We are experiencing an issue with an fuel tank level output on an AI card (Siemens AI card 331-7KF02-0AB0). We have a valid analog signal from the radar level transmitter in the field to the remote I/O PLC, but nothing from there to the control PLC.
The fuel level signal is normally shared by two generators with identical PLC set-ups. We have shut down one unit for re-build (the master unit) and are now having issues with the other generator (the slave unit). All wiring terminations were matched from the master to the slave and every signal (input/output) has been working perfectly except the level signal from the fuel tank.
The fuel level is used to initiate an automatic 'fill request' and a solenoid opens to re-fill the fuel tank. At the moment our tank level, transmitter output and mA input into the AI card are all sound. There is no output to the supervisory PLC - on the HMI for fuel level we see a level of 0. There is a red LED on the AI card indicating SF, but when we enter 'Diagnostics' the hardware for the module is noted as being 'OK'.
Other outputs on the AI card are indicating correctly on the control PLC without issue ... it is just this channel that is giving us a problem.
The square range selector on the side of the card is in the correct position for our application.
My question is that when everything on the system is working except a single AI, where do we look next for the issue? Is there something in the program that may have been upset during the swap from slave to master? Do we need to re-download the program to the PLC?
I am not familiar with Seimens products, so any assistance or guidance would be greatly appreciated.
You haven't told us what kind of signal is coming from the level transmitter (presumably 4-20mA), nor how the level transmitter gets its power, nor how the original wiring which "shares" one signal between two PLC analog inputs was accomplished; all of could play a role in a missing signal.
There are '2 wire' loop powered level transmitters which draw their operating power from the same 2 wires that are used for the 4-20mA output signal and then there are '4 wire' level transmitters which get their power over 2 wires and use 2 different wires for the 4-20mA output signal.
I suspect you might have a 2 wire loop powered transmitter that is not getting loop power.
I recently discovered that siemens redefines otherwise standardized controls terminology on their 331 analog input card.
What the rest of the world calls a "2 wire transmitter" (loop powered), siemens calls a "4 wire transmitter" if the DC power for the transmitter is sourced from the 331 card's L+ terminal, as opposed to a stand alone DC power supply which is not powering the 331. (The rest of the civilized world does not take into account what else the DC loop power supply is powering when using the terms '2 wire' or '4 wire'.)
This newspeak definition comes into play when setting the rotary switch on the 331 AI card and in the programmed configuration for the channel.
The 331 manual states "when the supply voltage L+ is fed from the module, you must assign parameters to the two wire transmitter as a four wire transmitter in STEP 7."
And the card's (channel's?) rotary switch position has separate position settings: C for '2 wire' or D for '4 wire'; but you have to know siemens special definition of '4 wire'.
See the thread at
(2nd page of a 2 page thread)
for screen shots of wiring and the manual statement.
There are other possibilities, for instance, the master AI was wired and configured for 4-20mA, but the slave was wired in parallel and (remains) configured for a voltage input.
A true 4 wire current loop (worldwide, not siemens definition) transmitter signal could be wired in series through both PLC AIs with identical programming, but is that the case here?
If both PLC programs were truly identical (are you positive that's the case or making an assumption?) then it is essential to understand the technique of sharing one signal between two PLCs.
Or perhaps the master supplied loop power to the level transmitter with the slave AI was wired in series. In that case the slave input would be a '2 wire', and the master a '4 wire'.
The devil's in the details, as usual.
Thank you very much for the response and I apologize for the lack of information ... new to all this.
We have an Endress & Hauser 2-wire transmitter in the field. The 4-20 mA output signal is split in a field JB. One cable returns to one generator and another cable carries the signal to the other generator. This signal returns to a Red Lion IAMS Universal Transmitter. This transmitter outputs the 4-20 mA signal to 2 locations: the AI card and a Weidmuller current isolator (this sends the 4-20 mA signal to the DCS system).
We have a good signal coming in from the field from the transmitter as we do have a current indication on the Red Lion and we do have a level indication on our DCS. The 4-20 mA signal from the Red Lion up to the AI card is also good.
We compared the switch settings on the side of the AI card to a few of the other generator I/O panels (we have 8 identical generator cubicles) and they are all set for 4-wire current loop - "D").
We are now drawing a blank on where to look next as it seems the field side up to the AI card is okay.
It is obvious that we are still missing something.
Thank you again for your suggestions.
