Go Down

Topic: making a potentiometer with paper? (Read 1 time) previous topic - next topic

utkarsh

Apr 07, 2013, 02:50 pm Last Edit: Apr 10, 2013, 01:13 am by utkarsh Reason: 1
Hi! I have been trying to make a potentiometer with a paper, but it doesn't work. I took a strip of paper, attached two alligator clips to the two ends, connected them to 5V and ground respectively, and used a wire connected to analog input as the probe. But all it did was detect contact with the paper, and the read value did not change with position. I suspected that the resistance might be too high, so I blackened it with a pencil, and it worked. But even then, it was not really perfect. It worked fine for the values near the ground, but not for those near the 5v. For example, even the the wire was really close to the 5v clip, it only went up to about 900(when the read value should be near 1024). Is high resistance the cause of this too? And is it even possible to use non-blackened paper as a potentiometer? Also, if it is, how can I do it? If it is not possible, are there any alternatives? any precautions I should be taking?

fungus

#1
Apr 07, 2013, 03:03 pm Last Edit: Apr 07, 2013, 03:05 pm by fungus Reason: 1

i suspected that the resistance might be too high



You were correct.
No, I don't answer questions sent in private messages (but I do accept thank-you notes...)

i_luv_arduino

Yes try making paper bigger thus causing less resistance but it is easier just to buy a potentiometer if your trying to complete a project good luck and tell us how you did it in the end! :)

Shpaget

900 is pretty close to 1023, considering the technology involved.

utkarsh

#4
Apr 07, 2013, 05:57 pm Last Edit: Apr 10, 2013, 01:15 am by utkarsh Reason: 1
Thanks for the replies! I am making a pen tablet from scratch, and since I am a bit short on electronic components(I am pretty much the only electronics hobbyist in about a 50 mile radius), I need to keep it as simple as possible. Is there any other alternative method I could use?

For resolution, I am thinking of dividing the whole area into parts and measuring each of them individually. But that would require an lot of sensing and digital writes. Would the Arduino be able to track the whole process fast enough for it to be a smooth experience(each sample would take about 12 digitalWrites and 3 analogReads)?

Would the paper not hinder such quick switching of electric potential across it?

Please reply.Any help appreciated.

Boardburner2

#5
Apr 07, 2013, 10:20 pm Last Edit: Apr 07, 2013, 10:42 pm by Boardburner2 Reason: 1
In uni days I recall an experiment In electrical potential that used typing carbon paper and conductive paint.

is high resistance the cause of this too?

Do you know the impedance of your measuring probe ?
Can you work out the impedance of what you are measuring ?

If you can answer those , have a look at kirschofs law.

utkarsh

#6
Apr 08, 2013, 01:58 pm Last Edit: Apr 10, 2013, 01:17 am by utkarsh Reason: 1
I really cannot say what it is about,but I googled it and found a few papers on it. It seems that the experiment was used to study flow of electric charges through different conductors and the conductive paint was used as a sample(and assumed that it was a cross-section of the desired 3D object. So for a cylinder,it would be a circle). Did you mean that?
I do not know much about impedance either, but wikipedia says that it is a bit like resistance, and I am using a plain wire for the probe, so it should be zero for my probe.
How would Kirchoff's law help here? Please reply.

Boardburner2

#7
Apr 08, 2013, 03:12 pm Last Edit: Apr 08, 2013, 04:43 pm by Boardburner2 Reason: 1
Ok

Basically if the impedance of the measuring probe is say 1000 times the impedance of the thing being measured you will get a good reading.
As the ratio decreases the readings will become increasingly erroneous.
They can be mathematically corrected however using kirchoffs law.

The reason for suggesting carbon paper is that it is manufactured to closer tolerances than you can paint and should give a more linear performance.

The impedance of the measuring probe depends on the instrument, for a DMM the maker will specify it.
If you are using the arduino ADC then look in the data sheet.

Boardburner2

Example

Assume the potentiometer you have made is 10k ohms, also assume the ADC has an input impedance of 10k ohms also.

A measurement near 0 volts will be quite accurate but a measurement at 5v will under read by 50%.

Your measurments reading 900 and not the expected 1024 indicate this may be what is happening.

You can work with that, just do not expect position/voltage to be a linear relationship.

The above example presuposses a constant current supply to the pot which is not the case in your circuit and complicates things but ignore for the purpose of the example.

A possible solution may be to draw a graph of position against measured voltage and derive a lookup table to perform corrections.

utkarsh

#9
Apr 09, 2013, 05:17 am Last Edit: Apr 10, 2013, 01:19 am by utkarsh Reason: 1
Thanks for the replies!
I measured it, and for the piece of paper paper I was using(the blackened one), it came out to be near 15k. The intended impedance for inputs of arduino ADC is 10k or less. So that might be the source of error.
Would it help if I could reverse the polarity and and take a reading then? For example, if the relation was (reading)=(position/total width)*(1024)*(1-e), where e is a constant, I could just add the readings of original polarity and the reversed polarity,and divide it by 1024 to get (1-e). Dividing the original reading with (1-e), I would get the correct reading(but this example is linear too =( ). For values near the extremities, it might help me get the values. Also, what kind of relationship is it going to be?

Boardburner2

Draw a graph, if the extremities are strange just don't use that portion of the pot.

KeithRB

My old Gnome synthesizer used a vinyl strip for a resistive "keyboard", you might try that.

http://www.paia.com/manuals/docs/GNOME-howto-article.PDF

KeithRB

On second thought, that was special stuff, so you might look at getting some of the anti-static bag material.

fungus


On second thought, that was special stuff, so you might look at getting some of the anti-static bag material.


I thought that too but I tried to measure the resistance of a bag and it didn't seem to conduct at all.

Maybe I should try a different bag...

No, I don't answer questions sent in private messages (but I do accept thank-you notes...)

Boardburner2

I suggest you do not waste too much time on this.
It's not supposed to be conductive.
Although it is useful for static dissipation it's primary goal is to reduce diffusion/permeation of gases and moisture.
Helium is very mobile in this respect which is why helium balloons have a heavy coat of metal.
They may be conductive though I do not know.

Go Up