Want to build a shortwave radio from scratch and ordering bunch of components from aliexpress. Scrolled down and rather choose the ones that arrives 3 weeks sooner. Should I worry or are they likely made in the same exact factory? I want everything I build to last for decades
is it even possible to fuck up the quality of these components?
Because of their low price makes me think it is extremely easy to make
>not knowing about counterfeit electronic components
you buy from AliExpress it's 100% counterfeit junk rest assured.
And no it's not trivial shit there are things like specs and tolerances and max currents and leakages and many many other shit that could affect your project in a bad way
just test the components you get.
what kind of asshole are you that you're not already doing this?
what even is this question?
>what even is this question?
OP is asking about "quality," specifically mentioning longetivity.
So he wants to know whether cheaper components are more likely to fail sooner, or if he needs to spend more to get components that are less likely to fail.
>just test the components you get.
Are there tests OP can do which will identify how likely a component is to fail?
>whether cheaper components are more likely to fail sooner
A transistor either works or it doesn't. And it will continue to work until it doesn't any more.
>or if he needs to spend more to get components that are less likely to fail
Buying quality components made by known manufacturers from reputable vendors is the preferred course of action for just about anything in life. Transistors. Toothpaste. Condoms. Footballs.
>Are there tests OP can do which will identify how likely a component is to fail?
Yes, I'm sure there are.
>Buying quality components made by known manufacturers from reputable vendors is the preferred course of action for just about anything in life
These transistors and components don't really have a brand, only more popular and less popular vendors.
The radio I am building is sort of 70s-80s technology and I even have to buy an old 60s-70s toshiba transistor for it because the other part for it is more expensive and rare.
The problem with many electronic projects, especially old ones, is that some components are hard to find and rare.
>rare and expensive
Anon, you want longevity, right? You know what the best pals of longevity are? Affordability and commonality. You want parts that are affordable, ez to find and ez to replace. Shit will inevitably break, and it doesn't matter how resistant your fancy radio is if it's a pain in the ass to find spares.
Find a radio schematic that uses components which you could find in any electronics store or junk yard. This way, even if you do get a shitty component that decides to die when you turn the radio on, you can replace it quickly.
Wrong. I've bought complete fakes and non functional semiconductors...see 2N3055 die shots. Recent shit spread on the dirt sandwich is my stupid ass fell for it again with a "ftdi" usb-serial converter purchase. Mea culpa but I'm a cheap fucker and I'm hoping the workaround I used works if I soft brick the piece of trash.
ON semi for 3055s. They are the legitimate motorola lineage.
Si labs for uart rs232. FTDI can smoke a turd.
>Binary transistors in a nonbinary world
Ok, dude; back in the day, Phillips would rate transistors as A, B or C grade, according to gain; everyone else probably did too; probably still do...
Thing is, a transistor is nothing but a solid state vacuum tube, and no one ever doubted these degrade with time, JUST LIKE EVERY FUCKING THING ON THE FACE OF EARTH AND BEYOND, so why wouldn't a transistor?
Transistors don't degrade with time in any meaningful sense unless you're really abusing them and pushing them at the edge of their maximum ratings nearly all the time.
They're inside fully encapsulated plastic packaging that is extremely durable. If there are no defects in the wafer it isn't going to magically develop them through usage in a sealed environment.
I'm sure there is some mechanism that will very slowly degrade transistors over a very long period of time even with light or no usage, like fucking cosmic rays or some shit, who knows? But that's a timeframe on the order of hundreds? Thousands of years? Longer? Realistically ICs are just gonna magically die, and it usually comes from exceeding ratings or running a transistor with high power dissipation and poor thermal management for a long period of time like thousands of hours continuously. Then maybe it might fail prematurely.
(Chips are mostly made of transistors, btw...)
>(Chips are mostly made of transistors, btw...)
...and then, the ones that came out with defects are scrapped.
>What's a bathtub curve
>It's actually a major (and increasing) concern
>Bullshit. Source: trust me bro
Straight out of the horse's mouth:
>You generally don’t perceive this deterioration because semiconductor companies always play it very safe—they set the clock-speed rating of their microprocessors so conservatively that almost every one of their products will continue to operate flawlessly throughout its intended lifetime. That strategy works. But it’s kind of like never taking your Ferrari out of the slow lane because you’re concerned that its engine might throw a rod 10 years down the road.
>“At modern geometries, the primary root cause of all these effects is the stress effect of high electric fields across the dielectric,” said João Geada, chief technologist atAnsys. “As geometries have got smaller, but voltages have not scaled at the same rate, the electric field across gates has increased, resulting in worse aging behavior.”
