretty drat good and we've had decades of time to see all the fun ways C++ lets you shoot yourself in the foot. So long story short, see if you can turn on warnings and errors and see what happens. I have a strong suspicion you might find some weird warnings that give you a hint of ways classes and pointers might be being abused in unexpected (and unspecified behavior) ways.

It's using around 55% of flash and 20% of RAM so far at least for static variables, haven't checked recently but I think I had over 200k free heap too. I've also added more code since this issue first appeared so that's probably not it, but I'll check it out. Interesting point about inlining, I'll try to experiment with that too and we'll see.

Duck and Cover posted:

Yeah I checked out their page earlier and didn't see a price list which was suspicious. Clearly you're committed at this point :v:

mod sassinator posted:

I'd step back and think about what you really want out of the project. Is it just the cool look of a dot display, or do you want the sound and whole experience? I say this because old mechanical tech like that is just a plain bitch to run and maintain. Moving parts stop moving without oil/lubrication over time. Moving parts need a lot of power to... move. If it's just a neat eye catching display you want then a TV screen showing a web page that animates little fake moving dot displays (or whatever else you can imagine) would be a lot easier to manage.

Eh it's a dumb DIY project, like mine and I'd imagine most ITT, so I don't think "practical", "reliable" or "cost-effective" really enters into it

Maybe I'm missing something but the power issue doesn't seem that bad. 24v is inconvenient but you won't need to flip the entire board (unless you want) every time. Based on those specs earlier it seems like it'd need under 2 mA to flip a dot, and on average you might need to do what, 20? So 40mA for 160ms, or in an hour around 10 seconds of flipping if you refresh every minute. So like 100 mAh for refreshes and 40mAh in idle current per hour. Of course at 24V that's quite a bit if power and you won't get far even on a giant 6S lipo. Why does it need to be portable anyway?

# ? Sep 17, 2019 09:45

Duck and Cover: Apr 6, 2007

mobby_6kl posted:

It's using around 55% of flash and 20% of RAM so far at least for static variables, haven't checked recently but I think I had over 200k free heap too. I've also added more code since this issue first appeared so that's probably not it, but I'll check it out. Interesting point about inlining, I'll try to experiment with that too and we'll see.

Yeah I checked out their page earlier and didn't see a price list which was suspicious. Clearly you're committed at this point

Eh it's a dumb DIY project, like mine and I'd imagine most ITT, so I don't think "practical", "reliable" or "cost-effective" really enters into it

Maybe I'm missing something but the power issue doesn't seem that bad. 24v is inconvenient but you won't need to flip the entire board (unless you want) every time. Based on those specs earlier it seems like it'd need under 2 mA to flip a dot, and on average you might need to do what, 20? So 40mA for 160ms, or in an hour around 10 seconds of flipping if you refresh every minute. So like 100 mAh for refreshes and 40mAh in idle current per hour. Of course at 24V that's quite a bit if power and you won't get far even on a giant 6S lipo. Why does it need to be portable anyway?

I never said it had to be. The questions were for me to get enough information to decide if it's worth the trouble. I think it'd look nicer without a wire, and it'd give me better flexibility for placement that's why I wanted to do it. Charging it once a day would be too often, I'd probably hope for at least 3 days which seems unlikely/more expensive than I want to pay.

Anyway. https://www.hackster.io/iizukak/flip-dot-clock-3dd850 is using a regulator to make the 24v to 9v. My understanding is this will work because of the model of Arduino he's using (Arduino Pro mini 5V) can accept that voltage. I wanted integrated wireless, so I was going to use a different Arduino (probably 3.3) If I were to use a 3 terminal regulator would it generate too much heat as I do plan on having an enclosure made/making one? Should I use something like https://www.digikey.com/product-detail/en/cui-inc/P78A03-0500/102-3780-ND/6098489.

# ? Sep 19, 2019 18:25

My Rhythmic Crotch: Jan 13, 2011

Just my humble opinion, I think you ought to pump the brakes a bit and think about the goals for the project and how you're going to get there.

