PID AUTO TUNE not working for V570

[kvj]

HI,

I'm having a problem working with PID on V570. I have 2 units, all the same configuration. The first one works, and the second one couldn't pass through the AUTO TUNE, Status stuck at 2.

I am attaching the program, I would appreciate if you can help me. 

Again the same program works perfectly on the first unit, but get stuck on PID Status=2 on second unit.

Please advice.

 

REgards,

kvj

 

 

 

[AlexUT]

kvj,

Please provide more details, then community will help you better.

1. Post project file.

2. Publish sensor type, output /actuator type, settings, what kind of device you manage - heater/cooler, etc.

3. Other details.

B.R.

 

[kvj]

Hi,

1. File attached

2. 4-20ma signal type, heating process

3. How long does it take for a 3 stage process? The other system works fine, all the same using the same program.

 

Regards,

kvj

 

HTR1.vlp

[Simon]

Hi, these are just my quick comments, but one key thing with autotune and heaters is to choose the right setpoint.  The autotune can get "stuck" or just take a really long time if the wrong setpoint is used.  Autotune drives the process to the setpoint, then applied alternative step changes to the setpoint, about 5-10% above and below the setpoint.  It measures the step response of the process to these changes in setpoint.

For example, if your ambient temp is 23° C and you choose a setpoint of 30°C the process will ramp up nicely to 30° (or 33°).  However for the negative ramp (back to 27°) it will take a very long time, especially if the process vessel is large and/or insulated, as the difference between the process temp and ambient temp is very small.  On the other hand, if the process temp is 90°C and the vessel is relatively small and uninsulated, the autotune will cycle much faster, as the vessel is able to lose heat quickly once the heater is tuned off.

Hope this makes sense, it seemed like a complete mystery to me until I worked through it.  My desktop test system is a 5W ceramic resistor inside a plastic cup 🙂

[kratmel]

One tip for autotune....

Please check carefully is it heater for this zone  conected to the correct PID output (via hardware and via software). 

I found one installation where part of heater from one zone is connected to another one and PID AUTOTUNE  stuck

[kvj]

Hi,

 I’d like to ask, I’m just trying to do simulation using one of my spare unit and using memory to feed in SP and PV. I can mimic the issue I have and when I try to use the PID SERVER tool to START AUTO TUNE, it forces my program to get pass to STATUS-4. It works fine, but I’m concern when the system powers off and on, does it need to auto tune again? I tried on my unit to power cycle, it still retains the value and its works perfectly now.

 

My main question is, is doing the auto tune using PID SERVER tool will work for good ? What will be of any possible issues later on, if any?

 

Thanks.

 

Regards,

kvj

[AlexUT]

Hi kvj,
Your question include two different parts.

My main question is, (is doing the auto tune using PID SERVER tool will work for good ?)[1] (What will be of any possible issues later on, if any?)[2]

Part [1] can be answered:
PID Server is running on PC.
It can consume much more resources (CPU, memory), and it may make PID Autotune more precise. But it is dependent of other conditions - PID startup parameters, Sensors/Actuators models and phisical location, inertion of sensors/actuators, etc.
And good side of PID Server is possibility to save PID parameters graph.
You can use this graph for feature reference, send it to someone for rteview, etc.
And yes, you can change PID parameters and make optimization by compare results of an original PID Autotune calculated PID parameters and changed ones.

Part [2] is acadenic, so everybody can add answers in accordance to experience.

B.R.
 

[kvj]

Hi, 

Thanks guys so much for the info.

I tried to RESET and INITIALIZE and then do the auto tune again using the PID SERVER, and NOT WORKING.  What I am doing wrong in here?

Pls advise.

REgards,

kvj

[Flex727]

Can you swap the PLCs between the two systems?

[AlexUT]

Does PLC you refer now connected to real system, or to kind of "simulator"?

 

[kvj]

Hi,

I am not onsite, and using my simulator unit.

I'm checking the consistency of the PID server tool if I can use this one next week onsite.  I can't see anything that I have done differently except that I RESET and INITIALIZE and start over. Now the PID server auto tune isn't working too.

 

[AlexUT]

If it not work with simulator, then control loop does not similar to real one, or PID startup parameters are wrong.

First, you have to look for description of status=-2.

1. Clean MI 304, as it calculated by PID Autotune.

2. MI 222 should be 1000, and you can use it to indicate percentage.

3. In linearization swap 4096 and 1000.  *4096 is wrong. Should be 4095 (range is 0 to 4095).

MI 2000 will be in range 0 to 4095 and you should link it to Analog Output.

4. In PID Autotune select 2 cycles and run it again.

Does it work now?

B.R.

 

 

[kvj]

Hi Alex,

Thanks for the input. I am yet to see this tomorrow.

 

So far, I have not seen this status during the test. 

I'll try to follow yours and see how it goes. Much appreciated.

 

[kvj]

On 10/10/2020 at 9:21 AM, AlexUT said:

000 will be in range 0 to 4095 and you should link it to Analog Output.

Hi,

Unfortunately the issue is still there.

1. Adjusted all the ranges.

2. Tried Stage 2

3. The first unit that works only do auto tune in 5 to 10 mins.

I would like to try manual PID, how is this done?

 

Thanks

[kvj]

Any help? Basically, what I did is to get the parameters from the working auto tune unit, and encoded to the second unit for manual tuning.  It works for 24 hours and for some reason client made adjustment on the mechanical side and it trips again. And my question is, is this the right way shown below  for V570 forcing it manually for PID ?

 

[Joe Tauser]

I would not continually force a value to the PID status register.  The PID block should write that value all by itself to report what it's doing.

Here's how I do PID-

 

I also always make something on another screen to allow manual tuning-

 

As far as manual tuning goes, I posted this a while back-

Manual loop tuning 101:

Start with Kp = 100 (10.0%) and Ki = 0.  Then start lowering Kp until the process starts to just ring around the setpoint.  If it overshoots wildly you need a bigger Kp.  

Measure the time between peaks.  This is the natural cycle of your process.

Then take your Kp value and double it.

The units on Ki are seconds.  Enter the time between peaks value.

 

If you want a better understanding of the best way to manually tune a PID loop Google "PID Loop tuning methods".  My method is loosely based on Ziegler-Nichols.

Joe T.

 

 

[kvj]

Hi Joe,

I thank you for sharing the info.

Yes, it simply works on the manual tune. I simply cannot work with auto tune.

Much appreciated.

Cheers!