[sdiy] A little OT, Begging a hand

[Ian Fritz]

At 08:25 PM 2/8/2011, David G. Dixon wrote:

>It is if we know VA, VB, VC and Vout, in addition to the applied voltages
>V1, V2 and V3.  Otherwise, we have 9 or 10 unknowns, and the situation is
>quite hopeless.


I don't know about that. I'm sitting here looking at the node equations. 
They have symbols for the node voltages, yes.  But if I put in numbers 
instead of symbols I still see three equations in four unknowns.

I'm trying to see if there is a more physical way of looking at things to 
explain to you why the system is not solvable.  Please try this:

Call the nodes A (upper left), B (lower left), and C (on the right, driving 
R8).

Suppose I increase R1. Then I can decrease R4 to get the same current as 
before from node A to node C.

Also whatever happens to the voltage at node B, I can change R5 to get the 
same current as before from node B to node C.

This may need to be iterated to consistency.  But then the currents into 
node C from the left are the same as before so the current out of node 
C  is also the same as before.  So V0 is the same as before. (The current 
out of node C flows through R8, independent of the changed R1 in series 
with it.)  So I've changed three of the R's but gotten the same voltage out.

I didn't change V1, V2 or V3.  So it follows that is not possible to 
determine the resistors from V1, V2, V3 and V0.

Hope this lights the bulb for you.  :-)

Ian