Potentiometer problems

Status
Not open for further replies.

be80be

Well-Known Member
Potentiometer and ADC I been playing with some joy sticks ok here's the problem four sticks is 8 pots.
You hook them up to the power rails what do you have in my case a 125 ohm pots is what they ended up looking like. Why because they are basically in parallel. Then I ran into if I change one pot it messes up the value of the rest.
Whats the best way to over come this.?
 
Put them in series?

What are you trying to do? Do you need to uniquely ID each pot? Circuit?

John
 
LOL ok I'm reading the pots on joysticks.. But there are a bunch 8 of them and any way you look at it just hooking up 8 to vdd and gnd there paralleled.
But that's not the big problem. It's when you move one it changes the value of the rest. So I have 8 pots hooked to 8 adc pins I move one and all them change a bit.
 
That's because the 8 ADC's on a micro are all multiplexed into one.. You need to increase the reading time.. If the hold/ sample time is misreporting, you are not leaving enough time for it to settle What micro are you using???
 
I never thought this to be a problem I have 10 mS between readings so that's not the problem.

 
1) If the source is able to maintain constant voltage, then the paralleled pots should work independently. If the pots are 10K each, eight in parallel should be 1.25K, not 125Ω.
2) Rather than using Vcc and Vss could you use addition pins to supply and sink the potentiometer currents. For example, in a simple Charlieplexing scheme, imagine the switches would be replaced by pots.



You would need 4 "spare" pins to power your 8 pots, but then you could also expand to more pots with no more pins needed for power. Even without using a high Z state for some of the pins, a scheme like Charlieplexing would reduce the number of pots that were powered at any one time.

John
 
The first joystick had 1k Ok it's 125 ohms. Drawing I posted with the 10 k is 1.5 k I know that. But that's not the problem the problem is some how moving the stick so it changes just one pot changes all pots.
I think what happening is it somehow shorts vdd to ground.
 
Last edited:
Kirchhoff's voltage law says that if the rails are at constant voltage, then changing the current from one pot, should not affect the current through the other pots. Which voltage regulator are you using? Maybe a newer model VR with better regulation from step changes would help? I have used the LP2989 (TI) because it is specifically spec'd for ceramic capacitors. A step load change of 500 mA causes an output change of 60mV, which recovers within 10mS. There is no way the pots you describe are going to show that much change in current.

Was your model circuit on a breadboard?

John
 
It's a joystick from a playstation 2 But it has the 10k pots the first one had 1 k pots it was from a toy. It did short some how. And the total was 125 ohms.
But I'm using the better one from the playstation I added a 470 uF across the power on the board and the 10k from ground plus the 1 k to the adc It works better. It maybe the clamping diodes on the atmega328p. But you can short the pots without adding the ground resistor and the one on the adc .
 
Last edited:
Why is there a 10k common resistor to ground???? That would cause your issue...
 
The 10k to ground is almost certainly the cause, the input impedance of the pic is around 10k, so every time you move a pot your changing the load resistance, and with multiple paralleled 10k inputs the load is less than the source imp.
Link out the 10k to ground and see how it goes.
 
be80be said:
I never thought this to be a problem I have 10 mS between readings so that's not the problem.
Click to expand...

10mS could VERY easily be the problem, depending on the source impedance provided to the inputs (and joysticks are fairly high impedance).

But as Ian says, why the common 10K?

You might also like to check my joystick tutorial, which doesn't need analogue inputs (just as PC ones didn't)
 
I thought 10 k was to high myself I changed it to 1 k. The pots was able to short without some resistance.
 
be80be said:
I thought 10 k was to high myself I changed it to 1 k. The pots was able to short without some resistance.
Click to expand...
It shouldn't be there at all!!! With a resistor common to all the pots, you will always have issues..

Why is it there? The 10k pot coupled with a (before )10k would be more than the input impedance of the ADC.. However with the pot in situe, you have a non linear pot motion..

When you say "The pots can short" What do you mean? The pots should sit on the + and - rails..
 
That 10k to ground absolutely should NOT be there. No way. Those pots need an absolutely stable voltage to be putting out a valid level. They are designed to output a ratio of the power source - i.e. +V to GND. I've connected up to 32 pots in a similar circuit, and they work fine. They were all 10K pots. If you have a problem with the total current draw, then you need a higher current power source. If you are worried that the pots will overheat from the individual current in each pot, you need a lower voltage. And if that lower voltage is not enough to get a valid reading from the ADC, then you will need to individually amplify/buffer each pot. But I doubt any of that is necessary with typical devices. Just loose the 10K and you'll be fine.
 
Theres not a 10 k I changed it and the pots can short if you move the sick to the right place
 
be80be said:
Theres not a 10 k I changed it and the pots can short if you move the sick to the right place
Click to expand...

I've never heard of pots doing that?, nor can I imagine how it would be possible?.

When you say 'short' - short from where to where?.
 
If you move the sick to like the one o clock it was letting one pot be at ground and one is at positive 5 volts. I guess it's made the clamping diodes clamp one of the adc pins
 
But that was with the cheap joy stick I changed all that to the ones I got from the PlayStation they have 10 k pots and I added a 1 k from ground to the pots this seems to workout good. I read someone else had this happen must be the atmega 328 chips clamping diodes. It only happens if you are moving both pots.
But it does short it kicked the breaker on a 1 amp USB power supply got that nice beep beep lol and no power till I reset it
 
I use those "cheap" Playstation joystcks.... Unless you breach the voltage allowed on the ADC port the clamping diodes will be fine.... As the port is an input the extremities of a 10k pot cannot damage the micro.... Even if there are 8 in parallel..
 
Status
Not open for further replies.

Similar threads

  • Question
Digital potentiometer problem and questions
Replies
7
Views
2K
GH Crash
7
  • Question
problems getting M74HC259B1R to work
Replies
18
Views
2K
777xxx
7
C
  • Question
L293D MOTOR CONTROLLER PROBLEMS
Replies
9
Views
3K
tepalia02
T
J
  • Question
Where to add a potentiometer to vary input voltage?
Replies
15
Views
1K
Jim Kailey
J
C
Tektronix T932 oscilloscope problems
Replies
0
Views
1K
Cello
C
Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…