# Thread: ?Create an Temp. Ave. network with thermistors?

1. Is there any decent documentation or can someone tell me how to create a temperature averaging network with using thermistors? I get the bridge idea with 4 thermistors but the Architect has some areas with 3, 7, 9, etc. He's got so many sensors that are unneccessary, but that's the print and I don't want to buy controllers to just be able to take his excess sensors. I figured I could create a network and save 30 points and \$\$\$ on controllers. I've got 1.8k inputs on my controls. Thanks in Advance.

2. Permanently Removed
Join Date
May 2002
Posts
9,542
Post Likes
If you know how to make series parallel circuits in even numbers just have 2 sensors in one enclosure. It's cheaper and easier.

3. That's what I'll do. If I am doing all the math correctly I can create networks using 4, 9, 10 or 16 sensors all at 18k.
4 - 2 Par. sets in series (900 ohms)x 2
9 - 3 Par. sets in series. (600 ohms) x 3
10 - 4 Par. sets in series with 2 Par. sets
(450 ohms)x 2 + (900 ohms)
I am still wondering though as electricity is mathematically exact that if a guy had x number of thermistors all at the same rating and just wired them all in series or even parallel I would think there would be an equation one could figure and write into the PLC? Of course you would have at least x^2 and maybe another x due to thermistors being non-linear. Maybe that's thinking too much but anyway thanks for the input - I'll be doubling sensors up as I got to get this job going asap.

4. Professional Member
Join Date
Oct 2003
Location
Minnesota
Posts
1,411
Post Likes
Originally posted by crab master
That's what I'll do. If I am doing all the math correctly I can create networks using 4, 9, 10 or 16 sensors all at 18k.
4 - 2 Par. sets in series (900 ohms)x 2
9 - 3 Par. sets in series. (600 ohms) x 3
10 - 4 Par. sets in series with 2 Par. sets
(450 ohms)x 2 + (900 ohms)
I am still wondering though as electricity is mathematically exact that if a guy had x number of thermistors all at the same rating and just wired them all in series or even parallel I would think there would be an equation one could figure and write into the PLC? Of course you would have at least x^2 and maybe another x due to thermistors being non-linear. Maybe that's thinking too much but anyway thanks for the input - I'll be doubling sensors up as I got to get this job going asap.
Actually you could figure out an equation and use it. It'd not be very accurate over the whole scale, without getting very complicated. Better would be to create a lookup table. With the table consisting of KNOWN and tested values measured at regular points throughout the whole expected temperature scale. ie If one had a theoretical thermistor with a rating (for accuracy) of from -20 to +180 degrees. That's 200. Actually measure resistance at every 20 degree increment. Results would be a table that'd look something like:
-20 ... N1 Ohms
0 ... N2 Ohms
20 ... N3 Ohms
40 ... N4 Ohms
etc.

Then one could use well none and proven math formulas to look up actual readings in ohms, and calculate a matching temperature to be displayed and used. This methodology works well for nonlinear scales. And will produce a result far more accurate than trying to do the whole scale with a single formula with constant values. For more accuracy, increase your sampling. ie Measure at every 10 degrees of increment, or every 5, etc.

This is the methodolgy usually used for nonlinear scales if accuracy is important. And is in fact the reason many manufacturers provide in their product literature a chart showing their product's typical values at a range of conditions. So one can build a table and do the math if needed.

Makers of DDC and PLC equipment often support this by including canned routines in firmware or software which allow one to simply point an input at such a table, into which you've plugged in the values, and it'll do the rest. Such routines are called various names. Characteristic curves, curves, nonlinear response tables, piecewise curves, etc. There doesn't seem to be any agreed upon naming convention I've ever found.

Back to the subject of using parallel-series circuits to do temperature averaging. It can be done, certainly. And routinely is. But it has some issues.

The more sampling points (thermistors) used, the greater the possible error rate and faults.

ie Kele makes a line of averaging sensors using thermistors. The 8 foot model uses 4 thermistors. The 12 and 25 footer uses 9. To get reasonable accuracy the manufacturer actually tests and measures each thermistor that's gonna go into a single assembly. Because despite the best, most carefully controlled manufacturing eviornment, no 2 thermistors are gonna be exactly alike in their response throughout the scale. Just slap em together randomly and yah can end up with a bunch of additive errors, and a final output that sucks for accuracy.

So they test and match. If this one has a slight offset this way, it's paired with one which has a slight offset the other. Etc.

Get my drift?

Now, for a 4 thermistor averaging circuit, yah can usually get by with little worries without testing. And get reasonable accuracy. I've done it, and seen it done countless times. It was a common practice when I worked for a major teleco for monitoring equipment rooms with a lot of electronics and many hot and cold spots in the same room.

And actually wiring it, wasn't too complicated.

Now, doing a 9 thermistor averaging circuit starts to get more complicated in the actual wiring, plus you're starting to get into numbers where yah need to really worry about additive errors. The reason Kele, for instance, bench tests and matches thermistor sets.

At 9 thermistors, field wired for 3 sets (of 3 thermistors paralleled) in series, you've dramatically increased the odds that the installer is gonna make a connection error. And you've likely introduced more error due to conductor resistance because you've now got a total of quite a bit of wire out there plus any added resistance due to a less than perfect splice.

Can be done, I've seen it field installed a couple times. Installer being very careful to get it right. But most folks avoid doing this routinely. A situation too ripe for errors and problems. And it can be bloody hell to troubleshoot.

A 4 thermistor field installed averaging circuit? Fairly common. More than that? Rarely done. Often more trouble than it's worth. With the teleco I used to work for, and with the company I currently work for, past 4, we go to separate points and just have the controller collect all of them and do an average.

After all, just how much does a point REALLY cost? At your buying cost? Compared to the time it'll take an installer to pull and connect the wires to make a 9 thermistor averaging circuit ... without making mistakes. And then testing it to make sure you've got decent accuracy and then adjusting input offset to correct if necessary? Add, if someone cuts a wire, or a single thermistor fails or wanders significantly, troubleshooting to find and fix problem can be a bear.

