Interference from VFD Off to 4-20mA analog signal loop drive Actuator

[Bmw318be]

Hi

Hi

without knowing the control loop configuration and seeing the code it's next to impossible to determine the issue
what position the actuator? it goes to a predetermined position (100%) when the pump runs?
but if the actuator is at a commanded position and the feedback does not match/change I would start looking at the feedback sensor/code

without knowing the control loop configuration and seeing the code it's next to impossible to determine the issue
Ans: i would check this as my equipment is the pump


what position the actuator? it goes to a predetermined position (100%) when the pump runs?

It goes to predetermined location, for feedback and command also matched in scada monitor.

It drifted forward and backward if capacitor was not installed in the loop end before actuator board input

but if the actuator is at a commanded position and the feedback does not match/change I would start looking at the feedback sensor/code

 

[Ingenieur]

Did you measure the ac and dc current with and without the cap?

change the pre-determined position from 100 to 90% ?

4-20 is dc
most devices have an input cap to filter any ac
perhaps the induced ac is a harmonic of 60 hz (or 50, what is your standard frequency?)and the filter is ineffective

so the feedback and command are compared?
what happens if they don't match?
does the plc adjust the command/output and attempt to match?

is the cap in series with the loop or across the valve input (parallel)?
what value cap?

that fact that it still does it with the pump off is confusing

is only 1 end of the shielded cable(s) grounded?
Vfd, command and feedback

 

[Ingenieur]

Are the signal cables and vfd power in the same conduit?

 

[Electric-Light]

Using the AC function on the meter might tell you the presence of AC but its rather useless in telling you the nature of noise. The noise is a mixture of different frequencies and DMMs usually have trouble locking onto the Hz.

Can you get a few small crappy speakers like those from a clock radio and connect it in series with the loop? From the way you're connecting the meter it looks like you're connecting the meter in parallel.

Ideally it doesn't sing except for pops and clicks while making the connection and perhaps very slight background noise. If you hear stuff, that's the noise that's not supposed to be there. Our brain is a pretty good spectrum analyzer if the signal falls in the range of our hearing which a lot of industrial controls do. If it happens to be loud enough to be musical,post it. One of us would know what that noise is. Some noise are quite obvious as to their source.

 

[Bmw318be]

Hi thanks

Hi thanks

Did you measure the ac and dc current with and without the cap?
Ans: sorry, i am not into electrical, may i know which cap the dc and ac shall i measused.

I did measured the signal to my actuator board, take out the 2 cables and measured in series with fluke multimeter, it is stable. Current does not see fluctuation.this is the PLC. However as we hooked up to the board, it drifted forward and backward but did not changed the position in scada monitor.

change the pre-determined position from 100 to 90% ?

3 pattern without capacitor installed as filter in the input.

1. When actuator stopped, drifting fw and back but did not change the position.

2. Sometimes if goes from 40% to 100% by itself but takes overnight to happened.

3. Sometines goes to 0% from 50% but takes sometimes and overnight.

It does not happen in the local mode.

4-20 is dc
most devices have an input cap to filter any ac
perhaps the induced ac is a harmonic of 60 hz (or 50, what is your standard frequency?)and the filter is ineffective


Ans: My country is using 50hz. May i know if u referring to the plc device ?

so the feedback and command are compared?

Ans: It matched from plc, however the drifted was not captured in the plc and the command and feedback still matched but the position of actuator has changed.

what happens if they don't match?

If command and feedback match but actuator drifted to 100%, it caused errror to actuator board in the pump. Fatal error and need recalibrated.

does the plc adjust the command/output and attempt to match?

From what i observed, it does not have feedback that can control the actuator to do the adjustment. As we calibrated the board of actuator to PLc signal like initialize 0% and 100% and saved from the board program.

is the cap in series with the loop or across the valve input (parallel)?

Ans: need clarify where is the cap
what value cap?
And: need clarify
that fact that it still does it with the pump off is confusing
Yes the board of actuator just taking command, we do not see the issue with the handheld device fluke and local mode. Does it go to do with the feedback of the PLC ?
is only 1 end of the shielded cable(s) grounded?

