First electronics project: ESP32 + DHT22 + OLED — what do I actually need to get

[Cathy]

Hi everyone,

I'm pretty new to electronics and I'm getting ready to build my first small project rather than just following individual tutorials.

The idea is fairly simple: an ESP32 (or Arduino) reading temperature/humidity from a DHT22 and displaying the result on a small OLED screen.

I'm putting together a basic parts/tools list, but I'm not sure where I should spend money and where cheap stuff is perfectly fine.

A few things I'm unsure about:

Breadboard and jumper wires

Are the really cheap breadboards actually a problem for beginners? I've seen people mention intermittent connections and poor contacts. Is it worth buying a better-quality breadboard, or is a cheap one good enough for simple projects?

For jumper wires, is it better to just buy a set of pre-made Dupont wires, or is making your own wires actually worthwhile at this stage?

Power supply

For something like this, is USB power from the PC/USB charger good enough during development?

I'm wondering whether I should already have a small adjustable bench power supply with current limiting, mainly because I'm worried about accidentally wiring something incorrectly and damaging the ESP32 or other components.

Is a bench supply overkill for a project this simple?

Multimeter

I currently have a very cheap multimeter and the continuity buzzer is painfully slow. Sometimes I have to wait almost a second before it beeps, which makes checking connections pretty annoying.

For someone just starting out, what features actually matter in a multimeter? Is there any reason to spend around $20–30 rather than just buying another very cheap one?

I'm not trying to build a professional bench yet. I just want to avoid buying a bunch of things that I don't really need, while also not making the learning process unnecessarily frustrating.

What would you consider the minimum setup for a project like this?

Thanks!

[tooki]

I am of the opinion that beginners should buy the best quality they can afford, precisely because they do not yet have enough experience to diagnose and work around limitations or defects in cheap stuff. (That doesn’t mean buying top-of-the-line everything, but it does mean avoiding junk, especially among product types where some stuff is really, really bad.)

First off, I’d recommend using an I2C-based environmental sensor instead of the DHT22 which uses a weird protocol that is timing-sensitive (and thus easily disturbed by other code running on the MCU). For example, try the SHT45.

Modules from Adafruit, etc. cost more than no-name ones from AliExpress, but you know it’s genuine and will work properly.


Breadboards are a product where you absolutely want high quality. I originally bought a number of cheap breadboards (before later investing in some top-of-the-line 3M ones), and compared to high quality breadboards, all of the cheap ones have quite low pin retention force (so it’s much easier for components with thin wires to come out unintentionally). Some have badly shaped contacts that snag on the component leads, making it hard to insert components and wires. And some have contacts so bad that they have many ohms (!) of contact resistance, which can affect some circuits (especially analog), but also means that if you run significant current through them, they can heat up.

But the worst is if you have a poor connection or a component pulls out unnoticed. Then, when the circuit doesn’t work, you don’t know why: is the circuit design wrong? Or is it just a bad connection? You will avoid a lot of frustration by using good breadboards.

In the end, cheap breadboards are just not reliable enough. Some will be OK, others unusable, so if you do must cheap, order more than you need, and choose several types from different vendors, in the hopes that some will be good, but don’t be surprised if some are essentially trash. (At least the trash ones do have one purpose: I use them as holders for soldering, like when soldering pin headers onto a module.) There have been many previous threads about breadboards, and people have recommended various brands that are good, but not as crazy expensive as the 3M breadboards now are. I recommend some mid-priced brand like Global Specialties.

At this stage, buy jumpers. You can get high-quality jumpers from Schmartboard on Digikey, etc. Others here have reported good results with silicone DuPont jumpers from AliExpress. The cheap standard AliExpress ones are usually OK, but the female contacts wear out quickly, and the wire can become brittle with time, so you just have to consider them as semi-disposable.

I would not recommend trying to make your own DuPont jumpers yet, because DuPont contacts are actually unusually fussy contacts that do not crimp properly with nearly all cheap “DuPont” tools, no matter what the tool vendors say. I have a big thread specifically on this topic if you are interested.

There is one exception: if you have Lidl where you live, and you see Lidl’s ~€20 DuPont & XH crimp set that they sell once a year, buy it: it is the ONLY cheap DuPont crimper I have tested that actually produces decent crimps. (They’re not as perfect as the crimps from the expensive tools, but the results from the Lidl tool are actually good, i.e. they do not pull out and do not mangle the contact.)


What you can make yourself are the non-DuPont wire jumpers for the breadboard itself: just get 0.5-0.6mm diameter (0.2-0.34mm², 24-22 AWG) solid wire and strip 6-8mm from each end.


