Member Login
Search
Jump to a Date
Sponsored Communities
Cool stuff
Neat Stuff

Visit our shop for nerds in control lifestyle products.
Thermal Overload
The threads that wouldn't die...
- PC reliability?
- Windows, real time
- PID loops
- PCs vs. PLCs
- Replacing people
- MS 'monopoly'?
- Software quality
- Where do we go from here?
- Why pay?
- PC reliability?
- Windows, real time
- PID loops
- PCs vs. PLCs
- Replacing people
- MS 'monopoly'?
- Software quality
- Where do we go from here?
- Why pay?
Fortune
I have yet to see any problem, however complicated, which, when looked
at in the right way, did not become still more complicated.
-- Poul Anderson
at in the right way, did not become still more complicated.
-- Poul Anderson
RSS Feed
www.control.com/rss/
To get a personalized feed, become a member at no cost.
I have to set the total flow value, and system has to close the valve after total flow sensed by magnetic flow meter.
Do I need a flow totalizer in addition to flow meter, and can a PLC handle the batch function as mentioned above? (to get the flow rate from flow meter, totalize and compare it with preset value and send out command to valve).
Do I need a flow totalizer in addition to flow meter, and can a PLC handle the batch function as mentioned above? (to get the flow rate from flow meter, totalize and compare it with preset value and send out command to valve).
Hello arfp1343;
It all depends on how precise you have to be in your calculations.
Total flow is (Sum of [flow/time *time]); So it is not too complicated to take a reading every 10 seconds, calculate the flow (for the past ten seconds), sum 6 of these readings and multiply by 60 to get total flow for 1 minute, and increment a register every minute with the latest flow value.
The problem is in a PLC the time readings tend to drift (which means your 10 second interval cannot be guarantied). So readings will be slightly off, and over time the actual value of total flow will not match your information.
To minimize the error, you could use a scheduled task for example, if available in your PLC.
If you can get a reading of total flow on your flowmeter (even if you don't have a transmitted value) you could have the operator read the actual total flow and "correct" your totalizer regularly to make it match with the flowmeter calculations.
Some PLCs offer alternatives: Rockwell's Controllogix has a hardware module that is dedicated to totalizing applications (1756sc-CTR8
8-Channel Counter/Flowmeter Module; Siemens offers a downloadable totalizer application program for their S7-300/400 PLCs that can be implemented in a scheduled task (OB35) to ensure stability of the calculations.
Hope this helps.
Daniel Chartier
It all depends on how precise you have to be in your calculations.
Total flow is (Sum of [flow/time *time]); So it is not too complicated to take a reading every 10 seconds, calculate the flow (for the past ten seconds), sum 6 of these readings and multiply by 60 to get total flow for 1 minute, and increment a register every minute with the latest flow value.
The problem is in a PLC the time readings tend to drift (which means your 10 second interval cannot be guarantied). So readings will be slightly off, and over time the actual value of total flow will not match your information.
To minimize the error, you could use a scheduled task for example, if available in your PLC.
If you can get a reading of total flow on your flowmeter (even if you don't have a transmitted value) you could have the operator read the actual total flow and "correct" your totalizer regularly to make it match with the flowmeter calculations.
Some PLCs offer alternatives: Rockwell's Controllogix has a hardware module that is dedicated to totalizing applications (1756sc-CTR8
8-Channel Counter/Flowmeter Module; Siemens offers a downloadable totalizer application program for their S7-300/400 PLCs that can be implemented in a scheduled task (OB35) to ensure stability of the calculations.
Hope this helps.
Daniel Chartier
most plcs these days have some means by which a very precise timebase is available. Integrating off that time base is quite accurate as long as you are careful about rounding errors.
Could you explain your calculations?
If available, connect a pulse output from you flowmeter to a PLC input/counter. The only error involved is precision of your flow meter itself. Be sure to configure the pulse rate for the On and Off period at max flow rate to be longer then the PLC Input specs and scan time.
Sorry to deviate slightly, but can you advise on where I can find the downloadable Siemens totaliser software? I may have an application for it in the very near future.
Thanks.
Thanks.
Hello Shaun;
You will find the file Totalizer.exe on this Siemens support webpage:
http://support.automation.siemens.com/WW/view/en/23061633< br>
Look at all the other applications available on the same webpage, at the top. If you have a few minutes, open the Applications and Tools menu at the page header; there you can access many functions developped by Siemens IT group (for PLC, HMI, communications, drives, etc); some are free downloads, some require that you pay a minimum fee (around 25-50 euros per download) through a Simatic payment card. But these applications are tested and supported; and they can save you a lot of development time.
