Sign in to follow this  
Followers 0
mplummer

RSlogix 5000 and PIDE

10 posts in this topic

I've never posted before and I hope someone can help me with my lack of knowledge. I've used RS Logix 5000 before but I never had to use the PIDE function blocks. I am trying to run heats with a contactor, but all the info that I've seen has been for analog outputs. I need to fire an output on/off to control the heats. I am using a 1769-L32E and a IT6 T/C card for me control the temp. Any help would be great. Thanks.

Share this post


Link to post
Share on other sites
No problem if you can cycle the contactors quick enough so that the temperature doesn't change too much just by switching off-on-off. Think of the CV output of the PIDE as the percentage of time the contactor should be closed.. then either use some logic or a specialty card to cycle the contactor to realise this percentage. eg. CV = 0.6 means (picking 10 seconds just for example) a 10 second cycle with the contactor closed for 6 seconds Further reading: http://en.wikipedia.org/wiki/Pulse-width_modulation

Share this post


Link to post
Share on other sites
For more precise control, or heat/cool/idle control, look at the SRTP Function block. It is designed to take an analog CV output, and convert it to a split range, time proportional control.

Share this post


Link to post
Share on other sites
It is perfectly OK to convert PID loop output into timing function. You also need to establish duty cycle time and apply PID output to it. Also limit your min. value as you do not want to have your contactor to switch as crazy. Personally I would never recommend to have a traditional contactor in this case - only solid state relay. Your contactor will not survive for long time.

Share this post


Link to post
Share on other sites
Sounds good except that it is hard to find very large solid state relays, at least pre-built ones. It was not clear from the description if this was electric heat or something else. A vaccuum contactor generally has much higher survivability when it comes to cycling in the mechanical designs. With an electric heater you can turn down the voltage if you have a way of doing this...if you can control voltage, then you could use direct proportional output without cycling. DC drives can do this but they are getting both rare and expensive. It also isn't too hard to take a basic "lamp dimmer" circuit (which is just a wave chopper) and scale it up with larger triacs to make a truly huge solid state relay even if it has to be pieced together in a custom rig. Just look for a triac big enough to do what you need and ask the company for documentation on a proper driver circuit. You can look on the internet for 555 timer circuits to generate the proper timing based on an input zero crossing pulse.

Share this post


Link to post
Share on other sites
I am using solid state relays to run my heaters, they pull about 7 amps per zone. I am try to hold them on for about 15 seconds max and about 2 seconds min. I can get them to come on but they are not wanting to shut off at temp. I can't turn down the voltage, all I can do is to turn the contactors on and off. I am try to heat a large metal barrel for a plastic molding machine, that is why I can't just leave the heats on until they are up to temp, the temp will run over. I am now using the PIDE with a SRTP function it seems to be heating but it is hard to control. Thanks for the help.

Share this post


Link to post
Share on other sites
Is the PLC output going logicly 0 and the electronics are not shutting off? 1/ Check wiring, particularly pull up/down resistors if needed Is the output of the PIDE staying high when you hit high temperatures? 1/ Check tuning 2/ Try enabling PIDE block with PV = SP for the first scan it's enabled (just use SEL blocks to put zeros in for the first scan it's enabled).. it stops the PIDE overshooting due to the starting conditions (PIDE does this, PID doesn't due to different algorithms) Edited by MikeSmith

Share this post


Link to post
Share on other sites
Why? Those times are probably too long. The PWM period is at least 15 seconds. The PWM period must be much faster than the thermal time constant. Do you know what that is? I double the barrel has a very long time constant since it must be able to maintain temperature between shots. In your case the thermal time constant may even be shorter than the PWM period. If so it is no wonder that you can't control your system. The PWM periods should be at least 10 times less than the thermal time constant. I would by getting a idea of how fast the barrel temperature responds to changes in the heaters and figuring out a time constant.

