>> Yes, more exactly, there are devices which do not have constant 
>> resistance/impedance.
>
> Are you saying the resistence doesn't determine the current? Or are you 
> just saying that the current can cause the resistence to change?

If you want to look at it like that, yes, the resistance will depend on what 
the applied voltage is, rather than always being constant.  Nobody really 
talks of "resistance" with such devices though, it's just accepted that V is 
not linearly proportional to I, and that for any given V you can look up the 
I from the chart.

> Now, see, I always thought electricity just takes *every* path, with the 
> current being determined by Ohm's law.

It does, and if use Ohm's law to work out the voltage at a point that is 
connected to 0V with a 1kOhm resistor and to 5V with a 1Mohm resistor, you 
will see the voltage is essentially 0V.  OK so it will actually be very 
slightly above 0V, but that's why there is a tolerance range on the input of 
such devices.

> So, if you have a high-resistence path from A to B, and then you add a 
> second, lower-resistence path from A to B, does the current flowing 
> through the first path change? Or does it remain the same?

It remains the same because neither the voltage across the high value 
resistor, nor the resistance of the resistor itself have changed, and then 
I=V/R.  Of course you get some additional current flowing through the lower 
resistance path now, but still the same amount through the high resistance.

Post a reply to this message

>> Are you saying the resistence doesn't determine the current? Or are 
>> you just saying that the current can cause the resistence to change?
> 
> If you want to look at it like that, yes, the resistance will depend on 
> what the applied voltage is, rather than always being constant.

You you mean stuff like if you put power through a wire, it heats up, 
and resistance varies by temperature?

> Nobody 
> really talks of "resistance" with such devices though, it's just 
> accepted that V is not linearly proportional to I, and that for any 
> given V you can look up the I from the chart.

Hmm, how odd...

>> Now, see, I always thought electricity just takes *every* path, with 
>> the current being determined by Ohm's law.
> 
> It does, and if use Ohm's law to work out the voltage at a point that is 
> connected to 0V with a 1kOhm resistor and to 5V with a 1Mohm resistor, 
> you will see the voltage is essentially 0V.  OK so it will actually be 
> very slightly above 0V, but that's why there is a tolerance range on the 
> input of such devices.

Hmm... my head hurts now! Heh.

>> So, if you have a high-resistence path from A to B, and then you add a 
>> second, lower-resistence path from A to B, does the current flowing 
>> through the first path change? Or does it remain the same?
> 
> It remains the same.

Oh thank God for that! At least something about my understanding of 
electricity is correct...

The whole "electricity chooses the easiest path" makes it sound like 
when an easier path becomes available, all the current goes down that 
and ignores the other paths! o_O

Post a reply to this message

> You you mean stuff like if you put power through a wire, it heats up, and 
> resistance varies by temperature?

Yes pretty much, now imagine that in your design the resistance of the wire 
changes enough to have a big impact on how your circuit works.  If you 
connect an LED to a voltage source and gradually turn up the voltage from 0 
volts, you will see that almost no current flows, it's like nothing was 
connected inside.  But suddenly, you will get to around 2V or something the 
current will shoot upwards, and from then on you only need make tiny 
increments to the voltage and the current will keep getting larger and 
larger very quickly until the LED blows haha

> Oh thank God for that! At least something about my understanding of 
> electricity is correct...

Recipe for solving circuits with resistors:
1) Label each unknown voltage node as V1, V2, etc
2) Write out Ohm's law for each resistor in terms of V1, V2 etc
3) Solve the set of equations you wrote out in 2) for the unknown voltages

> The whole "electricity chooses the easiest path" makes it sound like when 
> an easier path becomes available, all the current goes down that and 
> ignores the other paths! o_O

If the "easiest" path is of *much* lower resistance than the others, then 
yeh you can usually ignore the others.

