Resistor calculator to find parallel values to make x-ohms

[augustinetez]

I used to have calculator/spreadsheet where you enter a required value of resistor and it would calculate two (or more) parallel values to make up the value - anyone seen one anywhere?

Being a cheapskate here rather than going out and buying E96 resistors (I want to use whats on hand) and E96 in 2512 2W packages seem to be a bit thin on the ground from all the big suppliers anyway.

 

[crutschow]

Sounds like you want something like this app.

 

[augustinetez]

Well, it would be if it ran on a PC

So to amend the question - needs to run on a Windows PC

 

[augustinetez]

OK, after a lot of digging around, found one that does the job - Resistor CAD 2.0 - search using "Resistor CAD 2.0" (include the quote marks) and it comes up as the first result (not posting a direct link as it is on another forum).

 

[crutschow]

Attached is a simple, small Visual Basic app executable I wrote many years ago that calculates two 1% resistor values for a desired parallel resistance along with the % deviation from the desired value.

I'm running it Windows 10, but I've used it on several older additions of Windows, starting I think with XP.

Below is a screen shot of its window with the solution to a sample parallel resistance of 1234Ω.

On my machine, the Print button will generate a PDF file with the results.

 

[augustinetez]

Thanks.

It won't run as MSVBVM50.DLL is missing - I'll find a copy and register it on this machine and see if I can get it to run.

 

[crutschow]

If you are interested, I also have a similar app that calculates the best two 1% resistor values to give a desired resistor ratio or a desired attenuation.

This often determines values that are very close to the desired so the error in the actual value is due to the resistor tolerance.

 

[Nigel Goodwin]

I wrote a number of programs years ago, under DOS using Turbo Pascal, all of which used (or produced) standard component values.

For parallel resistors you simply entered the value you wanted, and it listed the best matches using standard values, and gave you the percentage error range using the tolerance of the resistors.

It (and the other programs) worked really well - no idea what happened to them all - I seem to recall I wrote them on a 386 DOS laptop? (I've no idea what happened to the laptop either?).

 

[Dick Cappels]

Once you understand the formula to find the resistance of parallel resistor you should be able to start to figure these things out in your head with pretty good accuracy.

Here is an example:
/Begin example



Kip Ingram
PhD in Electrical Engineering, The University of Texas at Austin Cockrell School of Engineering (Graduated 1992)Author has 9.1K answers and 2.4M answer views Mar 17

For parallel resistors you add the reciprocals. So, to get 30, given a 50, here is your equation:

1/30 = 1/50 + 1/x

which is easy to solve:

1/x = 1/30 - 1/50 = 20/1500

so x = 1500/20 = 75 ohms.

This process works for any number of parallel resistors.

Stay safe and well!
\end example.

 

[augustinetez]

Got crutschow's program to work, unfortunately it appears to be based on E96 series values for the calculation whereas I need it to work with E24 series values at the moment, Thanks anyway.

That one I posted in #4 does what I need and even lets me select which series of values to select the match from, so I can mix any from E6 thru E192.

you should be able to start to figure these things out in your head with pretty good accuracy.

My brain is not wired for maths, that's what computers are for , I was one of those at school that when told to find x -> there it is on the blackboard sir!

​

 

[AnalogKid]

rather than going out and buying E96 resistors

Note that while multiple parallel/series resistors can achieve almost any specific value by calculation, the overall tolerance always will be that of the individual parts. An expected value might be 2.731 K, but the tolerance with E-24 parts still will be +/-5%.

Things get a bit messier when you combine parts of different tolerances, but the worst tolerance part always degrades the overall tolerance of the group.

ak

 

[Pommie]

I'm in school tomorrow and daughters (babysitting) after but will have a go at making this a html page when I can so it works on any device.

Mike

 

[Pommie]

Note that while multiple parallel/series resistors can achieve almost any specific value by calculation, the overall tolerance always will be that of the individual parts. An expected value might be 2.731 K, but the tolerance with E-24 parts still will be +/-5%.

Things get a bit messier when you combine parts of different tolerances, but the worst tolerance part always degrades the overall tolerance of the group.

ak

I thought resistors were made and then sorted into tolerance ranges so 10% resistors were always 10% wrong so if you picked a high and low one you got close to actual value.

Mike.

 

[DrDoggy2]

The 1/r1 +1/r2 equation is easy to plug in to excel spread sheet:

cell a3 = 1/((1/a1)+(1/a2))

 

[Pommie]

The 1/r1 +1/r2 equation is easy to plug in to excel spread sheet:

cell a3 = 1/((1/a1)+(1/a2))

But then you have to put all possibilities in until you get a value that works.

Mike.

 

[crutschow]

Once you understand the formula to find the resistance of parallel resistor you should be able to start to figure these things out in your head with pretty good accuracy.

It's not calculating the parallel value that's the problem, it's determining which two resistors give a value closest to the arbitrary value you want.

If you look at my post #5 example, you will see many combinations that give close to the desired value, but some are much closer than others.

 

[Nigel Goodwin]

I thought resistors were made and then sorted into tolerance ranges so 10% resistors were always 10% wrong so if you picked a high and low one you got close to actual value.

Mike.

I've never thought so, and certainly over the decades I've had plenty of resistors very close to their specified values, and relatively few close to the tolerance limits.

I've always presumed they manufactured a specific value, then rejected any which were outside the specified range? - and certainly as time has gone by I imagine that manufacturing techniques have improved, so less and less are outside spec.

Same with transistor gains - you don't see many near the bottom tolerances.

 

[Nigel Goodwin]

It's not calculating the parallel value that's the problem, it's determining which two resistors give a value closest to the arbitrary value you want.

If you look at my post #5 example, you will see many combinations that give close to the desired value, but some are much closer than others.

That's pretty well what my old Pascal program did - but (as far as I can remember) I displayed the list in order of accuracy, not in order of the first resistor.

I would also suggest you knock a load of digits off the accuracy values (it's not needed), and get rid of the spurious scientific notation?.

But nice work

 

[DrDoggy2]

edit: oops

 

[crutschow]

I would also suggest you knock a load of digits off the accuracy values (it's not needed), and get rid of the spurious scientific notation?

I did that program probably at least 20 years ago, and have neither the tools nor the ability to now modify it.