Duck and Cover posted:

Clock. Maybe temperature/humidity. I'm more interested in a finish product then I am in the DIY aspect but the programming (hey other people have done that for me in theory I can totally use other peoples code/edit it for my purposes) is simple enough and the other stuff doesn't look too complicated. If it is I have brothers who I can probably get to help me (totally code for them doing it for me).

... snip ...

I wanted integrated wireless, so I was going to use a different Arduino (probably 3.3) If I were to use a 3 terminal regulator would it generate too much heat as I do plan on having an enclosure made/making one? Should I use something like https://www.digikey.com/product-detail/en/cui-inc/P78A03-0500/102-3780-ND/6098489.

It sounds like you are waffling between "just straight up copy this" and "I'm gonna freeball it / have someone else do it for me". It's nearly impossible to provide any advice on these kind of things because you're not asking a single, specific question but something more like "how do I start a fire to trap a buffalo to make him fall down a cliff so I can kill him for the fall harvest so that I can become chief of the tribe and ..."

A more answerable set of questions might be:
- What is the max input voltage that <particular arduino> can accept?
- How many watts of waste heat will be generated from dropping 24->9 volts at X amps?
- How do I dispose of that waste heat (heat sink, passive, active)?
- Given the challenge of the waste heat disposal, how would I select a switching regulator?

Not trying to be a dick, just something to think about.

# ? Sep 19, 2019 22:18

Splode: Jun 18, 2013; put some clothes on you little freak

Do not use a linear regulator to drop 24 volts to 9

# ? Sep 19, 2019 22:21

taqueso: Mar 8, 2004

Splode posted:

Do not use a linear regulator to drop 24 volts to 9

Of course you *can* if the current is low enough, but this is real good advice generally. No reason to add a burn hazard (burning your skin not starting a fire hopefully) to your project.

e: There are lots of switchers designed to work as pin-compatible 7805 replacements

taqueso fucked around with this message at 22:32 on Sep 19, 2019

# ? Sep 19, 2019 22:28

Cojawfee: May 31, 2006; I think the US is dumb for not using Celsius

Also, that guide lists a 7805 which drops to 5 volts, not 9 volts.

# ? Sep 19, 2019 22:29

Ambrose Burnside: Aug 30, 2007; pensive

If you're willing to pay beaucoup bucks for a brand-new flip-dot panel and don't have the ability to make this thing yourself, my thinking is itll likely make sense + you wouldn't have to pay THAT big a premium to just commission such a clock from someone. Odds are they'll source a used matrix panel and do the legwork to drive it, which will be s i g n i f i c a n t l y cheaper than a new module from [the only place in the world that still makes them new], so even with their labour et al premium tacked on you might still be pleasantly surprised at the final quote

# ? Sep 19, 2019 22:35

Sagebrush: Feb 26, 2012

Splode posted:

Do not use a linear regulator to drop 24 volts to 9

Just consider it a way of ensuring the system runs at a stable temperature

# ? Sep 19, 2019 23:34

Rexxed: May 1, 2010; Dis is amazing!
I gotta try dis!

Since he probably doesn't understand the issue based on the previous posts, a linear regulator will give you a stable output voltage but it gets very hot if it's used for a large voltage drop (depending on how much current you're running through it and how much power it will be sinking). What you want to do is use a small switching dc to dc power unit which can be configured to drop that 24V to 5v or 3.3v for your microcontroller without using a ton of excess power that gets converted to heat. These are usually called buck converters (boost converters go the other way, low voltage to high).

# ? Sep 19, 2019 23:51

Cojawfee: May 31, 2006; I think the US is dumb for not using Celsius

The linear regulator gives you your desired voltage and converts the rest into heat which goes out into that metal tab on the top. You're meant to solder that tab to a heatsink of some sort.

# ? Sep 20, 2019 00:30

Duck and Cover: Apr 6, 2007

My Rhythmic Crotch posted:

Just my humble opinion, I think you ought to pump the brakes a bit and think about the goals for the project and how you're going to get there.

