Why the circuit in the thumbnail is wrong? Do you know?

[@codedesigns9284]

PCB’s that “just work” is much different than a PCB that functions efficiently, lasts longer, costs less to run, optimized components and more! It’s the difference between buying an electronic device from a cheap outlet store versus from a company that invests research & development. Excellent video Robert!

[@theIpatix]

Sometimes I am quite surprised how some of my very first circuits I designed even worked. I remember building an Arduino accessory with long cables, without any capacitors at all and it just worked. Several years later I tried to power up the board again and it just would not function at all. Luckily I had access to an oscilloscope so I quickly found out a lot of the signals were extremely noisy and unstable. I added two or three capacitors and voila, it finally started to work again.

[@carlosgarcialalicata]

I agree. I actually don't even understand how I am alive after the way I have done things

[@xrfreedom]

I'm very impressed how you break down power integrity into something very understandable so quickly. It's quite amazing.

[@RobertFeranec]

Thank you Jason

[@japedr]

So basically, the models are: 1. Perfect voltage source with perfect wires. 2. Slightly more realistic source with output resistance and wire modelled as a resistor. 3. Any wire is not ideal but actually a transmission line, especially at higher frequencies. Among other things, voltages are not necessarily constant over any segment of the wire, so there are no real "short circuits". This can be modelled as "an infinite sequence of tiny passive components". 4. Lumped components like RLC are not real and Kirchhoff's laws are not valid. In fact, there are no such things as real "linear" currents in an infinitely thin wire, but 3D "current densities" flowing through a medium; moreover, there are no true voltages (except perhaps in pure vacuum) as that requires some mathematical properties on the electric fields that are no longer true. This is the realm of the Maxwells equations, where things like the geometry of the materials matter, besides their behavior in presence of EM fields. Design and simulation are incredibly complicated so simplified models of the most relevant effects are required. I would add a couple of things: - Everything that is above absolute zero temperatures injects "white" Gaussian noise, except it is not really white for a number of reasons, but approximately true in a reasonable frequency range. - Quantum physics magic: Maxwell equations are "wrong", no material is continuous, electron flow is unpredictable, and depends on atomic energy levels considerations, etc. But this only really matters for optics (lasers) or semiconductor materials theory, and of course, in those cases, tons of simplifications are made to make them treatable. Maybe quantum computing is looking like the most realistic short-term use case requiring the most complicated models. Of course, beyond that, there are the effects that can only be measured with particle accelerators or in black holes, but that is obviously way outside the scope of Electrical Engineering.

[@ElectricNimPedalsPhilippines]

You have no idea how your page has helped me a lot as a self-study Electronics hobbyist.

[@marcdraco2189]

Excellent point. We often forget the basics.

[@dankodnevic3222]

Thanks Robert, you just demonstrated, how much decoupling capacitors, next to the IC (or other variable load), are important!!! DC voltage drop can be fixed with broad power planes and LDO's close to the load, but you still need high bandwidth decoupling capacitors (low ESR and inductance), next to it.

[@nihar0689]

Hi Robert, thanks so much for the video. In regards to your view of energy and fields. I would suggest that you contact Dan Beeker from Sierra circuits. I recently attended his EMC seminar and found it quiet useful and his take on PCB layout and circuit design is completely based on how fields move in the dielectric of the PCB. If you can make a video with him to share his knowledge and expertise then it would be of great value to the electronic design community. 😊

[@RobertFeranec]

I am talking to Dan :)

[@user-dt3ub4wn7e]

All right... Here my master again.... Thank you so much my Master!!!!

[@markwebcraft]

Hands down, the most detailed and complete electeonics design information on youtube! "We're not worthy, we're not worthy, we're not worthy"

[@mc4ndr3]

Batteries tend to discharge as quickly as possible, resulting in fires and explosions. That is why a battery management system is so commonly used between the battery and the rest of the circuit. In a mechanical vaporizer, the human user acts as the battery management system, manually connecting and disconnecting the battery at the right times. This carries risk, of course, of the aforementioned fires and explosions.

[@Masudrana-vi7sy]

Interesting. It's always a pleasure to learn from you. Thank you very much.

[@fernandoi8958]

I had a problem with an ESP32 board where it suddenly stopped working after a few hours of performing simple data aquisition, sending over wifi to a server and then going into deep sleep mode... it turns out it had a problem with the power supply line during wake up / sarting wifi modules...

[@robertoaraiza7434]

Excellent refresh from my University classes.

[@victorchorques4893]

Very well explained. Thanks for sharing.

[@fetboy4444]

wow this playlist is awesome i learned so much. Thanks for sharing.

[@harrisoncreates]

This is awesome!! Thank you Robert :D

[@SteepFreeman]

Before this, i didn't think about power delivery like this. It's just a madness soup of really important things!

