# Integrating Gefran Linear Potentiometer PY1 with Beaglebone Black

simulate this circuit – Schematic created using CircuitLab

I want to connect one Gefran Linear Potentiometer with Beaglebone Black for displacement measurement purpose. The signal conditioner of the Lin. Pot. gives and output of 0-10 V. I am using a simple voltage divider circuit of 8.2K 0.1% resistor and 1K 0.1% resistor. Now I can either feed the voltage across 1K resistor to the analog input pin of beaglebone. Otherwise, as I understand, I can connect an external ADC to the output of the 1K resistor and then the ADC can be connected to Beaglebone Black. In the second case, I do not lose out in resolution provided I use a 24bit ADC. Beaglbone Black's internal ADC has a lower resolution. Which ADC should I use for this application? Can anybody help me with the circuit diagram for connecting the ADC with beaglebone and the voltage divider circuit?

• A schematics would be apreciated, and you don't need 24bit adc, because the voltage reference in gefran (I used them) are not super trooper, they are for industrial use 1%. – Marko Buršič May 23 '16 at 10:51
• Hi Marko, I have provided a block diagram. Hope this helps. My application is of monitoring load vs distance travel for a pressing application. That is why, for accurate monitoring, higher resolution of ADC is required. – Ritjeet May 23 '16 at 11:14

I hve made dozens of force vs. displacement meausurings, so I will explain in short form: The resoultion of the curve doesn't really matter, what really matters is the shape, so don't make yourself a trouble adding an external ADC for the first time. The Gefran potentiometer is very poor, I have rapleced them everywhere, but OK it's up to you. You have two possibilities:

simulate this circuit – Schematic created using CircuitLab

The other one:

simulate this circuit

You should sample with adc very fast and then implement low pass filter. When force Vs. displacement changes, $d=\sqrt{v_x+v_y}$ you store the point in array buffer, since sqrt is time wasting (probably yur MCU can cope with that)is enough to look a threshold difference in both channels.

• Hi Marko, thank you so much for your kind help. BTW, in the second circuit diagram, is the buffering of Vref really necessary? Has it been done with the purpose of impedance matching? Also, what exactly is transil ? (as you mentioned along with the diodes.) – Ritjeet May 24 '16 at 5:16
• Ritjeet, buffering the reference is always suggested, and almost always necessary. It is necessary on the BBB. Also, the ADC on the BBB has a strict voltage limit of 1.8 volts, so protecting the input with Zener diodes probably won't be adequate. I will try to post a opamp circuit that I have used on the BBB. – Mark May 24 '16 at 5:19
• Oops. Nevermind. Since the top of the pot is driven by Vref, the ADC input can never exceed Vref, and the protection isn't necessary. – Mark May 24 '16 at 5:52
• True. Thank you so much for the inputs. As you said, you have already worked with many load vs distance measurements, I wanted to ask one more question. Do you use a similar circuit for feeding load cell/strain gauge output to Beaglebone also? I selected HX711 interface from Sparkfun for the load cell integration purpose with Beaglebone. As the load cell output is in milli volt range, now I feel I can do way with those additional circuits. – Ritjeet May 24 '16 at 6:02
• @Ritjeet I used industrial equimpent, no Beagleboard. You need adequate load cell amplifier/ADC. That is more complicated than potentiometer. – Marko Buršič May 24 '16 at 7:18