Notifications
Clear all
Electronic Components & Theory
9
Posts
3
Users
0
Reactions
354
Views
Topic starter
2026-09-09 3:50 am
First, is there a legend on this site (or another) explaining all terms and acronyms related to Arduino, ESP32, Electronics, IoT, Raspberry Pi, and Robotics? Or is Wikipedia the better choice to lookup "VCC"?
I see pins on most components labelled with "VCC" but found in their specifications different volatages related to it, in one case 5V, another "VCC" indicates a range 1,65..3,6V with 'Absolute Maximum Ratings' of -,5..4,6V (note the negative value for "Absolute Maximum"). Thus 5V would be inappropriate.
Question: when I piece a circuitry together, are there systems apt to check my sketch (computerised sketch of course) for such faults? Which is the best way to prevent wrong voltage mishaps?
2026-09-09 7:32 pm
Hi @amphitryonoff,
Vcc is a designation commonly used for the positive supply line for digital logic circuits.
The value of that positive voltage depends upon which family of chips it is powering. In the late 1960s, this was often 5 Volts, to match the TTL Transistor Transistor Logic family introduced by Texas Instruments. 5V remained the norm for perhaps a decade, and is still used today by some of the devices that have remained in production for a long time. The 1980s saw the introduction of 3.3 Volt logic, and that remains a popular choice for new devices to this day. However, be aware that certain types of devices, such as the highest speed processors and memories may now require a lower voltage, such as 1.8 volts. So basically, read the specific data sheet for each new part, every time.
Absolute Maximum Ratings should always be treated as the title implies. For example, 3.3V devices are designed to be normally powered by 3.3V, and modern fixed voltage regulators usually produce a voltage very close to the value implied by their part number, providing the surrounding environment is appropriate for normal operation. However, even in well designed circuits, there will be some voltage fluctuations, and the Absolute Maximum Ratings are a guide as to how much variation the device will accept without damage. Note that in some cases, exceeding that rating by even the briefest period of time may damage the device. Thus, if it is 3.3V device, it is wise to aim as close to 3.3V as possible.
The 'Absolute Maximum' implies that having a voltage below the 'normal' voltage of say 3.3V is safe, which it usually is, although this is not always true. However, the operation of the device will probably be degraded, if the voltage is reduced towards zero. Hence, you may see statements to suggest a voltage will work normally with supply voltages covering a small range, such as 3.0V to 3.6V. Commonly, the maximum operating speed reduces with decreasing voltage.
The Input/Output Pins (in particular) of logic devices may have Absolute Maximum Ratings like
- -0.5V .. 3.8V OR -0.5V .. Vcc + 0.5V)
The -0.5V limit is often due to diodes built into the pin circuit to discharge electrostatic voltages. These diodes are designed to discharge electrostatic charge, if the voltage goes below -0.5V. These diodes have a very low current rating, and will be damaged if driven below -0.5V by a higher current pulse.
Similarly, other similar diodes will conduct if the pin voltage exceeds the momentary Vcc voltage, which for a 3.3V device, will be expected to be 3.3V. Note that "Vcc + 0.5V" is emphasising that the maximum will only 3.8V, when the incoming voltage is 3.3V. If the supply voltage fell to 3.0V(say), then the device might still function normally, but exceeding 3.5V on an I/O pin may damage it.
---------
I have never used a system that thoroughly checked for such design faults. In the rapidly expanding world of AI, I make no predictions of what a particular design system might offer, now or in the future, but I would still regard checking as the designer's personal responsibility. At the very least, the designer should ensure any automated check they utilise, actually functions correctly, to the point of deliberately injecting a fault for the checker to find.
Best wishes, Dave
Topic starter
2026-09-09 11:57 pm
Hi @davee !
TY for the explanation in length. This is a perfect example how an entry in a legend of terms for beginners could look like. Since I'm new to this kind of h/w-f*ck it is very instructive.
I have never used a system that thoroughly checked for such design faults.
Decades ago we had at work a system that indicated collisions between 3D elements, e. g. piping and steelwork. That is why I'm wondering that there are no similar systems for checking clashes of voltage mismatch in a wiring sketch. In principle it's the same task as in a plant schematic to prohibit joining a pipe with acid and another with alkaline solution.
2026-09-10 12:57 pm
Question: when I piece a circuitry together, are there systems apt to check my sketch (computerised sketch of course) for such faults? Which is the best way to prevent wrong voltage mishaps?
