Analog Proportional valve control

[irondesk40]

Have a question about controlling a 4-20ma steam valve by the use of a temperature process controller instead of a plc.
This is something that we have always done through the PLC, but we are going to put some new equipment in a plant in El Salvador and the management folks that we have been discussing this with do not want a plc to control the temperature, only use the plc to control the machine functions,etc, but use a temperature process controller to control a proportional steam valve.
Each machine will have about 6 areas to control the temperature in, so each machine would require 6 temperature controllers.
The maintenance folks in the El Salvador plant say it is easier for them to change out a temperature controller mounted in the front of a panel. I have to agree with them based on the fact that so far, I have not had any conversations with anyone from the plant that would be able to use a plc if they had a problem or had to autotune, etc.. With the panel mounted temperature controller, all that would be needed would be a simple document to explain how to enter the values you needed in the controller through its front display when needed.
Have used some simple Watlow controllers in the past to control simple electric heaters, and was wondering if there was any recommendation on which brand or model of process temperature controller would be the best for this type of application based on experience you may have had with them.

 

[Clay B.]

Honeywell makes several Temperature controlers. I know From doing some work in Mexico they had strong tech support south of the border.

Watlow and Honeywell are the 2 most common flavors of Temp controls I have seen.

 

[danw]

What you describe is cake for a Honeywell 1/4 DIN UDC Controllers (not the 1/8, 1/16 or 1/32 DIN), with current (4-20mA) output. I suspect a model UDC 2500 would do it.

reasons I like 'em:
- The displays are 12 segment displays so alphabet characters are readable, 7 segment icons of alphabet characters.

- Languages other than English can be selected.

- UDC's are as close to bullet proof as that category of controls gets. Very noise resistant front ends. Snubbed relay outputs.

- keypad has tactile feedback, you can feel the keys click. The rubberized bezel holds up well and protects the keyboard switches from dirt (not from acid vapor fumes though, as I found out)

- universal input is convenient - one input fits all - T/C, RTD or 4-20mA from a transmitter, same hardware, just wired and configured differently.

- replaceable parts. After a decade of considering the previous generation of UDCs 'disposable', the current generation has a complete parts list of spare parts boards, so controllers can b e expected to give 10+ years of service.

The auto tuning algorithm works well on thermal processes and offers a choice of 'fast tune' (allows some overshoot for quicker approach to SP) or 'slow tune' (no overshoot), depending on what the performance demand is.

- IR port works with a serial IR adapter for loading/retrieving config setups with a software package called P.I.E. My service guy who does boilers loves it.

The one thing the is missing is 24Vdc to run a loop powered transmitter, but I typically use a standalone power supply for that anyhow.

Dan

 

[TConnolly]

Sounds like a paradigm problem. Temperature controllers are not easier to change out than a PLC module. For one thing the temperature controller will have several parameters that must be reset if the unit is changed out. The PLC on the other hand will reconfigure the module on power up, and for that matter you can load the program onto a CF card or EEProm and have that reload the PLC program on power up. There is a PLC anyways, so how does increasing the component count increase reliabality?

That said, when it comes to temperature controllers IMHBMAO Yokogawa Green Series controllers are the only way to go.

 

[irondesk40]

Thanks

I agree about the plc, but some of the folks that I have been dealing with here lately as we are moving some of our production off shores, seem to have a fear of anything that involves using a PC.

Just the way it is, maybe if we all get lucky, after a few years of equipment not running, these plants that have been moved off shores will have to come back to the states, but I seem to doubt it.

I took a tour of one of our plants, and they paid the operators of the equipment about 90 cents an hour, so looks like it is there to stay.

 

[Andybr]

I would agree with everything that DanW said about the UDC's. They will do exactly what you want and will do it well. Unfortunately they will also do many things which you and your management may not necessarily want. This type of controller will typically provide facilities for manual control, set point ramping, auto tuning and many other functions over and above basic temperature control. My experience is that the operators will quickly discover these functions and will find an almost infinite number of ways of upsetting the system using the buttons on front of the controller. The built in security is basic at best and is no match for the average operator. You might find the PLC is far less trouble in the long run.
Andybr

 

[Mickey]

I have used these (see link below 500 series). Very easy to set-up and even have hot back-up models.

http://www.miinet.com/products/sg_process.shtml

 

[ghaulk]

I would suggest using the Series 96 Watlow Controllers. If you are just needing a single loop control scheme. These are inexpensive and very simple to set up. They can be ordered to output 4-20mA.

 

[_Woody_]

I have always liked the Omron E5_K series process controllers. For most of the same reasons Dan likes the UDC. The Omron has a wide proportional band and also programmable filters which make it great for PID control of fluid and gas processes. Also you can purchase it in bits (outputs types / comm's / Special inputs etc)