Surplus electricity

Not actually true.

A google search on "The conductivity of a one atom thick layer of gold in ohms per square"

gives a result that points out that a 10nm gold layer offers

Evaporated gold does end up as adjacent crystals on the surface. True single atom gold films - "goldene" - do rather better, but they are hard to make.

Would you want to design with wound film caps that have kohms (or megohms) of ESR?

And with gold metalization?

Absurd.

The most common film cap metalization is aluminum, sometimes zinc. A single atomic layer of either will oxidize after deposition. The AlO2 or ZnO monolayer will be a wonderful insulator.

AI Overview A natural, passive aluminum oxide layer typically forms instantly upon air exposure, with a thickness of roughly 1–10 nm (4–5 nm is common). This amorphous, self-limiting layer protects the aluminum from further oxidation, though it can grow thicker over years or via accelerated processes like anodizing (0.5–25+ µm).

Key Aspects of Aluminum Oxide Layer Thickness:

Natural Passive Layer: Immediately upon exposure to oxygen, a 2-5 nm layer forms. It can reach up to 10 nm over extended periods, but often stays around 4-5 nm in standard environments.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

Don't be silly. The resistance per square is across the length of the capactor - the way is woudn and connected puts the cuurrent paths across the capaictor in series rather than in parallel

There's very little gold laid down.

The absurdity is all in your ill-informed snap judgement

But the oxide layer is about one atom thick. It protects the metal layer underneath remarkably effectively.

One or two atoms thick.

So you are actually agreeing with me, but didn't understand that what you were posting supported my point of view.

Do you stand by your one-atom-thick-metalization claim?

You always snip your dumber lines.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

The reason it would be practically infinite is that you won’t have a continuous film because the surface is too rough, and because it won’t survive much handling at all.

Some of us have actually owned evaporators. (A Denton 502A in my case.)

Cheers

Phil Hobbs

Pretty hard to do in a single layer, but not impossible with several layers. SiO2 has an epsilon of about 4 at ordinary frequencies, so a 2 cm x 2 cm cap with a 10-nm SiO2 dielectric has a capacitance of

C(cm) = 4 cm**2/(4*pi*1e-6 cm) = 3.2e5 cm or 0.38 uF. (I use Gaussian units--sue me.)

I'm not up to speed on advanced packaging anymore, but BITD that would have been separate passives inside the package.

Earl 'Madman' Muntz would be proud. (Pease used to encourange Muntzing as well.)

Cheers

Phil Hobbs

Making a stack requires next to no litho resolution at all. In the thin-dielectric limit, interlocking arrays of rectangular boxes have the same capacitance as parallel plates occupying the same volume, because the plate areas are the same.

With thicker dielectric, it's worse, because you need extra dielectric thickness to deal with the 1/sqrt(r) field divergence at the rectangular corners. With tapered boxes, you lose more still, because the plate areas become more and more unequal as you move along their length..

Cheers

Phil Hobbs

The one atom thick metalisation is a mostly impractical idea but my Google search - quoted above - which you don't seem to have processed, says

" A google search on "The conductivity of a one atom thick layer of gold in ohms per square" gives a result that points out that a 10nm gold layer offers 2.9ohm/square, while a 1-2nm gold layer comes in a at

The critical thickness seems to be a couple of atoms thick rather than a monolayer but once you have that there's no point in making it thicker.

You've got to have enough gold atoms there to let them organise themselves into little crystals of gold that are still in electrical contact.

Calling that a monolayer kept the idea simple enough that even you should have been able to grasp it. but you clearly haven't.

You said (and of course snipped)

Metallised film capacitors are made by evaporating metal to create metal vapour which you let condense onto the surface of a plastic film

- the metal layer is about one atom thick. Thinner and it won't conduct electricty. Thicker and you are wasting metal. You wind two layers of metalised film together to make your actual capacitor.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

The problem with creating a metalised film one atom thick at it thinnest point by evaporation/condensation is that it ends up having to average four or five atoms thick. It's lumpy.

This takes time to explain.

What you posted was

Which is decidedly absurd.

Search on "goldene". It's remarkably difficult to create a continuous layer of single atom gold - you don't seem to be able to do it by evaporation, but it has been done and it is about a couple of hundred ohms per square. If a layer of gold 30 atoms thick has a resistivity of

So you still score a lot higher on the absurdity stakes.

A single atom layer produced by evaporation/condensation wouldn't be uniform enough to be continuous. Getting up to an average of four or five atoms thick give you a pretty much continuous film.

The surface of the metalised film in a metallised film capacitors doesn't have to survive much handling at all. The layers of film get wound together or stacked into a solid block. The metalisation only goes out to one edge of each metalised tape and in the simplest wound film capacitor you wind two tapes together so that the matalised edge of one tap is on one side of the roll and and the metalised edge of the other is on the other side. Metal-spray the ends of the roll and that's what the two wires go onto.

I learned this back in 1971 when my Australia employer threw me at the problem for want of anybody who knew much about it.

Lucky you. The Welding Institute in Cambridge had one and did odd bits of stuff for us.

In my case not because I haven't used the big A/C in years. It's swamp cooler only.

Ideally one would supplement when prices are low. It Texas you can, in California that doesn't make economic sense. The ToU rates they offer are a bad deal. You only get a miniscule discount during solar excess supply but when you then want to cook or bake a meal in the evening you get severely punish by huge upcharges. Meaning your total electricity bill would now be higher. Therefore, many people like myself are not signing up for that.

That can be done if you are mostly retired. For example, run the dryer or the vacuum cleaner while it is sunny outside. For working people it's just not feasible.