Power supply: well, I certainly do consider a lab power supply to be essential to electronics, and current limiting is great for protecting against beginner mistakes (and non-beginner mistakes! ). Luckily, you can get great bench supplies affordably. I highly recommend the ones from Korad, particularly the ones that have preset memory and an output on/off button, like the KA3005 series. (The KD3005 doesn’t have those.) I own the KA3005P and it is my primary power supply at home.

With that said, for pure MCU projects like this, you’ll be connecting USB anyway to program it, and in many cases you can just power it from that. An inexpensive, simple precaution is to use a powered USB hub (i.e. one that comes with its own power supply) in between, so that if anything goes horribly wrong, at worst you’ve fried the hub, and not your computer’s USB port.


There are many threads about cheap multimeters. I think $20-50 is a great price range in terms of bang-for-your-buck. For example, the Aneng/Zotek meters. FYI, you might be able to speed up your current meter’s continuity test by very thoroughly cleaning your probe tips with alcohol. (Or getting better probes. And still, clean them.)

[PGPG]

I'm playing with electronics since more than 55 years.
I'm shocked what people do now as their 'first electronic project'.
My first project (I have build when I was 10) consisted of LC+diode+C+headphones and (with antenna and GND) I could receive long-way radio station.

I have never had breadboard with jumper wires and I don't see any need for it. I have one prototype PCB and when I need to check some circuit I solder it at these PCB and after checking what I need I disassemble it to have PCB be ready for the next case.
In my opinion the multimeter you have is good enough for you for next few years. It's a waste of money to get a different one.
From my point of view must have is the supply with regulated voltage and current. Such supply was my second project. As my solution was linear and I wanted to limit power dissipated in my regulation element I needed to be able to set voltage I get from trafo. I took big trafo from some tube radio receiver. I unwound the anode winding and wound in that place winding giving me 6+6+6+6+6+1+1+1+1+1 voltages. At front of my supply I have banana sockets and two wires with banana plugs allowing to select 1V, 2V, 3V,...,33V, 34V, 35V to be used by following supply stages.
At work I have professional laboratory power supply, but at home when I need to supply something I'm still using this 50+ years old supply. Output voltage is regulated from 0V to 40V (in four 10 V ranges) and current limit from 0 to 2A (in two ranges).
The next must have is the oscilloscope. It took me 14 years (from 10 to 24) to reach enough skill to build oscilloscope, but nowadays the only rationale solution is to buy one.

[tooki]

I'm playing with electronics since more than 55 years.
I'm shocked what people do now as their 'first electronic project'.
My first project (I have build when I was 10) consisted of LC+diode+C+headphones and (with antenna and GND) I could receive long-way radio station.

OK, so? What’s wrong with people choosing different first projects? There isn’t one “right way” to start an electronics hobby. In my opinion, any approach that gets someone into it, and keeps their interest, is a success.

If systems integration is the avenue that attracts many people to electronics today, that’s fine! They’ll still learn, and will learn other things that you didn’t learn 55 years ago.

Radio just isn’t a very popular electronics hobby anymore. It used to form the core of electronics as a hobby, and today it doesn’t.

Chances are, if 10-year-old PGPG was starting with electronics today, you’d start with something else, too! (And the same is absolutely true for me, too.)

I have never had breadboard with jumper wires and I don't see any need for it.

Both approaches are valid. With that said, you might feel differently on this issue if you’d actually had a breadboard in the past. Or not. Some people love them, other people hate them, and most of us just see them as a tool that is useful in some situations and not others. What I don’t understand is the attitude of “I never needed it, thus no one else should need it, either”.

I have one prototype PCB and when I need to check some circuit I solder it at these PCB and after checking what I need I disassemble it to have PCB be ready for the next case.

To me that is wild, insofar as that’s pretty much exactly what a breadboard is for. It just gets the job done faster in many cases.

In my opinion the multimeter you have is good enough for you for next few years. It's a waste of money to get a different one.

For measurements, sure. But if they’re annoyed by slow continuity, that’s a perfectly valid reason to get a different one.

[ledtester]

About multimeters: Having one with a good continuity buzzer will save you a lot of time. Get one, though, with a traditional dial that selects the measurement mode - not anything that has "smart" in its description.

If you give us a list of meters that are available to you and within your budget we can give you feedback on which one is better.

Power Supply: For beginning microcontroller projects you won't need a bench supply - using USB power will work just fine.

Breadboards: I have not had any problems with breadboards but they are not well suited for large microcontroller modules. Consider getting a breakout board (if one exists) for your ESP32/Arduino like these:

https://www.aliexpress.us/item/3256808504815670.html

https://www.aliexpress.us/item/3256805839783502.html

[tooki]

Yes, that’s a very good point that I actually wanted to mention above, and forgot: if at all possible, DO NOT insert standard (0.1”/2.54mm pitch, 0.64mm square pin) PCB headers into a breadboard. Doing so WILL deform the contacts, and in cheap breadboards, this often leaves them fatally loose for anything else. The breakout boards ledtester suggested are a fantastic solution, as are multi-wire systems like SeedStudio’s Grove, Adafruit’s STEMMA and STEMMA QT, and Sparkfun’s Qwicc. (They’re mostly compatible with each other, see https://learn.adafruit.com/introducing-adafruit-stemma-qt/stemma-qt-comparison .)

Male DuPont crimp pins are fine, as their folded sheet metal construction makes them rounder than the pins used in standard PCB headers.

Machined-pin headers (which use round pins) also work well; I use these when I want breadboard-compatible PCB headers. (However, just know that female DuPont jumpers will not hold onto these at all. It’s kinda one-or-the-other.)

[Peabody]

Remember that you are asking these questions on a forum populated mostly by engineers.  You may get different answers from engineers than from hobbyists like yourself.  I'm a hobbyist, and almost always go for the cheapest option, usually with no regret.

The main exception is breadboards.  If I told you that you could stick a jumper into spring-loaded metal, but not make a connection, you would tell me that's impossible.  But your friends on the mainland have achieved that, and at very low cost.  So in that case, to save you hours of debugging and redesigning when the only problem is a bad connection, get a good breadboard.  Highly recommended by everyone are boards made by BusBoard Prototype Systems.  And they are not expensive.  The BB830 is US$8.95 at Digikey.

I think the same would apply to dupont jumpers, but I don't know of a source for good ones.  But you will have to initially test each one in the batch for resistance - no more than a few ohms.  I have taken to using 22-gauge solid wire for jumpers, although the sharp ends are probably not good for the breadboard.

Your meter is probably ok for now.  But be sure to clean the probe ends.  I just received a new $30 meter, and there was a major problem with the probe ends.  I ended up needing to use mild abrasive to remove whatever coating was left by manufacturing.

For this project, I think any 5V USB charger wall wart would be ok.  It will have its own current limit.

I think your choice of processor for this might be best determined by the needs of the sensor and the display.  If they need 5V power, and can communicate at 5V over I2C, then something like a Nano might be a better choice than an ESP32, which communicates at 3.3V.  You will be able to avoid the need for voltage translation if everything operates at the same voltage.

If you find that you enjoy electronics and want to do more, then at some point you will want to consider getting an oscilloscope.  The engineers will probably recommend some four-channel bench scope for only a few hundred dollars.  But I think most hobbyists can do well with one of the hand-held, battery-powered scopes made by Zeeweii and others for a fraction of that cost.  But that decision is down the line a bit at this point.

Good luck with your project.  By the way, there is also an Arduino forum you might want to check out.

[Terry Bites]

HaHa,
I helped out at a local school electronics club. The kids thought they could build a mobile phone.
"Well, there's not much inside one is there?"
Many of them couldn't see why you had to learn about ohms law,passives and transistors as a first step.
I have to say that even undergrads now are told by their dumb-as tutors to chuck a micro-controller in as a first step.

Bolting modules together like Leggo is not electronics.
And while I'm on the moan, assembling a pc is not computer engineering.

[PGPG]

OK, so? What’s wrong with people choosing different first projects?

Nothing wrong. I've just said that I'm shocked how complicated (in my opinion) things people do as their first project.

Chances are, if 10-year-old PGPG was starting with electronics today, you’d start with something else, too!

Of course, but I don't believe I'd start from something uC driven. Rather 2 transistor (or one digital gate) generator.

What I don’t understand is the attitude of “I never needed it, thus no one else should need it, either”.

I feel like you read what was not written.
I've just said that I don't need it to let OP know that not all people are using it.

To me that is wild, insofar as that’s pretty much exactly what a breadboard is for. It just gets the job done faster in many cases.

I’ve simply always been wary of contact-based connections. I imagine that such a circuit would behave differently after every shock. But I never had the chance to test it. When I was building lots of prototypes, such things weren't available to me, and by the time they became available, I only needed to assemble something once every few years.

For measurements, sure. But if they’re annoyed by slow continuity, that’s a perfectly valid reason to get a different one.

How often someone uses continuity tests?
In your design you see what is connected with what so what to test.
I use continuity test only when I get new ceiling light to find which wire is connected with 3 bulbs and which with 2.
Do using prototype boards side effect is that you need to use continuity test?

[tooki]

OK, so? What’s wrong with people choosing different first projects?

Nothing wrong. I've just said that I'm shocked how complicated (in my opinion) things people do as their first project.

You didn’t say that before. A ton of people complain that such MCU projects are too simple and shield people from learning “real” electronics. So that’s kinda what I assumed you meant.

Chances are, if 10-year-old PGPG was starting with electronics today, you’d start with something else, too!

Of course, but I don't believe I'd start from something uC driven. Rather 2 transistor (or one digital gate) generator.

You say that now, but it’s hard to know how you’d really behave in that situation; it is of course unknowable.

However, note also that OP said that was their first project as opposed to just following a tutorial. But surely the simple radio you made at age 10 (or the 2-transistor circuit you hypothesize you might make today) was not something you designed yourself, right? You built it from a schematic you got from a book or magazine, no?

To me that is wild, insofar as that’s pretty much exactly what a breadboard is for. It just gets the job done faster in many cases.

I’ve simply always been wary of contact-based connections. I imagine that such a circuit would behave differently after every shock. But I never had the chance to test it. When I was building lots of prototypes, such things weren't available to me, and by the time they became available, I only needed to assemble something once every few years.

Yet solderless connections are everywhere and are overwhelmingly reliable. (And some types of solderless connections, like crimping or press-fitting, are decidedly more reliable than soldered connections.)

Would your suspicions be true in the case of the worst of the cheap breadboards? Absolutely. But not when using a good quality one. Breadboards have their limitations (they’re not good for very high-speed circuits, for example), of course, but a good quality breadboard has no problems making reliable connections.

FYI: when solderless breadboards were introduced, they were not made for hobbyists (in fact, they were too expensive for them!). They were targeted at professional engineers, for quickly developing and prototyping circuits. Many, many devices built around discrete logic ICs were prototyped on breadboards — sometimes dozens and dozens of breadboards connected. Soldering and wire-wrapping were the other methods widely used in development.

For measurements, sure. But if they’re annoyed by slow continuity, that’s a perfectly valid reason to get a different one.

How often someone uses continuity tests?
In your design you see what is connected with what so what to test.
I use continuity test only when I get new ceiling light to find which wire is connected with 3 bulbs and which with 2.
Do using prototype boards side effect is that you need to use continuity test?

I’m sorry, but where on earth did you get that crazy idea? Do you really not know what a continuity tester is used for?!?

I’m a working electronics technician, and continuity testing is possibly the most often used function in a multimeter. Some examples:
- testing for the absence of short circuits on a circuit board or cable you just soldered
- checking for shorted diodes on a failed PCB
- testing whether a ceramic or sand-filled fuse has blown
- identifying which pin in a connector comes out to where on the connector (or bare wires) on the other end
- testing a cable assembly you’ve just built
- checking whether a switch or relay has failed
- tracing where a signal goes when reverse engineering or troubleshooting a PCB (often to verify that you’re on the right pin, based on a schematic)
- locating faults in a machine or appliance due to broken wires, traces, or similar

[iMo]

OK, so? What’s wrong with people choosing different first projects?

Nothing wrong. I've just said that I'm shocked how complicated (in my opinion) things people do as their first project.

Chances are, if 10-year-old PGPG was starting with electronics today, you’d start with something else, too!

Of course, but I don't believe I'd start from something uC driven. Rather 2 transistor (or one digital gate) generator..

They start with the MCU stuff because it is extremely easy today. I know in person XX young makers who mess with all possible mcu modules you may imagine, with almost zero understanding of electronics basics and related math. They are usually good with programming.
Working with transistors and related circuitry may require some need for measurements and math, and that is difficult..

[Renate]

I'll get in on the continuity rant.

I bought a cheapy DVM (25 €) and the stupid thing had no beeper. That really pissed me off.
Tracing circuits, tracing PCBs, ringing cables, you need something quick reacting and non-visual.
I picked up another DVM that annoys me too, but at least it beeps.

I made a scanner to check USB cables. Yeah, they sell them too.

[PGPG]

Because of:
https://www.eevblog.com/forum/beginners/502-bad-gateway-errors/
since yesterday I was not able to answer losing what I wrote...

Nothing wrong. I've just said that I'm shocked how complicated (in my opinion) things people do as their first project.

You didn’t say that before. A ton of people complain that such MCU projects are too simple and shield people from learning “real” electronics. So that’s kinda what I assumed you meant.

I supposed it is clear from telling that I'm shocked and then to cite the banal circuit I started with.

But surely the simple radio you made at age 10 (or the 2-transistor circuit you hypothesize you might make today) was not something you designed yourself, right? You built it from a schematic you got from a book or magazine, no?