The answer is for you to decide. Me, depending on which line of DDC equipment I'm dealing with, I can get another 8 inputs for as little as \$25, my cost. A straight 2 wire pull for each thermistor, which means little chance of connection errors. Plus if one cable is cut, or single thermistor fails, no biggie, you'll see it right away and know just where to look.

For a different line of controller, with intelligent and addressible sensors, it'd be a single pair pull with thermistor sensors all in series, and then tell controller to average. For this system, individual sensors and the controller itself are significantly more expensive than above example, put the wire pulling is simpler and easier. (6 of one, half dozen of the other)

IMHO, I wouldn't routinely attempt more than a 4 thermistor, field installed averaging circuit using parallel-series wired thermistors to a single input. The more often you do, the more chances for trouble. Wouldn't take but a couple times of fixing a problem to cost yah more than you saved.

Disclaimer: I am NOT an expert. The above is simply my opinion and experience. I'd be interested in knowing other people's experiences if they've done this with more than 4 with regular success.

In either case one can still look at individual point if troubleshooting and see troublesome hot or cold spot.

5. Builing is 2 floors at 192' x 56' ea. There are approx. 16 coils on one floor and 19 on the other. Heating coils in the ductwork. For every coil there is an average of about 2.5 sensors per coil that was spec'd. I have about 85 sensors total, which is crazy to me but that was what they spec'd. It isn't like there is a bunch of little offices either. Most is pretty open. Anyway I'll have enough extra points available to do the averaging with 4 thermistors for a majority of areas. Thanks for the input on going greater than 4 as we were commenting on how you would get a proper correction factor and offset. I too would be interested in hearing what other guys have done.

6. Professional Member
Join Date
Oct 2003
Location
Minnesota
Posts
1,411
Post Likes
Originally posted by crab master
Builing is 2 floors at 192' x 56' ea. There are approx. 16 coils on one floor and 19 on the other. Heating coils in the ductwork. For every coil there is an average of about 2.5 sensors per coil that was spec'd. I have about 85 sensors total, which is crazy to me but that was what they spec'd.
<Shrug> Heck, as long as they're paying the bill, I wouldn't argue with them. I'm not sure of their rational for it, tho. I can understand equipment rooms full of high value electronics. Or a customer of ours who is in the bug business. Don't laugh. They grow and sell pedigreed, sterile (as in no parasites, no bad bacteria, etc) bugs for medical research, biological research, etc. And it evidently pays well ... VERY well. Customer obviously has a lot of money to burn, and spares no costs when it comes to climate control in the various sections of their establishment. EVERYTHING is monitored, with backup monitoring for critical items. Plus they pay us to log into their system daily, examine trends and so forth in an effort to identify problems before things go wrong. Plus we've a standing maintenance and repair contract that's generous, but with a provisio that we have someone on site within 2 hours of any alarm or call.

I guess there's gold in them thar bugs!

In any event, in the two examples I gave, I can understand the logic of extra sensors and monitoring. But an office building????

One of our customers is a branch of what is one of the largest, most successful law firms in the world. The kind of place where, if you're one of their most sought after customers, you're greeted by a personal hostess. You CAN smoke inside if you wish, and they'll provide genuine Havanna cigars if that's what yah want. And some 100 year old limited edition Scotch, or a rare Madiera if that's more to your taste. Some \$100 an ounce caviar? Here, have a pound. A whole floor of their establishment is devoted to their Department of Suck Up and Ass Kissing. (What the gals who work there call it.) There's a whole staff of gals, and a virtual warehouse of expensive and rare goodies to eat or drink, ready to come running to fulfill any request by anyone on the "upper floor". (Where the prime customers are taken)

Heck, they had us outfit the place (7 floors) but settled for plain old one sensor per room, except for a couple conference rooms. Seems to be adequate for them, they're happy.

Interestingly, I found out that a bunch of the junior legal partners on the "lower floors" are a bunch of snobbish, snotty snits. But the senior partners on the upper floors? Pretty regular bunch of guys. Plain talking, a lot of common horse sense, weren't the least bit snotty or snobbish.

Numerous times on the lower floors I and my guys were given a hard time by the junior lawyers. Wanted us to work at night. Don't make a mess, and they'd nit pick a speck of broken ceiling tile left behind. Move a desk to get at something? Hell, they'd measure to make sure it was back exactly in the original place. Several rigged cameras to operate while my guys were in their offices to make sure they stole nothing, or didn't even touch anything they didn't absolutely have to. Couldn't use their restrooms, had to use the one on the lower floor that was reserved for the 'ordinary people'.

LOL ... I'm not exaggerating. It sucked, it really, really sucked.

When it came time for work on the two upper floors, occupied by the senior partners, I was worried. Already had a 3 inch stack of papers that represented the complaints formally filed by those on the lower floors.

Those 2 floors with the senior partners? A pleasure to work in. We were treated cordially. Were told we could do any work during the day that could be done relatively quietly. No hammer-drilling, etc. Some secretary was given the task by THE senior partner to check with everyone, and give us a schedule of when this or that lawyer would be out of the office. We were welcome to work in any such day or night. THE senior partner himself dropped by and chit chatted with us. And mentioned that the best coffee in the place was in the Executive Partner Lounge, and that we were to be his guest there. Grab all the coffee yah wanted, any time. After hours we could use it for a lunch room. But kindly leave it decently neat. Etc. We could even use the executive bathrooms.

Chuckle, he said, "Hey, we're all working men trying to make a buck and feed our families. And it's best if we mutually cooperate and coordinate so we can each do our jobs while interfering with each other as little as possible. We'll each do a better job that way, won't we?"

One night as we were gonna have lunch, next thing we knew a couple folks showed up. From some high class catering place. With pizzas. High class pizzas. Made by someone who really knew his or her business. With all the condiments, assortment of drinks, large salsds, etc. The head of the pair mentioned that the "Big Guy" has ordered it and was footing the bill.