For businesses, 3-5 years is fine. You make sure not to invest too much or too longterm. Then, should things turn socialist you can leave.

That's what business service providers are for. They scout such things and you can buy a targeted report. Those reports usually cost thousands of dollars but are often worth it.

Well, everybody knows that :-)

Nowadays with remote work it doesn't matter so much. I have worked remotely for decades. I have never personally met many of my longterm clients yet we have done lots of projects together, from concept to production. I could live on a Caribbean island as long as Fedex lands there.

Exactly. Just like it happens with other costs or red tape. Costs get too high -> hightail it.

It's the consumers who play those games. There is a reason why many people in our area have their huge RVs registered in Florida, Montana, Oregon, et cetera. Sometimes their whole fleet of cars.

I remember people in WA state who got together on Saturdays, rented the biggest U-Haul truck they could find and then drove down to OR. There they bought whatever was needed. Big ticket items such as washers, dryers, fancy dishwashers, leather sofas, video games, computers, and so on. No sales tax in OR.

And they do even though Internet sales are now taxed. Some people look for small mom and pop shops out of state whose biz volume is below the threshold so they can sell sans tax. That's an instant 10% discount in your area.

The subject was and is film caps. They use aluminum or zinc or combo metalizations. Many atoms thick. Look it up.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

We've had the ACbrrr running for several weeks, already. It's been at or above 100F for much of that time.

While I prefer the cooler in "Summer" (the hot season that is also dry), SWMBO isn't a fan. And, the cooler is just not capable of providing comfort during Monsoon (the hot season that is humid).

Exactly. The cost for the energy that you are providing is offset by the added "transportation cost" of any that you *purchase*. Utilities aren't fools. They realize they are on the wrong side of this equation and are just fighting a delaying tactic to try to remain relevant.

If you are home much of the day -- or have pets -- then you are paying to keep a large volume of space "comfortable" (heated/cooled). I suspect the calculus is different for people whose employers keep them warm/cool during "business hours".

These changes require behavioral changes and discipline. In addition to opportunity.

The purpose of my original post was to try to identify ways to use "surplus electricity" without an explicit effort (and, across a range of environments/users).

I presently "do extra work" when I know that I *will* have a surplus instead of defering that work given the possibility that it might NOT be necessary. Or, to exploit periods of lower consumption.

E.g., I can power up every node and have them run diagnostics while I have surplus power available -- instead of waiting to *suspect* the need to run diagnostics.

Or, retrain recognizers and models instead of doing those things less frequently (because they have diminishing returns).

Or, recording TV/radio/cable broadcasts and eliding "commercials" and "commentary" -- on the off chance that the user may want to consume said (if not, then you can just discard your efforts!)

But, these are relatively small loads, given that they have to be supported "regardless"/eventually.

I plan on moving my servers out to an off-building and leaving them powered (and cooled/silenced) there, reclaiming "habitable space" indoors. Eventually, powering and heating an endless pool as I start retiring the electronics kit.

But many people don't work remotely, despite the lack of a real NEED to be on-site. Yes, there are some efficiencies from being able to interact with a variety of individuals spontaneously and *instantly*. But, there's a high cost to providing this (travel, scheduling, etc.).

If I had to interact with clients and peers "in person", my work and sleep schedules would be subject to THEIR requirements. Why should I

Or, let your business own those assets and write off your oil change, cost of insurance, cost to garage, etc.

But, not everyone has those luxuries.

Your habits are unlikely to change for a few dollars on a grocery bill or other sundries. But, could easily be coerced for big purchases, regardless of how infrequent they may be.

I bought a tape deck in the late 70's and had it shipped to my parents' home to save the local sales tax (considerable on a $2200 purchase) as the cost of shipping was a small fraction of that. When the equipment arrived at the local shop, I went down and picked it up and paid the vendor for the shipping (that never happened but *appeared* to on the invoice). So, he made a few extra bucks and I got the kit earlier (and without having to make the drive out-of-state to fetch it).

We have "outlet stores" here, in the middle of nowhere (notaxwhere).

But, this is inconvenient for many people to exploit.

I used my resale license to shift those taxes as I was making large purchases for clients. Inconsequential if buying a case of 9T tapes -- but, quickly add up when buying $20K pieces of kit!

Where? I could do it for myself, but if you want to pose as an expert, here's your chance to shine.

That can still be directly above the equator. The counter balancing mass above that height has to stay at a point that is as far enough below the equator to compensate for drift of the anchor point above the equator (or vice versa).

Arthur C. Clarke didn't invent the idea and rather more technically able people have looked at it in detail over the past century. It seems to be practicable. Practical is a different question.

For those who aren't convinced that Bill knows enough about search to provide suitable links:

The basic question was whether the metalisation was any thicker than that required to get more than one atom thickness at the thinner points, and none of your links seem to answer that explicitily.

There was some emphasis on the self-healing properties of very thin films - when the there is an arc you want it to be able to vapourise all the metal in the vicinity without getting the film hot enough to get carbonised into a state where it too can conduct - but no discussion of metalisation thickness as a pile of atoms.

You didn't shine.

The basic google query says

AI Overview Comparison of Metallized Film Capacitors and Film/Foil ... Film capacitor metallization typically involves an ultra-thin layer of aluminum, zinc, or an alloy, measuring roughly 0.02 to 0.1 µm

The self-healing caps usually have a geometric pattern of square cap regions connected by tiny fuses. Those are usually made with aluminum/zinc composite layers.

You are allowed to look that up and learn.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics