Siemens S7-200 Analogue I/p

I am using S7-200 PLC but this is a general question.I have an analogue input scaled from 1 to 6 .I am using a level meter 4-20mA to measure water level in a tank. I want to use a high level alarm so as when water reaches 5.5mtrs to have a warning. My problem is that when level reaches 5.5 the level meter o/p plays around this value thus my alarm goes on and off continously.I do not want to latch the alarm so I increased the samples and deadband .There is some improvement but the problem is still there. Any ideas???

You could use a debounce timer i.e
The level has to be Equ to or Grt than 5.5 metres for so many seconds before the alarm appears.

----|EqU|----|Grt|----------Ton.

T/Dn Alarm
---||----------------------(L)

Frageo.

You need to define your problem a little better.

The reality is that the level is going to bounce around a little bit, going above and below 5.5 when the level is there. The debounce timer that Gerry suggests is good, in that it will help ensure that the level is really above 5.5, not merely at it, but it won't solve the problem completely.

You don't want the alarm to go on-and-off.
You don't want to latch the alarm.
You don't want a deadband (or you already have as big of one as you want)

So what do you want? And why don't you want the alarm to latch, at least until the level drops to a reasonable level (like 5.4)? Are you looking to inhibit the alarm once it's fired, until the level drops below a deadband?

If you can describe it, we can help you code it.
But we've got to understand it first.

Are you controlling the level? Or simply monitoring it?

It sounds like you might have your Set Point (or Target Level) too close to the High Alarm Level.

It also sounds like you might have a high Slosh-Factor. Slosh is the effect you see on the ocean when the waves are choppy as opposed to when the waves are virtually non-existent; Rough Seas (high Slosh) vs. Smooth Seas (low Slosh).

You might need to control the Alarm Condition in Ladder instead of in the Alarm Function.

An aside regarding Slosh...

This effect can affect Pressure Transducers as well as Float-type Transducers...

This is a general description in terms of Float-type.

The rougher the surface and the smaller the surace area of the detector, the harder it is to determine an average level. The level of the surface, at any given point, in a tank with a high Slosh-Factor, will be chaotic. The wave action might be periodic, but the timing of the readings are not likely to be in-sync.

Even if you increase the number of samples taken before judgement is made (without taking so many samples that your response is too far behind), there is no reason for you to believe that the readings will be reasonably and properly distributed across the range of possible readings so as to produce a reasonable and consistent average.

Float sizes can range from a very large percentage of total fluid surface area to a very small percentage.

At one extreme, if you had a large float that covered the entire area of the tank you would get a very good average indication under any Slosh conditions. However, you might have problems if the average level changes by large percentages in small periods.

At the other extreme, if you used a pin-float to measure the level of a fluid with a high Slosh-Factor, you would be reading the crests, troughs and points in-between. At the same time, the float would actually be accelerated upward or downward at times, thus giving an even more bogus indication.

The lower the Slosh-Factor, the smaller a float you can get away with. If the Slosh-Factor is at or near zero, you could get a good reading from a pin-float.

If you have a tank with a high Slosh-Factor, you could install a float with a larger surface area, or you could (might be able to) install a baffle to reduce the Slosh-Factor.

Thank you for the suggestion GARRYT1 .It seems a good idea.

"You don't want the alarm to go on-and-off."
This happens so fast Allen that is not even displayed on my HMI display.I just hear the PLC o/p relay closing /opening quite fast.

Allen
"You don't want a deadband (or you already have as big of one as you want) "
As I wrote in my question : "I do not want to latch the alarm so I increased the samples and deadband .There is some improvement but the problem is still there. "

"And why don't you want the alarm to latch, at least until the level drops to a reasonable level (like 5.4)? "
After some thought and small changes to my program I latched the alarm
and reset it when level drops to a reasonable level.

Thank you all for your help

What you want is HYSTERISES

Turn the alarm ON when the level exceeds your setpoint PLUS say 5%

Turn the alarm OFF when the level falls below your setpoint by say 5%

Make the hysterises percentage a variable, so you can expereiment for best results.