LOL ... I'm not surprised he was THE Senior Partner. And the snots and snits and wannabes were on the lower floors.

He knew how to get things done. My guys were especially careful and courteous while working those 2 floors. And didn't have to be told to be so. He'd respected them, they were paying him back for it.

Geez, if only I had more customers like that ....

7. yeah swing some my way if you get too many customers like that.
Anyway,back to reality, what control system is doing all the monitoring for the pet place?

8. Permanently Removed
Join Date
May 2002
Posts
9,542
Post Likes
Originally posted by crab master
Builing is 2 floors at 192' x 56' ea. There are approx. 16 coils on one floor and 19 on the other. Heating coils in the ductwork. For every coil there is an average of about 2.5 sensors per coil that was spec'd. I have about 85 sensors total, which is crazy to me but that was what they spec'd. It isn't like there is a bunch of little offices either. Most is pretty open. Anyway I'll have enough extra points available to do the averaging with 4 thermistors for a majority of areas. Thanks for the input on going greater than 4 as we were commenting on how you would get a proper correction factor and offset. I too would be interested in hearing what other guys have done.
ASK them. They probably would lower their number from a practical standpoint.

9. Professional Member
Join Date
Oct 2003
Location
Minnesota
Posts
1,411
Post Likes
Originally posted by amigo
yeah swing some my way if you get too many customers like that.
I only wish.

Anyway,back to reality, what control system is doing all the monitoring for the pet place?
At the Bug Palace? Automatrix.

The original system, the beginning, was put in years ago. 1997. Customer has never seen any reason to change it to something else. They only add.

Yes, it's a dinosuar system ... in some people's minds. But it gets the job done, is highly reliable ... as reliable as any I've seen, including the newest DDC controllers. More reliable than many. And AAM has maintained a remarkable degree of backwards compatibility.

ie We do a lot of add-ons and expansions for previous customers. It's common for one of our installers to open a box and install an additional controller alongside an existing, older item that's not even made any longer. And they co-exist on the network and in the database on the front end, no problem. And when a customer updates front end, as most have done, no problem. Front end understands old, legacy stuff just fine.

There is still a demand for propriety systems. Automatrix is what we use in this role. Altho we also do some BACNET using AAM devices.

As concerns the company for whom I work, we've tried a bit of this and that over time. And have finally settled on concentrating on Automatrix and TAC.

Actually the main company made that decision. We're affiliated with, but separate from, a larger firm.

However, we agreed with the decision fully. Agreeing that it was better to become VERY good at a couple alternative systems, as versus trying to do everything half-assed.

TAC was decided upon because we've past experience with them, and with Andover. And it fits with the fact we also do card access, security and surveillance, and fire detection and protection systems. And we want to be in the LON side of things, also. Not many customers demand or even ask for LON (or any supposedly 'open protocol'). But some do. And, in particular, some customers we'd like to do more business with.

We do some other DDC systems for service and repair. But no longer support new installs using them.

There are a variety of reasons for our decision. But not least among them are issues of cost and practicality.

Cost:
As with most things, the more yah buy from the same vendor, the better discount you can usually get. We like discounts. We'd rather a strong relationship with a few vendors, than a tenuous one with many.

Plus, inventory and ordering is simplfied. Techs and engineers have a standard "approved" item list they order against. This has greatly reduced all the "orphan" items in the back storeroom. Leftovers or overstocks from previous jobs, which no one really wants to use on new ones. Yah know; the oddball type relays, special purpose sensor, different actuator that only useable in a few cases that'd maybe justify it's expense and difficulty of installation, so on and so forth. Inventory was killing us. Keeping track of it made more difficult because previously we might have a dozen techs who favored two dozen different cube relays. We got hard nosed and reduced that down to a small handfull. Big emphasis on multi-usage. ie We don't stock or order single pole cubes at all. And all have LED indicators. This also reduced the number of different bases we required.

Etc, and so forth. The idea being to be merciless about reducing shear numbers of different items. If we can identify ONE item that can do the job of 3 or 4 others, that's what we're gonna use, in most cases. Even if, per unit, it costs a bit more. We feel that in the long run, we actually save money. More bulk buys with attendant price discount, less time on paperwork to order and track things, stock em, count em, etc. Less of the endless shelves and bins in the back room holding "orphans" that just collect dust for the most part. While representing a surprising amount of frozen cash if yah actually inventory it and count it all up.

Which we did. And were blown away with the final numbers. NOBODY, inhouse, suspected we had that much cash tied up ... and useless ... in that stockroom.

Practicality:
Fewer parts a service tech needs to haul around. And fewer misc consummable items a person needs to wire, attach, install, etc. The number of different screws, nuts, bolts and other fasteners; mounting plates; adaptors; fittings, etc that a guy had to have was drasticaly reduced.

Fewer install errors. Tech has and uses fewer different devices and assemblies. He soon becomes very familiar with them. Learns the tricks to making install faster and easier. Knows what will fit and what won't in a given amount of space. Knows exactly what auxilliary fittings or items he need to get it done. And he soon knows all the terminals or other connectors by memory. It's old hat, he knows what to do ... and what not to do. It still happens in special cases where it can not be avoided that he might have to use an odd-ball item. But it's few times and far apart. We still make sure he has ALL the cutsheets or manuals for every item he's gonna use. But mostly, he doesn't need em.

Fewer surprises, fewer gotchas. Faster, more reliable work.

The idea extends to the controllers and their accessories and add-on items. To actuators. Etc.

This also increases productivity of PM's and engineers. They KNOW what they're working with, from long familiarity. Fewer and smaller endless parts lists a PM has to make up, order, track, distribute, or find a source for. Fewer ordering mistakes. Tech calls and says he needs 2 pole cube and base, 24 volt coil. PM knows exactly what guy wants. Duct mount temperature and humidity sensor? Only question needs to be, for AAM or TAC? And both PM and tech knows exactly what is being talked about, what to expect, how it's gonna be mounted and terminated, setup parameters, etc. No surprises, no mistakes.

