detect if a path is drawn in pencil on paper

Thread Starter

-live wire-

Joined Dec 22, 2017
922
Hi what's up. So basically my goal is to detect if someone drew a path on a piece of paper between two points in pencil. Although the graphite is conductive, unless you press down really hard, the max length I want to be able to draw appears to have a resistance in the 100s of M Ohm to GOhm range.

I had to make the pencil line significantly thicker between the two points (about 6 inches apart) to even get a resistance reading on the multimeter of 100MOhm.

Also when applying a voltage of 30v and measuring it in series with the pencil line, even though 30v should still be measured on the output theoretically, I only measured 1V (I suspect due to the internal resistance of the multimeter, which is usually negligible but not with such a massive resistance of the pencil line).

So suppose point A is connected to V+, point a to point b thru the pencil line drawn, point b to some circuit, and that circuit connects to ground.

The goal will be to integrate this with an arduino. However an arduino does not seem like it can be directly used as pins cannot be left floating and the internal resistances might be enough to get in the way.

So my idea is basically I gotta make some circuit that can tell when points A and B are connected thru this high resistance path. However since the resistance is so large accounting for the resistance of everything and noise becomes important.

So is there maybe an op Amp or 555 circuit for this? Or maybe the arduino adc could work?

Whats everyone's ideas?
 

djsfantasi

Joined Apr 11, 2010
7,860
My first thought, which may not provide you with a complete answer, is why your approach leaves the inputs floating?

The Arduino has a state for inputs, where they are pulled up internally. The other end of the measured circuit would be at ground.

Or you can add a pull down resistor and the input won’t be floating.

As you imagine you’re solution, are the inputs digital or analog!

I’ve generally described yiur options. If you find any of the listed options interesting, please reply with additional details. I’ll describe how to use the Arduino.
 

DickCappels

Joined Aug 21, 2008
7,829
You might find that your noise problems do not materialize. If they do, a simple single pole low pass filter may help.

1631426173697.png
I don't know whether accuracy in the resistance measurement is important, but with nothing to go on I assumed a resistive divider would do the job for you. R1 is a dropping resistor you select, R2 is the resistance of your pencil line. C1 filters out the noise if there is any. Find C1 experimentally because not only would a certain amount of noise attenuation be desired, but response time is probably important too.
 

Ian0

Joined Aug 7, 2020
3,459
Is it your choice of pencil? If so, does a 6B pencil work better?
The longer you can integrate the current the less of a problem noise would be, but if C has to be electrolytic its leakage would spoil it.
Try buffering it with a specialist op-amp - there are a lot of bits and pieces connected to the a/d input of a microcontroller - all of which might contribute noise at high input impedances - such as FETs that switch pull-up resistors and the hardware for the digital input and output.
 

Reloadron

Joined Jan 15, 2015
6,114
A Google of "conductive pen for circuit board" or just conductive pen will get you a few dozen pens which are used for repair of traces on circuit boards. Drawing a simple line of about 1.0" on plain white paper using a #2 pencil I get about 118K Ohms. Provide more detail as ro exactly what the project is about and pretty sure you will get better help. As was asked, analog or digital input to your Arduino?

Ron
 

djsfantasi

Joined Apr 11, 2010
7,860
Depending on whether you’re using an analog or digital input on the Arduino, some of these comments are inappropriate.

So we need to know.

Also, you haven’t mentioned, but the characteristics of the pencil are important, too. What’s the hardness of the lead? How sharp is the pencil?

And the paper characteristics are important, too. What’s the paper thickness (it affects the width of the line)? What’s the paper finish, matte, semi-gloss or gloss (affects the graphite continuity of the line)?

Proper selection of these parameters will affect the success of your design. Some projects may not need this level of detail and for others they are critical.

You may not need to be so specific, but if you are it makes it easier for people helping you to make recommendations that are appropriate and repeatable.
 

MisterBill2

Joined Jan 23, 2018
9,463
Back about 1985 A client asked us about using an industrial process controller that did follow a line drawn on paper. It allowed programming temperature/time profiles for a number of different recipes. The goal was to duplicate a legacy system that had been in service for many years.
My finding was that the original device was not available new any more.
Additional research revealed that a much cheaper approach used a PC and an A/D card as well as a D/A card and allowed much better control and much longer cycles..
So yes, it is possible to track a line drawn on paper, but it is not normal paper nor is it a normal line.
 

MisterBill2

Joined Jan 23, 2018
9,463
Before going any farther with this we need to know the purpose of the whole idea. There are possible ways to do it with current technology, but without n explanation of the purpose and intent, nothing..
 

Thread Starter

-live wire-

Joined Dec 22, 2017
922
You might find that your noise problems do not materialize. If they do, a simple single pole low pass filter may help.

View attachment 247748
I don't know whether accuracy in the resistance measurement is important, but with nothing to go on I assumed a resistive divider would do the job for you. R1 is a dropping resistor you select, R2 is the resistance of your pencil line. C1 filters out the noise if there is any. Find C1 experimentally because not only would a certain amount of noise attenuation be desired, but response time is probably important too.
Okay thank you
 

Tonyr1084

Joined Sep 24, 2015
6,339
I won $20 from a guy when I bet him I could turn an ordinary pencil into a light dimmer. Shaved half the wood off and used paper clips. Sliding the clips closer together the light got brighter as the resistance dropped. Further apart the dimmer as resistance increased. The rig was a 12 volt wall wart and a 12 volt auto lightbulb. Small one. But the pencil got hot. Not hot enough to erupt in flames, but it did get hot to the touch.

I would imagine a pencil line could be a potential heater on paper and could start a fire. So be careful. As for your high ohm readings, that could be because of how you probed your pencil lines. Thin point contact will raise the resistance whereas greater contact surface area will result in a more accurate reading. Perhaps a piece of copper from a copper pipe flattened and pressed tightly on the paper might yield a more accurate reading. Surely, if my light dimmer had meg-ohm readings then the light bulb would not have lit. I suspect your high resistance measurements is a result of how you are making that contact. Staples or paper clips might yield some better results too.

Experiment.

Back in the early 80's I was trying to lube a motor using graphite lube. It was a 110V motor, and when the graphite created a conductive path between Line and Neutral there was a loud pop as the graphite blew out. Since it was in a metal frame there was no fire. But on paper? Could have been a more exciting learning experience for me.
 

MisterBill2

Joined Jan 23, 2018
9,463
Those sketches in post #11 describe the theory well enough but the reality of strain gauge sensing is a lot more complex if it needs to be even slightly close to accurate or even useful.
First, the change in resistance is a very small percentage of the total, so you will not see it with a common ohm meter. This means that the voltage delivered to an amplifier is just a very few millivolts, at best. In addition, the change is not linear over much range.

AND I am still wondering what the purpose of detecting if a line is drawn between two points is. No matter what, the effort to connect to a line at two points is going to be real. An optical means, or perhaps some variation of touch screen is more likely to provide a working solution.
 
Top