0
\$\begingroup\$

I know the topic is immensely broad but I dont have the practical knowledge on the following issue:

Let's say I have a very long like 1km BNC cable carrying DC signal around 1V analog voltage signal coming from a transducer. And let's say the BNC goes all the way straight to an ADC with 100 Meg-ohm input impedance at the end. If the measurement should be very accurate, do we need to buffer this signal?

BNC cable even 1km long has very very low resistance comparing to ADC's huge input impedance. Theoretically form this point I would think one wouldn't need a buffer.

But how would that be implemented in practice for 1V DC level and 10mV DC?

\$\endgroup\$
  • \$\begingroup\$ Certainly I would first match the source(output)/load(input) to cable characteristics impedance. This has not so importance if you have a DC signal, but still...an low impedance buffer at source and amplifier at end is needed. \$\endgroup\$ – Marko Buršič Jun 14 '17 at 10:05
  • 1
    \$\begingroup\$ There is no such cable BNC. BNC is the connector, while the cable is coax with its characteristics impedance as parameter. \$\endgroup\$ – Marko Buršič Jun 14 '17 at 11:11
  • \$\begingroup\$ @MarkoBuršič Thanks for the correction I actually didn't know this and many of my friends ignorantly using this BNC term for the cable then. But if it is not too much for you could you add your answer how a DC signal got affected by in a 50 Ohm transmission line. How can I simulate this and see how DC is being affected? C and L are frequency dependent I cannot picture in my mind how DC voltage could be affected by L and C parameters of a coaxial cable. \$\endgroup\$ – atmnt Jun 14 '17 at 15:27
  • \$\begingroup\$ You didn't mention the transducer. How it is generating (sourcing) the signal, the 1km coax has pretty big capacitance, not sure if the transducer can drive the coax. The DAQ high input impedance and the long coax will act like an antenna collecting all the garbage signals from environment. You can try to mount the transducer on coax and you'll see what you will get. \$\endgroup\$ – Marko Buršič Jun 14 '17 at 16:59
1
\$\begingroup\$

texas instruments buffer example: link

enter image description here

Your source is the transducer's voltage and the 50 ohm should be replaced to 10M (for example). The output resistor 50 ohm matches the characteristics impedance of your cable (50 or 75, for example). At the end you should place the same resistor (50 or 75 ohm).

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
  • \$\begingroup\$ I have 3 questions: 1-) But isn't Rc=0 for DC voltages? Why would I need 50/75 Ohm termination? 2-) What is R1 for? OA1 already has a huge input impedance 3-) I still didnt get conceptually what would attenuate the DC signal without a buffer \$\endgroup\$ – atmnt Jun 14 '17 at 11:02
  • \$\begingroup\$ R1 is the matching resistor as you said the DAQ has 10M input, so it is basically the 10M input impedance to the amplifier. Not sure if it is adequate. With such large capacitance and impedance you will get just noise on the DAQ. The characteristics impedance is constant (L/C ratio of the transmission line), does not change with frequency. \$\endgroup\$ – Marko Buršič Jun 14 '17 at 11:08
0
\$\begingroup\$

Given your sensor Rout is low (1 ohm) and tolerates 1 mile of coax capacity --- at 100pF/meter, that is 170,000pF --- then you don't need a buffer.

If the coax shield is used by interferers as a return path, you will have errors in your measurement, because GND at one end of the shield will not be same as GND at other end of shield.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.