This suits most of our customers, too. They like it.

Just the other day I dropped by a customer for whom we'd recently completed a job. A casino. One of those customer care things. Stopped by to glad hand, ask em how they liked what we did. Any problems? Any concerns? Etc.

They have an inhouse maintenance and repair staff. Foreman of the HVAC division mentioned 6 VAV's seemed to be behaving oddly. What he saw on front end graphic screen didn't match measured readings displayed. Temps and such being maintained were fine. But, for instance, he showed me one which said reheat valve was shut. And staying shut. But you could see by DAT that heat was being modulated open and closed. And when he changed SP, everything stayed operating as before. Well, the short of it was that after I investigated, dispensed with pretty pictures and looked at raw insides of controller I found that front end screen display was showing data for floating point motor control. But that actual control of reheat was via analog channel, on an analog output. A few minutes, and a ladder revealed that in fact VAV had a 2 to 10 VDC Belimo motor on it.

As did 5 other units Casino HVAC guy was puzzled by. All 70 something others had floating point motors. VAV controllers were set up correctly to operated 2-10 VDC valves. Not a problem. But front end graphics guy hadn't changed screens to indicate 6 VAV's were different from others. Just a matter of what SP was being shown on screen, etc.

I told guy I'd get it fixed. By end of day I had a guy show up and we changed motors, rewired harnesses, setup controls, etc. To make those 6 VAVs identical to the others. Job had been originally managed by another PM. I managed to get hold of him (the account had been turned over to me) and inquired. Answer I got was "Oops", essentially. They'd somehow been 6 valves short, he'd had 6 suitable valves and motors sitting around collecting duct, but with 2-10 VDC motors. So he'd sent them to the site installer and told him to use them.

When I informed HVAC foreman for customer that problem was fixed, and not a screen fix. VAV's now worked exactly like all the others, same motors, etc. He was extremely pleased.

And we chatted about how he just friggin hated it when hired contractors had no standards, did things this way here, another over there ... in the same installation. With no rhyme or reason. Used different parts to do exactly same thing, in many cases. Making his job more difficult. ie Relay failed. Maybe one of 20 doing very same job. But perhaps 4 different sorts used. Grab spare relay, and find that instead of just unplugging one and installing another, had to change terminal ends, move wires, maybe add pigtail to extend a too short wire, have to drill new mount holes, etc. Or, he had to stock 4 kinds of relay for same duty purpose. Same with valve or damper actuators.

A simple example. I know. But I hear this sort of thing a LOT. From almost every customer we have.

They like standardization, and things being the same. Even the customers for whom we do LON work (TAC). They DON"T WANT a mixture of TAC and other vendor's controllers. They want the fewest possible different sorts of controllers used. To the most extent possible, they want exact. plug-in replacement. Not just same points listed on front end. Literal unplug and plug in replacement. Wiring harness the same, mounts the same, tools needed the same. PID loop parameters on new controller should be same as last, and produce identical results. Etc, and so forth.

Or, as close as one can get. And a damn good reason for any variation.

So I read some of the discussions in this group with interest. Sometimes amused interest. Not amused because I think I'm smarter than anyone else. In most cases, I see other people's points and think their reasoning sound, as far as it goes. But much of the discussions aren't of any use to me. And I'd not use them.

Not because they're not good ideas and ways.

Because it doesn't fit our philosophy nor what our customers want and ask for.

What do they want? First, forget debates about LON vs BACNET vs Propriety. Most of our customers know little of that debate, most couldn't care less. Or nearly so.

And even those who are well versed in that discussion have their hip waders on when the subject is brought up, because they're pretty sure all sides are gonna be feeding em some really deep BS. Generally they see pros and cons in all 3 sides of the argument. And don't think any one side is a clear and uncontested winner.

Mostly what they want is reliability and good workmanship. Known, reliable equipment, which has no unexpected surprises or gotchas. KISS. Reasonably long term supportability. Most have been burnt in the past by buying something that was outdated, no longer made, and unsupported with 3 to 5 years. Most have seen the worst examples of folks in our business. Who hacked together a system that only did part of what was spec'd, and that not well. With cooperation and followup support ending shortly after the warranty period.

Mostly, they want something that works EXACTLY as the spec reads. That's it. If they get that much, no arguments or excuses, claims of add-on costs not made clear up front, etc they're reasonably happy.

Got any idea how often I go to a site and listen to customer ***** about last guys who never made what they did work the way they said it would ... exactly, no excuses, no reqests for more money? Damned near every last one of em. Whether they're talking about older legacy systems, or the newest, latest and greatest LON miracle system.

LOL ...

That's why I get amused at some of the discussions or debates here. Where guys argue about whether LON or BacNet is better, or hotly debate merits of this or that wham-bam gizmo as concerns which is best at talking on the widest variety of comms methods, speaks how many different propriety and open protocols and lingoes, can or can not act as a web server,and if it can ... which does it better and has the easiest front end. Oh, and which can do all the above while baking cookies on the side and shining your shoes. So on and endlessly on. Lots and talks about the latest shiney toys and all the tricks they can do.

All the while, all my customers spend their time talking about, "Just make it work, and work RIGHT ! And we'd really like it to work as well after the warranty expires, for at least a few years.

Chuckle, no lie. One customer this past year is an example. Have had their control system reworked extensively, or completely replaced ... 5 times in 12 years. Several different controls systems, from pneumatics to DDC. Work done by several different contractors. But whole system never worked to customers satisfaction. Always had problems. Some major. All irritating.

Last work, previous to ours, installed a "latest and greatest", made by a major manufacturer you'd all know the name of, installed by their people.

We ended up tearing it all out and starting from scratch. Working under the doubtful and jaundiced eyes of the customer.

