Infinite Resistor Problem - Confirmation

It would be, and you can bet I'd sell tickets (at $2/minute.) And hotdogs, meals, and probably rooms for the week-long extravaganza, because I wouldn't be done in a single day.

If enough folks want to pony up $5k/wk for this wonderful event, be sure to write. I'll include transportation to and from the airport, the option of either A-frame housing or separate rooms with their own bathrooms, commercial kitchen services and food, horse riding, evening hand gun, rifle, and automatic weapons practice with my picture on every target, and wall sized video monitors to watch me solder 0201 resistors with a Weller D550 soldering gun using pieces from a rounded hanger wire as a tip. If I survive the abuse, I'll personally sign certificates of completion, too.

If this 'entertainment' catches on, I'll make it a yearly event! ;)

Jon

P.S. Actually, I'm already working on providing regular retreats not unlike this -- four times a year is the current goal. Very soon, I hope to see. Different function, of course. But I'm adaptable when the price is right. ;)

Ball Industries? Is making IC chips that are on a spherical shaped die?

Anyway. Make a ball shaped array. Build layer after layer with tines connecting the proper inner nodes together with higher layers.

Hell, just drawing it in CAD would be fun, and challenging.

I'd say 0805 has the right size and strength for it, as far as terminations go.

I wonder how easy it would be to configure a quad array into a multi-layered spheroid.

Not much different then the guy wanting to put it on a cylinder.

A LOT more fun to draw. Screw electronic layout packages. This is a job for regular old CAD.

Let's see... toothpicks for skeletal needs as the layers grow... super glue... solder, small, tfe coated hook up wire.

Naaaaah!

A real man doesn't need perfboard. ;-)

...and what is wrong with taking gerbers from a scematic capture program, and using a n assembly house with a Pick-and-place machine?

By the way, for those wondering if there is a practical application for the infinite resistor grid problem, there is this patent:

Which claims ownership over using a crossbar on a 2D resistor grid as an analog computer solving a range of commonly found finite different method problems.

Jon

Oh, s*it. Now that I read it, the damned patent writer didn't even get their equations right.

Oh, well. I guess that invalidates the whole thing. ;)

Jon

"finite difference method"

Jon

Please note that anyone can patent almost anything if they have the money; the item does not have to be practical, it does not have to be workable, nor does it have to be buildable. And the "explanation" does not have to be reasonable. Also, a patent is "good" until it either expires or some court has invalidated it. Hell, if you were clever enough in the write-up, you could patent the paper clip...

Enough to wine and dine the examiner and pay the $50k/mo patent atty fees, yes.

The patent law specifies that the subject matter must be "useful." The term "useful" in this connection refers to the condition that the subject matter has a useful purpose and also includes operativeness, that is, a machine which will not operate to perform the intended purpose would not be called useful, and therefore would not be granted a patent. In most cases, the usefulness requirement is easily met in computer and electronic technologies.

Whatever all that means. Sounds like "practical" but I take your point, anyway.

See above note.

See above note.

hehe. That seems to be the case, often enough.

This, I know. Once issued.

Hmm... The wheel would be more profitable to me, if I'm going to go bother enough to be clever about it.

Jon

P.S. I've already watched the process of one company patenting all uses of a binary search for temperature lookup tables by software in cpus used in instrumentation. That patent was preceded by other such uses for exactly that purpose and wasn't the least bit novel and was besides the obvious approach to be taken by any practitioner looking at that problem. Yet it was issued. Competitors of that company use binary searches in their products (I don't think any publicize it) and the company doesn't bother suing them, either. Everyone just acts like it doesn't exist. It's all Alice in Wonderland stuff to me.

You make four rounded corner rectangles. That makes the standard surface tension self seating mode (but you knew that). Then you have to come up with a way to connect those together. On a single layer, you will have to make a ring that connects between the short side of the rectangles. Ifyou are using more than one layer, you probably know what to do. :-)

Or, if you really do want to Siamese them all together, it would look like a clover where the leaflets are just far enough from each other not to break a very low voltage rule.

I would also use like ten ohm resistors instead of 1 ohm, because your results should have a bit more precision if you match and cull your stock well enough. (I guess the same can be said for the 1 ohm set, but for some reason I think ten would be easier to work with the results of, and easier to match).That doesn't mean that it would be any different though.

It would work with a non-rotated grid as well.

Probably expired.

I see a small differentiation. :-)

Why do you say so?

Jon

my

I have seen patents that "met" all of the disqualifications i cited' "patent law" be damned. On another note, i saw two patents, sequential serial numbers that covered the SAME THING and worded almost exactly the same way - different authors and art units.

Maybe 100K 0.5% SMT resistors would be better?

Note especially sheet 1 (of drawings) where economy is used.

my

So why is the elephant still in the room? Is this some kind of Emperor who runs around naked and no one dares point it out or admit it?

I'm just flummoxed.

Jon

I do not know. It depends on what instruments will be doing the real test reading, and what readings are going to be taken and at what excitation levels.

The precision is to emulate an ideal circuit as closely as possible. That way, the readings can be counted on as representative or closely so, to the ideal circuit.

I really do not know if a high resistance value resistor can be trimmed tighter, or if the low value job is sufficient.

Either way, matching them, and culling out the extreme variants would be recommended.

I guess pumping a known voltage and taking current readings would yield the best the best data. I just do not trust ohm meter readings as being tightly toleranced compared to a good current reading.

Pulse testing would probably be best so that no heating of the parts happens. Then again, a dielectric fluid immersion would probably be the way to go for temp stability.

How is that working out for you? And the wife?