Controling tempature in a residential area
 

Hello

I have done a lot of programing with AD and some Allen Bradley in the past but have nothing anything for last couple of years due to a disability

Now looking to get my mind right again by designing a program using AD to do some home automation

I want to start with the HVAC in the house by zoning different parts of the house using thermocouples in different rooms. then have the PLC TO control dampers in the supply air vents.

Plan to use a DL205 with a 250 CPU will have 24v outputs to operate the dampers a nd the HVAC system. Use a touch screen to input the data.

My biggest problem is how to do the program using thermocouples to provide temperature inputs. Done analog programing in the past but doing thermocouples is not in my knowledge.

Anyone have any information out there?

This sounds just like a project one our younger EE's wanted to do on a new house he was building. He got overwelmed with all the other stuff that needed to be done so he never got as far as the design. He was going to use AD also. We did look into some devices and one was temperature sensors. I would stay away from thermocouples. An inexpensive one would not have the accuracy out of the box. Go with an RTD. An RTD would be much closer in accuracy from one to another and if you don't use a transmitter you could run the correct wiring direct to the PLC input card. A thermocouple needs T/C extention wire from the T/C to the PLC input card or a transmitter that would just need a twisted pair. Add at least $90/point for a cheap transmitter. Do a search into the differences between a T/C and RTD. We get higher grade T/C's and there still can be a difference of 3°F from nominal, cheaper grade easy 6°F. Not good for a house. I think he may have found an RTD in a nice looking room enclosure.
The programming isn't any different once you get the signal into the PLC. If it is a T/C or RTD card each will have specifics to correctly convert the sensor signal to a real temperature number. Once you have that number, your process variable (PV), you just use it in comparitors to do what is needed. If PV1 is above SP1 open damper1. You could do that for each room and if no room requires heat/cool shut down A/C or heater. Lots of options.
Good luck on your project.

my plan is to take temperature readings in the main part of the house, to control on/off part of the HVAC system in parts seldom used try to maintain them at a set temperature

think will need a pressure device to monitor the air flow in the main trunk maybe compare air pressure of the return to the supply air pressure monitor when one needs to change the filter.

want to be able to use a touch screen monitor temperatures in each room, outdoor temperature, and the temperature of the pool. be able to change temperature by room

whole plan is eventually set up sensors on doors and windows for intruder alarm. Set all the lights in the house to be controlled by the touch screen


yes I realize way to much time on my hands, but after 2 years sitting in a wheelchair and watching law and order on TV. that there has to be something more productive to do with my time

Hello, Joe,

I am glad that you are feeling like tackling some jobs again.

I have gotten involved with a couple of air conditioning projects on this forum in the past. One of the big difficulties in using a PLC is that most HVAC systems use 24 volts all right, but it is 24 volts AC, not DC. There are not too many PLC's that can accept that voltage directly as an input. There are some that will allow 24VAC as an output.

I thought it was a smart idea to make the design fail-safe, due to the fellow's inexperience in working with PLCs. My soultion was to keep the normal 24 VAC control thermostat, then use relays to intercept that signal between the thermostat and the HVAC unit. This allowed the HVAC to continue in normal mode and keep running, even if the PLC was dead, not in run mode, or disconnected.

It must have worked okay. He sent me donation credits for this church that had the 11 HVAC units.

Gee you could become a lawyer??? Nahhhh I think your project is a better idea and probably more fun.

Dan Bentler

Home thermostats incorporate something called a heat anticipator. People are surprisingly sensitive to heat variations in a home and what happens is that people feel comfortable before the temperature sensor actually registers that the air temperature has reached the comfort value. So a very small current is passed through the bi-metallic spring to warm it up a little in antipation of reaching the desired temperature. The heat anticipator can be adjusted by a service technician. Modern digital thermostats adapt to anticipate the heat and don't need adjustment. In any heating control system that you design you should build in some kind of anticipator, otherwise your room will feel either drafty or too great of a hot/cold swing to be really comfortable.

I have similar plans using a micro controller. Two things to watch for are:

1. Prevention of short cycling of the furnace or AC. AC needs to be off for at least five minutes before starting again.
2. Prevent overheating of furnace or freezing of AC evaporator when the air flow is too low. I am planning to install two additional sensors in the furnace. One measuring the temperature of the air entering the furnace and one measuring the air coming out of the furnace. If the temperature rise is too much or too low, turn off the furnace or ac.

