Analog input offset 1761-IF2OF2

wildcatherder

Member
Join Date
May 2009
Location
Morro Bay
Posts
168
I have an MKS 1179A Mass Flow Controller connected to a Rockwell Automation analog expansion module 1762-IF2OF2. The output from the MKS1179A feeds into a differential analog input. It is necessary to tie the signal common to the minus side, since the output from the MKS 1179A is single-ended. Unfortunately, when connected this way, I get a 0.25VDC offset on the analog input. If I remove the positive analog input lead I do not see this offset, either on the MKS output nor on the disconnected analog input.

Can someone explain why I'm getting a voltage offset and what I can do about it?
 
You would connect the "Flow Signal Out" of your flow meter to the"+" side of your analog input card (channel ?) and the signal common of your flow meter to the minus side of your analog input card(channel ?).

I have to ask, do you have the dip switch on the analog card set for voltage?
 
I double-checked. Both DIP switches are set to OFF which should be voltage.

The connection is as you describe. I should note that there is an analog output set point voltage connection from the expansion module to the flow controller, as well. IT is a single-ended output and connects to the set point input and signal common.
 
It sounds like a ground loop offset.

If you completely disconnect the PLC AO, does the PLC AI still see an offset?

Is the power supply common grounded?

If messing with the earth grounding isn't suitable, you can use a voltage "loop isolator".
 
When I disconnect the analog output of the PLC, I still see the offset on the analog input.

The two power supplies in the instrument are grounded to earth ground.

I see less than a millivolt difference between signal common and earth ground. I was an electronics tech for ten years and this one is a real puzzler.
 
When I hear 'offset when I connect' I automatically assume ground loop. But it might not be electrical.

Some flow meter technologies need a periodic zero adjustment, others don't. Thermal dispersion does. Have you done the zero adjustment in the manual?


MKS_MFC_Zero_Adjustment.jpg
 
Analog input offset

Your suggestion will be kept in mind because I know I will need to tweak zero at some future point. The manual says that even mounting the unit vertically rather than horizontally can change the zero point. (Ours is mounted horizontally.)

However, this is a new factory-calibrated MKS Flow Controller.

More importantly, there is no offset when the output of the MKS Flow Controller is connected to a single-ended input on the main terminal of the MicroLogix 1100 PLC.

I cannot use that input because it only has a resolution of 1% of full-scale , full scale being 10V. Unfortunately, the MKS Flow Controller has a full-scale output of 5V, meaning that the effective resolution is only 2% which is not good enough.

The IF2OF2 module analog inputs have 12-bit analog inputs and have a rated resolution of 0.5% of full scale, which gives the required 1% over the MKS Flow Controller output range.
 
More importantly, there is no offset when the output of the MKS Flow Controller is connected to a single-ended input on the main terminal of the MicroLogix 1100 PLC.

Interesting. That says that the grounded (-) of the analog input on the single-ended AI does not create an offset. is 'grounded' and that the offset does not appear when the (-) is grounded.

If the signal is OK at the single-ended input, then either the offset at the module AI is either
- calibration zero issue (a 2.5% zero offset seems high for a new unit, but who knows?)
- a ground loop.

So simulate a single-ended input on the module's AI. Note that the module's common terminal is not connected to earth ground, according to the installation manual:

"The analog common (COM) is not connected to earth ground inside the module. All terminals are electrically isolated from the system" (pg 13)

So run a jumper from the (-) on the module differential AI to ground and see if the offset disappears.

Dan
 
I did connect "-" of the differential input to earth ground. The DC offset disappeared. I was surprised because I thought I had tried this before.
Then I noticed that the valve on the flow controller was "motorboating", at least that what we call it in audio. The DC offset has been converted to an AC offset with a frequency of about 1 Hz and was causing the flow valve to flutter on and off, because this same noise was now on the output to the flow controller, as well as the input. I can see the slow oscillation in the counts display of RSLogix 500 for the analog input.