Share this post


Link to post
Share on other sites
mplummer, We have integrated CompactLogix control retrofits of injection molding machines. You are right; an improperly tuned PID will have massive overshoots at cold-start. This is typically due to integral wind-up. The construction of injection molding machine barrel is where electric heat is applied to the "skin" of the barrel, but the measuring point (thermocouple or RTD) is typically buried to the core. This physical installation has a large thermal lag time between when heat is applied, and heat reaches the measurement point. Small machine = small problem, big machine = big problem. We migrated away from standard PID to PIDE. There is an AUTO-TUNE feature to the PIDE. We used the auto-tune and experienced satisfactory results. VERY IMPORTANT NOTES to consider: 1. You will NOT decrease the cold-start time using the programmable controller alone. 2. The heat up to temperature without overshoot required time is dependent on the existing physical installation. 3. Auto-tuned PIDE will allow you as the integrator to get the fastest response without over-shoot achievable with the existing physical installation. 4. Make sure each heater band is functioning in each heater zones. You can do this with an IR image of each band, and amp draw measurement on each band, or simply touching a sprue of plastic to the exterior of the band during heating. It is VERY important to have 100% healthy heating zones, prior to the auto-tune, to obtain the best results. Check the AB web-site for documents relating to auto-tune application examples for PIDE. We actually incorporated auto-tune on engineer accessed screens at the HMI for our convenience. Here is an overview of how the auto-tune works. It is absolutely mandatory that the thermal load is at a steady state. On a large injection molding machine, this typically is only after the barrel heat has been turned off for about 24 hours. If you invoke an auto-tune, you will only have one shot at it in any 24-hour period. The key reason for the barrel needing to be at steady-state (room temperature) is so that accurate data measurements will need to be recorded of system dynamics and capabilities during the tune. Auto tune steps. 1. Set each zones manual output to 0.00%. Set each zone to expected run time temperature (several hundred degrees above ambient) 2. Set each zone to MANUAL mode 3. Turn on all zones. They will sit at 0.00% output 4. Enable all zones to auto-tune The auto-tune feature will go on-line with each zone, and ensure there is a "steady-state" temperature, (neither rising nor falling). About 30-60 seconds later the auto tune will kick the output to your pre-defined max output level, (we used 100%) The auto tune will then internally start a timer. The auto-tune will be collecting key information about the dynamics of each zone. The first thing the auto-tune will document is the time from the start of applying heat, to the time heat is seen at the point of measurement. This is the "thermal-lag" time. The auto-tune will also find the "slope", or rate of change actual with a known percentage output. On a typical injection molding machine when several zones are auto-tuned simultaneously, the nozzle will finish first. (Least thermal mass). The center sections of the barrel will typically finish tuning later, (up to 10 minutes on a 2,000 tone molder, and the rear zone most likely will finish last, as it has the throat pulling heat away from it. Once all of the zones have auto-tunes complete. The auto-tune feature returns (3) set of suggested PID parameters. (Slow / Medium / and Fast). We could often use the Fast parameters, but if your IMM is 3,000 to 4,000 Ton, you may want to use medium gains. Once we read the auto-tune returned suggested PID parameters, we would enter those into the PIDE. We were quite satisfied, as we were using MATH, and not "tweaking" to achieve a "tuned loop". How do you know if you have done a job? It is important to collect data BEFORE tuning and AFTER tuning. If you have an HMI with trending feature, you should be able to graphically plot the temperature approach to set point before. (Sluggish? Aggressive, but with overshoots that take hours to recover?) Then after the auto-tune, and using the suggested parameters, you should graphically see a confident temperature approach to set point with minimal overshoot. Again, You will NOT be able to heat the barrel any faster by pushing buttons alone. The key takeaway here is that you will have the best values to heat-up and maintain the barrel temperature for molding operations. With SSR control of electric heat you do not need to have a MAX on of 15 seconds. During cold-start the outputs will be held on until the PIDE throttles back. SSR's can be switched frequently, I suggest a (5) second period for your main barrell zones, and a (2) second period for your nozzle zones. ie 50% CV for barrel would be 2.5 sec on and 2.5 sec on. 50% CV for Nzzle would be 1 sec on and 1 sec off.

Share this post


Link to post
Share on other sites
Plastic, that is a pretty good explanation but I think it misses a vital point. What are the PWM periods relative to your time constants? Does your auto tuning system pick the update times? How often are the temperatures sampled during auto tuning?

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now
Sign in to follow this  
Followers 0