Ohhh, previous job had given them this really neat and nifty, really pretty front end. Nice pic work. Really. And points could be viewed from anywhere on their internal network, if yah had the right password. You could even acces over the Web. Etc and so forth.

None of which made equipment work any better, nor did it control building temps as it should have, and so forth.

What they wanted was for the equipment to work right. As per spec. Wanted energy savings. Wanted reliable operations. IAQ scheme to work as it should, with proper OA taken in, not too much, not too little. So on and so forth.

System being taken out, hadn't done that well. Oh, I'm sure it could have. I'm equally sure that guys who did that job weren't nearly as good at proper equipment control and setting up proper sequences of operation ... plus checking it, tweaking it, and making damned sure it worked right ... as they were at making pretty pictures and setting up nice looking Web pages.

And everytime customer complained, contractor wanted more money, Saying it wasn't his fault or problem.

Anyway, we put in new system. And this time it worked.

RIGHT.

Maintenance staff and other staff were flat amazed. Many comments like, "Hey, it's actually confortable in my office. What'd you do? Replace the air conditioning system completely?"

Answer is, "Nope." Original mechanical and electrical equipment was just fine. Needed a bit of tuning and balancing. A little fix here and there. Mostly it needed proper controls. And to be operated way original designer intended.

I still hold that the biggest problem in our industry is NOT which widget is best, fanciest, or has the most features. It's more a matter of installing, setting up, testing, calibrating, tuning and adjusting whichever equipment yah have and use ... PROPERLY.

Quality workmanship, by a man or woman who really knows his or her business and particular choice of equipment, very well. Who throughly understands how the stuff works.

A very good pneumatics man is gonna make a customer happier than some other fellow who puts in the latest and greatest available in DDC, who does a shabby, half assed install job, and no better at the set up, programming, and testing of the various controls.

Just my opinion.

10. Permanently Removed
Join Date
May 2002
Posts
9,542
Post Likes
Osy- Just an observation - you continually downplay the protocol debate, and I can understand some of your points.

However, you spend a great deal of time talking about standard graphics, consistent use of identical parts... etc, to streamline your operation.

As with most things, the more yah buy from the same vendor, the better discount you can usually get. We like discounts. We'd rather a strong relationship with a few vendors, than a tenuous one with many.
Well, that's a total "given". Obvious.

Well, if you had a single communication system, you would streamline your techs education/implementation. And, if you got into a bind with the manufacturer, with an association like Lonmark you would minimize damage moving to a different line or controller because of their standards. (if necessary)

Additionally, even your architecture would be identical, although some of the names may change.

Your network tool could be the same...

That's the amusing part to me. You see, we do the same thing with minimizing relay styles, transducers, etc. [*]But, we also have standardized architecture layouts so our guys can walk on the job and know what's going on.[*]We use the same networking setup tool.[*]We use the same network analysis tool.[*]We use the same web servers and routers[*]We are also able to more easily plug in some other manufacturers part in if we have any problems.

Those are both hard and soft cost savings.

You can continue to utilize your proprietary, but if that company decides to dump on you there is no decent solution. We minimize those risks.

I wonder how many LON jobs your firm rescued from a bad contractors install? You don't have to tell me how much easier it is to right that ship than a proprietary, or worse yet an older proprietary.

Frankly, I think you benefit from Lonworks far more than you care to admit.

11. Permanently Removed
Join Date
May 2002
Posts
9,542
Post Likes
... to spin this even further - standardization has it's limits also.

Lonworks has pushed to the manufacturers to a competitive market and pushed innovation. There's an example of an open standard benefit.

That's why you get companies like Circon cranking out product. Even right now they have a LNS based Card Access system TAC is probably falling all over themselves (absolutely green with envy) to try and re-badge. Being LON it won't be difficult to integrate. (don't do it Circon, let them sweat I say)

Also, this environment has allowed the smaller installtion control company to keep the bigger, (ie complacement - know any?) guys to stay on the ball more. More of an opportunity for smaller integrators to develop "google" like ideas.

So, I say if you are a company hiding behind BACnet non-solutions or proprietary dead ends you are doing your customer a dis-service. Not to say there aren't LON companys doing a little hiding, (ie TAC) but it's much harder.

12. Professional Member
Join Date
Oct 2003
Location
Minnesota
Posts
1,411
Post Likes
Osy- Just an observation - you continually downplay the protocol debate, and I
can understand some of your points.

However, you spend a great deal of time talking about standard graphics,
consistent use of identical parts... etc, to streamline your operation.
Standardization which will benefit both us and our customers.

I downplay the protocol debate only in as far as to say that it's of little interest to me. Some. Open protocols, of course, are a good thing in our business. But open
protocols, or propriety ones, and which is best ... is but one aspect of our business. One issue. And I don't know that I'd say it's even the most important one.

Do you know how often the subject of LON vs BacNet vs Propriety comes up in a discussion with one of my customers?

I could count the times on the fingers of both hands. Without running out of fingers. And that's covering a period of about 5 years. The time period during
which I've worked for this company. And I've personally worked well over a hundred projects in that period of time, and talked to a lot of customers and prospects.

Wanna know how many times a customer has expressed a desire for some networking ability LON can do but which CAN NOT be done with other systems?

Once.

That's not to say LON can't do some really nifty things that others can't, or don't do as well, or as cheaply, or whatever the case may be.

I'm just telling yah how many times I've had a customer who WANTED a feature that was unique to LON or which LON did so much better or more cheaply, etc such that LON was clearly the only valid choice.

Discounting another customer who specified LON for a very simple, but valid reason. He, actually his organization, wanted to go to an open protocol in all their facilities. And they wanted ALL Honeywell controllers. Honeywell did
LON, ergo they adopted LON. This fellow was a PE, FWIW, and a HVAC specialist. Personally didn't see either LON or BacNet as a clear winner in the open protocol debates. But they, his organization, was gonna standardize, dammit.
And since he had a fondness for Honeywell, LON won.

