I have a simple question. This is a qustion that I was asked in a job interview.
If there existed and infinite two dimewnsional array of 1 ohm resisters, what is the resistance between two point seperated by a knights move (i.e. up two, over one, or over two and up one.......etc...)
--
Posted via a free Usenet account from http://www.teranews.com
not sure about it, but if you start from the result for a "simple move", that is the resistance between two adjacent points, you find 1/4 Ohm as a result. This comes from the simmetry and infinity of the structure, which in turn demands for the current to split in four equal parts. Now, it seems to me that the same argument should hold for your case, because also here the generator current can't distinguish between the four directions: in every direction the resistance has the same value. Therefore, I would have answered 1/4 Ohm.
I'd say that such a lattice has the property that resistance is always R/4 irrespective the two points taken.
Maybe tomorrow I'll regret giving this answer...Did I get the job? ;-D
I would say no. It is a lattice (I think this should be the right word in english), that is a net of resistors. Every node joins 4 resistors toghether. One could also visualize it as a "mosquito net", where the threads are chains of resistors of some unit value, crossing with perpendicular threads of the same nature.
Maybe we should find that company a book titled "Meaningful Interviews" or something like that. If our HR folks would have asked me to pose such questions I would have chewed them out.
If I'm not slipping up, the resulting equation would be:
PI PI / / 1 | | 1 - cos( 2u + v ) R = ------ * | | ----------------------- du dv 4*PI^2 | | 2 - cos( u ) - cos( v ) / / -PI -PI
But thinking as a programmer, I'd probably just solve it numerically, not through numerical or closed integration of the above, but instead by:
(a) create a matrix V[] to represent the node voltages (b) select some central point for A, call it V[x,y] (c) arbitrarily initialize the matrix node values to 0.5 (d) initialize matrix element V[x,y] to 0.0. (e) initialize matrix element V[x+2,y+1] to 1.0. (f) divide the grid into a checkerboard arrangement with white and black (or red and black) squares (g) process all the black squares, those not V[x,y] or V[x+2,y+1], so that their new value is the mean of the surrounding four nodes. (h) process all the white squares, those not V[x,y] or V[x+2,y+1], so that their new value is the mean of the surrounding four nodes. (i) until satisfied, go back to (g). (j) add up the four node voltages adjacent to V[x,y] into 'sum' (k) print 1/(sum-4*V[x,y]) as the resistance.
Which looking at the near center of a 80x80 grid (probably close enough) comes out to 0.7728 ohms.
I don't think so. In terms of an exact calculation, it is (2/pi) for an "over one, up one" move. That's exact for an infinite grid as I understand the OP's question. The "over two, up one" has slightly higher resistance. But not so high as to be sqrt(5). That figure is wrong on its face. At a distance of "over two, up two" the exact answer is (8/(3*pi)), which sets an obvious upper limit. The actual figure for the knight's move must be between (2/pi) and (8/(3*pi)) or else between 0.6399 and 0.8488. And that is what I found when I did it using an entirely independent numerical method, from scratch.
Chris Foster wrote in news:Xns9845848B07020eddolan@66.150.105.47:
This question was posed by a technical guy, not HR. I did get the job, but I have no idea if my answer was correct. The fellow who asked the question moved to California very soon after I stared working there, and I never found out the correct solution
--
Posted via a free Usenet account from http://www.teranews.com
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.