Right, but....
Communistic country do not have many advantages, but shortage of all kinds of goods teaches people extraordinary resourcefulness. In the case of electronics, it quickly forces you to adapt every schematic to the components you actually have on hand, since trying to acquire the exact ones specified in the diagram (even a resistors with specified value) could take years. Substituting different component values was the first step into circuit design—a path I was very quickly compelled to take. When I was 15, I designed a darkroom timer for a friend. The entire circuit used germanium transistors, as those were the only ones I had. The timing consistency wasn't satisfactory. My friend persuaded his mother to buy a silicon transistor - a relatively significant expense at the time simply because I suspected it would improve the consistency. And indeed, it did (after I redesigned the device to replace one germanium PNP transistor (all germanium I had were PNP) with a silicon NPN one).

I’m sorry, but where on earth did you get that crazy idea? Do you really not know what a continuity tester is used for?!?

I’m a working electronics technician, and continuity testing is possibly the most often used function in a multimeter. Some examples:
- testing for the absence of short circuits on a circuit board or cable you just soldered
- checking for shorted diodes on a failed PCB
- testing whether a ceramic or sand-filled fuse has blown
- identifying which pin in a connector comes out to where on the connector (or bare wires) on the other end
- testing a cable assembly you’ve just built
- checking whether a switch or relay has failed
- tracing where a signal goes when reverse engineering or troubleshooting a PCB (often to verify that you’re on the right pin, based on a schematic)
- locating faults in a machine or appliance due to broken wires, traces, or similar

It looks that I really do live in a different reality
I am a main designer in our very small company. From all your examples the one brings to mind... To connect at desk a set of our devices having RS485 we use cable made by 4-conductor (12V supply + A,B) ribbon cable with 6-pin 2.54 mm pitch headers crimped on (having 2 pins not used clearly shows how to insert pin header from device). My brother made new such cable with more sockets but it doesn't work so I am considering to use continuity testing to find where is the problem. I am considering and considering and about 2 years passed. May be it is why I think the function is very rarely used
About your other examples:
- when (5 years ago) we had a problem with rising whiskerses they were not giving enough good short to be detected by multimeter continuity test function. We had to use microscope to find them,
- I am designing everything with big margin. I don't remember ever having shorted diode (except my school times, but then I didn't had any digital multimeter),
- each our device have PTC fuse, none of our devices have any other fuse,
- when in design I use new connector in most cases I don't have one to test anything- I just believe in datasheet,
- we use relays in our devices, but I have build testers so you connect the whole device and brrrrum (sound made by fast switching relays) and you get info if all relay contacts work,
- last time I was reverse engineering a PCB was when I was a student and was asked to repair PCB (about 10x30cm full of elements) from car wheel balancer. It took me 3 afternoons to get schematic and next three to understand how it works (when wheel rotates and then how it shows on analog (moving-coil) indicators at what position and what weight to put on wheel - it was using AC mains signal to recreate from two samples at capacitors the amplitude and phase of signals from rotation time when not rotate),

I really think that continuity test function is very rarely needed by hobbyist when what he is doing is designing and making devices. When reverse engineering PCBs it can be used often, but designing is much more satisfying then reverse engineering, I think.

[lichurbagan]

For someone just starting out, what features actually matter in a multimeter? Is there any reason to spend around $20–30 rather than just buying another very cheap one?


Thanks!

It's always better to buy a good one. Cheap one will work but after a few days, the probes will get disconnected.

[PGPG]

It's always better to buy a good one. Cheap one will work but after a few days, the probes will get disconnected.

Am I different then others also in an aspect of using probes? Who uses these probes... I almost never use them. Years ago I have made cables with banana plug at one end and small gripper at the other. I have a dozen or so of them. I am always using them. When I want to touch something it is possible with much less risk of shorting with anything else and in most cases I just want to grip something and I always can do it. You need less hands if you can grip your test wires.

[rstofer]

It seems to me that everything that can be done with an Aruino has been done and the completed project is on the Internet.  I prefer to buy prebuilt modules and then string them together with SPI or I2C.  Or have a PCB built to spec.

Google AI for some variant of 'arduino esp32 environmental monitoring example' and jump to any of the projects or just follow the AI instructiolns.

There are a few chip variants of Arduino including ESP32.  I would probably want to use a variant that iincludes WiFi hardware and a web server library.

https://www.seeedstudio.com/Arduino-Uno-Rev4-WiFi-p-5717.html
https://docs.arduino.cc/retired/library-examples/wifi-library/WiFiWebServer/

ETA:  I don't know what I'm going to build with it but I bought the Arduino board above from Amazon for US$ 27.50.  Same day delivery...

