Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

Register Log in
  • Congratulations The Obturator on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Interfering signal on 4-20mA loop: origins?

Status
Not open for further replies.
ScottyUK

ScottyUK

Electrical
May 21, 2003
12,915
We're in the process of updating part of our steam turbine control system. While commissioning I've found a minor problem with noise on the 4-20mA signal from the Balluff BTL5 linear position transducer which we have just installed on the hydraulic servo valves. I've used these transducers elsewhere on the same site, and the feedback signal is deathly quiet.

The interconnect cables are multicores with individual screened pairs, with overall steel wire armour. Each transducer has one pair carrying the signal, and a second pair carrying the DC power. The screen is terminated at the 'quiet' ground in the equipment room housing the control system. The field end is floating. Power to the transducers is 24V DC and is a shared supply to all eight transducers. The noise problem does not improve when isolated supplies are used. The noise signal is also present on the power supply pair, but at a much reduced amplitude.

The interfering signal is what is really puzzling me: a brief burst of HF, or possibly RF, noise of a couple of microseconds duration, repeating every 4ms. On the 'scope the envelope of the waveform has the vague appearance of HF commutation noise from a large rectifier, but I can't account for the 250Hz. I haven't been able to trigger reliably on the HF interfering waveform itself, only on the envelope of the waveform, so I'm not sure of the nature of the HF signal.

I've proven that I can suppress the signal on the 4-20mA loop to an acceptable level with a 1uF polypropylene film capacitor, and the results are further improved with the signal pair passed a couple of times through a ferrite toroid.

We have a digital trunked radio system at site which I have some suspicion of as being the source of the interference, but that is only a guess. If any of you are familiar with the unpleasant noise that the signal from a mobile phone causes when it is next to the speakers on a PC, the radios have a similar effect. They're also implicated in a few other mishaps with electronic apparatus on site, but it is politically very sensitive to criticise the system.

Does anyone have any other ideas where my interfering signal might originate? Although the immediate problem is 'solved', I don't like mystery signals getting in to the control system and I am concerned where else it may be propagating to.

----------------------------------
image.php
I don't suffer from insanity. I enjoy it...
 
Replies continue below

Recommended for you

Sort by date Sort by votes
Hi Scotty.

Sounds like you have a well founded cable and layout. I suspect the BLT5 itself. It has circuitry that generates pulses, detection, power conversion/regulation, probably a microcontroller. Since you have checked around the plant and the noise does not fit most plant radio/communications I think you should start suspecting the transducer.

Can you remove it from the circuit and drop in something like a fixed resistor that will give a fed back current of 4mA or so and see if you still get the noise? That way you have the power out and a signal back still giving the least situational change for a test. I could easily see a micro in that device doing something regularly at 250Hz. It could be sleeping, waking up, running some code and sleeping again. There could be a problem with its filtering components or?

Keith Cress
Flamin Systems, Inc.-
 
Upvote 0 Downvote
That's a very valid question. The BTL-5's in question are the second generation design, and our previous experience was with the earlier design. I have a couple of spares of each design, so I will test them with a direct connection to the servo controller, eliminating the field cable, as a comparison. That way I should determine whether the noise is present due to the cable, or the transducer.

The transducer uses a magnetostrictive sensing element, sending out a 'ping' and measuring the time for the 'echo' to return from the location of the sensing magnet. It's rather like sonar in principle. You now have me wondering if the 250Hz is the rate of 'pinging' in the sensor.

I will have to investigate the earthing to the metal body of the sensor: this is intentionally not tied to signal or power ground, and by construction of the transducer it should, in theory, be at the same potential as the hydraulic valve actuator body, i.e. power earth. If the earth path through the actuator body is not adequate then I will have to install an earth bond to the transducer body. I don't want to bring this earth bond back to the 'quiet' instrument earth because it will very likely also be tied through some (un)fortutious path to plant earth, which is noisy as hell.

----------------------------------
image.php
I don't suffer from insanity. I enjoy it...
 
Upvote 0 Downvote
A noisy earth ground may imply earth currents. Do you have any idea why the plant earth is noisy?
respectfully.
 
Upvote 0 Downvote
Because it is a 1.8GW power station! There are a wonderful combination of 275kV transmission circuits, generator outputs at 16kV up to 16kA, a private 11/3.3/0.415kV distribution system, etc, etc. All on a common earthing grid.

I am starting to suspect the interconnecting cable as the source of the problem - loop impedance is about 7 ohms, and there is no local decoupling at the transducer. I am about to install an electrolytic and HF bypass capacitor at the transducer location, so I will know better this afternoon.


----------------------------------
image.php
I don't suffer from insanity. I enjoy it...
 
Upvote 0 Downvote
Hiya-

Although you might have shielded cables, it *MIGHT* be
interesting to rent/borrow a spectrum analyzer and sniff
around the site to look for "funny" RF signatures.

I'd also borrow/rent an attenuator for it and cart it both
around the site and off site too.

