designing with pots

pot

gain

from

and

If you want to do something like that, I would suggest a 10 turn linear precision pot, not a trim pot. There are ten turn precision dials available for these pots as well. These dials have a resolution of one part in 1000 or

On a sunny day (Sat, 27 Feb 2010 09:44:37 -0600) it happened "mook johnson" wrote in :

I do not think it is right to depend on the linearity of trimpots, or even normal pots, so as to add a scale perhaps? Trimmers should be designed to be somewhere midrange at the correct point, taking into account tolerances of at least up to 20% of the pot resistor value, and also providing limiting resistors if needed at both ends to prevent overload, or whatever weird gain situation. Also taking into account should be that the wipers may sometimes, due to many reasons, mechanical wear, oxidation for example, fail to make contact 100% all of the time. This is especially important in power supply applications, and then you need a resistor to hold that contact to a safe voltage, or design the circuit so that an open pot causes no damage, There exist expensive ten turns that you can put a scale on that are accurate and of better quality.

pot

in

om

nd

to

If you use a panel mounted 10 turn pot to get the needed linearity, beware of the capacitance issue. You will be putting capacitance onto the inverting node of the op-amp.

I suggest a good quality 10 turn panel mounted pot, if the frequencies are low enough that you can stand the weird bandwidth variations as you adjust the gain. At high frequencies, you may be stuck with a trim pot to keep the capacitance low.

You are dating yourself, John. :-)

I have doubts that electronic education these days even mentions the term. ( I know they do, but ask any recent grad, and they are like "What's a rheostat?". Much less the difference in utilization.

We always allow 1/2 turn from the ends with a 10 turn pot.

with a 20 turn, that should be a whole turn..

This factor is based on how much you want to pay for the pot..

It is used like a rheostat. it is the RF of a standard inverting opamp circuit.

The pot is this guy

This is exactly the type of answer I was looking for. Thanks.

Then you would need to hand plot the linearity curve over turns/position situation.

Multiturn trimpots aren't really any better than singles, but are more expensive and a nuisance to adjust.

John

Hmm. we pay on an average of 78.. 180 bucks for a 10 turn pot that isn't anything special looking how ever, the precision of movement is there which is where the cost factor is.

The long life one's for servos with large bodies with that kind of precision, we pay more.

You could use a $8.. $20 T10 pot that shows the same spec's in most regards but do not come close to maintaining position of return point, that much we have found.

Then there are those that are just simply over priced for no apparent reason other then paying for the expensive union help (in the corporate eyes) and filling the pockets of the greedy business owners.

Greed is the source of all evil and normally the down full of businesses that we all knew and used to love.. And those that are following quickly behind...

You have a choice. If you don't like the price, don't buy it.

John

Methinks the abrupt change is due to the mechanical (making it electrical) connecting "tab" or wrap-around used for outgoing leads. Whatever you see in terms of partial turn(s) for the constant resistance just add (say) one tenth of that range. BTW, have you tried ONE turn pots?

On a sunny day (Sat, 27 Feb 2010 17:57:45 -0800) it happened John Larkin wrote in :

pot

The other choice is to look for a digital solution, like encoders, up down buttons, or keypad, and display.

No, I'm trying to make the adjustment low sensitivity in terms of gain change per turn. A 1 turn pot would be a nightmare in this circuit. :)

Right now if the pot operating in the linear range. There is a +/- 1/4 turn tolerance on the setting. Most techs can get in that ball park reliably. The problem I'm having is when it is at the extreme low end of the pot and the resistance "fall off a cliff" as the wiper reaches the end. On the first board the setting for .5 gain was fine. On the next board, the setting for .5 was right on the edge of the cliff. just blow on the pot and it would change gain. Admittidly I designed for 1/4 turn in from form the ends which in hindsight was a bad idea.

Use two then.

One for coarse adjust, and another for fine adjust. D'oh!

pot

gain

from

and

Yup. A couple of dollars worth of rotary encoder or mdac can replace a few hundred dollars worth of precision pot.

I know one company that still buys the concentric-dial kelvin-varley voltage dividers, the ones with switches on the outer rings of a knob and a 10-turn pot in the middle. They go for aroubd $2K these days.

John

I've done dip switch + pot, coarse and fine.

Most multiturn trimpots aren't better than singles for setability. They have cheap, sloppy screw drives. The multis are also more inclined to jump value when mechanically shocked.

But the best pot is no pot; use software calibrations, stored in eeprom, and automate calibrations.

John

Thats the next step. This is a all analog solution. One PIC and I could elimiate the pot and the tech to tweak it.