[tooki]

OK, so? What’s wrong with people choosing different first projects?

Nothing wrong. I've just said that I'm shocked how complicated (in my opinion) things people do as their first project.

Chances are, if 10-year-old PGPG was starting with electronics today, you’d start with something else, too!

Of course, but I don't believe I'd start from something uC driven. Rather 2 transistor (or one digital gate) generator..

They start with the MCU stuff because it is extremely easy today. I know in person XX young makers who mess with all possible mcu modules you may imagine, with almost zero understanding of electronics basics and related math. They are usually good with programming.
Working with transistors and related circuitry may require some need for measurements and math, and that is difficult..

What’s wrong with starting with “extremely easy”? Seems like the ideal place to start.

People who refer to themselves as “makers” are usually not the ones interested in going deep in electronics. But what’s wrong with that? And some will get into “real” electronics eventually anyway.

[tooki]

Because of:
https://www.eevblog.com/forum/beginners/502-bad-gateway-errors/
since yesterday I was not able to answer losing what I wrote...

Ugh, yep, I hear you!!! It happened to me, too, in another thread.

Nothing wrong. I've just said that I'm shocked how complicated (in my opinion) things people do as their first project.

You didn’t say that before. A ton of people complain that such MCU projects are too simple and shield people from learning “real” electronics. So that’s kinda what I assumed you meant.

I supposed it is clear from telling that I'm shocked and then to cite the banal circuit I started with.

It was clear that you were shocked, but what wasn’t clear is why: to you, the MCU project is too complex for a first project, but to others, it’s too simple, as iMo’s response is a perfect example of.

But surely the simple radio you made at age 10 (or the 2-transistor circuit you hypothesize you might make today) was not something you designed yourself, right? You built it from a schematic you got from a book or magazine, no?

Right, but....
Communistic country do not have many advantages, but shortage of all kinds of goods teaches people extraordinary resourcefulness. In the case of electronics, it quickly forces you to adapt every schematic to the components you actually have on hand, since trying to acquire the exact ones specified in the diagram (even a resistors with specified value) could take years. Substituting different component values was the first step into circuit design—a path I was very quickly compelled to take. When I was 15, I designed a darkroom timer for a friend. The entire circuit used germanium transistors, as those were the only ones I had. The timing consistency wasn't satisfactory. My friend persuaded his mother to buy a silicon transistor - a relatively significant expense at the time simply because I suspected it would improve the consistency. And indeed, it did (after I redesigned the device to replace one germanium PNP transistor (all germanium I had were PNP) with a silicon NPN one).

My point was that a 10-year-old would likely not be able to design that radio circuit from scratch. Adapting is not the same as conceiving.

Similarly, people today begin by following instructions, too. And it’s not uncommon to substitute parts, though the pressure to do so is certainly far less in many places today. (Though we regularly see posts on here from people who say they cannot easily get parts in their country.)

I’m sorry, but where on earth did you get that crazy idea? Do you really not know what a continuity tester is used for?!?

I’m a working electronics technician, and continuity testing is possibly the most often used function in a multimeter. Some examples:
- testing for the absence of short circuits on a circuit board or cable you just soldered
- checking for shorted diodes on a failed PCB
- testing whether a ceramic or sand-filled fuse has blown
- identifying which pin in a connector comes out to where on the connector (or bare wires) on the other end
- testing a cable assembly you’ve just built
- checking whether a switch or relay has failed
- tracing where a signal goes when reverse engineering or troubleshooting a PCB (often to verify that you’re on the right pin, based on a schematic)
- locating faults in a machine or appliance due to broken wires, traces, or similar

It looks that I really do live in a different reality
I am a main designer in our very small company. From all your examples the one brings to mind... To connect at desk a set of our devices having RS485 we use cable made by 4-conductor (12V supply + A,B) ribbon cable with 6-pin 2.54 mm pitch headers crimped on (having 2 pins not used clearly shows how to insert pin header from device). My brother made new such cable with more sockets but it doesn't work so I am considering to use continuity testing to find where is the problem. I am considering and considering and about 2 years passed. May be it is why I think the function is very rarely used
About your other examples:
- when (5 years ago) we had a problem with rising whiskerses they were not giving enough good short to be detected by multimeter continuity test function. We had to use microscope to find them,
- I am designing everything with big margin. I don't remember ever having shorted diode (except my school times, but then I didn't had any digital multimeter),
- each our device have PTC fuse, none of our devices have any other fuse,
- when in design I use new connector in most cases I don't have one to test anything- I just believe in datasheet,
- we use relays in our devices, but I have build testers so you connect the whole device and brrrrum (sound made by fast switching relays) and you get info if all relay contacts work,
- last time I was reverse engineering a PCB was when I was a student and was asked to repair PCB (about 10x30cm full of elements) from car wheel balancer. It took me 3 afternoons to get schematic and next three to understand how it works (when wheel rotates and then how it shows on analog (moving-coil) indicators at what position and what weight to put on wheel - it was using AC mains signal to recreate from two samples at capacitors the amplitude and phase of signals from rotation time when not rotate),