Post a reply to this message

scott wrote:
>> You you mean stuff like if you put power through a wire, it heats up, 
>> and resistance varies by temperature?
> 
> Yes pretty much, now imagine that in your design the resistance of the 
> wire changes enough to have a big impact on how your circuit works.  If 
> you connect an LED to a voltage source and gradually turn up the voltage 
> from 0 volts, you will see that almost no current flows, it's like 
> nothing was connected inside.  But suddenly, you will get to around 2V 
> or something the current will shoot upwards, and from then on you only 
> need make tiny increments to the voltage and the current will keep 
> getting larger and larger very quickly until the LED blows haha

Is that why LEDs display a non-linear brightness response too?

>> The whole "electricity chooses the easiest path" makes it sound like 
>> when an easier path becomes available, all the current goes down that 
>> and ignores the other paths! o_O
> 
> If the "easiest" path is of *much* lower resistance than the others, 
> then yeh you can usually ignore the others.

I presume you mean you can "ignore" them for the purposes of figuring 
out the total current flowing through the system. Clearly the current 
flowing through each path only depends on the resistence of that path 
and the potential difference...

Post a reply to this message

> Is that why LEDs display a non-linear brightness response too?

Non-linear with respect to what?  The brightness response is pretty linear 
with respect to current, double the current you get double the brightness to 
within a few %.

> I presume you mean you can "ignore" them for the purposes of figuring out 
> the total current flowing through the system. Clearly the current flowing 
> through each path only depends on the resistence of that path and the 
> potential difference...

I meant if you have a point of unknown voltage connected to two (or more) 
other known voltages via resistors.  If one of the resistors is drastically 
lower than all the others, then you can just assume the unknown point is 
directly connected through the lowest resistance and treat all the other 
resistors as essentially not connected.

Post a reply to this message

scott wrote:
>> Is that why LEDs display a non-linear brightness response too?
> 
> Non-linear with respect to what?  The brightness response is pretty 
> linear with respect to current, double the current you get double the 
> brightness to within a few %.

Well, I had a diagram for a circuit where you connect a lamp to the 
battery through a linear potentiometer. The brightness of the lamp 
varies roughly linearly. But the brightness of an LED... does not.

(I'm showing my age now! Can you imagine trying to buy a 3V lamp today??)

>> I presume you mean you can "ignore" them for the purposes of figuring 
>> out the total current flowing through the system. Clearly the current 
>> flowing through each path only depends on the resistence of that path 
>> and the potential difference...
> 
> I meant if you have a point of unknown voltage connected to two (or 
> more) other known voltages via resistors.  If one of the resistors is 
> drastically lower than all the others, then you can just assume the 
> unknown point is directly connected through the lowest resistance and 
> treat all the other resistors as essentially not connected.

Right. Gotchya.

Post a reply to this message

"Invisible" <voi### [at] devnull> wrote in message 
news:49196e5f$1@news.povray.org...
>
> Yeah - maybe I should get said to write open source software?
>
> Oh, wait...

You may not always get paid to write open source, but it's good practice at 
writing real applications and working with (lots) of other people, and it 
looks very good on a CV

Post a reply to this message

Gail wrote:

> You may not always get paid to write open source, but it's good practice 
> at writing real applications and working with (lots) of other people, 
> and it looks very good on a CV

You're not the first person to point this out.

Unfortunately, it requires real talent. (And, usually, C/C++ skills.)

Post a reply to this message

> Well, I had a diagram for a circuit where you connect a lamp to the 
> battery through a linear potentiometer. The brightness of the lamp 
> varies roughly linearly. But the brightness of an LED... does not.

LED brightness is proportional to current not voltage!

Post a reply to this message

Invisible <voi### [at] devnull> wrote:
> losers they were! I still remember the weekly paper fights. Oh, and the
> C'paddy races. (The objective being to say the word "c'paddy" the most
> times in 60 seconds.) Like, WTF?

Was there running and wrestling at the same time? If so, could be this:

http://www.kabaddi.org/History.htm
http://en.wikipedia.org/wiki/Kabaddi

Post a reply to this message