Fairly simple.

for parallel resistance, 1/r = 1/r1 + 1/r2 + 1/r3 + ...
for serial resistance, r = r1 + r2 + r3 + ...

 

Fairly simple.

for parallel resistance, 1/r = 1/r1 + 1/r2 + 1/r3 + ...
for serial resistance, r = r1 + r2 + r3 + ...

???

The recent posts are on how many keystrokes the example computation requires on various calculators.

 

This morning I was trying to recall where I learnt about RPN and math operations using a stack. I suspect was programming something in Assenbler. Timex Sinclair and the Z80? I cannot believe that is so hard to retrieve from my deep memory...!

 

Fairly simple.

for parallel resistance, 1/r = 1/r1 + 1/r2 + 1/r3 + ...
for serial resistance, r = r1 + r2 + r3 + ...

Have you looked at the title of the thread?

 

13 on the Casio fx83

I must be having a "dead brain day" just done it again and including turning on the calculator 22

 

So what were the 13? Seems a bit hard to imagine since 1.33, 39, and 17.4 account for 10 and I would think that there has to be at least one key stroke after each numeric value, so that's 13 right there.

This is like counting the steps up a lighthouse! can't get the number of strokes right, by dropping the Zero's and mentely moving the decimal point, it's either 23 or twenty four! my brain aches now!

 

This is like counting the steps up a lighthouse! can't get the number of strokes right, by dropping the Zero's and mentely moving the decimal point, it's either 23 or twenty four! my brain aches now!

I know the feeling. Counting them while I push them increases the chances that I will mess up what I'm doing. So I generally find that just writing them down works best.

 

I must be having a "dead brain day" just done it again and including turning on the calculator 22

I think we are discounting turning on the calculator, so let's call it 21.

 

I know the feeling. Counting them while I push them increases the chances that I will mess up what I'm doing. So I generally find that just writing them down works best.

Have you ever tried the Stroop effect by reading colors' names painted in a different color?

 

What about it?

Sorry I was not clear. I was just trying to point out that if you wanted a 'non standard' resistance value, you could design one from standard value resistors by using the reciprocal of the difference of the reciprocals between a standard value (a higher value) and the one you want.

 

Sorry I was not clear. I was just trying to point out that if you wanted a 'non standard' resistance value, you could design one from standard value resistors by using the reciprocal of the difference of the reciprocals between a standard value (a higher value) and the one you want.

Mind showing an example?

 

you could design one from standard value resistors by using the reciprocal of the difference of the reciprocals between a standard value (a higher value) and the one you want.

Yes -> that's explicitly governed by the formula for the parallel resistance.

 

Mind showing an example?

Another approach is to recognize that a given resistance is also a given conductance and conductances in parallel add. So just like it is pretty easy for most people to find a combination of series resistors that combine to give something close to a desired resistance value, it becomes just as easy to find a combination if parallel resistors that combine to give something close to a designed conductance value.

 

Mind showing an example?

Example:


You have a LM317 adjustable voltage regulator and desire an output voltage of 3.3V and you are fresh out of variable resistors.


The fixed voltage between the Voutput and the Vadj pin is 1.25v. If you use a 240 ohm resistor between Vout and Vadj, the current avaiable to R1 will be 1.25v/240ohms (5.16 ma)


If Vout is to be 3.3v, then 2.05V will have to be dropped across R2.


The value of R2 needs to be 379 ohms. (Halfway between the 5% values of 360 ohms and 390 ohms)


At this point you could go for it and use the 5k pot and go crazy trying to adjust the thing to 3.3V, but you could also use a 2.5k in parallel with 470 ohms for R2.


I use open office (free) It allows doing lots of 'what ifs'. Just 'plug in' standard values until the result yeilds something close to a standard value


=(B12^-1-B14^-1)^-1 = 2573.33 ohms (B12 = 397.4 ohms and B14 = 470 ohms.)


The value to put in parallel with 470 ohms to make a 379 ohm resistor is 2.5K.


If you really wanted to, you could use a small value pot to 'tweek' the output voltage even closer, if you wanted to.

 

What about the difference of reciprocals?

This is a trivial example, but:

Say you need a 200 ohm resistor, and have the standard 220 ohm part.

1/200-1/220 =4.545e-4

= 1/0.0004545 = 2.2K

220 // 2.2K = 200 ohms.

Hi,

Yes that is right, and not a bad idea to bring up in this thread because sometimes we have to calculate a second resistor that when put in parallel with the first will result in the desired resistance for the circuit we are working with.
This idea comes from the same formula (which i like the way Wendy wrote it so i will follow that form here):
1/RT=1/R1+1/R2

and then solve that for 1/R2:
1/RT-1/R1=1/R2

or:
1/R2=1/RT-1/R1

and we have the required formula.

If we want this to get even a little more interesting, if we need a very precise value we might consider using two resistors in parallel to the first, of which the second and third differ by a factor of 10 or so. We then have simply:
1/RTa=1/R1+1/R2 (course adjust)
1/RT=1/RTa+1/R3 (fine adjust)

I had to parallel three resistors a few times in the past because i didnt have two values that resulted in a value that came out close enough to the required value for the circuit and i did not want to use a potentiometer. Resistor #2 gets you close, then resistor #3 gets you closer.

While we are on the subject, it is also interesting to look at ways to parallel a resistor and a potentiometer. Keeping the pot fully functional, it doesnt work out so neatly unless the resistance change doesnt have to be too great. For the pot just partly functional (as a single variable resistor) it works better.