[@IncaTrails]

The circuit is correct. The parasitics effects are only seen when the circuit is realized ;-)

[@m4l490n]

Excellent video!! It is a big problem when people design PCBs without absolute disregard to this PDN considerations and those boards work. People tend to think that because that board worked all of them will. Then when one board doesn't work and you explain why, they immediately think you are crazy because that's how they've been doing it all along and they haven't had any problems. It is a pain in the butt trying to explain people why a board may not work.

[@mehdikorch8505]

Hey Robert I want to say thank you for all this very helpfull Videos i flo you for a long Times. And i learn realy a lot.

[@UnbornWarrior123]

Hi robert, any update on the fourth section? I am eagerly waiting.

[@mrechbreger]

My pick and place machine is using exactly this kind of power supply :D Don't even think about bypass capacitors on the pick and place head. The company is called Mechatronika in Poland. I fixed it already and it's not crashing (and rebooting) anymore.

[@mrafayshams3089]

I have so far designed 2 PCBs for a project based on the PIC18F45Q43. But both times the PIC18F45Q43 burns out when I power the circuit. In the first design, I did not have decoupling capacitors, but in the second design I did; but the result was the same. I'm using 7805 regulator with decoupling capacitors as well. If you think there is something I'm missing or something I can fix; please help. Thank you in advance sir.

[@sakshimehra31]

Hi Robert do you take one on one some sessions for some queries .

[@seanm8030]

I wouldn't say this is "wrong". It doesn't include certain things, and this may not be sufficient for some applications. However, if it works, then what better criteria are there?

[@jimcrowjoe451]

What type of drawing tablet do you have for your videos?

[@RobertFeranec]

Wacom Intuos S Bluetooth Black

[@ZoeyR86]

also the LRC effect can have harmonics this is a Common mistake for new designers with RF. this is why decoupling capacitors and controlling return paths are so important.

[@maks886]

Of course we will be interested! :D

[@Ozzy3333333]

A "thumbnail" does not need to have PI circuit analysis, SO THERE IS NOTHING WRONG WITH THE CIRCUIT, ONLY CLICKBAIT TITLE. If your going to look at trace loss, you forgot about the ground path IR losses, and battery IR.

[@Ozzy3333333]

99.9% of the people on YT do not need to go into this type of analysis, majority will be under 10KHz, so again, nothing wrong.

[@tahsindegirmenci5667]

Thank you.

[@meysamramezani3087]

Thanks for the video robert

[@AI_Image_Master]

What type of a 3.3v cmos logic level chip is drawing 2 amps. Example please.

[@RobertFeranec]

it doesn't have to be exactly 3.3V / 2A. That was just an example to make it easy for everyone. In reality the numbers can vary but the idea is the same. For example, try to search for intel i7 datasheet and check the max current. The CPU core voltage may "only" be 0.7 - 1V, but peak current can be easily 150A or more - so lower voltage but much higher currents, still the same idea.

[@AI_Image_Master]

@@RobertFeranec Thanks. I have only been playing with simple stuff.

[@vit777vit]

Comrades, blocking capacitors have not yet been canceled, and, of course, start wiring the PCB design with power circuits.

[@voicebangla360]

Good

[@erikmjelde4428]

Nice video

[@spudhead169]

Cheers Latka

[@leeminho2024]

Thank sir

[@edwardneuman6061]

The math has to be wrong, your not going to lose a whole volt with 1 ohm resistance.

[@guatagel2454]

It is an schematic. It is not wrong, but incomplete. The title of the video should be "why this won't work?". Only then your analisys is right. For example, the SIM800L module states that the module must be powered directly with a lithium battery with the shortest possible wires. If you use long wires, or just a voltage regulator as the power supply, the module won't work or will fail. So, if you draw a battery and a SIM800L, you must write in the schematic "closest possible". If you don't write that your schematic is incomplete, not wrong.

[@gustavowkaiser]

Devil hides itself on details! 🙈🤦‍♂️ Sometimes, only a VNA can make it come up...😄😄😄

[@richmiller9844]

My girlfriend helps me with my "impedance"😎

[@tchiwam]

Hmmmmm 3.3 +/- 10% = 2.97V to 3.63V... Better power supply is needed...

[@tchiwam]

Also something to consider, square wave with a rise time of 300ps is pretty much all frequencies up to 3.3GHz

[@tchiwam]

LOL... posted just before your comment :)

[@Lehpurdzzz]

The tolerance you just quoted is what the ic can operate on. How does changing any aspect of the power supply affect the impedance of the path between power supply and ic

[@LAClucas1980]

I like your explanation but you focus in the problem and you do not speak about solutions ..practical solutions ..practical examples

[@RobertFeranec]

I am working on it. This video is just a small part of decoupling video series which is much much much more complex than what I was originally expecting to record. But I recommend to watch the following video from description - that is a practical example and if you follow that video, there is also solution. The whole topic is broken into number of videos as it is very complex.