<Shrug> Made sense to me. As good a reason as any, I suppose. LOL ...

We were doing Honeywell at the time, so it was just fine with me.

Mostly my day to day conversations with customers and prospective customers, however, involve other issues that aren't really related to the LON vs BacNet vs Proprietary protocol debate.

It's the kind of things I mentioned before in my posts. Costs, of course. And just how reliable is the particular hardware to be installed. What's it's failure rate? Quality of craftsmanship in doing the installation. How detailed, accurate, complete, and honest is the testing and commissioning of the system? How well does it match the specs? This last is so important to many that more
and more of our customers are hiring third party groups to come in and do extremely detailed acceptance testing and validation.

They're not NEARLY as concerned that manufacturer X's controller can talk to manufacturer Y's controllers, as they are that the DAMNED CONTROLS are installed and operating correctly. The equipment being controlled is operating correctly and as efficiently as possible. That each and every physical point has been checked, calibrated, and is working well. And that the sequences of operation, for normal ops, for special conditions, and for fault conditions operate EXACTLY as the customer or the customer's engineering rep specified that they should. No exceptions, changes, etc. Close, or kinda-sorta not
being good enough.

In any event, the points you listed in your posts as to why going LON are good. And I can't really find anything in them to quibble about.

But the fact is, I'm not married to LON, nor to BacNet, nor to AAM's system. (I am, however, very fond of AAM's hardware. And Andover's, for that matter.)

So I do all three. And which one in any particular circumstance is dependent upon the customers wants, needs, desires, and ideas. (And pocketbook)

We don't do the "one answer" fits all thing. We present, normally, a customer with a couple alternatives, and the customer chooses.

So far, few are insistant upon LON. A few more are interested in some sort of "Open Protocol", but aren't the least convinced that LON is significantly better
than BacNet, or vice versa.

If our customers ... as a general rule ... have a decided preference they express, it' s usually for a particular make or two of DDC equipment. ie Honeywell, or Trane, TAC, JCI, Andover, etc. And AAM. Our AAM customers tend
to be sold on AAM, like it. And give us quite significant repeat business.

Mostly what we get is a list, usually pretty precise, of what they want the system to do and be able to do. To which we'll respond, most often with at least two different proposals. Customer picks one. Often one is a LON system
the other is AAM/BacNet. (AAM can do either BacNet or it's own proprietary system)

Except in the 2 cases mentioned above where customers expressly specified a LON system, I've not seen a spec generated by a customer or customer's rep that we
couldn't do with AAM, for instance. Which is admittedly a dinosaur. But a hard working, reliable dinosaur that's more flexible and capable than many folks realize it to be. I'd have said the same back when I did Andover about
Andover's stuff.

It's the customer's choice. Sometimes it's TAC and LON, sometimes it's AAM. More often AAM as we can usually sharpen pencil and bring down the price on that more than with the TAC. Tho, we're trying to increase the TAC business. Particularly now with TAC having joined Andover under a new umbrella corp. We're hoping to see a merger of the best ideas and strenghts of the two. <Shrug>
Time will tell.

to spin this even further - standardization has it's limits also.

Lonworks has pushed to the manufacturers to a competitive market and pushed
innovation. There's an example of an open standard benefit.

That's why you get companies like Circon cranking out product. Even right now
they have a LNS based Card Access system TAC is probably falling all over
themselves (absolutely green with envy) to try and re-badge. Being LON it won't
be difficult to integrate. (don't do it Circon, let them sweat I say)
Certainly Lonworks has had a part to play in it. But, myself, I wouldn't have phrased it quite as you did, as if to imply that Lonworks was THE key role player in the push.

Time, technological progress, and so forth were pushing changes whether Lonworks ever came about or not. I've been doing this digital controls thing, on and off again, for some time. While DDC systems for BAS usage is relatively
new to me, ohhh ... past 10 years, PLC's and other digital logic are not. A hint. I used to work with stuff when the state of the art meant yah had discrete flip-flops, inverters, AND, OR, etc circuits each separately on their
own boards, the size of the average hardbound book. And was working with PLC's regularly in the early 80's.

For many years there has been an increasing push for standardization. I'm sure you've heard of ModBus, right? And most readers in this group, I'm sure, have noticed a lot of various stuff has a serial port interface, speaks ASCII, etc. Ethernet itself is a standard, of sorts, as is HTML.

In the industrial controls world, an increasing push for standardization has been going on for many, many years. And is the reason we have things like standardized motor forms, relay footprints, and even ... increasingly
descriptive terms for various control functions that most have agreed to use the same way.

If something has changed, it's that prices for digital and other controls have dropped dramatically, and they're being used in ever more places. By more people. And digital technology has drastically decreased in size, increased in
abilities per dollar spent, and so forth. So it's a natural function of things that a clamoring for standardization exists, and is increasing. When a single
function electronics board used to cost \$4000 a pop. One thought little of the idea that you'd need a highly trainned and experienced tech to make it work. Now one can buy an electronics package which accomplishes the same, or more, than the item I'm thinking about for \$5 or less. Needing an expensive tech just to wire and program the thing, a single purpose thing, would be ridiculous.
So besides the size and price coming down, increasingly part of the design is for it to be increasingly plug and play. And that means, conforming to some sort of standard.

Certainly Lonworks is a player. And a major one. THE MAJOR driving force? That I might quibble about. If I were actually all that interested in that particular debate. But certainly it is one of the major driving forces.

[QUOTE]Also, this environment has allowed the smaller installtion control company to keep the bigger, (ie complacement - know any?) guys to stay on the ball more.
More of an opportunity for smaller integrators to develop "google" like ideas.
/[QUOTE]

Of course. Nothing I can quibble about here.

Innovation is and always has been the the arena where the small time player has an advantage.

Big, major companies, corporations, etc ... are not very good at innovation or change. Never have been. Likely never will be. This is something one learns if one studies management, sociology, etc. Very large organizations are BEST at one thing, maintaining a status quo. Their very organization and structure restrict and inhibit innovation or change. Slow it down.

