Problem with Instrument Control Loops

Appreciate any help on solving a long standing intermittent problem. We have an instrument control panel that serves a treatment plant. We
monitor gas pressures on seven analogue control loops. All the loops and PLC digital I/O's are supplied from a 24 VDC supply, and the 0 Volts is floating (not connected to earth). To explain one of the loops: 24 VDC Positive to an Intrinsically Safe Barrier (MTL 787SP), out to field instrument, back to Intrinsically Safe Barrier, then to analogue input module on PLC, out of PLC, back to 0 Volts. There are seven of these loops. Intermittently, the PLC sees +20mA coming from the field instrument, and they all do this at exactly the same time. On checking at the field instruments, they are giving out nothing like 20 mA. As a quick check, I bypassed the IS barriers, and everything works fine. At first I thought that the zenner diodes may be leaking. So, changed all barriers and the DC power supply. Rechecked, PSU supplying correct voltage and not overloaded. On connecting the inputs to the IS barriers, with the outgoing loops disconnected, the plc is picking up milliamps and it is erratic. My next thought is that something in the plant is leaking to earth, and returning back through earth to the IS barrier earth. This is what the PLC is picking up. The IS barriers are commoned together with an earthing strip. This earthing strip is connected to the cubicle earth. If I remove this earthing connection from the barrier strip, everything returns to normal. Or if I earth the 0 Volts so that it is not floating, everything returns to normal. The PLC sees 0 mA with the outgoing field loops disconnected. I will probably never find the earth problem causing this, as it is a very large plant, and the PLC controls everything. Would installing a permanent earth to the 0 Volts on the PSU be the answer to my problem, or is there anything else I haven't considered. TIA
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
BIGEYE wrote:

I'm trying to visualize this. Does your PLC analog input card have isolated channels? Or do they share a common? That could be an issue but I'll have to think about it a little.
Are your circuits wired as follows??
+24 VDC to pin 1 of MTL Pin 3 of MTL to transmitter + Transmitter - (minus) to MTL pin 4 MTL pin 2 to analog input + Analog input - to DC power supply -
On second thought, tie the neg terminal of the power supply to earth (IS ground per ISA RP12.6). If it floats, if the negative terminal is less than zero wrt earth, you will get current flow thru the zeners that connect IS ground to pins 2 and 4 to the power supply (-) thru the resistor in the analog input card. This is probably what's happening and the plc is seeing the max signal (20 ma) even with nothing connected to the field side of the barrier.
ANSI/ISA RP12.6 has some info on proper use of zener barriers.
hth
Tim
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

1. I haven't looked at the analog card, but I think they share a common. 2. This is how the circuits are connected. 3. The only way I could get this problem to go away is to connect a temporary earth to 0 Volts on PSU.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
BIGEYE wrote:

I see (finally) that you are in the UK. I referenced ANSI/ISA. There should be a similar document in your area. And MTL should have some installation guidelines.
Tie the negative dc terminal of your 24 volt power supply to earth (ground). In the US, we are required to have a dedicated isolated ground for zener barriers that ties to the main building ground point (IS ground). The negative of the power supply is connected to the IS ground bus.
See page 8 of http://www.mtl-inst.com/products/hazproducts/barriers/700/Datasheets/MTL700%20series.pdf
for info on connecting the 787SP.
hth,
Tim
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

On
At
earthing
to
Equal potential is what you need. You need to ground one side of your system and "float" the other end. I like to ground the PLC and all the shields to the cables at the power source. Shields are taped back on the termination on the remote device. Are you using non shielded control wire?
Triac I/O in the PLC also can have some bleed over. ( in my experience) I do not use triac I/O
Bleed over/ induction can be a factor in some high current situations.
Your terminology is a bit strange to my ears, so I hesitate to point you to a direction. Do you have a drawing of the system?
What changed recently?
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.