1
\$\begingroup\$

A floating DC voltage output sensor will be powered locally and the signal will be sent outdoors 200 meters far away to a data acquisition board. I didn't receive the sensor yet and don't have the datasheet. But my question will be about something else.

For such long transmission I was first thinking to use a differential line driver at the sensor output or convert the voltage signal to current and send it as current or send as digital ect ect.

But I can use a differential ended input data-acquisition board has the following architecture:

Omitting the voltage divider effect, if I use this module and data acquisition as in below diagram would that be a adequate to eliminate common mode noise issues or capacitive coupling noise issues?Or still a differential line driver or current conversion is needed?

to be updated soon...

\$\endgroup\$
  • \$\begingroup\$ Seems like overkill... \$\endgroup\$ – BeB00 May 31 '18 at 0:16
  • \$\begingroup\$ The approach where the isolation module sits upstream of the diff amp isn't the purest. But it may be good enough for environments with not too much interference. The manual mentions 8B input modules. Do you have this DAQ board nearby? What ICs are these modules made with? If you can post a photo of an actual 8B input module, that would help too. \$\endgroup\$ – Nick Alexeev May 31 '18 at 0:53
  • \$\begingroup\$ Questions you need to answer include Source impedance Spectral BW , ambient noise level and spectrum worst case and desired min. SNR network topology and Access to earth ground \$\endgroup\$ – Sunnyskyguy EE75 Jun 1 '18 at 4:21
1
\$\begingroup\$

Because I have built so many remote sensors I used a SSM2142 600 ohm Diff line driver, but it needs +/- 15 volts to drive a 1K diff load 200 meters away. It needs +/- 18 volts to drive a distant 600 ohm load. Not needed here.

schematic

simulate this circuit – Schematic created using CircuitLab

I used 600 ohm STP cable with 22 awg wire as a twisted pair. Gives excellent AC (to 50 KHZ) and DC performance. You may need an op-amp to drive the SSM2142 as it does not have offset and gain trim built in, and its input impedance is only 10 K ohm.

In the schematic the 1K load divides the signal by 2, so the +/- 10 volt drive signal becomes +/- 5 volts at the isolation modules.

I am assuming you have DC power options at the sensor, but you have not clarified that.

\$\endgroup\$
  • \$\begingroup\$ SSM2142 has Rout of 50 ohm, and will drive 0.16uF without oscillation. Thus lots of high frequency noise gets differentially shorted out, if a cap hands across + and - leads. \$\endgroup\$ – analogsystemsrf May 31 '18 at 3:26
  • \$\begingroup\$ @analogsystemsrf. The OP's diff isolated inputs have a voltage bandwidth of 1 KHZ, so local 60 HZ/120 HZ noise can show up as jitter. I think diff all the way with STP is still best bet for low jitter. \$\endgroup\$ – Sparky256 May 31 '18 at 3:34
1
\$\begingroup\$

Here's the problem as I see it and it's fundamental when sending analogue signals some distance over cable: -

enter image description here

The isolation module won't help - the only solution is to balance the impedance of the driving source - it doesn't need to be a differential driver but it does need to have a balanced impedance - this is why you use a resistor in both legs as well as STP cable. Added to this is the likely requirement to have balanced decoupling capacitors to ground at the input of the DAQ and a broken shield at the sending end i.e. only ground the shield at the DAQ end.

You should also use 10 kohm balancing resistors to ground at the DAQ end to keep the signal common-mode range within that required by the DAQ.

\$\endgroup\$
  • \$\begingroup\$ If it is a passive wheatstone bridge type pressure sensor, the resistor you drew called Rs can be assumed to be equally distributed on both outputs thus the output is inherently impedance balanced. However you need to power it and that power supply may not be impedance balanced to earth. The effect of this is that as you apply pressure to the sensor you may get an increase in differential noise. If your bandwidth is low this can be largely filtered out of course. \$\endgroup\$ – Andy aka May 31 '18 at 10:14
  • \$\begingroup\$ I can only recommend you contact the supplier and be prepared to add capacitors to ground at the input to your DAQ. It's all about how powerful the nasty noise source is and how that couples to your cable. If you can't impedance balance naturally you have to consider brute force capacitors to ground to act as nose suppressors. \$\endgroup\$ – Andy aka May 31 '18 at 10:23
  • \$\begingroup\$ Ratiometric generally means it's a passive bridge. You can only mitigate by adding capacitors as I have suggested. \$\endgroup\$ – Andy aka May 31 '18 at 10:33
  • \$\begingroup\$ It usually means that but read the data sheet to be sure. \$\endgroup\$ – Andy aka May 31 '18 at 12:26
0
\$\begingroup\$

Your inputs must be in the common mode range of the amplifier, so you do NOT want to isolate them. Ground one input at the receive end, and (if pickup is a problem) use a common-mode choke on the pair. Consider a shielded pair cable (ground the shield at one point) if that cost is acceptable.

This presumes that (as shown) the power at the sensor end is a battery, and floating (not grounded). It also presumes the wiring doesn't go outdoors and won't be struck by lightning.

\$\endgroup\$
  • \$\begingroup\$ @GenzoWakabayashi The amplifier is only an amplifier for signals in a limited range around ground; this is the common mode range, and SOME connection to ground is necessary to keep the common mode input condition satisfied. There is no 'sensor ground' intrinsic to the sensor as shown. \$\endgroup\$ – Whit3rd May 31 '18 at 7:59
0
\$\begingroup\$

The approach where the isolation module sits upstream of the diff amp isn't the purest, I would say. But it may be good enough.

The isolation modules may have some variation from one module to another, and that would show up as an offset after the diff amplifier. You can calibrate some of that variation, then subtract it from the data.

These 8B isolation modules have 1kHz bandwidth. They will let the common mode noise from mains AC through. The difference amplifier will remove the main AC noise.

Additional ideas

Place the diff amp upstream of the isolation module. Run single ended signal through one isolation module.

Convert the signal to 4-20mA at the sensor. Run the 4-20mA through the cable.

\$\endgroup\$
0
\$\begingroup\$

Problem has not been sufficiently defined in terms of interference. Your biggest challenge may be to reject CM induction noise over a wide spectrum from line f to AM/SW band where the modulation can get rectified in the receiver non-linearity. So it is the CMRR of the cable matching for each line and source E or B field that determines the noise level x CM impedance. Thus choice of STP cable or coax with Transfer Impedance characteristics that are analyzed to determine the optimal SNR over the signal and noise BW is the professional way to design a network. So it depends on your network ambient noise whether you use STP or double shielded 50/75 ohm coax both with CM PI filters

Current loops are often used but also diff. Voltage sources with impedance control such that CM impedance .

Until you define background E, B fields in uV/m or uA/m over the spectrum, we can only guess which is an adequate solution. For long cable telephones a hybrid transformer is used on paired lines. For base-band and sub-band and up, 75 Ohm coax with excellent Forward Transfer Impedance is used both with match impedance and possible equalization.

Another approach is modulated FM combined thru inverse tree networked combiners on shared coaxial cables using the lowest carrier necessary to give say 10:1 carrier deviation ratio and a huge boost in SNR from the deviation ratio.

Again, the professional way is analyze worst case SNR requirements and ambient spectrum the choose ideal DM impedance , carrier and CNR to achieve SNR requirements , then cable and transceivers.

Can you define your network topology and ambient noise and signal energy worst case?

\$\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.