I really think that continuity test function is very rarely needed by hobbyist when what he is doing is designing and making devices. When reverse engineering PCBs it can be used often, but designing is much more satisfying then reverse engineering, I think.

I would agree that continuity is not a super common function when designing a product. But it is as fundamental as it gets when it comes to troubleshooting and repair.

I do lots of repairs (of commercial devices), and thus have no influence on the product design.
You might design in lots of margin in a diode or relay, but someone else might not have. By the time the broken device hits my bench, the failure has already occurred, and I need to find what has failed.
PTCs are great for some applications, but cannot be used in others. I don’t think I’ve ever seen a device with a mains AC input PTC instead of a fuse.

As for your brother’s cable: I can’t even remember the last time I built a cable that didn’t work perfectly. However, I test them after I make them, just to be sure, because I don’t want to risk damaging some expensive piece of equipment because a cable is wired wrong. So a simple continuity test (testing for shorts, and testing for correct pinout) is a worthwhile step.

As for reverse engineering: sure, designing something new is usually more interesting. But if I need to repair a device I don’t have a schematic for, or I have been tasked with copying a circuit because we need more of them, then I have to reverse engineer it.

[PGPG]

Ugh, yep, I hear you!!! It happened to me, too, in another thread.

This time I edited the answer (didn't made a copy) and when I pressed 'Post' EEVblog started to verify if I'm a human and then nothing happened. When I tried to get back I found that session expired and was informed that system left message body empty

So now I am adding sentence by sentence to already posted message (to not lost everything).
Our devices are 12V or 24V (10..28) powered so that doesn't contradict what you wrote about PTC.

We never had a problem with connectors crimped onto ribbon cables. But the old ones were gray, whereas the current ones are black—maybe they’re of inferior quality. The cables had been made but not tested (they had always been fine). We discovered the problem several months later when we needed them for something. We used older ones (with fewer connectors) instead, and for now, we don't have time for trivial cables.
We don't worry about the wiring because (given the 12V power supply we use for bench testing) our devices are immune to any wiring errors.

On two occasions, the installer informed us that lightning had struck directly the building where our access control system was installed. They had to repair all systems with long cable runs—telephone exchanges (this was a long time ago), alarm systems, and network cards in half the computers. The message ended with a complaint that, since our system only required a reset, everyone else would now make money while he got nothing.

[tooki]

Ugh, yep, I hear you!!! It happened to me, too, in another thread.

This time I edited the answer (didn't made a copy) and when I pressed 'Post' EEVblog started to verify if I'm a human and then nothing happened. When I tried to get back I found that session expired and was informed that system left message body empty

So now I am adding sentence by sentence to already posted message (to not lost everything).
Our devices are 12V or 24V (10..28) powered so that doesn't contradict what you wrote about PTC.

We never had a problem with connectors crimped onto ribbon cables. But the old ones were gray, whereas the current ones are black—maybe they’re of inferior quality. The cables had been made but not tested (they had always been fine). We discovered the problem several months later when we needed them for something. We used older ones (with fewer connectors) instead, and for now, we don't have time for trivial cables.
We don't worry about the wiring because (given the 12V power supply we use for bench testing) our devices are immune to any wiring errors.

Yeah, ribbon cables in general don’t need testing beyond visual inspection (especially if you mean IDC connectors).

The color of the connector doesn’t matter, but (assuming you are talking about IDC connectors) the connectors and cable need to match: wire gauge, pitch, and wire type (stranded or solid). If you have a mismatch (for example, a connector made for 26AWG stranded, but you use 28AWG solid) it can be unreliable.

[PGPG]

Yeah, ribbon cables in general don’t need testing beyond visual inspection (especially if you mean IDC connectors).

I didn't know how to name them. Google translated written by me flat cable into ribbon cable.
Looking for IDC in Mouser shows the kind of connectors I had in mind.
Cable (about 1m) comes just from the drawer. Can be 30 years old. Connectors (10 per that 1m) were send us from our company other location (200km away from us). If they were bought for it or found also in their drawer - I don't know. Specification of cable, connectors - probably no one knows

[tooki]

Oof!