What have been apparent exceptions to this, usually aren't. One might point out Mr Ford. But the fact is Ford did most of his innovation when he was young and the company small and hungry. ATT and it's Bell labs might seem an exception. But not really. What made a difference there was that the Bell Labs used to be allowed great lattitude from normal company rules and structure.
Was almost a separate entity. Even then some of their folks had a really, really, rough time getting the parent corporation to adopt their ideas. Some never were until the researcher/developer quit, formed his own group, and
proved the concept. Microsoft used to do a lot of innovation in their earlier days. These days, more often than not, their new "innovation" is something they bought from someone else. There are an awful lot of the various
underlying components in Windows that I recognize from earlier days that used to carry someone else's name buried in the code.

Absolutely, the small players will drive the majority of the coming changes, Sysint. The small controls companies, AND the smaller customer ... who might well be more willing to take a risk and gamble on a new concept than would a
major, big corporate customer. Larger company and corporate customers tend to want to wait and see, as concerns new developments and ideas. Prefer "Tried
and True". But will watch and observe the smaller player to see what works for them. Once concept seems to be proven and valid, THEN the big time guys get more excited and interested.

Of course, the majority of smaller companies ... are gonna fail and go tits up. But this is normal no matter what business one is in. It's the nature of the beast. Others will thrive and survive. Becoming successfully independent,
maybe becoming large. Or in many cases, they, their ideas, and personnel will be bought by a big player. At a handsome price. Big players aren't good at growing innovation inside their own organization. But they are pretty good at buying it. Once the innovation has shown itself to be valueable.

So, I say if you are a company hiding behind BACnet non-solutions or proprietary dead ends you are doing your customer a dis-service. Not to say there aren't LON companys doing a little hiding, (ie TAC) but it's much harder.
ROFLMAO, yah just couldn't resist the temptation, could yah? Yah just couldn't resist playing the game of "My toys are better than your toys."

<Shrug> Maybe they are, maybe they're not. I surely wouldn't know. First time I ever heard of Circon was in one of your posts. And when I asked a fellow in
our automation enginnering department who's been doing BAS since the early 80's "What's Circon?" He guessed, "A French car of some kind?"

It's okay, take a swipe at TAC any time. Won't offend me in the least. Take a swipe at AAM all yah want. Or Trane. Or whomever and whatever.

Myself, I make every attempt to avoid that particular game however. Some people like Fords, some prefer Chevvies, or Toyotas. Some might even drive a Circon. Whatever floats your bubble.

Some folks are enamored with brands and names, the newest latest and greatest toys, etc.

And that's fine. To each, his or her own. Myself, I tend to pay more attention to a person's skill and knowledge in using whatever he or she has. Using whatever is his or her favorite tool or device of choice. In my experience,
this has made more difference, in reality, than has having the newest, greatest, shiniest, most featured gadget.

It's kind of like fishing. My favorite pasttime. One year I ran across a new neighbor at the lake place. Obviously a fellow of substance and money. And he showed up with the very best, most advanced stuff yah can imagine. Gotta admit, I sure enjoyed it when he spent considerable time showing it all off to me. It was impressive, truly impressive, I was awed. There wasn't a new innovation,
or "best of" that he didn't have.

It was kind of a sucky time of year for fishing. Too hot. Too late in the summer. Natural feed too plentiful. Lots of folks went out fishing and came back with nothing or little. My new neighbor took out his fancy boat, and even
fancier gear. And came back with a better catch that the others had, so far. And crowed about it, more than just a little.

I took out my aging, and rather sparse, simple 16 footer Lund. Standard V-hull. Couple old Johnsons on the rear. The kind of thing that's common as ice in a Minnesota winter, around here. And was using my old fish finder, and my 1970's Garcia 5000D rod and reel, my favorite. Came back after less than half the fishing time, with a better stringer than that fellow. He kinda grunted and
muttered something about "luck". I didn't say anything, yet. But a bit later a friend of mine came in with a prize stringer, in anyone's book. And the equipment he'd used was similar, and of similar vintage, to mine.

Now a blind and deaf person could've known that the new neighbor was now really miffed and put out. He couldn't help but let it show. And made yet another remark about "luck". This time I couldn't help but comment. "Luck hasn't a damned thing to do with it guy. It isn't the toys you've got to play with. It's who uses their toys best, to the best advantage, with the best skill and knowledge. He's simply better at fishing than you are. Knows his equipment and how to get the most out of it, and knows more about fish, their habits, and habitats. The latest and greatest technology can be a help, but it's not gonna make up the difference when the other fellow knows more about fishing and fish, and getting the most out of what he's got to work with, than you do. I'm not shabby, but that fellow is GOOD. All the latest technology in the world isn't
gonna help you beat him at this game. You've gotta learn a lot more about the art and skill of fishing itself, and how to use the equipment yah have skillfully, to stand a chance at besting him."

There's a point in there, somewhere ... I think. Maybe not. More likely I'm just rambling.

I'll note, however, that I've got a whole storeroom full of neat and nifty gadgets. At one or another point in the past, each was touted as the newest, latest, and greatest innovations. MUST HAVE's. I'm talking tech stuff one
might use in our trade, or a related one. Some, I've adopted for use, finding them worthwhile over the long run. But many ended up being no more useful, really, in a practical world, than an older item that's long tried and true.

And in still other cases, item works just fine, as advertised. But how many gadgets can a fellow haul around at one time? How many can yah remember how to use properly? How many can yah perform proper maintenance on? Keep repaired? Keep fresh batteries in so it actually works when you need it? How many are kinda fragile, so it becomes a major PIA to keep em stored and protected proper? Etc.

So in numerous cases, item is a good one, but really not worth the hassle if I can make do with another, older item, that's sturdier, more reliable, easier to use, and perhaps has more appliations to which I can put it. So item sits on a shelf. It's no less neat and nifty. Really a marvel of technology. But I can make do with a different, older item. Have more room in the truck. And it's one less item I've gotta buy batteries for, remember the instructions for,
maintain, replace, etc. It's better. But not better enough to be worthwhile in the long run.

