Skip to Content

Register Login

Industry Online Support

Technical Forum

Navigation

floating point range woth 4-20mA

Created by: automait at: 8/29/2007 10:49 PM (3 Replies)

Rating (1)

Thanks 0

4 Entries

8/29/2007 10:49 PM	Rate (0)
automait Posts: 5 Rating: (0)	Can someone shed some light on floating point ranges. I have a flow transmitter 4-20mA with a rage of units 0 -300 metres cube/ hour. I need to accumulate the value periodically say everysecondand compare against a set point value of 20,000 metres cubed. The problem is what value in floating point will be generated? i know the range in positive numbers is from x10 to the -38 to x10 to +38 but do these extreme values correlate to the range in mA from the transmitter? If so i would need to scale the values down enough for the accumulator to be able to reach the equivalent value of 20,000 metres cubed? Any info / clarity graetly appeciated.
automait Posts: 5 Rating: (0)
Suggestion To thank Quote Answer

8/30/2007 7:29 AM	Rate (0)
Dorftrottel? Silver Member Joined: 8/10/2006 Last visit: 8/13/2024 Posts: 298 Rating: (29)	I hope I am answering in a helpful way. I would first look at what accuracy do you need ? If you were filling a tank, and your compounded accuracy (instrument error + time error + analog resolution, etc...) was approximately 0.005%FS (very accurate for a turbine flowmeter), you would have 10m³ extra at the end of your cycle. Can you handle that ? Then ask, what is the cost of my waste? For example, if in the course of a year, 52 batches, with 10m³ waste/batch at € 10/m³ costs € 5,200/year. You are well on the way to affording a coriollis flowmeter, which you can talk to via profibus, is extremely accurate, and (generally) bypasses your need to do FP conversions. I have a preference for not using floating point unless necessary. I would convert my setpoint value into the scaled integer value, and then do simple double word adds when accumulating the volume. I would then every 1 or 2 seconds convert the volume to floating point for display on the HMI. HTH, Jim

Suggestion To thank Quote Answer

8/30/2007 5:23 PM	Rate (1)
Denilson Pegaia Platinum Member Joined: 9/23/2005 Last visit: 11/14/2024 Posts: 4352 Rating: (1452)	Hi Some time I thought in this type of problem (errors accumulation with sucessive float point summs), and I suggest one FAQ. This FAQ, revised, was published in internet at: http://support.automation.siemens.com/WW/view/en/13406965 With some trics, it is possible to expand the mantissa capacity of the float point. I suggest take a look at the SCL source of the sample program.
	Denilson Pegaia Siemens Industry Sector Technical Support
Suggestion To thank Quote Answer

9/2/2007 3:43 PM	Rate (0)
automait Posts: 5 Rating: (0)	Guys, Thanks for the info, scaling the value is much easier as dealing with actual engineering units of the process. Cheers.
automait Posts: 5 Rating: (0)
Suggestion To thank Quote Answer

|

Share this page:

Share this page on...

Follow us on