Installing fuses on PLC IO's

We have a standard to install fusible TB's at PLC AI/AO channels and knife TB's at PLC DI/DO channels. Also the 24 vDC field power supply to all PLC cards is also through fusible TB's.

Now we have a client whose system has got all Isolated channel cards(DI,DO,AI and AO). He is asking me to remove fusible TB's from the entire panel and used Knife TB instead. His point is that installing Fusible TB at the power of cards is creating an additional risk. This can bring the whole Plant
down.

Also since the cards have isolated channels they will be protected in case of short circuit.

Should I take his feedback and do so or keep the original design?

There's no rules about anything like that. It's just a preference. If the customer requests something, I usually do it unless there is a good reason not to.

There is a side benefit to installing fuses on 4 -20 mA analog circuits. With the right style fused terminal, you can put the leads from your multimeter on the terminals. Then, when you remove the fuse, your meter is in series in the circuit without having interrupted the signal. A real boon in situations where an open circuit generates an alarm or shuts things down.

Why so much fusing? Why not just fuse the branch of the power going into the Output card common?

No national or international standards would be violated, and the customer wants it. I would just change it, and mark the reasons for the change.

If there is a risk to your kit, make sure the customer agrees that they don't care about the risk, then do it.

His point is that installing Fusible TB at the power of cards is creating an additional risk. This can bring the whole Plant down.

Click to expand...

If he is right that a short circuit can brink the whole plant down, then there is something wrong with the electrical design.
The way to do it is to segment the power supply in such a way that a single error cannot shut down everything.

That being said, most (but not all) PLC outputs from PLCs of the current state of the art have integral current limiting for outputs. So if properly designed and with properly selected PLC parts there will not be a short in the field that can shut down the whole system.

We do use knife terminals for isolating groups of field items, but that is for isolating sensors only (PLC inputs) during commisioning and troubleshooting, not for solenoids or actuators (PLC outputs).

Steve,

As much 4-20mA trouble shooting that I do it's with bottom feeders who don't understand what a 2 wire loop powered circuit is. I can't really recall seeing fused 4-20mA in the field. So I'm curious.

Given that 4-20mA is inherently current limiting, so a short from the regulated output, typically the (-) terminal of the field transmitter, to ground point common with the negative terminal of the power supply would only conduct whatever current the transmitter was attempting to supply to the receiver.



Do you fuse the DC (+) line running to the Tx (+)? Along there is where a short to DC(-) shorts the power supply, nowhere else.

The fuse that I was referring to isn't so much a protective device as a means to cut the meter into the circuit without interrupting current flow. When I came across the tip a few years ago, my first impression was, "I wish I'd thought of that on xxx project". When the OP mentioned fusing his analog circuits I thought I'd point the fact that doing so provides an unanticipated bonus. Of course, if the OP's analogs are all voltage rather than current, there's no bonus.

So you're using a fused terminal block?

When you want to measure the loop current, I assume you connect the milliammeter, which reads zero until you lift the fuse out of the circuit, at which point the meter conducts the current.

Close the fuse, back to normal.

Cool.

We fuse every single 4-20ma input with 250ma fuses. It allows the normal loop to be short circuited without blowing, however they do blow when we have lightning strikes, ground faults on the high voltage system, when AC is accidentally connected to the 24vdc system,and when fires melt the comm cable together. It's funny though because most times the new guys we get have never heard of fusing analog inputs. We also use an external resistor for the same reason. If it gets blown up replace a $.50 reaistor instead of a $2000 card.

So you're using a fused terminal block?

When you want to measure the loop current, I assume you connect the milliammeter, which reads zero until you lift the fuse out of the circuit, at which point the meter conducts the current.

Close the fuse, back to normal.

Click to expand...

That's the idea.

We fuse all our analogs individually and each di / do card supply individually. Most of our field devices are outside, in pits or tanks or other environments where cables can get damaged, water ingress happens from time to time, etc.

The OPs client suggests that fuses will mean the whole system gets taken out by a single fault. That seems reverse logic to me! If there is no fusing and a field device develops a fault, there's the potential for the DC supply to go into current fall back mode and drive your 24V rail down to zero volts if the fault impedance is low enough. Now EVERYTHING on that supply is down, instead of just a single sensor.

That's why i like the LED fuse carriers... can open a panel and immediately see the one with the lit LED