glizzyguzzler @ glizzyguzzler @lemmy.blahaj.zone

Posts

223

Comments

510

Joined

2 yr. ago

Are you Steubs or Wheeling?

Well we’ll have to wait for the next edition of the map of Ohio and it’s surrounding regions, but I have a feeling that the labeling is slightly off and Toronto actual is farther south and east.

Excellent question, but I don’t have an answer, sorry :(

Si, siempre

Do not worry, I never assumed you gave even one shit about Michigan. I assumed you gave two, maybe three, shits about essential geographic context in the art of cartography

Valid criticism, but out of the scope for the map. Maybe in the next edition, if my publisher funds it

This map is fully accurate, Michigan is just off the map

Yes, as noted in the original post Toledo is now in the correct location

Hey I don’t make the rules, I just report the geography

Tbh this is just wrong, it’s clear Ohio is larger than Canada even if you include their puppet state (Toronto).

Clearly if New York wanted instigate intrastate warfare they would conduct a lightning assault through Canada and catch Ohio by surprise, who would likely be thinking that their Appalachian Line on the border with New York would keep them safe. Canada’s so small they could cross it in hours!

Our guiding light in these dark times

Tiger, you’re very similar to many of the semiconductor EEs I know :) and I mean that in a teasing-but-you-know-cause-you-work-in-the-industry way. Yeah, we only really care about whiskering in the context of electrical devices. That’s what it’s saying. Read the “Mechanics” section, it tells you nothing about actual electromigration doing it; they describe an E field encouraging metal ions in a fluid to make a reaching whisker and link to electromigration because it technically is “electromigration” making the targeted whisker occur. But IC-style electromigration is not causing the whisker, clearly cause no currents are flowing, which is why I took the time to write the explanation in the first place.

But just because the semiconductor community called it whiskers so it shares the name with the Big Whiskers, does not make the process anywhere close to similar. The current densities that cause absolutely not present for the stress ones, which the wiki article is about.

Tiger I think you're being pedantic, they linked to Whiskers (metallurgy) not Whiskers (electromigration). There is a difference! But it's not super clear cut, which is why I took the time to write about it.

Electrons do not always move at the same speed in a given metal. A lot of things affects mobility, but the E field is very important too. Both things combine so that electrons do not always move at the same speed in a given metal. But you can simplify in an IC world because there you're riding the saturation velocity basically always, which is why I assume you keep claiming that.

I want you to know that your experiences from your education and job are valid - you do deal with whiskers in ICs, not denying that; the fact is that whiskers due to stresses and strains aren't called electromigration which is what the original comment says.

"A similar thing also called whiskers can happen inside ICs and has been a known failure mode for high frequency processors for many years. I work in chip design, and we use software tools to simulate it. It’s due to electromigration and doesn't rely on stresses but instead high current densities."

Check out https://en.wikipedia.org/wiki/Hot-carrier_injection hot carrier degredation, it's in the vicinity of electron mobility but in a semiconductor setting. Key link is it's electrons with momentum doing the work. In this case electrons (much hotter than in electron mobility, which are limited by the saturation velocity) smash into the gate dielectric, making it a worse dielectric. Hot carrier injection doesn't have to end in damage to the dielectric, but when it does it's hot carrier degredation. There's a lot going on though, semiconductors are really complex - like electron tunneling also exists.

The metal moves due to very different reasons. I would not say whiskers due to mechanical/residual stresses are due to "electromigration" - electromigration isn't even there since the wiki definition is "transport of material caused by the gradual movement of the ions in a conductor due to the momentum transfer between conducting electrons and diffusing metal atoms". You build stresses and strains into semiconductors for better mobility profiles, and I'm sure that can cause whiskers - but again, it's not electromigration.

Electromigration, as noted, plays a role in the form of encouraging stress whiskers to grow in a direction (with a very relaxed definition).

But in ICs, with their very unique extremely small scales, electromigration can directly form whiskers by moving individual ions via electron collisions. But the generation mechanism for those whiskers shares nothing with Big Whiskers generation mechanism. That's my point.

Electrons in metal do not always move at the same speed; they move at v=mu*E where v is the velocity, mu is the electron mobility, and E is the electric field. Crank the E, you go faster. At very high E fields you reach the electron saturation velocity where slowing factors limit the maximum speed - I assume in your IC world you're basically always there due to the extremely small regions (E = V/m; any V with m at nanometers is big E) which is why you claim that. But even then the electrons are accelerating due to the E field, smashing into ions and losing their momentum (mass static, so it's just velocity), and then re-accelerating. The saturation velocity is the average bulk motion of electrons but it's not a smooth highway, it's LA traffic (constant crashes).

Electrons can gain significant momentum, which is just their static mass times their velocity. Limited at velocity by the saturation velocity, current density is important for significant momentum exchange. Luckily ICs are so tiny that the currents they drive are massive current densities.

What you said originally is correct; it's just in ICs electromigration can cause whiskers. In the Big World it can't. But it can influence Big Whiskers to grow to the worst places and fuck up things optimally if you take an extremely relaxed view of electromigration that defines it as "movement of ions encouraged by an electric field".

This statement is not fully accurate. Whiskers in OP’s case are about (usually) tin whiskers that grow, often visibly, and then can connect (short) to unintended areas.

Electromigration is effectively when a large potential difference encourages ions to relocate to reduce the potential difference.

Big Whiskers have two methods of formation. The first way is that tin ions are able to move by becoming soluble in some form of water so they’re mobile. The other way whiskers can form is from stress alone. (Stress being force per area that compresses or tensions the metal in question, applied through a multitude of ways) Whiskers can be directed by electromigration so they form tendrils to a differing potential, basically purposefully ruining stuff instead of randomly shorting things.

Now in integrated circuits (ICs), there are extremely high currents running through extremely small regions. Electromigration in ICs is caused by electrons getting yeeted at extremely fast speeds, giving them significant momentum. They collide with ions in their path and dislodge the ions from their matrix. This can result in voids of ions preventing current from flowing (open circuits) or tendrils of ions making a path to an unintended area and connecting to it (shorting it). The tendrils here are also called whiskers, but are generated in a very different way (e.g., no water solubility or inherent stresses required) and on a significantly smaller scale. And probably not in tin.

The more you know!

Mine clicked just after a year :( so it’s waiting to get switches soldered cause I ain’t buying a new one but don’t have the will to do it yet cause I got my ancient G5 still. (I live in America, so doing warranty shit is a hassle, and I surmised it would fail again so I should just fix it now)

That’s what I mean, even e-waste quality razer isn’t double clicking!!

I had to replace the cable on my G5 after it frayed after about a decade, but after that it was back to it. Sorry you lost yours, and I hope you never double click on the G502!

(All the weights gang, build that wrist strength)

Mine is a G5, which looks like it lost the MX518’s sick ass faux metal and instead gets what I can best call “cracked lightning?”. I was too young to figure out mouse buying so the fam’s resident nerd chose that for me - I thank her to this day