5V parasite problem on long wires and environmental effects

ufkyldrm · Joined: 21 May 2018 Posts: 27

Hi guys.

I'm using pic on some industrial projects. Getting pulse from sensors with reed switch is one of them.

I'm getting wrong pulse signal because of environmental effects or cable length.

Is there any way to get this signal without parasite?

My wires are 10 meter but are surrounded by power cables for big machine. So they have magnetic field.

24v circuit with relay can be solution for that?

I heard capacitors can be used for that, do u have any filter solution?

Ttelmah · Joined: 11 Mar 2010 Posts: 19215

Big question. What type of wire?. Shielded cable?. Twisted pair?. Makes a huge difference.

temtronic · Posted: Fri Jul 27, 2018 11:49 am

I can run 10 to 15 MILES of solid copper wire and not get 'corrupted data', have so for 30 years.Industrial buildings, some with rooms FULL of arc welders and other 'noise' sources.
As Mr. T asks, what type of wiring AND what data speed / protocol / signalling are you using ?

Mike Walne · Joined: 19 Feb 2004 Posts: 1785 Location: Boston Spa UK

Also, have you had a look (with an 'scope) at the interference signal from your big machine?
Will give you an idea of the size of problem you're dealing with.

Mike

ufkyldrm · Joined: 21 May 2018 Posts: 27

ufkyldrm · Joined: 21 May 2018 Posts: 27

Mike Walne · Joined: 19 Feb 2004 Posts: 1785 Location: Boston Spa UK

Can you draw a simple diagram of your wiring arrangement?
Grounds, terminating components etc.

Mike

ufkyldrm · Joined: 21 May 2018 Posts: 27

elcrcp · Joined: 11 Mar 2016 Posts: 62 Location: izmir / Turkey

That is a problem commonly seen on industrial applications. Although there are several solutions suggested, what I would suggest is what I did.
-Use shielded cables and connect shields to common ground at one end (prefably on main circuit)
-Connect filter capacitors (in parallel) from picofarads to milifarads at main power input on your circuit
-(Optional) Use filter coils at both ends of signal cables like these ones https://www.ebay.co.uk/p/10pcs-5mm-Clip-on-Noise-Ferrite-Core-Ring-Bead-Filter-RFI-EMI-Cable-Clip/8014413332?iid=173142890423
First two option together solves the problem generally.
_________________
There is nothing you can't do if you try

Ttelmah · Joined: 11 Mar 2010 Posts: 19215

ufkyldrm · Joined: 21 May 2018 Posts: 27

Ttelmah · Joined: 11 Mar 2010 Posts: 19215

So from the circuit, you have the wire shown coming out of the top of the Arduino board as one core in the cable, and the ground as the other?.
What resistor value?.
Which end of the cable is the shield connected?.

The resistor value wants to be low. The shield should be connected just at the PIC end. Then given this is not going to be a fast signal, a reasonable capactor between the signal and ground at the PIC end.

Mike Walne · Joined: 19 Feb 2004 Posts: 1785 Location: Boston Spa UK

Mr. T. and I suspect that you are not understand the nature of the problem you're facing.
I do not think you are being devious in any way.

You started by complaining of problems caused by a "big machine".
But you are not supplying the necessary detail of your system.
You've told us that your components are as below :-

Ttelmah · Joined: 11 Mar 2010 Posts: 19215

I'm particularly suspicious there may be connections at both end of the shield, which can then result in current flowing in the shield, which can make it worse than useless. Also that the actual drive may be something like 4K7 or 10KR, which means that the line when pulled (up?) by this, is quite high impedance.

temtronic · Posted: Sat Jul 28, 2018 7:16 am

I had a look at his link..