adccommunitymod (AutomationDirect) asked a question.
Created Date: February 17,2009
Created By: twitte
**** This post has been imported from our legacy forum. Information in this post may be outdated and links contained in the post may no longer work.****
Hello, I am using a AD 250-1 PLC with analog TC and RTD input cards to measure temperature. I am having problems with the signal being affected by a VFD located nearby. I have tried to filter the signal by using the logic in the manual, but this had almost no effect. Is there any type of filter (lowpass) on these input cards? Other manufacturers (Allen Bradley,NI) have noise rejection built into these input cards, so this was not a problem for me. I have tried moving the VFD as far as possible away from the PLC and mounting it close to the motor (3ft), but this did nothing either. If anyone has any input on this problem I would greatly appreciate it. Thanks Tyson
Created Date: February 17,2009
Created by: KPrice
Tyson, I can make some suggestions that you may already be doing.
Use ungrounded shielded thermocouple probe.
Use a metalic T/C head and run conduit such as metalic sealtite.
Use twisted shielded T/C cable, grounding shield only at one end.
(Usually a good ground is only available in ground bar in control cabinet.)
Install PLC in metalic grounded control cabinet with branch circuit grounding conductor.
Connect -T/C wire to 0V at module.
Hope this helps.
Created Date: February 17,2009
Created by: twitte
Thanks for the info.
I am using shielded wire, but I do not have it grounded. I will try this now.
Thanks
Created Date: February 18,2009
Created by: FACTSTech
When using a VFD and Thermocouple and RTD's you MUST follow good wiring practices. For example, VFD must have input reactor, T/C cable must be shielded, where this shield is terminated depends on the application. The RTD must use Belden 8403 or equiv. shielded cable. These cable runs must be separate from mains and switching lines. The VFD if speed controlled by PLC must be 4-20mA not voltage. All inductive loads must be surpressed with a transorb or equiv. PLC ground, panel ground, 24V power supply, negative terminal of 24V power, should all be properly connected to the same ground stud on the panel.
Created Date: February 18,2009
Created by: twitte
I have concluded that the noise is coming into my RTD and TC lines and then into the PLC. I concluded this by placing a RTD with a 4 " wire lead directly inside my enclosure and closing the door. The reading from this RTD stabilizes, but still is a little affected my the VFD (+-1 deg F). Does AD sell anything to filter this noise out? I am already using shielded cable that is grounded inside the box. The reason I am frusturared is that the devices I have used prior have all been able to filter this noise (VFD) without me doing anything. I still have not installed a line reactor, I suppose this is the next step.
Created Date: February 18,2009
Created by: KPrice
twitte, sorry you are having this problem, I know you do not want to do this, and that you want what you have to work. I am just suggesting for possible future installation to look at T/C-4-20ma transmitters. You still need good wiring practices, twisted-pair, shielded cables, good grounding, etc. but it might save you some hassle. I just noticed that ADC offers RTD/4-20ma transmitters.
Created Date: February 18,2009
Created by: DenisOZ
Unfortunately this is a well known problem with the Facts RTD input modules - 06, 205, & Terminator.
We have been using these for several years with very mixed results. Their immunity to RF is pretty terrible.
I still use the RTD modules for convenience when necessary, but there is no rhyme nor reason to their performance, despite all manner of grounding techniques.
Generally, the shorter the run, the better. I have had great success with 06 PLCs configured as RIO slaves on refrigeration compressors where the runs are a couple of metres, but I have also had vicious noise from other installations when less than 5 metres from the PLC.
I have done extensive tests with common set point temperature controllers (Omron, RKC, Shimaden) on the same circuits that are unreadable with the Facts units, and invariably the readings on those devices are rock solid.
On one 16ch Terminator installation, I fitted ferrite cores to all inputs to try and settle the inputs down - with little effect.
The convenience of the temp input modules can not be denied - and where they work they are a great cost effective solution.
As KPrice stated, the obvious way to overcome the noise inputs is to use 4-20ma inputs. Their lower impedence seems to work well - but it is an expensive hassle for the PLC and field end of the installation.
My usual solution to the problem is to program in heavy averaging - easy with the new Filter Over Time ibox - but this can lead to instablilty and errors if the timebase is too long. A filter time of 2secs and divisor of 30 will work OK without becoming unstable, unless there is severe input excursions.
I have posted about this every couple of years, but I am surprised that it does not come up on the forum more often
I don't think that our EMF here in Oz is much different to that of the US. Our supply voltages are higher, but that generally works in our favour.
I have tried all manner of extra filters at the VSD sources, and in a couple of cases removed the shielded output cable and fitted expensive replacement with little effect.
On one site, the noise can not be coming from VSDs, because there are not any on site, but RF from somewhere is causing havoc with a 25 ch RTD installation in a freeze store. I guess the metal sandwich panel building acts as a giant antenna. I am at the end of my tether with this one, and another PLC vendor (Allen Bradley) has been invited in to prove that theirs will work at the expense of the ADC gear.
My next tack is to try some 3rd party RIO vendors using modbus ( Wago, Phoenix, Weidmuller) and see how they perform in the same environments.
Best of luck with this, and if others are having the same problems as we have had, then lets see if it can at least get acknowledged at Facts end. I love ADC gear, and it burns me up to see other vendors with a possible solution to this issue.
Created Date: February 19,2009
Created by: twitte
Yes, I feel the same way. It is a shame because everything else on my setup seems to work great, but beceause of my application (industrial dryers) i need a lot of temperature inputs. These units are always installed in a production line environment where I have no control over what other noise generators are present. It is important to me that I have a system that is reliable. It seems evident to me that I will have to look at a different PLC. Any idea if the cards in the 405 setup have the same issue?
Created Date: February 27,2009
Created by: DenisOZ
Response???
Bumping - Where are you Facts?
Created Date: February 28,2009
Created by: a agnone
Interesting, we have several ovens using J type thermocouples. The PLC is the 405 series. Nothing special, just high temp J wire welded at the tip. They run straight to the plc card, no junctions. They run alone, no voltage wires in the wire duct to the panel. Some of the runs are 30 feet. But again it is stated this problem seems to be with the RTD units.
Created Date: March 02,2009
Created by: twitte
I installed a line reactor prior to the VFD and this substantailly reduced the amount of fluctuation I am getting in my reading, to the point where I can use this to control my gas furnace. This worked fine in my environment, however, I am still a little worried about using this in a customer's environment where there are motors that are seperate from my equipment. Is there any possibility that FACTS will develop a card in the future that will come with an analog filter (can be adjusted for frequency) like some of the competitors offer? Thanks.