It sounds like you are waffling between "just straight up copy this" and "I'm gonna freeball it / have someone else do it for me". It's nearly impossible to provide any advice on these kind of things because you're not asking a single, specific question but something more like "how do I start a fire to trap a buffalo to make him fall down a cliff so I can kill him for the fall harvest so that I can become chief of the tribe and ..."

A more answerable set of questions might be:
- What is the max input voltage that <particular arduino> can accept?
- How many watts of waste heat will be generated from dropping 24->9 volts at X amps?
- How do I dispose of that waste heat (heat sink, passive, active)?
- Given the challenge of the waste heat disposal, how would I select a switching regulator?

Not trying to be a dick, just something to think about.

I can read specs, like they all say what the max voltage is, some are 5 (I think I saw one with a regulator that had a max of 9v), the wifi ones seem to tend towards 3.3v. Which you probably know.
-
-
Uhh passive unless it's too much then aluminium heat sink. Like this question is fine, but it wasn't necessary for any of the flip dot clock designs I've seen and I don't think I'm producing any more wasted heat then they are, but I don't know that for a fact. Of course this is stupid, as if I'm using wifi (increased power consumption) and using 3.3v that would increase my wasted heat (right? I think I'm right).

Really I can't ask the good questions until I know which Arduino I'm using, and I wasn't sure which Arduino I wanted to use until I got an idea of how much worse 3.3v was to step down to compared to 5v. Like if I use the 5v one that's in the tutorial I can just follow the tutorial, and add wifi and maybe temperature/humidity (possibly generating too much heat through the increase in power consumption) but I think I rather have those things included on the board just for simplicity sake.

Rexxed posted:

Since he probably doesn't understand the issue based on the previous posts, a linear regulator will give you a stable output voltage but it gets very hot if it's used for a large voltage drop (depending on how much current you're running through it and how much power it will be sinking). What you want to do is use a small switching dc to dc power unit which can be configured to drop that 24V to 5v or 3.3v for your microcontroller without using a ton of excess power that gets converted to heat. These are usually called buck converters (boost converters go the other way, low voltage to high).

I think I did understand that as evidence by linking to what I thought would be a solution a dc to dc converter. Guess I wasn't clear communicating that fact. Unless you're talking about something different? I mean it does say it's dc to dc converter. https://www.digikey.com/product-detail/en/cui-inc/P78A03-0500/102-3780-ND/6098489

Edit: I see my link got truncated. Thus some confusion. https://tinyurl.com/yxg8pa64

Cojawfee posted:

Also, that guide lists a 7805 which drops to 5 volts, not 9 volts.

Probably something else I looked at and I got mixed up.

Duck and Cover fucked around with this message at 02:41 on Sep 20, 2019

# ? Sep 20, 2019 02:02

Rexxed: May 1, 2010; Dis is amazing!
I gotta try dis!

Duck and Cover posted:

I think I did understand that as evidence by linking to what I thought would be a solution a dc to dc converter. Guess I wasn't clear communicating that fact. Unless you're talking about something different? I mean it does say it's dc to dc converter. https://www.digikey.com/product-detail/en/cui-inc/P78A03-0500/102-3780-ND/6098489

Edit: I see my link got truncated. Thus some confusion. https://tinyurl.com/yxg8pa64

Oh yeah, I did miss that, but that may work if you only need 500mA. I use these a lot for stuff like using 12V fans on my 24V 3d printer and knocking down the LIon batteries on my robot thing to 5v for an arduino:
https://smile.amazon.com/LM2596-Converter-3-0-40V-1-5-35V-Supply/dp/B01GJ0SC2C/
https://smile.amazon.com/LM2596s-Converter-Step-down-Regulator-Stabilizer/dp/B07CVBG8CT/

The one you linked will probably work fine depending on the current draw of your microcontroller, and it's manufactured to be pin compatible to a linear regulator. The ones I linked are little separate circuit boards intended to be tossed into an enclosure with some other boards and wired up. The ones with the LED displays are kind of neat for seeing what voltage you're getting but it's a bit of an extravagance because you usually just set the voltage with a multimeter before hooking it up to whatever output you're using. They're also bigger.

