burner temprature & time control

hi every body
i'm facing problem in completing plc program which control burner through servo motor (control gas flow) analog output (0-10V) so i have to rise temperature but within specific time .... i think in pid control put i didn't found in parameters anything to control time ... if any one have any idea i'll be think full my plc is EATON PA Model

My first thought is do you have a good understanding of the thermodynamics of the heated object (s). In other words how many BTU it takes to heat it XX degrees. Also you will have to add increasing heat loss to ambient as temperature rises. From all this you will then be able to get to BTU per time or gas volume per time. When you know this it may be a bit easier to do the PLC.

Dan Bentler

you mean i have to modeling the system ??? i think that's difficult.... all i need to have is closed loop to bring temp from existing degree to desired smoothly !

Sounds to me like you're trying to achieve a ramped setpoint.

I'm not familiar with your make of plc but I assume that if it has built in PID function then there will be a specific register which holds the setpoint for the PID loop.

This register will need to be incremented on a periodic basis, e.g. for 100 degrees per hour the register needs to be incremented by 1 degree every 3600/100 seconds = 36 seconds.

This can be achieved with a self resetting timer which, in my experience, is normally accurate enough for this type of application.

You may also want to stop the pid loop and make the setpoint register track the process value until the burner is lit.

mek_control said:

you mean i have to modeling the system ??? i think that's difficult.... all i need to have is closed loop to bring temp from existing degree to desired smoothly !

Click to expand...

Yes you have to know how the machine operates and what you want it to do to install effective controls.

You could do this emperically ie turn it on and record temperature vs time. YOu would have to know BTU input (or gas flow rate whichever you prefer)

Dan Bentler

At this point you don't even know if what you want to do is possible with what you have.

mek_control's project sounds like a cascaded loop application where the servo controls the error between the valve set point generated by the outer temperature loop and the actual valve position. The inner loop part is not the problem.

mek_control said:

so i have to rise temperature but within specific time

Click to expand...

Do you even know if the system is capable of heating with the specific time?
You don't need ramps. You need closed loop control with faster closed loop time constants.

you mean i have to modeling the system ??? i think that's difficult....

Click to expand...

A crude model will do if you are lucky. If the plant time constant(s) are long you will not be able to heat the plant up quickly. If the plant gain is to low you will not even get to the higher temperatures. Dead time will also require the controller gains to be lower and this will slow response.

If the dead time is long then a Smith Predictor will be required. Unfortunately a Smith Predictor does require a model to predict with.

If you know the plants gain, time constant(s), dead time and the required response time I can calculate the gains for you and tell you what kind of PID to use in the PLC. Not all PIDs are created equal. You want a PID that will handle output saturation nicely and not suffer from integrator wind up.

all i need to have is closed loop to bring temp from existing degree to desired smoothly

Click to expand...

I thought you to do it a a certain amount of time. So you need to increase the temperature quickly and smoothly without over shoot.

You can see you haven't provided enough information for good answers but now you know what to look for.

I have several of these systems in my plant that we have refitted. I must ask is this an existing system or a new install. On a system like this you will require an external flame controller that will give an output to the plc. If you skip this you will eventually blow up whatever you are heating.

I believe the poster above is correct. Find the pid function in your plc and read the help file on adjusting the parameters.

On a side note we use seperate pid controllers hooked up to the flame controller which also controls the servo and take the valves encoder input, as well as blowers and ignition. The plc will handle the remaining sensors and outputs including pilot lights. I am not a huge fan of using thermocouples or load cell inputs on a plc for whatever reason. I prefer a separate module even if I'm using a hmi.

califflash said:

On a side note we use seperate pid controllers hooked up to the flame controller which also controls the servo and take the valves encoder input, as well as blowers and ignition. The plc will handle the remaining sensors and outputs including pilot lights. I am not a huge fan of using thermocouples or load cell inputs on a plc for whatever reason. I prefer a separate module even if I'm using a hmi.

Click to expand...

Why include a plc at all, if only for lights?

He said
"The plc will handle the remaining sensors and outputs including pilot lights".

I believe he may be referring to the flame ignition type of pilot lites.

Dan Bentler

Trust your PLC, but use a watchdog timer per NFPA 86C and FM approved High Temp Limit controller in addition to your FSR (Flame safety relay).

All the equipment vendors are pushing for safety relays which are hyped up overpriced BS but at a minimum take an output from a card with critical controls and keep a simple ATC/ IDEC etc. off delay relay alive whose output is in series with the master safety relay, safety gas valve. That way if PLC locks up it drops out the skid.

Cascade the temp control so the servo position is inner PID loop and the the outer "loop" or more simply a timer and proportional only control algorithm pushes setpoint to inner loop.

Usually you don't have to do this as the servo (honeywell motor etc. ) is electric and so slow there is plenty of lag in the system.

Don't forget you can just as soon play with the linkage on the valve to get some response tuning.

Just remember when the system is tuned, get some bright orange spray paint and liberally mark every piece of field valve actuator linkage so "joe night shift" doesn't screw up the combustion controls.

hey every body

thanks to give attention to my case
i also want to add something here because all of u talk with his own range I'm working in eaton plc so this is my pid wizard

http://img710.imageshack.us/f/pidr.jpg/

how to prove your ideas with this wizard that's the point .

thanks again

From your screen shot I see the PID interface is set for auto control. You want forward control and set the Integral and differential to either 0 or 1% and adjust proportional untul you get a response you are looking for.

the 2 loops are temperature (SP)- actual process temp (PV) = error

this error is cascaded into your gas flow loop which becomes flow SP - actual flow rate (or valve position) = error

for burner applications you will need some small minimum flow so your flame is stable so you need to have a low fire minimum setting tha can be done vith linkage or minimum output from PID loop (i think for your PLC that would be LOWER BOUND OF ....)

the problem you fall into with PID loops in a PLC is converting engineering units into PLC units. 0-2000'F must be scaled into 0-4095 (or whatever the units are for your PLC. perhaps the max number for you is based on the screen shots 16 or 32 bit which like allen bradley would relate to 16383 or 32767. (this is what I imagine the 30,000 % means). so if you are scaling your thermocouple make a note of the min - max that is scaled to and the PID % or range is a division of that range. That is. if your PID controller is deviating by 10% proportional and the units are 0-16383 then a 10% deviation is 1,638 points (or in this case 0-2000'F would be 200'F proportional band.

Every PLC's proportional control algorithm is different. most define the deviation as the max number you set is the deviation (value / units) at which the output is 100%. For example if the temp is 0-2000 'F and prop band is set to 20'F (in this case still in 'F units) then when the temp deviates by 20'F the output is 100%. Deviation by 10'F iand output is 50%. etc.

Unfortunately PLC's don't think in 'F so the scaling issues.

Don't forget the OUTPUT of the PID loop is also NOT in 'F so 0-100% output = 4-20mA = ????????

often you need a rescale of the CV (aka output) so you can scale and this allows compression or skew of the output and can makeup for sluggish or overactive loops.

even if your PLC did NOT have UPPER BOUND or max output parameters you can do a compare of the output and clamp it with a conditional that replaces a GRT than value with a different value.

you can use a compact controller pid for the temperature, all you need is something to measure the temperature , then if you adjust on 80 degrees you adjust the time as well all this in 10x10 cm small instrument , no plc's

I agree with adel... no need for plcs, and specially no need for Servo Motor, why would you use it?

On both ways if you need a PLC, just take the the reading of the thermocouples and enter it into control loop to acheive the ON/OFF control for the high and low temperature required.

I have done this on a burner and it worked fine. no need for plcs at all. the PID controllers does it all.