Ans: may i know is it referring to analog signal or VFD ?
Vfd, command and feedback

My response above

 

[Bmw318be]

Hi

Hi

Are the signal cables and vfd power in the same conduit?

This cable are actually under a platform passing VFD Cabinet, we see thay it is mixed of power cable mixed with the control below photo.

Howver tried to by pass this long wired and to the main source to the actuator board, still having the drifting

signal

 

[Bmw318be]

Using the AC function on the meter might tell you the presence of AC but its rather useless in telling you the nature of noise. The noise is a mixture of different frequencies and DMMs usually have trouble locking onto the Hz.

Ans: yes i just like to confirm with a device it is interference not board error or PLC error. So can narrow down
Can you get a few small crappy speakers like those from a clock radio and connect it in series with the loop?

Ans: good idea, how about i using the usb speaker ? . I would compare the 2 source

1. PLC which we suspect has interference
2. Fluke which very stable

From this we can see the different.

May i know what should i expect from this test ?

From the way you're connecting the meter it looks like you're connecting the meter in parallel.

Ans: Ok let me correct it to series

Ideally it doesn't sing except for pops and clicks while making the connection and perhaps very slight background noise. If you hear stuff, that's the noise that's not supposed to be there. Our brain is a pretty good spectrum analyzer if the signal falls in the range of our hearing which a lot of industrial controls do. If it happens to be loud enough to be musical,post it. One of us would know what that noise is. Some noise are quite obvious as to their source.

Ans: are u referring to using the speaker ?

 

[Electric-Light]

Ans: are u referring to using the speaker ?

Yep.

 

[Electric-Light]

Your videos suggest you're connecting the meter in PARALLEL. It will check the presence of the current but it shorts out the circuit and prevent communication.

To place it in series, you need to CUT one side and slip it in between.

You could daisy chain the mA meter and the speaker at the same time if you want. The 8 ohm speaker only adds 0.16v at the most drop which has no effect on industrial 4-20mA loop so you don't need to remove them right away. You obviously need to address the cause of noise if they make a sound.

Where did you get the idea to place an electrolytic capacitor in parallel to a current source or current input?

I have a feeling your tuning is off too as others are saying. When you change directions on a machine handle, it moves a small amount before the table moves. If loosen to tighten has 0.1 turns of "dead turn" before engaging and switching from tighten to loosen has 0.05 but the controls apply 0.1 to both directions, it may seem ok for a while but as it repeats the it over and over, it will creep.

 

[Bmw318be]

Your videos suggest you're connecting the meter in PARALLEL. It will check the presence of the current but it shorts out the circuit and prevent communication.

To place it in series, you need to CUT one side and slip it in between.

You could daisy chain the mA meter and the speaker at the same time if you want. The 8 ohm speaker only adds 0.16v at the most drop which has no effect on industrial 4-20mA loop so you don't need to remove them right away. You obviously need to address the cause of noise if they make a sound.

Where did you get the idea to place an electrolytic capacitor in parallel to a current source or current input?

It is at current input before just my bord + -

I have a feeling your tuning is off too as others are saying. When you change directions on a machine handle, it moves a small amount before the table moves. If loosen to tighten has 0.1 turns of "dead turn" before engaging and switching from tighten to loosen has 0.05 but the controls apply 0.1 to both directions, it may seem ok for a while but as it repeats the it over and over, it will creep.

Sorry are u referring to mA that was tuned ?

 

[Saturn_Europa]

My response above

Have you tried using a mA output on your fluke to calibrate the analog board on the actuator?

 

[Electric-Light]

Ans: good idea, how about i using the usb speaker ? . I would compare the 2 source

You have to rip it apart and separate just the speaker element, so it will work as long as you're willing to ruin its function as a "USB speaker".
Headphones might or might not work. Some have 100s of ohms of impedance and those might affect the loop.

is the cap in series with the loop or across the valve input (parallel)?
what value cap?

I looked at his video and it looks like he's putting a tiny electrolytic capacitor and his meter across the current source. This doesn't seem to be proper.

 

[Bmw318be]

Have you tried using a mA output on your fluke to calibrate the analog board on the actuator?

