0

Below is a floating signal source is connected to an instrumentation amplifier. The floating signal source stays in a metallic enclosure.

enter image description here

To make best out of it for this connection in terms of noise, the circuit should be balanced. Here is what I understand so far:

1-) The source impedance Rs1 must be found empirically so that we should add Rs2 in series to the negative terminal of the source. So I have to measure all source's output impedances and add proper resistors. Rs1 must be equal to Rs2.

2-) We should use twisted shielded pair. This will ensure that magnetic coupling to pairs be L1=L2 because of twisting. And the shield will protect the capacitive coupling of electric fields.

3-) The shield should be connected at one end to the earth. In this case at the receiver since the source is floating.

I'm planning to follow the above procedure. But what I don't know about the following:

1-) Should the point A be connected to the Metallic case?

2-) Should the AI GND be connected to the earth?

3-) Is 10Meg recommended value? I read it from just one document.

GNZ
  • 1,559
  • 4
  • 25
  • 52
  • How well balanced are the two lines as shown and the Rx does nothing to shunt stray EM noise as the impedance is too high. – Tony Stewart EE75 Jan 18 '18 at 21:49
  • I dont know Rs1 yet but I will measure it empirically and add Rs2 equal to Rs1. Please see this document they recommend 10Meg for Rx http://www.calex.com/pdf/4ground_shield.pdf I dont know why. – GNZ Jan 18 '18 at 21:52
  • You mean Rs1 does not need to be equal to Rs2? – GNZ Jan 18 '18 at 21:55
  • That 10 MEG could be any value from 10 K to 10 Meg. It is a *static* reference only. If your source is high-impedance you need to buffer it with a op-amp and a SSM2142 600 ohm line driver, or something equal. What are you measuring? –  Jan 18 '18 at 22:02
  • @Sparky256 I will use maybe 10 different transducers. Here three examples https://www.vaisala.com/sites/default/files/documents/PTB110-Datasheet-B210681EN-D.pdf and http://www.prelectronics.com/pdf/2202-1749-US.pdf and http://www.hbm.cz/wp-content/uploads/b0615.pdf – GNZ Jan 18 '18 at 22:30
  • Btw I think Tony means that since the Inamp has huge input impedance because of its buffers total impedance seen by the each source lines are around 10Meg level so source output impedance will have insignificant effect on the imbalance. Is that correct? – GNZ Jan 18 '18 at 22:31
  • But what I dont get is if the InAmps eliminates the need to impedance matching why do we need to use twisted cables? – GNZ Jan 18 '18 at 23:07
  • ok, so you'll never get the correct answer here again... so you can go with the idiots that know nothing about this. Since people want to downvote the correct awnser. – drtechno Jan 20 '18 at 00:10

0 Answers0