# ? Sep 20, 2019 03:09

Duck and Cover: Apr 6, 2007

7$ + shipping Arduino MKR WiFi 1010 first 500 subscribers

https://store.arduino.cc/digital/create sign up for the monthly get a discount code. There should be a window saying as such. So I'll be using that I guess.

# ? Sep 20, 2019 17:39

Cojawfee: May 31, 2006; I think the US is dumb for not using Celsius

What even is that? Create virtual arduinos in the cloud or something?

# ? Sep 20, 2019 17:47

evil_bunnY: Apr 2, 2003

it�s to IoT all th things, which surely isn�t going to blow up in everyone�s faces. no sir

# ? Sep 20, 2019 18:25

My Rhythmic Crotch: Jan 13, 2011

It's an online IDE which strikes me as kinda late to the party, I mean mbed has been around for ages and platformio is good and free.

# ? Sep 20, 2019 18:42

Splode: Jun 18, 2013; put some clothes on you little freak

Web based IDEs are such a bad and pointless idea. Everyone who uses mbed just immediately exports it out of that horrible web ide and uses a real one.

# ? Sep 20, 2019 23:55

mobby_6kl: Aug 9, 2009; by Fluffdaddy

It kind works for some cloud development since the code can live and be immediately deployed there but Arduino?? I don't really see the point, but as long as I can just use Visual Studio they can do whatever.

# ? Sep 21, 2019 01:05

CarForumPoster: Jun 26, 2013; â¡POWERâ¡

Splode posted:

Web based IDEs are such a bad and pointless idea. Everyone who uses mbed just immediately exports it out of that horrible web ide and uses a real one.

No.

Jupyter Notebooks own. Also back when you could prototype easily on it, circuits.io got me into making sure an Arduino would, in theory, do what I wanted.

# ? Sep 21, 2019 01:59

Hadlock: Nov 9, 2004

Now that GitHub has free unlimited private repos I've switched entirely to git/GitHub and vscode. Newer Chromebooks can even run vscode natively via crostini

Web ides are a neat idea but don't offer much plus text rendering is noticably slower

# ? Sep 21, 2019 03:00

ante: Apr 9, 2005; SUNSHINE AND RAINBOWS

Web IDEs own when I'm jumping between three or four different computers. Although I prefer VSCode with an FTP file system to my server

# ? Sep 21, 2019 03:02

Ambrose Burnside: Aug 30, 2007; pensive

e: nvm i just misunderstood that datasheet

Ambrose Burnside fucked around with this message at 02:35 on Sep 24, 2019

# ? Sep 24, 2019 02:07

shame on an IGA: Apr 8, 2005

You're trying to make a watchdog timer

# ? Sep 24, 2019 02:39

Ambrose Burnside: Aug 30, 2007; pensive

shame on an IGA posted:

You're trying to make a watchdog timer

that's the magic phrase that unlocks the decent search results, thanks

# ? Sep 24, 2019 02:48

Fanged Lawn Wormy: Jan 4, 2008; SQUEAK! SQUEAK! SQUEAK!

I have found that is often the case with my projects - I want to do �thing� but don�t know the trade term for what it is. Then, one day, you find it and a whole world unlocks.

# ? Sep 24, 2019 04:22

Adbot: ADBOT LOVES YOU

# ? May 29, 2024 18:54

Ambrose Burnside: Aug 30, 2007; pensive

silver lining?? of having to do the beam-break thing w continuous sending/checking of packets: p sure i can assign a unique code to each emitter-receiver pair, as an extra check against reflections causing false positives

otoh the second half of this is two more pairs acting as a ballistic chronometer, and i'm p sure this is too sluggish an approach to hope to get reasonably-accurate (say +/- 5 fps) velocity measurements from, so i'll play around w the beam parameters (and the breaking thereof) once i've got a functional mk1 counting module on the breadboard

Ambrose Burnside fucked around with this message at 05:20 on Sep 24, 2019

# ? Sep 24, 2019 04:53

The Something Awful Forums > Discussion > Hobbies, Crafts, & Houses > Arduino: Eh, just grab some pliers and bend the pins

«‹›89 »