For next time, it’s probably better to just spend the €20 to order good cable and connectors!

[negativ3]

A diy continuity tester would be an awesome first project.
I have one which has a high-bright LED as well as a buzzer and with decent probes, it's instant.

I'm sure the probes are the expensive part of the project.

[paulca]

Since you mentioned Arduino OR ESP32.  Firstly.  Go for the former unless you NEED the later.  The ESP32 is the less beginner friendly of the two.

The ESP32 is really a Wifi/BT wireless module.  Thats it's entire purpose in life.... until makers discovered it has (usually) two excessively powerful cores for that soul task and started running arduino style sketches on the Wifi module itself instead of a separate MCU.  Espressif of course doubled down on that and released a whole range of different powered and "kitted out" modules and ICs.

If you don't need Wifi or Bluetooth though, the ESP32 is 90% pointless.  As an MCU it isn't all that great either. It has mediocre peripherals.  Although so does the Arduino.

For your stated requirements and no extensions the ardiuno wins hands down.

I would even embrace that eco system more heavily still.  Consider using an "Arduino shield blank".  You can buy them cheaply on Amazon/Aliexpress.  It's basically a "solder type" breadboard which mounts directly on to the aduino pinout.  You add your DHT22 and OLED on it, solder to the right pins.

The arduinno already harbours some protection from newbie mistakes.  Like a resetable fuse for one, assuming you get a good clone.  It has it's own power supply for 5V which can be derivived from teh USB port directly and 3v3 if you need it.  It will put out dozens of milli-amps.  It will still allow you to kill the chip with a shorted pin though.

[tooki]

Since you mentioned Arduino OR ESP32.  Firstly.  Go for the former unless you NEED the later.  The ESP32 is the less beginner friendly of the two.

In what way is it less beginner friendly? The only difference is that you need to add the ESP32 board definitions URL -- a one-time thing. After that, nothing about it is any more or less complicated than an Arduino.

The ESP32 is really a Wifi/BT wireless module.  Thats it's entire purpose in life.... until makers discovered it has (usually) two excessively powerful cores for that soul task and started running arduino style sketches on the Wifi module itself instead of a separate MCU.

That claim maaaaaybe could be made for the old ESP8266. Certainly by the time the ESP32 came out, it was not a wireless module with incidental MCU.

Espressif of course doubled down on that and released a whole range of different powered and "kitted out" modules and ICs.

And this is... bad?!?

If you don't need Wifi or Bluetooth though, the ESP32 is 90% pointless.  As an MCU it isn't all that great either. It has mediocre peripherals.  Although so does the Arduino.

I use ESP32 series boards for almost all of my MCU projects. And the vast majority of them don't use Wifi or BT. Why? Because the ESP32 is cheaper than an Arduino, but has FAR more processing power, much more memory, and better peripherals than an Arduino. (And nearly all peripherals can be routed to any pins, which is much more flexible than Arduino.)

As someone who isn't a gifted programmer, the excess RAM and CPU resources of an ESP32 let me write code that forgoes optimization in favor of explicitness. This lets me focus on the functional logic without having to worry about saving resources.

And while other MCU families (like STM32) have FAR richer peripherals, they also carry with them with FAR greater complexity to get them running.

For your stated requirements and no extensions the ardiuno wins hands down.

Nonsense. They're completely interchangeable for the described application.

I would even embrace that eco system more heavily still.  Consider using an "Arduino shield blank".  You can buy them cheaply on Amazon/Aliexpress.  It's basically a "solder type" breadboard which mounts directly on to the aduino pinout.  You add your DHT22 and OLED on it, solder to the right pins.

Arduino shields are indeed great. That's why I often use ESP32 boards in the Arduino Uno layout. Best of both worlds.

I think there are reasons why ESP32 is, by far, the most common non-Atmega MCU family used with the Arduino framework, even if you don't see them. This means there is also HUGE community support for these MCUs.

The arduinno already harbours some protection from newbie mistakes.  Like a resetable fuse for one, assuming you get a good clone.  It has it's own power supply for 5V which can be derivived from teh USB port directly and 3v3 if you need it.  It will put out dozens of milli-amps.  It will still allow you to kill the chip with a shorted pin though.

All of that is true for ESP32, too, with the sole exception of the polyfuse.


Like, seriously, dude... your post seems like a bunch of whining over nothing.


IMHO, ESP32 has only a few real downsides:
1. It is not designed for very-low-power applications: its "deep sleep" mode is a joke.
2. It doesn't have Schmitt triggers on its inputs.
3. On most ESP32 models, the ADC is not awesome.
4. Compiling takes longer.