Control System Design Philosophy

[ClydeMule]

Gentlemen,

First of all, I would like to thank everybody has participated in this forum. I have been involved for the last year or so and have learned a great deal about HVAC from this forum. You can only get so much from books.

I kind of work on my own and am self taught with a lot of things, so I would like to throw this question out to the group.

When faced with a simple control system to be designed/built I am often faced with making the decision between using a small inexpensive PLC or just using a number of relays.

For example, I build small (up to 30 ton) process water chillers, many for the medical market. MRI and CT Scanning. We build a city water switchover device, that will shut down the chiller and switch some solenoids to run city water thorugh the MR and down the drain for emergency cooling. A couple of time I have had engineers spec to have dual pumps in the chiller and the city water switch over so I had to design logic for when to switch over to city water. Without going into design details, what I decided to do was use (2) alternating relays, and a programmable relay, one of those ones with a timer setting and a function setting. It was set to pulse the relay 5 sec on and off when the input contact was closed.

THe cost of these parts was over $200.00. I could have done the same thing with simpler wiring for 120 using a er, small fixed IO plc. However, I shied away from the PLC (I am an experienced user and programmer) because it has an "air" of complexity that would freak out some people. Even though it makes the wirng a million times easier and the whole thing cheaper, it seems that when people see this little black box they tend to look suspiciously at the equipemnt. This is epsecially true (at least I have seen) in the HVAC industry. If it does not say Trane, York, Johnson Controls, Honeywell, etc, then it must be junk and/or from Mars.

In many ways, a more sophisticated controller can make a job more reliable, and often cheaper. But it can come at the expense of usability.

My question then is where do you draw the line between complexity and usabity? What criteria would you use on deciding on a control scheme for something as simple as I described above? Do you agree that there is a perception problem with controls?

Everyone wants something that is simple, but powerful, with an ifinite lifespan that will cost $2.79.

Thanks for you input.

Clyde

 

[MintJulep]

Not a bad question. Here a few things to consider:

Does the client have a preference?

What else does the client have, if they have other equipment with PLC's they are not likely to be scared of one.

How good are you at writing documentation. A schematic is pretty simple to produce, a good System Functional Description explaining the program may be more difficult.

A PLC is percieved to be more sophisticated than relay logic, even though the cost can often be lower. Thus you can charge more for less and put more $$$ in your pocket.

How reliable is the power supply? Some low-budget programable devices can lose their little minds if the power goes off for too long.

 

[ClydeMule]

Mint,

The client is not anyone specific, just in general. We build over a thousand chillers a year with a wide range of end users. We are discovering that what the Fabrication and machine tool market requires is much different than the HVAC market. When can I get it vs. when can you get me all of the data on the COP for this chiller during the full moon?

Different customers have different needs, but sometimes you have to dance between what the engineer and owener wants vs. the person actually pushing the buttons.

Our documentation is so so. I hate to admit this but it is true, and I write most of it myself. It is difficult for the one who wrote the program, wired it or designed the wiring, and conceived the whole idea to translate the details for someone who is just walking up to a machine for the first time.

You are right about the power supply issues. At first glance the PLC's are cheaper but it costs a lot of money to keep the from getting fried or from losing their program.

Clyde

 

[ChasBean1]

I'm mechanical and HVAC much more than electrical and electronic. That said, I would tend to favor protection for a $300,000(±?) piece of equipment on a macro scale (electric) rather than a micro scale (electronic). What do you think would be more reliable in terms of catastrophe such as a fire or flood or a worker accidentally running a sawz-all through a feeder? Will the cooking of a small component on a circuit board prevent protecting the equipment? Military ships still use sound powered phones because they don't rely on electricity... I would think in these types of terms for protective features. But then again, relays fail. Reliability is the key - if greater reliability means more compexity, so be it, but usually simplicity promotes reliability.

 

[dpoppeli]

If you used a PLC for the chiller switching application, how would you change time-values in the field? At least with discrete relays the customer can do it as required by turning a knob - you can even describe on the phone which knob to turn. With a PLC, they would need to figure out how to interface their laptop, get "y" companies proprietary software, learn to use the software, or pay someone $x per hour to do it for them. (Do they make PLC's with knobs yet?)

 

[twainagain]

One cost to be figured is that of special service calls and trips where the user is needlessly baffled by the PLC. I love the little Omrons myself, but the $100 savings won't buy much of my time.

Also, when the PLC fails, they're starting over, especially if the manufacturer has gone belly up or, worse, merged with someone. The relays can be quickly replaced by equivalents no matter the brand.

Don