Hope this helps,
Daniel Chartier
You will find the file Totalizer.exe on this Siemens support webpage:
http://support.automation.siemens.com/WW/view/en/23061633< br>
Look at all the other applications available on the same webpage, at the top. If you have a few minutes, open the Applications and Tools menu at the page header; there you can access many functions developped by Siemens IT group (for PLC, HMI, communications, drives, etc); some are free downloads, some require that you pay a minimum fee (around 25-50 euros per download) through a Simatic payment card. But these applications are tested and supported; and they can save you a lot of development time.
Hope this helps,
Daniel Chartier
I have found one solution:
http://plctalk.net/qanda/showpost.php?p=247913&postco unt=3
I'm not sure if it is valid...
http://plctalk.net/qanda/showpost.php?p=247913&postco unt=3
I'm not sure if it is valid...
You will need to sample your flow rate at an accurate, constant rate (dT). Depending on the accuracy required, you should also consider using something like trapezoidal rule numerical integration to get more accurate results:
Total(n) = Total(n-1) + (Flow(n) + Flow(n-1))/(2*dT)
Finally, you need to consider round-off errors as your total value becomes large in relation to the small increment in flow you add every calculation period. Using double-precision floating point for the total calculation can help.
Rockwell Automation's ControlLogix and CompactLogix controllers have a built-in Totalizer instruction made to handle this exact application and which uses the above techniques to improve accuracy.
Total(n) = Total(n-1) + (Flow(n) + Flow(n-1))/(2*dT)
Finally, you need to consider round-off errors as your total value becomes large in relation to the small increment in flow you add every calculation period. Using double-precision floating point for the total calculation can help.
Rockwell Automation's ControlLogix and CompactLogix controllers have a built-in Totalizer instruction made to handle this exact application and which uses the above techniques to improve accuracy.
You know, there's a relatively easy way to do this. If what you are trying to do is to totalize flow, many flowmeters have a scaled pulse output (one count equals n volume units). A simple low speed counter then works as a totalizer.
If you also need to display flow rate, you may have to also pull a 4-20 mA pair. At that point you need to decide what the aggregate work burden is, and decide accordingly.
Walt Boyes
Editor in Chief
Control and Controlglobal.com
www.controlglobal.com
Mailto:wboyes@putman.net
Read my blog SoundOFF!! At www.controlglobal.com/soundoff
If you also need to display flow rate, you may have to also pull a 4-20 mA pair. At that point you need to decide what the aggregate work burden is, and decide accordingly.
Walt Boyes
Editor in Chief
Control and Controlglobal.com
www.controlglobal.com
Mailto:wboyes@putman.net
Read my blog SoundOFF!! At www.controlglobal.com/soundoff
All these fancy timing methods are OK, but don't forget the accuracy of your flowmeter is probably +/- 0.5% at best. So what's the point in using super accurate timers?
Before real time clocks we used timers & counters with a few seconds correction at midnight with quite accurate results.
Roy
Before real time clocks we used timers & counters with a few seconds correction at midnight with quite accurate results.
Roy
From Control Engineering magazine...
Related articles from Control
Engineering magazine- Liquid level sensor: Non-mechanical optical technology extends application possibilities
- Wireless manager for instrumentation and control devices
- Level sensor: Magnetostrictive device extends mounting options
- New podcast: One road to wireless instrumentation
- Tutorial: Instrumentation / DCS integration languages, part 1, EDDL
- FactoryTalk industry-specific applications launched
- Torque: Hall-effect sensor, reaction torque sensor, torque data acquisition
- Transient blocking circuit protection via acquisition
- Automation Fair event focuses on convergence, sustainability, and technical talent
Above articles copyright 2008 Reed Business Information.
Subject to its Terms of Use.
Your use of this site is subject to the terms and conditions set forth under Legal Notices and the Privacy Policy. Please read those terms and conditions carefully. Subject to the rights expressly reserved to others under Legal Notices, the content of this site and the compilation thereof is © 1999-2008 Control Technology Corporation. All rights reserved.
Our Advertisers
Help keep our servers running...
Patronize our advertisers!
Patronize our advertisers!