You may want to use a differential pressure (dP) transmitter across the filter to determine the need to replace the filter. We do that on our large systems. Now, if you are limiting the air flow you may not get a proper reading. We have 11 make-up air system heaters on one building with variable speed drives to maintain a specific building pressure. The OEM tried to monitor the filters but with the variable speeds you could only do it at max speed. Just something to keep in mind. Check with omega.com for some inexpensive sensors.
I think a lot of us would like to do something like this.

Who knows, being familiar with Fred Thompson on Law and Order might turn out to be valuable. Ole' Fred has never played a role where he said things that he didn't actually believe. For me, his greatest role was playing himself in Marie. Marie Raghetti's life expectancy wasn't worth 2 cents until Fred got involved. I hear Fred is looking for campaign workers all over the country.

the part having a coundry about about 2 years ago put in a brand new Lenox system its a 12 seer conditioner and a 2 stage gas furnace. Know the thermostat now controls the low stage when more heat is required switch to the high stage.

know that I will need to design this in the PLC program. I don't see any difficulties but maybe just being optimistic.

Already figured on the built in heat anticipation in the current t/stat figured with the PLC would have it automatically adjust the user set temperature by a couple of degrees and will put in 5 minute time delay to prevent short cycling.

Also think this will be good to get my mind working and have the house on the market, would think done right looks good would be a good selling advantage, plus we are planning on building a new house next that I will be able to incorporate the PLC into the house design so if everything goes right will have all the bugs worked out

Ok here is question was looking at the RTDs was wondering if any suggestions of placements of the probe in a room. preferably on a wall, and a way to camouflage it?

also planning on using a AD 205 PLC with a 250 processor, was looking at the rtd input card looks like maximum of probes is 4, so can you use multiple cards in a rack?

You might do some research in the HVAC world as to proper position of the sensor. I'm sure there are good guidelines for that. Also look for designer covers for the sensors.
I've used the (as TI)AD405 and 305's and I believe the 205 has a means for additional cards. Whether a real rack or just adding a card on the end of the stack to some kind of limit. Their web site has tons of info. The catalog has good information. Or just call them; they seem to be quite helpful. I have used their touch screens and they do a nice job for the cost. The older ones were a little dim, but that was 6 or more years ago.

My 1.5 cent...

This in reality is a lot more difficult than it seems. One the reason that has been already mentioned is that the BMS and PLC world diverged many years ago and two world has quite a different standard. Using PLC will work but it cost a quite bit more.

I did find one vendor that sells a damper controller that utlize modbus protocal. the issue I had with it is that it need a gateway to program it which cost about $1k, I think. Of course, it would be a lot cheaper if you can find just a open/close damper.

A good resource is the Controls Forum at hvac-talk.com

realize there are cheaper systems out there, do appreciate the suggestion, but really want to do this with a PLC.

More for the challenge of designing the program, the mental part, ever since becoming disabled the physical part of doing anything has to be done by a contractor. Will just need somebody to do the wiring and install the dampers. Just need to do the actual program, then to over see the wiring.

Money is not a problem, just the actual boredom of just sitting around. Have the direct soft software and the software to do their touchscreen design. know I have the experience to do the programming, done a lot of analog programming in my previous life, just the RTD and thermocouple programming is a new ball game.

You've got your goal set on doing this, so I'll chime in and advise you to stick with thermocouples, versus RTDs.

Thermcoouples are much less expensive than RTDs. Yes, the
ANSI spec shows that a standard limit of error T/C can have 3° or 4° error, but that is not typical, particularly when the T/Cs have not been subjected to high heat (>600°F). The easy work around is to hang a thermometer that you trust next to the TC and compare the thermometer's reading with the PLC reading, then either add or subtract an offset/bias value or put the offset value in as part of a calibration. The offset will not change over time because the TC isn't being stressed at room temperatures, like it is in high heat.

You can make or fabricate your own TC from TC wire, stripping the insulation and twisting the bare ends together. PT100 RTDs are purchased in a sheath, and just cost more.

Yes, you have to run TC wire through from the measuring point to the PLC, but presumably you'll have to run copper wire for the RTDs, so it's running one or the other.