Yes, there are such beasts. My brother-in-law works for such a company that wrote one and sells it to well-healed customers and also contracts its use by in-house engineers to do the work for other companies. I pinged him and he replied:
He works for: "Cadence Design Systems". Previously, "Synopsys". Analysis is a many-colored horse, so to speak, basically lots of different levels of abstraction. Spice is the electrical equivalent of FEA and simulates at the transistor level. On the far other end are things like Verilog behavioral simulation - abstracts a nand-gate into a functional model and then simulates billions of nand-gate equivalents in the same circuit. Verilog is actually a language, kind of equivalent to C or C++, and the simulators are things like VCS (I think that's a Synopsys product), verisim and so on. There are several different levels of abstraction in between those, and different focuses - behavior vs timing for instance. Everything from Synopsys, Cadence and Mentor (Siemens) are commercial products and very expensive. There are open source type options for some of these. There's an org called "open road asic" (iirc) that tries to make a full suite of freeware chip tools available. I have no idea how good or bad those are, but of course they can't be getting the level of software development thrown at them that the big 3 eda companies put in.
However, I wouldn't be surprised at all if you simply submitted a sketch to an AI (I use both the free versions of Gemini and ChatGPT) that it could answer very specific questions or even say... "No this will not work and here's why..."
I'm doing some work on the Physics of a self-balancing robot and I've sent MP4 videos to illustrate a behavior that was hard to describe. It was able evaluate a video. In another case, it used some extensive trig equations to approximate an angle of something. The value my simulation was settling on was not that angle by about 10%. I hand sketched it on paper and showed the geometry and Algebra only solution I used that was coming up the same to 5 significant digits. It admitted my solution was both exact to theory (vs an estimate) and "more elegant" than its solution. They love to blow sun-shine. But the point being in this case was that it deciphered my sketch and hand written solution.
But also, another point... verify, because its not always right, but it might catch something you didn't.
Inq
Topic starter
2026-09-10 11:26 pm
Hi @inq !
I pinged him and he replied:
Tell him my thank you. Now I see it's not just a cakewalk to analyse a circuit diagram completely. It's much more difficult than a P&ID where pressure, temperature and medium define the pipe class, and the flow rate and max flow velocity define the pipe diameter. In circuitry you may even have to consider interspersions, I learned this week.
I'm doing some work on the Physics of a self-balancing robot
In a first approach I would regard it as a physical pendulum with the center of mass above its pivot point -- actually alternating pivots, and almost constatly changing inertia. A nice challenge. But from aeronautics we know, sufficiently powered everything will fly 😉
Your hint, AI may help is hip nowadays. Interestingly the "three best simulators" advertised in this clip offer AI support when inserting enough small coins. I still have some AI-aversion, 25% empty words to hide the holes. My thinking is, a factory like JLCPCB likes orders, so they may facilitate defining PCBs they will produce, an in fact they offer such a system for free.
Well, I'm still not in the position to setup a working circuit diagram, currently only the keyboard wiring and the TCA8418 keypad scan device are in rank of 'increased probability'. ESP32-C6 one notch below (would S3 fit better?), display uncertain, battery indefinite. And it's already week 3 of my one-year-only project.
2026-09-11 10:23 am
Tell him my thank you. Now I see it's not just a cakewalk to analyse a circuit diagram completely. It's much more difficult than a P&ID where pressure, temperature and medium define the pipe class, and the flow rate and max flow velocity define the pipe diameter. In circuitry you may even have to consider interspersions, I learned this week.
The chasm between what he does designing CPUs (hundred billions of transistors) and what we do here connecting an MPU development board to a few breakout boards is... the Pacific ocean. I've never looked into something like you're describing: a simulator or analyzer. But I can't help thinking there would be something publicly available. Personally, I use the Smoke, No Smoke Rule - If smoke gets out: Bad. If smoke stays in the chip: Good. I buy in bulk and get the discounts. 😎
Seriously though, I didn't really embrace AI, but was kind of forced into it since Google searches now AI summarize everything up front. At first, I found it surprising how it seemed to consistently get down to the gist of my questions. Then, I started doing follow up questions and again, it did a far better job than me drilling into some datasheet that would have been two pages down in the old Google search after all the YT videos. Besides, I was more paranoid about opening some datasheet website, that might have been a re-post instead of the OEM and have hundreds of advertisements... if I was lucky.
Case in point, for my robot, I wanted 40V worth of batteries, MPU: ESP32-S3, two DRV8825 stepper drivers, two stepper motors and a sensitive 6DoF IMU: LSM6DSV. I needed some step down voltages devices to play with this. I'd never run anything much above 12V before. It first suggested some high-end things ($30+ a piece). I suggested I wanted something cheaper. It walked me through why the high-end was very good, but that I could get near as good with two different step devices. One from 40V to 5V that was good enough to run the into the MPU's 5V and a separate 5V to 3.3V to isolate and clean up the noise for the IMU. Both of which I had in my Smoking Inventory. It gave me a circuit diagram and everything with all my pieces. Because I had mentioned I was concerned about all the motors and MPU making too much noise and screwing with the IMU, it included capacitors and resistors to further filter the noise and went into detail about why and how the parts mattered. Since I had given dimensions, it suggested things like twisting certain bundles of wires and which device would be better to have close versus having long wire runs. IOW, I learned a lot from that interaction.
