Sudden Voltage Drop on Profibus PA

Have you tried replacing the existing transmitter with a known good level transmitter?

What does the manual say is the issue?

What does the manufacturer say is the issue?

Try some basic troubleshooting techniques and then come on back and let us know what the results were please!

Have you tried replacing the existing transmitter with a known good level transmitter?

• I'm using a non-contact radar level transmitter.
• based on my findings, the LT's measured value/trend seems normal, and no sudden voltage drop occurs as can be compared with the inventory software

• Trend Comparison between LT and Inventory system.

Both trend data were sampled between 17/8/2015 to 20/8/2015.

Figure 1: local Transmitter's trend

Figure 2: Inventory system's trend

Refere to Figure below, Red line represent the secondary value 2, a distance between radar sensor to product level. In other word it is Ullage value. While trend in inventory system recorded in innage value.

As can be seen from both readings on 17 Aug, LT's trend doesn't occurs any signal drop to zero as compared with Inventory system's trend (happens once/ yellow arrow).

• from this comparison, might be the transmitter is working fine.

What does the manual say is the issue?

• in LT manual doesn't mentioned anything about this issue

What does the manufacturer say is the issue?

• I'm still waiting for their feedback on this issue

Are you certain the level transmitter is not configured to "auto-calibrate" mode?

Are there other devices in the control/reporting loop?

If yes;

Are there too many devices in the loop?

Is the polling table correctly configured?

Is the device priority high enough in the hierarchy assignment?

Are you experiencing a high number of collisions and/or db loss in the communication channel signal?

Have you verified polarity at every device in the loop is correct?

Does the baud rate of the meter match the system required rate?

Whatever is happening is common to all three level transmitters because the drop-to-zero artifacts occurs at exactly the same time on all three devices.

It seems unlikely that one transmitter can 'pull down' the others, because Profibus is a master/slave protocol and is unlikely to report a bad value for all transmitters, unless it is a network issue where values cannot be read. SO I think it unlikely that is a transmitter.

Any Profibus system has communications error logs and that's the first place I'd start.

There's likely some Profibus setting that determines what the 'system' reports when Profibus is in a known fault state, similar to fail-safe settings on analog HART transmitters which can fail high or fail low to indicate a problem in the circuit. If so, then the drop-to-zero artifacts might be an indication that the system detects a fault and for some period of time issues fail-safe zero values until communications is reliably re-established.

It might not be a Profibus problem at all.

• The type of transmitter is not mentioned. Ultrasonic level transmitters can fail to see the surface of foaming, steaming, turbulent or crusted liquids on occasions, leading to a false zero level signal being transmitted. Condensation on the ultrasonic horn face can also cause this.
• Some instruments of various measurement techniques can be configured to "hold last value on signal loss" rather than "drop to zero on signal loss". In the case of liquid in a tank, level changes are going to be slight, and not catastrophic level change as suggested by the downward spikes. Reconfiguring the instrument might be a solution.
• Given a generally slow rate of change of level, it is possible to predict the approximate value of the next measurement point from the previous two or three level scan points. If the actual level measurement point is significantly far from the prediction (by a value to be determined by local experiment), it is possible to ignore the actual and substitute the predicted, then substituting a corrected averaged value for the predicted one the next time a valid actual level signal is received, thereby removing the zero spikes at the receiving end.

It's only setting up and coding, after all!

yes i'm agreed with you Mr. Carl_E.

I'm using a non-contact Radar Level Transmitter.

Obviously this drop to zero phenomenon happens on the same time.

since this system's hardware (OLM, PLC and CPU) are located outside of the building and exposed to open environment which might can be struck by lightning or cause by improper sharing the input/power supply with other heavy devices for plant process.

thus, i suspected the transmission cables or connection might be the main root cause for this issue.

• Instrument grounding
• Profibus PA screen cables
• DP/PA coupler
• PE (dirty earth) and IE (clean earth) grounding

What brand transmitter are you using? It seems to me your transmitter is starting to malfunction. I have experience the same issues, the transmitter will either go to zero or a hundred% once you replace the transmitter the issue will disappear.

I'm using VEGA.

based on your experience,

1. is it happens during the tank is empty?
2. my case, its happens during all affected tanks are filled up with product (oil)
3. and it is connected in a loop. for about 6 other LTs are facing this same issue at on the same time

what's the ambient temperature where the tanks are located when the issue happens?

since this phenomenon happens randomly, i can say that the ambient temperature is around 30~34 degC

Given that they happen to several devices all at the same time, I'd be looking at all things that they have in common, Mildred. Profibus is one. Power supply might be another. Is there an intermittent fault in the power supply?