How Much Current Will Destroy a PCB Track? | Track resistance measurements explained by Eric Bogatin

[@eropis]

I love these Videos. They rarely tell me anything really new but its always nice as sort of a podcast on the side.

[@miguelangellopezdibenedett9783]

Eric provides the best lectures.

[@biswajit681]

Excellent Robert!!This type of session required for the isolated Converter Design and magnetics design..

[@KostiantynKostin]

Thank you Eric! Thank you Robert! I'm learning so much from both of you and I'm trying to put this knowledge to do something useful.

[@egg99]

This is a topic i was really curious about!

[@jazzbwooy8889]

Amazing and exactly what I needed, thank you so much!

[@igalkagan]

Thank you very much!

[@peterlarkin762]

Great to see theory in practice!

[@theovannieuwenhuizen5756]

Great demo! Does anyone wonder why even with the power supply in off the voltage measurement still sits at around 3 Volt? Must have a huge capacitance build in. And why not measure the supply current via a precision shunt resistor to know what the supply current really is before assuming it's x Ampère?

[@robertbox5399]

I'm doing similar testing with a 350A PSU, a 500A current probe and a 200A contactor to switch the power. I would not put a PSU directly across a short - the strain on it is huge, so I'm using about 5kW of power resistors. The current is set by adjusting the PSU voltage between 2 and 28 volts. A 2mm trace can pass 150A for a fraction of a second before it stinks the room out. The power supply itself uses a precision shunt BTW.

[@robertbox5399]

You know when thermal runaway is approaching as the current graph becomes a downward slope rather than a flat horizontal line. You can also see the PCB trace change colour each time a pulse (20-500ms) is applied before damage occurs. Vias are incredible and often last longer than the trace they are connected too so I don't think they are important for current handling.

[@IHDIHD-q8z]

Great content, but what calculator can be used daily and what should the settings be? Saturn? IPC-2152 with modifiers??

[@hidennseek1483]

I don't trust that power supply for 60$ from amazon. Prob most uncertainty is generated by it. You should measure the current with another meter for so small measurements. Nevertheless, the video is brilliant, I did enjoy it.

[@samlachance2]

Honestly, it's probably fine, unless you're talking about ripple or something. I have a different, but similarly cheap power supply, and they're basically an adjustable supply with voltage and current meters strapped on the front. So when you adjust the current, you're adjusting it from 0%->100% of its capability, but it's up to you to turn the knob until the ammeter is showing what you want. The display isn't measuring the state of the knob; it's measuring the current that's actively coming out of the supply. It's kind of an elegant solution tbh. 10% on the knob could mean wildly different things from one device to another, but as long as the voltmeter and ammeter are good, you know exactly what's coming out of it. 🤷‍♂

[@arekx]

He has a good voltage meter, so he most likely also validated his power supply.

[@envisionelectronics]

No it is not the most uncertainty in this circuit. Go back to school.

[@hidennseek1483]

@@envisionelectronics Can you please educate me? Which one of the devices used in the measurements has highest standard deviation? Please support your comment with datasheets.

[@swake0019]

Nice interactive training video. Here comes a trick question: If the track is half as long, it will have half the resistance. Does this mean we can have twice the current?

[@bene5431]

Only If you're limited by voltage drop

[@gsuberland]

For the same ohmic power dissipation you can have sqrt(2) times the current, because P = I^2 * R.

[@jorgealzate4124]

I want to take the opposite route and build a heating element, similar to PCB based hot plates. How much temperature can I use reliably? There are better PCB substrate, mask, or other PCB materials for this purpose? What documentation is recommended for this kind of research? Thank you

[@tylerreid5863]

Look into ceramic heating elements or nicrome wire. Depends what size area you want to heat and how hot.

[@jorgealzate4124]

@@tylerreid5863 thanks for your suggestion, but that's not what I'm looking for. What I want is something similar to some 3D printers heated bed that is built basically by a PCB finished with an anti sticky layer. It has the benefit of being easily manufactured, cheaper to replace, and in the case of heating elements for example for a TCO, a flexible substrate can be used. Even more, there is a small project of a small PCB reflow plate built this way, that project is the reason I asked my question, because perhaps there are better materials/finishes *FROM PCBs* to do this, with the benefits I mentioned above

[@bene5431]

There's a video about that on this channel: This You Need To Know About PCB Materials from 2 months ago

[@gsuberland]

If you're looking to do this with standard PCBs, what you want to know is the glass transition temperature (Tg) of the fiberglass. Standard cheap FR4 is usually somewhere around 130-140°C, but you can get Tg 150-160 and Tg 170-180 too.

[@sodium.carbide]

if you wish for more international viewer. please consider using international measuring system. love the burning part anyway..

[@theovannieuwenhuizen5756]

I second that.

[@AllTheFasteners]

I used to think like this, but like it or not the American influence is so significant that if you want to make progress in electronics, it is pretty important to become fluent with US units.

[@swake0019]

@@AllTheFasteners Have to disagree with you. The real stuff in electronics is always expressed in international measuring system units as a base. Some people stubbornly refuse to use it due to they're not capable of evolving or due to commercially drive marketing. The conversion to imperial units is prone to errors on all stages, the USAmericans alone made enough catastrophic errors related to this conversion to prove this.

[@vinays6674]

37:15 is it 1oz copper here?

[@RobertFeranec]

I believe it is 1oz + plating

[@vinays6674]

@@RobertFeranec considering 1oz copper roughly how much current can a 6 mil trace carry on an outer layer based on this experiment and not as per IPC.

[@gillsejusbates6938]

@@vinays6674 try to think for yourself. you dont need to ask any small detail, you can very easily look something this trivial up