Yes, the TC wire is TC wire, but the RTD extension wire should be twisted pair, not romex, so the cost difference isn't huge.

Dan

Any type of ventilated box will work, maybe an old thermostat with guts removed, or a small electronic project box from Radio Shack or Allied Electronics, or you can have your own custom made from wood, metal, or plastic.

Yes, as many cards as the rack has room for. If you get the DL-205 9-slot rack, with a 250 CPU, 2 digital Input cards, 2 digital Output cards, then you would have room for four of the 4-channel Thermocouple or RTD cards (16 total temperature inputs). Normally I would say that you need about twice as many digitial inputs as ouptuts, but I assume that most of your inputs will be the termperature measurements.

I agree with Dan. Thermocouples would be much cheaper to buy (or make), and will provide enough accuracy for HVAC.

I have welded up many thermocouples using a homemade rig. Many thermocouples used to be made this way. Use a fused circuit, wear gloves and goggles, and don't try this if you are not agile with steady hands.

Pour a half-inch of mercury in the bottom of a quart mason jar, then add 2" of mineral oil floating on top of the mercury. Twist your bare T/C wires tightly together for 1/8" to 1/4" length, forming a small ball at the end. Wire-nut both the other ends to the hot black wire in a 120 volt cord, drop the 120 volt white neutral into the mercury near one side of the jar, apply power, then lower the T/C wire ball into the mercury (about 1.5 to 2" away from the neutral wire). The small arc will blow away the mercury, killing the circuit. It will weld, then you pull it out, with the oil quenching and protecting the weld from oxidation. It works every time.

Be warned again, the lowest AC voltage you can input into the DL-205 is 110 volts AC. You WILL NOT be able to connect it to your existing 24 Volts AC HVAC control circuit (thermostat circuit) without using a relay. See this list:

http://web1.automationdirect.com/adc...c_205/AC_I-z-O

If you do want to tie to your existing controls without relays or diodes, you can actually get a DL-105 that has 6 inputs that accept 24 VAC. The model F1-130DA is one such, and has 10 inputs, 8 outputs, and a built-in 24 VDC auxiliary power source:

http://web1.automationdirect.com/sta...cs/f1130da.pdf

If you want interface directly to 24 VAC with many Inputs and Outputs, then consider buying an Alerton Bactalk VLC (a PLC made for HVAC control systems), or one of the other Building Management System controllers. Learning to program and set up one of these systems would be a valuable trick in your tool bag.

http://www.alerton.com/Products/BACt...el/VLC-550.asp

The junction of the two wires is the sensing point of a T/C. Being an exposed junction, it will have a fast response. That's not always a plus. If you open a door in the winter time (in a northern climate), the in rush of cold air will be seen by an exposed junction T/C almost immediately.


The more you hide the bare junction the more thermal response lag you get.

I like Lancie1's idea of hiding a TC in a gutted thermostat. Thermostats have openings to allow the guts to sense air temperature, but stats look like they belong on a wall.

Dan

Hi Joe,

I am well into a similar project.

It's been a few years in the making. I planned on doing it about 10 years ago, but never got aroun to it until this year.

A SLC5/03 in 7 slot rack is in the furnace room. There are quite a few analog temperature sensors (LM35s), which connect to a NI4 module via several 16-1 analog multiplexers (ADG507s). A thermocouple (via a AD595) monitors flue temperature. There is a RIO scanner in the rack, which interfaces two drops of Flex IO. An outside temp sensor provides "outside reset" for control.

We have hot water radiant floor heating, with a Viessmann Atola boiler. The original controls have been replaced by the PLC, except for a pair of hard wired aquastats on the boiler.

A PID loop to controls a 4 way mixing valve to set the floor loop temperature. Individual relays control zone actuators.

A VB program, on top of a MySQL database monitors and provides a setup interface. See attached pictures for screen shots.
http://www.plctalk.net/qanda/uploads/Main.gif
http://www.plctalk.net/qanda/uploads/Mix.gif
http://www.plctalk.net/qanda/uploads/Boiler.gif
http://www.plctalk.net/qanda/uploads/FamRoom.gif

If anyone is interested I can elaborate further on this.



Cheers,

Peter

Funny you should say that. A friend of mine moved out to DC to help out on the campaign.