Measure the voltage to ground at the AI terminals. It sounds to me like the voltage is outside what the AI can handle. There should be a spec in the module docs. I have seen this when operating current inputs in series and sometimes when the driver and receiver simply don't share the same idea regarding the voltage on a current loop. A current input is most often a voltage input across a resistor. Sometimes these are single ended or at least, not fully differential. And not isolated, so if either end or both or the resistor is outside what the voltage input can handle it stops working. But if you use a battery and a pot(isolated), it works fine.
> The 4-20 mA signal from the Red Lion up to the AI card is also good.
Does this mean that a milliammeter can read a reasonably correct 4-20mA current signal at the pair of wires that connects to the Siemens 331?
Or can you measured the voltage drop across the 331 input connections, for instance, across terminals 2 and 3 or terminals 4 and 5 (wherever your signal is)?
Siemens uses internal 250 ohm resistors to create the IR drop that the analog input A/D uses. So you should get a 1 to 5 volt signal scaled equivalent to the 4-20mA signal.
If you can measure either the current or the voltage drop as reasonably correct then the signal is powered and good. The fault has to be in the configuration or the hardware AI.
If there is not a signal, then the tortuous path from the E&H transmitter back to the 331 is not clear to me; there are too many devices listed.
Is the 331 supplying the loop power to the signal that comes to the 331 from L+ and M? With all those other devices, it might not be supplying the loop power.
It sounds like there might be a splitter or an isolator in-between the E&H and the 331 and any of the in-between devices might be powering the signal going to the 331.
If some other device is powering the signal (not the 331's L1+ and M), then Siemens considers the incoming signal a 2 wire signal, not a 4 wire signal.
> they are all set for 4-wire current loop - "D")
I believe that position C is for a 4 wire transducer; position D is for a 2 wire transducer.
If you've set the position at D, then the signal should connect only to the input + and - terminals, with no connections to L+ or M and then Siemens (and the rest of the world) considers it a 2 wire transmitter.
Let's clear something up.
How is the transmitter powered?
If the only wires are the loop, two wires on the transmitter and those two wires are connected to two terminals on the receiver, then it is a 2-wire sensor. The AI is configured for 2-wire and it sources the loop power internally. (Measuring range selector on [D])
If there are four wires on the transmitter, two for power and two for the signal loop. It is a 4-wire sensor. If the transmitter has only three wires; power, signal, and common, it is still a 4-wire sensor. If the transmitter has only two wires, but you put a power supply in series with the loop, it must be configured as a 4-wire sensor so that the AI is not injecting a second power source. (Measuring range selector on [C])
The only time that the 331 AI is configured for a 2-wire sensor is to make it the power source for the loop. Using L+ and M-, even on the AI's terminal strip, makes it a 4-wire sensor for the purposes of the AI's configuration, even if only two wires are leaving the panel to go to the transmitter.
Hello and thank you for your response.
These are some of the things I have tried and the results;
With all connections as found:
Voltage from +6 to ground = 0.312 VDC
Voltage from +7 to ground = 0.010 VDC
Voltage from L+ to M = 26.75 VDC
Voltage across +6 and +7 = 0.454 VDC
I was curious and swapped the measuring range module from position C to position D;
Immediately the red SF LED went out.
Voltage from +6 to ground = 25.51 VDC
Voltage from +7 to ground = 0.031 VDC
Voltage from L+ to M - 26.73 VDC
Voltage across +6 and +7 = 25.49 VDC
At this point I connected a Fluke 789 directly into the AI card (avoiding all the other 'bits' we have in our system) and managed to adjust the output (fuel level) on the HMI. Progress? Not really.
I also tried to follow the field wiring from transmitter to the AI card.
The engineer's drawings indicate we have a 4 wire transmitter, the as-builts from the engineering firm show we have a 2 wire transmitter and nothing seems to match what we see in the field. There is also a mystery jumper in one of the panels - if it is removed, we loose power to the transmitter.
We need the generator for site power and only have a few hours here and there to investigate what is going on ... so, again, I appreciate your input and insight.
I have 6 identical units using the 331 and only one of them has this problem. I suspected a common mode problem, but when I used a DVM to read from ground to the loop, I read nothing and the AI worked while the meter was connected! (Disconnecting the loop and injecting a signal worked fine also.)
I connected a 1k ohm resistor from signal ground to the loop and it has been working fine for the last 12 years.