As a Jr. Engineer (way too many years ago) I was surprised
when we hooked up a little intercom system to help ring out
wiring for some insturmentation. The intercom pair was with
the insturmentation (differential) and well shielded,
although a couple of hundred feet. Only a couple of feet
was twisted pair unshielded. That was enough for a local
AM spanish radio station to come blasting on through!

Unexpected rectification (bad joint causing a diode
action) was the culprit.

It too was swamped out by a capasitor.

YMMV, but just a thought.

Best of luck!

Cheers,

Rich S.
 
Upvote 0 Downvote
Hi Keith, waross,

Problem solved: as normal, it is the simple things which are hardest to see!

Turned out to be a result of an insufficiently stiff power supply because of the long cable run, and lack of any local reservoir capacitor to prop things up a bit. The BTL draws power in pulses from the supply, hence the wierd 'interfering' signal. I assume that the current loop driver on the output must rely on a reasonably quiet supply, or at least isn't designed to deal with abrupt changes in supply voltage. The supply disturbance breaks through into the output signal.

There is also the small matter of the screen carrying power: the screen is tied to 0V at the source, and an incorrect connection in the field J-Box caused the negative return current to flow in the screen. This was a significant contributor to the problem, with a large reduction in noise once corrected.

Things are now very calm with a 10000uF/160V electrolytic and 1uF film bypass cap temporarily providing local smoothing. DC ripple is down to a handful of uV. Longer term we'll probably move a supply up to the turbine level and provide local power to the transducers. The caps were just what was handy, before anyone asks: I know it is overkill!

We've got a 1uF film capacitor plus a small ceramic providing RF suppression on the mA input to the system which has flattened the RF noise completely.

The bench test gave results with the noise below the noise floor of my scope, so the transducer itself is ok, provided the supply is reasonably stiff.

----------------------------------
image.php
I don't suffer from insanity. I enjoy it...
 
Upvote 0 Downvote
(I know I will have to take some beating for this...)

There are no mysteries in engineering - only problems not yet solved.

(I am ready to defend that standpoint)

Gunnar Englund
 
Upvote 0 Downvote
Okay I'll bite..

<begin beating attempt>

Good grief!

A 'mystery' is related to a person's 'knowledge'. It has nothing to do with engineering or medicine or prospecting or crime but rather with the lack or possession of some particular bit of knowledge. Until the knowledge is gained you have a mystery. Until the disease is identified you have a mystery....

<end beating attempt>

Keith Cress
Flamin Systems, Inc.-
 
Upvote 0 Downvote
Now, this will be "jb style":


A 'mystery' is related to a person's 'knowledge'. ///agree\\\ It has nothing to do with engineering or medicine or prospecting or crime ///has, too\\\ but rather with the lack or possession of some particular bit of knowledge ///yes, if you do not know the solution, you may think it is a mystery: which it is NOT!\\\. Until the knowledge is gained you have a mystery ///or an unsolved technical problem\\\. Until the disease is identified you have a mystery.... ///as I said in my latest comment\\\

Search "jbartos" in eng-tips to learn more about jb debate style ;-)



Gunnar Englund
 
Upvote 0 Downvote
Hehehe ..... jbartos-style....hehehehe [noevil]
Good one, Gunnar.

Glad you solved the mystery, Scotty. I followed the thread with interest.

[red]Failure seldom stops us, it is the fear for failure that stops us - Jack Lemmon[/red]

Make the best use of Eng-Tips.com
Read the Site Policies at FAQ731-376
 
Upvote 0 Downvote
I still disagree... (oh yeah I am familiar with jbartos-style though I haven't seen him in a long while and don't like the method really as it confuses me. lol)

You are postulating that anything you don't understand in engineering to be "a problem". Indeed many things may "not be understood" but not be problems... They are still a mystery to the ignorant until the ignorant makes the effort to become informed. No problem need ever have been present.

Q.E.D.

:)

Keith Cress
Flamin Systems, Inc.-
 
Upvote 0 Downvote
No Keith,

You are jumping to conclusions. This is how I see it:

There are very many things in this world that I do not understand. But that doesn't mean they are mysteries - or even problems. Just things I do not understand.

A small subclass of these things sometimes pop up and manifest themselves as problems. Things that need to be understood and solved before a machine or some other equipment performs as expected.

It is this subclass that I am referring to when I say that there are no mysteries in engineering - only problems waiting for their solution.



Gunnar Englund
 
Upvote 0 Downvote
skogs,

Do you believe in gremlins? If not, please visit my plant where I appear to have an infestation. You are welcome to have them all!


----------------------------------
image.php
I don't suffer from insanity. I enjoy it...
 
Upvote 0 Downvote
That reminds me of a problem that a customer had with a German photoswitch. This new design had a phase chopped power supply like a light dimmer. Trouble was that he had over 100 of these run by a single power transformer and all the SCRs turned on at about the same time. That created an enormous power surge. We solved it by selling him a 1uF RC noise suppressor at each sensor. Not the best solution, but he needed something UL that could be stuck in the wiring box. He said that experience caused the manufacturer to change the design.
 
Upvote 0 Downvote
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top Bottom