In the end, the items I use are the bare minimums to do a good job, of proven worth and reliability. And items I know how to get maximum usage out of, and make work at their best and peak ability. In short, best bang for the buck, Considering both item's inherent abilities, and MY abilities in making the most out of it.

I approach most things like that.

FWIW, SysInt, we do door access. A fair amount of it. And even integrate it with other systems. But not using Circon. That's not a slam at Circon, by any means. As I have no idea what their stuff is like.

Don't even know if they're front wheel drive, or rear wheel; how many cylinders; etc. <G>

Take NOTHING I've said as a slam on anything you have said. It's not. We're simply looking at the same things from a different point of view. Most of our customers are larger commercial, industrial, and institutional organizations.
Many of whom are risk adverse. More comfortable with "tried and true" as versus the latest innovations.

Personally, I think most innovations which will work out and will become proven and the bugs worked out, are gonna occur at a smaller level. Small to medium sized controls companies working with small to medium sized customers.

See? We do agree, from time to time.

But maybe not on the LON thing.

You think I do my customers a dis-service by offering anything else. I don't. I offer solutions which meet THEIR wants, needs, and ideas. And pocketbook. Just how does it become a dis-service to provide a customer with something that does exactly what they want, the way they want it, for the price they agree to?

Their system will operate EXACTLY as we say it will, and that they agree to and specify. Or we'll eat the costs to make it do so. We don't guarantee a system that operates according to OUR ideas, we guarantee one that operates the way the customer wants it to, and does everything the customer desires., the way they want it.

And it'll still be operating as well, after the warranty expires. And several years thereafter. Items with a reliability problem, we drop like hot rocks. Or press manufacturer to fix, then we replace said items.

Our particular rep is for reliability, and proper operation. Not cutting edge innovation. The later is simply not us. Not even the biz we persue.

As concerns LON, when asked, I usually tell folks it has many good points and shows a lot of promise. And I convinced it's BEST in all cases, or that it's gonna be a clear winner in the future? Nope. I tend to think it's an interim thing. A good one. And perhaps the next will incorporate many of it's features and advantages and philosophies. But I don't think we're there yet, as concerns a final, universally adopted protocol and methodology.

But we have certainly moved closer. And maybe the next step will be LON 2, or maybe it won't be. Could be a merger of ideas from LON and BacNet. Or perhaps an HTML/XML, with expanded defines, tags, and variables to cover BAS needs. I've noted that I'm seeing a few items here and there with HTML/Web interfacing built-in. Talk to it, program, make parameter changes, etc using standard
ethernet link and a web browser, local direct connection. Or permanently connect it to a LAN or WAN and give it an IP. Nifty. And at the price of modern electronics, memory, etc, probably not even that expensive to implement.

Besides, when my customers, and the engineers I work with who're from various design and architectural firms, who lay out systems and requirements, and what they want them to do and expect them to do, aren't sold solely on LON. I'm not
gonna do solely LON.

13. Permanently Removed
Join Date
May 2002
Posts
9,542
Post Likes
We sell Lonworks as the ability to control your future. We constantly have customers burned by the supposeadly open manufacturers like ALC, Alerton, Trane, etc...

They find out once they have problems there is nowhere to go but back to the source. That's what the customer is concerned about - Service/Support. So, for me it's a protocol debate as well.

So - you say:
I could count the times on the fingers of both hands. Without running out of fingers. And that's covering a period of about 5 years. The time period during which I've worked for this company. And I've personally worked well over a hundred projects in that period of time, and talked to a lot of customers and prospects.
So, you don't sell protocol by name. You sell quality support and install. Well, if you have a problem with the supplier, you have a problem with quality. That is minimized with LON because you have an easier time of alternates should you have a problem, and right down to the device level. That's a protocol issue. Frankly, it's probably the only thing that has pushed the BACnetter's to provide any form of interoperability.

I have a 2000+ device customer that will agree wholehearteadly with my assessment after going through Trane, JCI, ALC and other controls through the years.

You stated "They're not NEARLY as concerned that manufacturer X's controller can talk to manufacturer Y's controllers, as they are that the DAMNED CONTROLS are installed and operating correctly."

I agree. However, my aforementioned customer is extremely pleased that integration between his access/lighting/hvac/CCTV is now much less complicated and much less expensive. And, if "the DAMNED CONTROLS" aren't installed and operating correctly, the owner can more easily go somewhere else to have those damn controls fixed.

Don't misunderstand me, it wasn't my intention to proclaim Circon "god of controls". Actually like alot of that TAC product line. I am, however, pointing to how a company goes about being straightforward about their delivery to the customer. And, the ability to take something one company does well and more easily integrate that with another. That customer is me, you, and to whom we sell product.

I still think Lonworks is a bigger driving force to lower cost controllers. Modbus has never had good penetration in our industry to accomplish this. (or enough standards) Certainly isn't bacnet. bacnet is a farce in comparison. And, being displaced by Internet protocols.

I can go over any level of infrastructure to you and point out whereby LON installations are less expensive. Especially in future add-on conditions after initial construction.

And, it still remains that if your 3 lines were LON, your tech training requirements wouldn't be as much because of the standard language.

I'm sure you aren't saying your venerable AAM controllers would be less so with a LON communication structure. In fact, I think at device level you would be extremely hard pressed to find a better communication protocol for reliability and scaleability.

What I see that I don't like is everyone skirting the open. Now we have a proprietary framework option above LON and other protocols. Personally, in some ways that's a disease. It's a step backward. Adding complexity that need not be there.

At any rate, thanks for the response. Your'e say you are a little long - but been around awhile, got alot to say. I'd like to see your operation someday. Sounds like a class act. I'm sure I'd learn something.

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•

## Related Forums

The place where Electrical professionals meet.