Yes did, it worked perfectly, only when using the analog on scada at the site then the issue arise

 

[Bmw318be]

Hi

Hi

Your videos suggest you're connecting the meter in PARALLEL. It will check the presence of the current but it shorts out the circuit and prevent communication.

To place it in series, you need to CUT one side and slip it in between.

You could daisy chain the mA meter and the speaker at the same time if you want. The 8 ohm speaker only adds 0.16v at the most drop which has no effect on industrial 4-20mA loop so you don't need to remove them right away. You obviously need to address the cause of noise if they make a sound.

Ans: The use of electrolytic capacitor does help the drifted, however in the long run, I finf my actuator feedback does not tally with the position

Where did you get the idea to place an electrolytic capacitor in parallel to a current source or current input?
Ans: At the input parallel

I have a feeling your tuning is off too as others are saying. When you change directions on a machine handle, it moves a small amount before the table moves. If loosen to tighten has 0.1 turns of "dead turn" before engaging and switching from tighten to loosen has 0.05 but the controls apply 0.1 to both directions, it may seem ok for a while but as it repeats the it over and over, it will creep.

Ans:What
What is the dead turn ? .May I know if you are referring to my drifting. So do you think what info that you need to confirm this

 

[Bmw318be]

Hi

Hi

You have to rip it apart and separate just the speaker element, so it will work as long as you're willing to ruin its function as a "USB speaker".
Headphones might or might not work. Some have 100s of ohms of impedance and those might affect the loop.



I looked at his video and it looks like he's putting a tiny electrolytic capacitor and his meter across the current source. This doesn't seem to be proper.

Yes,

The capacitor are in parallel of the signal, but it stopped drifting, the issue after long run, it seemed my feedback does not match with the command of 4-20mA

 

[Ingenieur]

Yes,

The capacitor are in parallel of the signal, but it stopped drifting, the issue after long run, it seemed my feedback does not match with the command of 4-20mA

that's kind of what we figured
not really hardware, but control strategy/software

 

[Bmw318be]

Hi

Hi

that's kind of what we figured
not really hardware, but control strategy/software

Unity pro Xl scada is the softwared used

The VFD that we used is

https://s24.postimg.org/bnkg1lmg5/20170125_173413.jpg

This the local control panel

https://s30.postimg.org/hfq9bgxkx/20170126_182752.jpg

 

[Bmw318be]

Hi

Hi

Your videos suggest you're connecting the meter in PARALLEL. It will check the presence of the current but it shorts out the circuit and prevent communication.

Ans: Yes, because the terminal was connected to my actuator board, parallel it at the output signal allow me to see the feedback from my board.

I just realized that my actuator has a potentiometer which can be calibrated as output mA value

To place it in series, you need to CUT one side and slip it in between.

Ans: noted, so i do in input and output
You could daisy chain the mA meter and the speaker at the same time if you want. The 8 ohm speaker only adds 0.16v at the most drop which has no effect on industrial 4-20mA loop so you don't need to remove them right away. You obviously need to address the cause of noise if they make a sound.

Where did you get the idea to place an electrolytic capacitor in parallel to a current source or current input?

Ans: installer idea, any proper way of connecting the capacitor ?
I have a feeling your tuning is off too as others are saying. When you change directions on a machine handle, it moves a small amount before the table moves. If loosen to tighten has 0.1 turns of "dead turn" before engaging and switching from tighten to loosen has 0.05 but the controls apply 0.1 to both directions, it may seem ok for a while but as it repeats the it over and over, it will creep.

Hi

 

[Bmw318be]

Hi

Hi

What SCADA software are you using. Try trending the position command and the position feed back. If they are tracking it's as others have mentioned....a control philosophy or programming error. Tuning a PID is not for the faint of heart. I'd probably call in a system integrator.

If not you may hear the dreadful words "ever since you....."

It use Unity Pro X1 Scada

May I know how trends for this position feedback working principle is ?
Try trending the position command and the position feed back. If they are tracking it's as others have mentioned....a control philosophy or programming error.

 

[Bmw318be]

Hi

Hi

Have you tried using a mA output on your fluke to calibrate the analog board on the actuator?

The
board is having output that would need to be calibrated to the PLC display or the current input mA. I did calibrate this before. After sometimes, the output drifted up or drifter down

eg Input 40% Output 70%