>Printed circuit board failures
>Passive element failures
Literally a 5 second search...
they definitely do degrade in a meaningful manner even if you're running them fully within spec.
>If there are no defects in the wafer it isn't going to magically develop them through usage in a sealed environment.
They actually do, the electrons themselves can get embedded in the dielectric materials or physically move the atoms increasing the resistance. Obviously, this is generally much more of a concern with newer nodes/smaller features, but still occurs in everything. Higher currents and temperatures obviously cause this to happen more often as well.
>Realistically ICs are just gonna magically die
You may not notice the effects until it does just magically die, but there will still be some. In processors they'll often run slower to compensate for this damage as time goes on, as not doing so will lead to it not being useful at all much earlier.
If I had to guess, it's more likely to be quadratic or linear than exponential, since the power consumption of a processor is equivalent to the square of the voltage (10% reduction in voltage -> 19% reduction in power). But yeah, generally running it with a lower voltage will help quite a bit, as there's less current and less heat to cause problems. Of course, this will be slower, but usually it's not much and you can often still get advertised clocks even undervolting a fair bit. I think I have one of my 3900xs running at 1.13V and 4GHz. Varies by chip, though, my other 3900x is I think 1.14V and 3.85GHz.
Capacitors aren't really included in chips, though, it's usually just things like the inherent ones of mosfet gates or occasionally for analog or sram things which are probably femtofarads or less (picofarads at most). Any large capacitance (like nF or higher) is going to be off the chip on the motherboard.
yeah what i meant is that capacitors somewhere on the board will dry out/fail, which will then speed up the degradation of other components until eventually something releases the magic smoke
>OP is asking about "quality," specifically mentioning longetivity
this is what I meant.
The ones that arrive sooner are from smaller stores/sellers and rather buy from larger sellers that sold more if quality is something to be concerned about.
This is what I thought. They are all made in China so would not be surprised if the same type of transistors/components come from the same factory.
if ur design is component dependent then yes, u better check as many components may be faked/mislabeled, and in special cases even genuine parts maybe binned to get proper circuit performance u typically see this done for ADCs or Opamps, if ur design is not component dependent then it doesn't matter at all.
>OP is asking about "quality," specifically mentioning longevity.
when it comes to longevity, if OP has parts that are properly rated for his application and he is not driving them hard or straining them then they can pretty much last forever, but since we can't be a 100% sure of the component type or spec he has to take a black box characterization approach similar to what Andreas Spiess is doing here:
I wouldnt worry about it
most of those components are a bunch of trivial shit that you can't get wrong as of 2023
it's not so much longevity (unless we're talking electrolytic caps), so much as it's a question of tolerances.
If an expensive component says it's 1% tolerance, you can rest assured knowing every parameter is within 1% of its listed value.
Cheap components are going to be a random number generator in terms of component values, regardless of the specified tolerance.
This goes for part numbers too. Don't be surprised when your TL074 winds up being a CD40106B
>What's yield rate?
No, transistors never fail; ever. Electronics is that magical realm where never goes wrong, starting in manufacturing. No duds, ever; MTTF=∞
Yes that will send you newspaper wrapped in newspaper
>from mouser, digikey
No, you're gonna get premium components if anything.
Samsung wants to nickle and dime them for .25 cents off of a resistor. You're happy to pay a whole nickle for a pack or whatever
I use mouser for anything I need to be sure about. Otherwise I use tayda for cheap shit but it's usually okay. I bought opa2134s from them and they're real so that's a good sign.
Guys building a shitty shortwave abortion from a shitty 50yo schematic he probably found online, not desining a deep space probe that has to TX images of venus for fucks sake.
OP shitty ali express transistors are fine. They are not always 100% up to spec so some tweaking of the schematic might be needed. The seller might chink you and send you a defective lot, so set up a test jig according to the data sheet just to make sure...
Talk to your care team manager and ask to go back to grade school.
a transisior will either be dead on arrival or work forever*
it will actually lose performance with age but if you don't drive it at the absolute peak ratings this will be extremely slow - if you run a kabylake chip at 1.52V for a year it's going to shave say 50MHz off the max frequency
but it's exponential - lowering the voltage results in big savings
1.52V - 50MHz
1.51V - 25MHz
1.5V - 12.5MHz
1.49V - 6.25MHz
so if you use a sane voltage, like 1.3V, you degrade ~ 23 hertz per year
the capacitors are first to fail 99% of the time
numbers pulled out of my ass but you know what i mean