Author Options:

need advice for process control software Answered

I am new to process control software. We are working on a solar steam project that will require some form of automated control. We need suggestions for software (preferably open source and will run on a windows pc) that will handle temp and pressure data aquisition. The software should also be able to affect small electric motors based on the data parameters. Also any recommendations for temp and pressure (psi) sensors would be greatly appreciated. Thanks for any suggestions


What temp are you playing with, what PSI, how many sensors do you want, what type of environment are the sensors expected to work in...And you said a small electic motor how small is small ?? fractional HP ?? Or tiny DC motor kinda small ?? Is there any wiring constraints for the sensors, thing's like one is 900 feet away from the PC ??

Big, There will at least two temp sensors and two pressure sensors. The sensors will be placed in a steam boiler and or steam engine. The sensors will be 50 to 100 ft from the pc. The motors will be used to drive single axis mirrors. So I am guessing that small dc motors will suffice.

Maxim/ Dallas semi has some nice temp sensors, and there one wire sensors are easy to use, you could use some thing like the DS18S20 for temp, and they have some ADC for you to link a pressure sensor to your system like the DS2450 I don't know what PSI your looking at but you can link almost any pressure sensor to this ADC...

Now about the motor controller they offer a DS2406 which is an addressable switch and this can control transistors, etc.

About the software, I don't really know of a place with Free software made for your application (( This is why custom electronics controls are so expensive, software is not free )) Maxim does offer several software SDK's and they have includes several examples of basic thermostats and they offer technical assistance and a web forum all about programming so go HERE and take a look, honestly programming and using these devices is not that hard (( I'm not good at programming but I have gotten some one wire applications to work, With a little help, but you only have to ask for it ))

Oh I almost forgot you need to build or buy a one wire interface, to interface your computer to your devices, I built mine and it goes to my 232 port but they offer USB ones now and I just used the data sheet and built the interface like it said to..... only I added a few blue Leds...

Thanks Big Bwana -- this is very helpful.

No problem, and don't tell every one this but Maxim Offers Free Samples of all those parts too

. While I have used several process control packages, they were all specific to to a PLC system (eg, A-B PLC-5) or "mainframe" (eg, Rosemount RS-3). Nothing general-purpose or free. ;( Googling "process control" +software turns up 2.5M hits, but I didn't see anything that looked like it would be free on the first few pages. Maybe you can find something if you dig a little deeper.
. You may be able to talk someone out of a copy of obsolete (for them) software. Try asking at industrial supply stores and process control engineering firms.
. You may be able to find a very simple package by searching for "PID control". (Proportional-Integral-Derivative) or "PI control" (you may not need derivative).
. When it comes time to tune your loops, search for "PID tuning".
. Sensors are usually not that critical. For your purposes, just about any brand/style/&c sensors will work. Just don't get the cheapest you can find.
. Your final elements (valves, &c) will be more important. For controlling turbines, you will probably want some relatively fast acting valves, especially on smaller turbines.
. Your computer probably isn't going to put out the type of signal that the valves/&c will be looking for, so budget for some transducers. Same for the inputs.

Found free PLC software, for only the company's own PLCs of course, but it will do relay ladder logic programming and has a simulator. Heh, offhand it looks like RSLogix (program for Allen-Bradley PLCs). Relay logic is generally the "language" of process control in industry, so this looks like a good way to learn it.

I think a PID controller sounds like overkill, for actual running this sounds like a mere thermostat and pressure switch application, but then I've never used them before, so... This could be a split application, switches to actually run it, sensors to gather info while it's running. Sure that means you'll be analyzing the data then manually setting the switches, but that'll reduce the installation cost and should yield a more robust system, thermostats and pressure switches rarely fail and don't care about voltage spikes and other things that can kill electronic controls. Suitable thermostat controls can be salvaged off an old furnace for a hot water system as it is standard for such boilers to do both circulating hot water and steam applications. Pressure switches in theory could be taken off an air compressor, some of those go to rather high pressures and offhand I don't think the water in the steam will bother it, compressed air is "moist" and condensed water is a problem with air compressors.

Thanks forgesmith. I will check that out. When you mention manually setting the switches, do you mean doing this always or only during initial setup. We do need to make the system primarily automated so that ultimately it will start and run without manual intervention.

I mean set the switches for whatever range or certain value you want for initial start up, then as you gather data you can manually adjust them for optimum running under normal expected conditions. Then you're done, "set and forget."

There's a possible maintenance issue. Furnace temperature controls are robust and normally will outlive the furnace. They use mechanical electric relays but it's rare for the contacts to ever need cleaning. They're actuated by a temperature sensing "bulb" that goes into a short cylindrical chamber that goes into the boiler itself, thus is well protected, a thin copper tube goes from bulb to control, a fluid or gas inside the bulb yields pressure which does the switching. Practically nothing there that can go wrong. Application notes: In the US the standard is the controls run on 120VAC. There are terminals for the house thermostat, that part of the system will feed current to the circulator pump terminals. Your application would use the current fed to the burner terminals as the control will maintain boiler temperature regardless. You can use a relay with a 120V coil to switch a control voltage. Check with a heating parts supply store, there are simpler units available without circulator controls, some may even be all-mechanical and have contacts you can use with your control voltage.

Pressure switches as found on common air compressors are normally all mechanical, with mechanical movement of the electrical contacts. Over time the settings can drift, adjustment nuts and screws work loose. Which is why we have Loctite threadlocker, when you have them where you want them just put on a drop or two of the penetrating variety, it'll work itself into the threads and keep things from coming loose. Being all-mechanical they don't care what voltage they have as long as it doesn't exceed the ratings, so a low control voltage will be fine.

Both type of controls would prefer to be out of the weather, keep them in enclosures where they'll stay dry. If you can get them both as all-mechanical and only doing low-voltage low-current control line switching then DIY enclosures are fine, otherwise electrical codes may get involved which will rack up some extra costs rather quickly.

Thanks forgesmith. I new to all of this so the info is a big help.

. Yep. PLC software is usually tied to a particular PLC, but, if it comes with a simulator, makes a great ladder logic (AKA relay logic) learning aid. The eWare32 site mentions a simulator.
. I got out of the business about the time A-B switched the name of the software to RSLogix from referring to it by the A-B catalog series number (5200, IIRC).
. For some of the turbine controls, simple on/off and/or manual should work, but for speed and/or flow control, a PID block is, IMNSHO, the only way to go. I guess it depends on how much automatic control is desired and how accurate it needs to be. I do tend to look at this from an industrial point of view, instead of DIY, but I'm used to working on $1M+ systems where spending a few $K for software or special I/O modules wasn't a big deal. ;)
. A wall t'stat is a PD controller with very high gain (or low proportional band), if you really stretch your imagination (no I; anticipator providing the D in one direction).

What exactly are you doing? small steam turbines or a sterling engine? I'm also interested in solar, so please let me know if there is anything I can help with.

We are going to use the steam in a steam engine. While we are interested in a turbine (tesla or otherwise) we want the engine to be able to be built with basic tools. We need to software to automate and control the system.