I was an R&D electronics tech for 10 years and an electronics system engineer for 7 years. This is the weirdest ground loop problem I have encountered.

I have tried shunting the AC component of the noise to earth ground with a capacitor but values from .01 to 10 microfarads all result in the same fluuttering of the flow controller valve.

Although I have tried to contact both MKS and Rockwell Automation tech support about this problem, I think they are closed down this week, like many major tech companies.
 
Something I just learned today that surprised me a little bit:

The Micrologix 1400 DC powered models (and probably the ML1100) have their "DC Neutral" terminals connected to chassis ground internally. This might provide a path for a ground loop current to flow.

If you are using a DC powered Micrologix, try lifting the ground wire from it. I would not recommend leaving it that way.

As a better long term solution to your problem, I would recommend an isolated signal conditioner to convert that 0-5VDC to a 4-20mA signal and mount it as close as possible to the flow controller.

Example:
http://www.omega.com/pptst/DRSL-DC2.html
 
Last edited:
I don't know any way to "lift the ground" on the IF2OF2. It is connected to the Micrologix 1100 by a ribbon cable. The only signal common connection goes to the analog output of the PLC (must have that) and the "-" of the analog input.

It does seem like isolation is the only answer. Marketing will balk because we have been selling these at a particular price for years. The PLC is only six inches away (by wire) from the flow controller, which makes this especially annoying.
 
I don't know any way to "lift the ground" on the IF2OF2. It is connected to the Micrologix 1100 by a ribbon cable. The only signal common connection goes to the analog output of the PLC (must have that) and the "-" of the analog input.

I meant the Micrologix base power terminals could provide a path for a ground current to circulate and lifting the ground from the ground terminal of the base unit (if dc powered) might prove it.

wildcatherder; said:
It does seem like isolation is the only answer. Marketing will balk because we have been selling these at a particular price for years. The PLC is only six inches away (by wire) from the flow controller, which makes this especially annoying.

Wow, well at least you don't have a long run of potential induced current to study. Another thing you might gain with a signal conditioner is resolution. It is too bad the flow controller does not provide a 4-20mA option. I rarely have problems with analog current signal noise.
 
I see what you mean but this model Micrologix 1100 is AC powered.

I think we all agree it is a ground loop. The flow controller is never going to be used in the region below 5% flow, so I may be able to tweak the data, as long as the offset is constant, and still meet the 1% accuracy spec.

Thanks to all those who contributed.
 
Before you give up, how are you getting the ±15V dc power for the MKS? Two separate 15Vdc power supplies ? One p/s?

Have you tried lifting the common 0 volt connection of the power supply(s) to see if that eliminates the offset and the motorboating?

Dan
 
I found that the original design had a "optional" grounding lug. When I was called in to fix the project, I noticed it was completely disconnected and fastened it to a bolt through the mounting plate for the power supplies, flow controller and power supplies.

So, Danw, thanks for suggesting I don't give up and to check the grounds again. As soon as I removed this connection between the AC earth ground and the mounting plate, the offset disappeared.

Now I have to go find some diagonal cutters and permanently remove that grounding lug.
 

Similar Topics

On a Micrologix 1100 analog input, I appear to be getting an internal offset of about 11 counts out of 1023. Running the program from RSLogix...
Replies
5
Views
4,327
Good day all. I have a compact logix control that is receiving an analog input from a T/C. The program is written in FBD which I am not real...
Replies
1
Views
1,545
In this sample programming, what does U4 mean? Any assistance would be greatly appreciated.
Replies
8
Views
169
Can someone please explain to me in simple terms, the best practice with multiple 4/20ma pressure/flow transducers on multiple different 24VDC...
Replies
10
Views
339
Hello, I am trying to connect a 0-5v pressure sensor to the Micro 810, but I am not getting a clear input voltage reading. I connected the sensor...
Replies
14
Views
371
Back
Top Bottom