In a first approach I would regard it as a physical pendulum with the center of mass above its pivot point -- actually alternating pivots, and almost constatly changing inertia. A nice challenge.
https://forum.dronebotworkshop.com/neural-networks/building-a-neural-network-from-the-ground-up
https://forum.dronebotworkshop.com/user-robot-projects/inqster-the-re-awakening
2026-09-12 3:50 pm
Hi @inq & @amphitryonoff,
Thanks for your contribution Inq!!
Although I only have briefly passed Verilog and its relatives on the opposite side of a very wide, darkened street, my impression is that is essentially just for digital logic, albeit including capacitance, resistance and inductance analogue influences, as well as thermal, etc. many of which are predicted from the physical architecture within that universe ... apologies in advance if I am wrong.
Synopsys, I think is a massive suite for chip design. These are indeed very complex, smart and expensive simulation workhorses, which live in their own universe, constructed almost entirely of the same semiconductor, like silicon. Oversimplifying, that universe is about 20mm x 20mm x 1mm at its extreme maximum, albeit occasionally parallel universes are stacked on top of each other, with inter-universe transport mechanisms, all of which will be part of the simulation.
=======
The simulation that I presumed was being discussed previously was considering the a semi-random selection and interconnection of multiple universes and wandering itinerants from disparate and often unknown sources, constructed on a much larger scale. e.g. ESP32s, capacitors, resistors, wires, sensors, etc., etc. In most cases, only the outside appearances and traits of each universe is observable or documented. Models for some parts may be available, which are equipped with a simplified view of inner worlds, but most practical circuits are left with areas whose inner details are only known to very small and isolated tribes who are forbidden from sharing most their experiences for commercial and other reasons.
These simulations can sometimes be divided into (almost) digital only components or analogue only components. This can reduce the simulation complexity, but most systems are a mix of both types, and stitching the two types together is a challenge that I have yet to see a 'perfect' answer for. (There are real attempts, but they have compromises.)
Perhaps the most common analogue simulator family is derived from SPICE Simulation Programme for Integrated Circuit Emulation, originally written by a PhD student, who made a programme to estimate how a small (e.g. a few transistors scale) which was being hand taped for silicon fabrication, would behave, based on the physical dimensions, etc of each transistor, etc.
This code has been rewritten, expanded with a user interface from the punched card original, and added to, by many contributors, with programmes like LTSpice and a Spice simulator part of KiCad now freely available.
These tools are useful for instigating a specific aspect of a circuit, such as predicting the voltage-current-time relationship of a FET, but have many limitations when considering the typical kind of circuit that can be built in an hour on a breadboard with $20 worth of components.
Of course, I am out-of-date, compared to the state-of-the-art tools, so please treat my comments with caution and adequate pinches of salt.
Best wishes, Dave
Topic starter
2026-09-13 9:12 pm
Hi @inq !
The chasm between what he does designing CPUs (hundred billions of transistors) and what we do here connecting an MPU development board to a few breakout boards is... the Pacific ocean.
The initial step in technology was long before our time by picking up a stone and use it as a tool. The rest is merely a stringing together of experiences and skills resulting from that.
Personally, I use the Smoke, No Smoke Rule - If smoke gets out: Bad. If smoke stays in the chip: Good. I buy in bulk and get the discounts
Reminds me the story about small farmer and big farmer. The later will adjust his fertilizer spreader very precisely while the small farmer does it approximately only, one sack more or less doesn't matter while for the big farmer 100 sacks or more makes a difference.
Seriously though, I didn't really embrace AI,
Then I misconceived your narration, sorry.
https://forum.dronebotworkshop.com/neural-networks/building-a-neural-network-from-the-ground-uphttps://forum.dronebotworkshop.com/user-robot-projects/inqster-the-re-awakening
Wow! Not bad. But once more I was on the wrong track, when you say...
self-balancing robot
... I surmised somethng with two "legs" not two wheels. I did not read along both threads, sorry if my question is already answered there: what is the benefit of two wheels compared with three or four, why not just one only?
Topic starter
2026-09-13 9:55 pm
Hi @davee !
Of course, I am out-of-date, compared to the state-of-the-art tools, so please treat my comments with caution and adequate pinches of salt.
I can't judge that, I'm retiree and have no idea about hardware, I made my living on s/w, mainly under VM/CMS.
This idea to use virtual devices is the reason why I still cling to simulating my project. Here the current state:
The display shows the four-level stack, if I squeeze a bit also the Last X register may be displayed.
Next hurdle, may I use IBM Plex Mono as font?
Next huge hurdle is migrating my 'firmware interpreter' to this ESP32 platform.
Best,
M.