Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations waross on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

Question about using Cat5 cable to wire PID-tech sensor and DAQ 1

Status
Not open for further replies.

zeened

Computer
Dec 13, 2004
4
0
0
CA
Hello,

This is my first posting to this forum so I appologize if I am not observing any conventions.

I want to acquire data from a PID-tech (photoionization) sensor from Mocon. My plan is to use cat5 cable to connect the sensor to a DAQ. I am concerned about line noise interfering with the signal. The nominal range of the sensor's output voltage is 0 to 1 Volt. The sensors background signal noise does not exceed 30 uV. The manufacture suggests the electronics should have an input noise significantly less than 30 uV.

Is cat5 cable an acceptable choice? How long could the wire be between DAQ and sensor? If a large distance is required (>100m) what are some possible solutions?

Thank you in advance for your suggestions.
 
Replies continue below

Recommended for you

Personally, I would use shielded twisted-pair, with the shield being grounded only at one end. Twisted pair minimizes the loop area, thus reduces EMI pickup. If possible, configure the DAQ for differential floating inputs (if source is floating, else use Non-referenced single ended input configuration), to maximize common-mode rejection ratio. Set DAQ input scaling as close to 0-1 V unipolar as possible for best dynamic range.

However, would not suggest >100m run in the above case. I try to locate DAQ hardware as close as possible to the signal source, typically 3m max.

How precise does the measurement have to be? Averaging numerous readings together may lessen the effects of external noise pickup. Must also take into account the update rate required.

Wheels within wheels / In a spiral array
A pattern so grand / And complex
Time after time / We lose sight of the way
Our causes can't see / Their effects.

 
On further reflection, does Mocon supply a sensor with 4-20mA output? This standard is much more immune to EMI/RFI and better suited to longer wiring runs.

Wheels within wheels / In a spiral array
A pattern so grand / And complex
Time after time / We lose sight of the way
Our causes can't see / Their effects.

 
Thanks for your suggestions.

Strangely, PID-tech manual does not say wat the sensor output is, unless I am missing something? It does mention that the package type is "City Technology 4p" I am not sure what this is, nor have I been able to find any explanation on the internet. There is a certification sheet that gives the entity parameters as: Vmax = 5V, Imax = 30mA, Ci = 87uF, Li = 162mH, Pi = 0.15W.

If it does turn out that the current output is 4-20mA does this mean I could get away with >100m wiring?

Would it be possible to amplify the sensors output inexpensively?

Thanks!
 
Zeened,

Those parameters in the certification sheet are intrinsic safety limits. Are you dealing with a hazardous location? If so, you'll need a barrier -- consult MTL or Stahl.

As for the sensor output, if it's unconditioned you will have a fairly high impedance, low voltage signal and analog2digitalman was giving good advice in using a shield.

If you can condition the signal (ie, stick a transmitter on your sensor to give you 4-20 mA), you can use whatever type of twisted pair is available in general. Cat5 is avoided in most control apps because it's hard to reliably torque a terminal screw on it or crimp a connector to it without breaking a wire. If you're using RJ45 stuff, that isn't an issue though.

If you get your 4-20 mA signal going, you're only limited by how much load it can drive. I've sent 'em a mile and back.

By the bye, what gas are you sensing?

Good luck! Let us know how it goes...

Old Dave
 
The location is not hazardous, the sensor is to be deployed in an office building to monitor air quality. In particular the gases being monitored are carbon monoxide and nitrous dioxide.

Thanks for the tips about the transmitter and the cat5 issues. Using a transmitter seems like the best solution based on the information I currently have. Any suggestions as to a manufacturer? I did some poking around on google and came accross the API 4010G which looks like it would do the job, although it is a bit pricey (190 USD).

Thanks for all your help.
 
No prob, Zeened,

I'll disclose that my company distributes Vulcain and BW Technologies equipment for carbon monoxide and nitrogen dioxide, as well as lots of other gases. Also some of our own, and Macurco. But your prices will be in that neighborhood or higher, especially for the NO2.

Other places to check are Lab Safety Supply, MSA, and Sherlock.

Good luck to ya!

Old Dave
 
Hello, this is my first post in this forum so greetings to all.
I would suggest using an intrumentation amp to bring the signal up to a better range, say 0 to 10 volts. Its easy to build from an op amp and can be found by googleing AN-31.pdf + op amp. AN31.pdf is a good op amp circuit doc put out by National Semiconductor. Heres a link
Well, hope I helped, Im looking forward to interacting with and learning from the comunity here.
 
BTW current mode like the other post states is best for long runs, if your DAQ supports it. My suggestion is an alternitive only if it does not. It also provides common mode noise rejection if done correctly.
 
Here is one of many companies (another being analog devices) that provide modules for isolation, amplication, noise rejection, and filtering. I use them fairly often to isolate and filter signals prior to going into the DAQ card. Instrumenation class. One of these accepts +/-1V input and bumps it up to +/-10V range. Another takes a 4-20mA input and scales it to voltage.

Ideally, if available, I would go the 4-20mA route. As DRWeig stated, the loop can be ran for mile(s).

Wheels within wheels / In a spiral array
A pattern so grand / And complex
Time after time / We lose sight of the way
Our causes can't see / Their effects.

 
Thanks for all the feedback everyone.

If I use an intrumentation amp to get the signal between 0 and 10 volts what is my range going to be like? I think my resolution requirements are not going to tolerate that amount of noise. (Signal is between 0 and 1.2 V with maxium being 4.1 V) Also I am not familiar with the 4-20mA standard.. what kind of resolution can I expect with this? Futhermore it seems that most DAQs that accept current input are quite pricey. Ideally I want to put something together for around $300 ($1000 absolute maximum)

Based on the price of the DC transmitter and noise considerations, maybe it makes sense to put some ADC hardware with the sensor in a NEMA type housing. I was thinking about using Diamond Systems Prometeus EA.

Anyway this is getting away from my orignal topic so I will cut it short to avoid annoying anyone, but any input people care to give is appreciated.

Thanks again.
 
Well, I agree with Analogkid2digitaladult, the 4-20mA current mode idea is the best for the long cable run you talked about, by far. It has good resistance to noise and is used all the time in industrial environments.
I dont know what your requirments are, but if it was me, for my own personal use and cost was a factor and I couldent use current mode (4-20mA), I would use the sweet little four channel DAQ that Dataq sells for $24.95 and build a Differential instument amp. Since a diff amp has a differential input it will reject common mode noise. That is to say, noise, that is on both inputs at the same time,such as impuse noise from motors and relays.
As for the original question about CAT5, for the wiring, I would use shealded twisted pair and ground one end and leave one end of the sheald ungrounded. Also, in industrial environments I have seen voltages between two supposed grounded panels that would swamp a 5 volt signal, let alone a 1 volt signal, so you have to watch out for that as well. CAT5 is just multiple pairs and should work if thats what you have around. If your app is for gas detection in say a barn or warehouse, I know they make LON network devices that use slow speed data over the power lines. They may make sensors or devices that would work for you. LON NET is all the rage for that sort of thing in industrial buildings these days.
Anyway, some ideas, but the best idea IMHO is the 4-20mA mode if you can. Good luck
 
Status
Not open for further replies.
Back
Top