Skip to Content

Register Login

Industry Online Support

Technical Forum

Navigation

Thermocouple module update rate / conversion time

Created by: Dynamart at: 2/1/2012 12:20 PM (8 Replies)

Rating (1)

Thanks 0

9 Entries

2/1/2012 12:20 PM	Rate (0)
Dynamart Regular Member Joined: 8/8/2011 Last visit: 8/19/2024 Posts: 6 Rating: (1)	Can anyone clarify the module update rate for the following T/C modules as I find the information in the manual ambiguous: 6 x T/C 6ES7331-7PE10-0AB0 Manual states 'basic conversion time' 30/50/60/300. Is this the update time for all 6 channels? 8 x T/C 6ES7331-7PF11-0AB0 Again, manual refers to 'basic execution time of the module (all channels enabled)'. Is this the module convertion time? Can it really update 4 channels in 10ms? Thanks.

Suggestion To thank Quote Answer

2/1/2012 2:39 PM	Rate (0)
mralalog Regular Member Joined: 5/7/2010 Last visit: 8/6/2012 Posts: 13 Rating: (2)	Hello, Dynamart; Your question is not a simple one as there are so many dependents; The two modules you mention have analog to digital converters that sample for a period set in the hardware configuration of either 10HZ, 50HZ, 60HZ, or 100HZ. There is also overhead due to reading the converters, processing the data, wire break sampling, and self calibration.Overhead time adds to the sample time and results in the final update time you will see. There is a difference between update time and step response. Update is how often the module reports a new data point to the PLC. Step response is the time it takes for a step input on the module input to be reported at 100% of its value at the PLC. Step response is 3 to 4 times the update rate due to uncertainty of theoccurrence of thestep vs the sample time and internal filtering. If you select 10HZ update, you get very high noise rejection of 50, 60, and 100HZ interference. Alternately, if you select 100HZ sampling, your noise rejection of 50 and 60HZ will be not so good but your update rate will be faster. It is all a compromise. The 331-7PE10 has 6 A/D converters and will supply an update for all 6 channels in about 21ms at 100HZ sample ifwire break and self calibrationare disabled. Enable wire break and the update increases to about 52ms. The 331-7PF11 has 4 A/D converters and has an update rate for all 4 channels of about 21ms (wirebreak disabled) at 100HZ sample rate. Self calibration cannot be disabled on this module. If you must have aconstant data update rate for a critical process, I suggest you choose the 331-7PE10. If you have a more specific question on the operation of these modules, I would be happy to discuss. mranalog

Suggestion To thank Quote Answer

2/1/2012 4:26 PM	Rate (0)
Dynamart Regular Member Joined: 8/8/2011 Last visit: 8/19/2024 Posts: 6 Rating: (1)	Thank, that's a great reply. I've also redescovered the 8 T/C Ex module 6ES7 331-7SF00-0AB0 which will update all channels in as little as 10ms (400Hz, 9 bit) and is cheaper too.

Suggestion To thank Quote Answer

2/1/2012 4:47 PM	Rate (0)
0_Einstein Diamond Member Joined: 7/7/2010 Last visit: 10/3/2024 Posts: 15262 Rating: (2421)	Dynamart Thank, that's a great reply. I've also redescovered the 8 T/C Ex module 6ES7 331-7SF00-0AB0 which will update all channels in as little as 10ms (400Hz, 9 bit) and is cheaper too. Perhaps this link will help you spend your money wisely : http://www.omega.com/temperature/Z/ThermocoupleResponseTime.html Having a sample of 400 Hz when the response time may be no better than 1 sec is a rather expensive approach. RTD is for more accurate or up to moderately high temperature input (<800C). TC is cheaper and faster andless accurateand can read > 800C. Maybe this will help. http://www.jimpinto.com/writings/tempsensors.html I have had this discussion many times. We usually go with TC input and update @ 10-20Hz. We often want 1000Hz, but that's not realistic and can be misleading.
	science guy
Suggestion To thank Quote Answer

2/1/2012 4:56 PM	Rate (0)
mranalog Posts: 5 Rating: (1)	The 331-7SF00 has only one A/D converter. 8 channels are multiplexed. The manual is talking about the time to integrate one channel, not the time to update all 8.I think that with onlytwo channels enabled the ratecan be no better than10ms. (Multiplexed inputs can update no faster than 4x theintegration rate with this A/D converter) To get the fast update you must give up AC noise rejection, resolution and number of channels sampled. With 2.5ms sample rate, you noise will be pretty high if there is any AC noise present.
mranalog Posts: 5 Rating: (1)
Suggestion To thank Quote Answer

2/1/2012 6:24 PM	Rate (0)
Navnag Platinum Member Joined: 11/21/2007 Last visit: 9/25/2024 Posts: 2325 Rating: (308)	Hello In Most of the Processes ( That I am aware of )temperature always changes very slowly ( in Seconds ). Most of the Sensors also have an inherent delay in measurement especially when using Thermowells.. Ultimately having a T/C measurement in ms ultimately has no meaning. It may give you bragging rights but serves no useful purpose as far as control response is concerned. Hope this helps Cheers Navnag
	There is no such thing as the last bug in a program There are no stubborn bugs, There are only stubborn Debuggers.
Suggestion To thank Quote Answer

2/1/2012 7:09 PM	Rate (0)
mranalog Posts: 5 Rating: (1)	You are correct, temperature normally changes slowly. I do hear over and over again that fast (~10ms)update is needed and wonder exactly why. I also know that there is confusion about integration time, update time, and step response. They are not the same thing! Of importance is noise rejection. The AC line interference is always present on signal lines, even shielded ones. If you set the integration time to something other than the local line frequency, you will most likely see interference in your temperature reading.
mranalog Posts: 5 Rating: (1)
Suggestion To thank Quote Answer

2/2/2012 9:34 AM	Rate (0)
Dynamart Regular Member Joined: 8/8/2011 Last visit: 8/19/2024 Posts: 6 Rating: (1)	Thanks for all the input guys. I think I now understant the terminology in the manuals. BTW the application is gas turbine flame monitoring so T/Cs are needed as they are in the gas path (1200 DegC). I appreciate the comments about step response, but we need fast updates (<200ms) for rate calculations. We can't use constant bus cycle either as it's a H system. Ideally we'd want to use ET200S (as per rest of system) but the 180ms update + cycle variations give problems (can't go into more detail on SW as it's not my area). So I'm looking at S7-300/ET200M as an alternative. That, or use ET200S 2T/C as single channel (80ms).

Suggestion To thank Quote Answer

2/2/2012 2:02 PM	Rate (1)
mralalog Regular Member Joined: 5/7/2010 Last visit: 8/6/2012 Posts: 13 Rating: (2)	If you need to have fast updates and step response, wire break diagnostic before eachupdateand no missing updates, choose the 331-7PE10. There is an A/D converter for each channel with full isolation between channels. Set the integration time to the local line frequency, 50/60 HZ. Disable self calibration. You willhave 6 new updates in <<100ms and step response in <<200ms.(Thatmeans reporting -full scale to +full scale ifthe temperaturecould change that fast!) You get what you pay for. mranalog

Suggestion To thank Quote Answer

|

Share this page:

Share this page on...

Follow us on