# How to give specific range of resistances?

Discussion in 'Electronic Design' started by Michael, May 6, 2007.

1. ### MichaelGuest

Hi there - I need some odd valued potentiometers. I have two different
setups I need:

one that ranges from 178.2 - 207.1 ohms

and

one that ranges from 96.3 to 168.6 ohms

So - my first inclination was to just use a single resistor for the
smaller value and a potentiometer to cover the range. So for the first
I'd need a 178.2 ohm resistor, and a 28.9 ohm pot. However - 28.9 ohm
pots seem to be slightly less common than I would like. Same goes for
the second setup and the desired 72.3 ohm pot. Now - I can fudge
around on these values a bit - but not too much. These values are
setting the output voltage of a light, and if that voltage goes too
high I suspect the magic smoke of the device will be released.

So - any suggestions? I don't need very linear operation from the
potentiometer, if that matters. I don't think many pots have
mechanical stops to stop you from setting the value too high. What do
I do?

Thanks!

-Michael

2. ### John LarkinGuest

Why do you need such precise resistances? What does turning the pot
control?

You can do a series-parallel pad on a pot to force the limits, but
it's mathematically a pain.

John

3. ### CharlesGuest

resistance) OK? If so, find delta R and then pad a higher variable device
to that value using the familiar Req equation. Then, add a series resistor
to establish Rmin.

4. ### MichaelGuest

It doesn't need that level of precision - but it needs to be fairly
close (+-10%).

What is this series parallel pad you're referring to? Would that just
be putting a resistor in parallel with the pot?

-Michael

5. ### MichaelGuest

Can you give an example of the schematic that you're describing? (and
how to find the values for the resistors)I guess I don't entirely

Thanks,

-Michael

6. ### MichaelGuest

Err, is what you're saying that you'd just have a single resistor in
series = rmin, then a resistor in series with a pot whose max parallel
resistance would give rmax - rmin? Are there any better ways?

Thanks,

-Michael

7. ### CharlesGuest

Close. The pad is in parallel with the rheostat and decreases its Rmax to
delta R.

Better? Hard to tell as I am still not sure what you would like to
accomplish.

8. ### CharlesGuest

50 ohm rheostat shunted by a 69 ohm resistor gives an Rmax of 29 ohms. Rmin
is obvious.
Add a series 178 ohm resistor and you'll have 178 to 207!

9. ### John LarkinGuest

------+--------r1------------+--------
| |
| |
| |
+-----r2----pot--------+

As I said, mathematically a nuisance. If it's an opamp circuit, there
are usually cleaner ways to do this.

John

10. ### Spehro PefhanyGuest

o
|
|
+----+---+
| | |
.-. .-. |
Rp | | | |<-+
| | | |
'-' '-' Pot varies from Rh ~ Rl (measured)
| | with wiper tied to element as shown.
+----
|
.-.
| |
Rs | |
'-'
|
|
o

Say you want a resistance that varies from Rmin to Rmax, and you have
a pot (of greater nominal value than Rmax - Rmin, of course) that
varies from Rh to Rl.

You put a resistor in parallel Rp = 1/(1/(Rmax - Rmin) - 1/(Rh- Rl))

And a resistor in series of

Rs = Rmin - 1/(1/Rl + 1/Rp) ~= Rmin - Rl ~= Rmin

Eg, you have a pot that varies from 3 to 207 ohms and you want to go
from 70 ohms to 170 ohms.

Rp = 1/(1/100 - 1/204) ~= 196 ohms

Rs = 70 - 2.95 ~= 67.0 ohms

The resulting rheostat will be somewhat nonlinear depending on how
much you shunt down the element value. Don't use a higher pot value
than necessary.

You can use nominal rather than measured values for the pot, and
assume Rl = 0, if you want to simplify things, but typical tolerance
on pot elements is fairly wide, and minimum resistance may be
significant if you're going for values in the hundreds of ohms.

Best regards,
Spehro Pefhany

11. ### whit3rdGuest

The range is the specification of a two-terminal device (like
a potentiometer with wiper connected to one fixed terminal),
i.e. a rheostat.

Consider a parallel combination of fixed resistor R1 and
(series-combination R2 and Rvariable);

R1//R2 = 178.2 Ohms,
R1//(R2 + Rvariable) = 207.1 Ohms

is a pair of equations in two unknowns (R1 and R2) and
Rvariable :== 100 Ohms is a free chosen quantity.
That freedom is required because pots aren't available in
odd values.

Solve for R1 and R2 and if they aren't negative, that's
a good solution. Otherwise, go to the next standard Rvariable
value.

As others have noted, protection against overvoltage/overcurrent
usually uses other techniques- a current limit could be
a LM317 with program resistor, or a voltage limit could be
a TL431 with program potentiometer. Rheostats are kinda...
old-school. Sometimes that's good.

12. ### Tom BruhnsGuest

As an alternative to adding a shunt (parallel) resistor, just put
mechanical stops on the pot.

If your tolerances are 10%, it's a bit silly to be giving resistor
values to 0.1% --

One might suspect there's a better way to do things. Can you tell us

One thing to beware of with potentiometers is that the wipers
sometimes fail to make good contact with the element. In that case,
the resistance goes to "infinity" or at least some large value. It
can be worthwhile when you want a rheostat function (a variable single
resistance) and you have a potentiometer (a resistance with a variable
tap point) to tie the wiper and one end together. Then the resistance
will normally only jump to the end-to-end value if the wiper fails to
make contact properly.

Beware about DC current through the wiper. Some pots handle that
gracefully, and others do not. Wire-wound pots are probably OK.

Cheers,
Tom