opamp output protection from high voltage spikes

Go To Last Post
8 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I have got a design where this 0-10V analogue output from an opamp is taken over a distance and then to an analog input card of a controller.

It is understood that there may be induced voltage spikes (as this is an industrial situation) along the wire over the distance.

What kind of protection could I use on the side of the opamp ouput to make sure no spurious spikes can damage the opamp (output side)? Any recommendations? I can think of zener diod? I could probably have also a resistor. Inductors?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
over a distance

Perhaps a litle more information would be helpful.
1 meter vs 100 meters...

Ideally one would have the ADC located adjacient to the sensor, as this eliminates many issues. (One of the main ones being no noise gained in the digital link.)

Presumably you have a good reason for not doing this.

A zener alone doesn't do much, they are optimally used with an in-line resistor. This, with the impedance of the connecting cable, and the ADC input circuitry's input impedance, impact your reading.

Does the ADC have it's own op-amp buffer at its end?

Obviously you will want to use a shielded cable to send your signal from one board to the other, and ground it at one end.

There are prior threads discussing shields, and whether they should float, have a high Z connection to ground, be driven, be grounded at one end, or be grounded at both ends.

It is amazing, sometimes, how "simple" design considerations can be rather complex issues if one wished to analyze them in depth.

JC

What is the frequency response needed? i.e. Essentially a slowly changing DC signal, or AC?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

unfortunately the location of the hand built transducer (opamp ouptut) and the analog card of the controller are in two different locations. They are atleast 30meters aways. I think the practice is to use two seperate cables without any shielding. Dont ask me why? I have seen them do this. They just use the same wiring type for all signals (analog/status/IO etc).

Frequency is not an issue. All I am doing is monitoring voltage from 110VDC battery bank. I am thinking have a 3 diode solution. See diagram attached.

I dont think resistor is necessary in series with zener diodes... why would it be?

Attachment(s): 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

A current loop is far more tolerant to external noise ...

Ross McKenzie ValuSoft Melbourne Australia

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

When wired as a shunt regulator the resistor dissapates much of the energy. Without it the zener has to dissapate all of the energy.

Diagram is from Wiki Zener Diode.

Normal diodes clamping to V+ and Gnd would hold the line to ~ V+ + 0.6 V, and Gnd - 0.6 V.

An output resistor would then limit the output current when the output in in clamp mode.

30 meters is a long distance to carry an analog signal.

Assuming you can't redo the entire system, I would suggest a twisted pair within a shield for the cable.

It will be interesting to hear what others suggest.

JC

Edit: With the normal diodes clamping to V+ and Gnd you don't need a 10 V (or 12 V) zener clamping the line with respect to ground.

Attachment(s): 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Rather than zeners, I'd be using TVS diodes and I don't see the reason for the upper zener. Therefore, if you have a fault (transient etc), the TVS diode conducts and shunts the voltage to gnd. What limits this current? The impedance of the source of course! So if someone applies 24VDC, then there will most likely be smoke from the TVS diode.
What TVS diode?

Maybe a SA12.0A

What do you want the schottky diode to do? I dare say it may not assist in normal operation.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

luvocean1,
A bit of clarification please:
A. How many wires in the cable? What is their function?

B. What is the input impedance and termination of the signal at the "analog input card's" end of the cable?

C. Is your "input analog card" single-ended or differential?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

luvocean1,
A bit of clarification please:
A. How many wires in the cable? What is their function?

B. What is the input impedance and termination of the signal at the "analog input card's" end of the cable?

C. Is your "input analog card" single-ended or differential?