Software could compensate for the exposed TC if Joe wanted it to.

Just a follow up on temp sensors...

I've been looking for an easy way to weld TCs, but I think I'll pass on the "mercury in a bottle" method. I'm sure it works well, but is it worth the hazard.

I've been using National LM35's for years now... Epoxied to a magnet on the oil pan of my '88 Taurus way back in '89. Son still uses the car, the oil temp sensor still works fine, after some 19 years.

The only thing that you have to watch for on a the LM35 is cable capacitance. For any cables longer than a few meters, you need to use a RC snubber across the output. Something like 100R in series with 1uF. If you don't, it'll oscillate on you.

For temp control "anticipation" (PID rate) you want to have high resolution. That's a nice thing about the analog sensor ( as apposed to a digital one). I've used an op-amp to give a X10 gain, which gives an output of 2.0V at 20C. Couple that to a 16bit analog input, and you have 1/100degC resolution.

I solder pigtails on each lead, heatshrink individually, then over all three. One more heatshrink over the whole thing does it. Use epoxy for an ugly location.

I've fitted them in existing thermostat boxes. That works fine too.

Another trick is to epoxy one to the back of a small heatsink, such as one for a 14 pin DIP IC.
The heatsink is mounted exposed on a flat bakelite switch plate cover. Doesn't look too bad, and works great.

For about $2.50 each or so, they're hard to beat.

Cheers

Peter

Any errors you encounter with the thermocouple inputs can be corrected with scaling controls on the "engineering page" of the HMI.

That is a perfectly practical application of code to hone your programming skills, and apply indefinitely.

I am most curious about your damper actuators. Are they going to be controlled by digital outputs?

Paul

was planning on digital outpus for the dampers, then use ice cube relays to control the hvac equipment control relays.

So, you'll have an open and a close command?

What are the actuators you are using?

yes planning a straight on / off for the dampers

not planning on using any type of analog in the program, just thermocouple temperature input so I can control the dampers

actuators i take you mean the dampers, thinking of using the basic 6" 24v. motorized damper

I'd recommend that whatever the current heating/cooling system is in place in your house be left there, just for resale purposes. Many buyers nowadays hire an inspection service and HVAC controls that are way out of the ordinary are likely to raise a red flag.

My Dad has had experience with such. He retired 25 years ago when he moved to Arizona and built a solar house. A former work colleague had moved there a year earlier and had drawn up the plans and had built two such houses, one for himself, one to sell. The colleague built his own circuit boards for heating/cooling control, using TTL and analog chips. My Dad did the heating/cooling control in his house with standard thermostats, several relays and a toggle switch. Not because he couldn't do it the other way, but he prefers the simplicity and reliability of relay logic.

The custom circuit board system had persistent reliability problems due to lightning strikes taking out the analog temperature sensor circuits (the T/C wire runs were long antennae). The owner who purchased the for-sale solar house hired my Dad to replace all the special circuit boards with my Dad's thermostat/relays/toggle switch components about 2 years after he purchased the house, due to the unreliability of the custom system. That house has sold twice since without problem - the heating/cooling system looks ordinary with its thermostat on the wall and a toggle switch labeled 'summer/winter'. The small panel that hosues the relays has a wiring diagram pasted on the inside of the door that looks like an appliance wiring diagram that any heating/cooling guy can figure out.

The colleague's house recently went on the market, but according to the real estate gal who called my Dad about changing out the heating controls, potential buyers are wary of the maintainability of the heating/cooling system, having gotten reports from the instpection services that point out that the whole thing is a one-off, custom system with who-knows-what heating cooling controls. Granted, this is not a hot market for real estate in Arizona, but I can understand the reluctance when there are other properties for sale . . .

Now, I'm not saying a PLC won't be reliable and that relay logic is a better way. I am saying that a PLC is 'custom' to the extreme. Having a PLC for heating control would mean that an ordinary property owner could not call any of the heating guys in the yellow pages to fix his heating or cooling. In fact, it even rules out the commercial HVAC guys, too, because they don't use PLCs either. It would take someone with very specialized knowledge AND the development software/cable/etc.

But, if you keep that heat/cool thermostat on the wall with the wires disconnected, so you can re-connect it whenever it is you decide to sell the house down the road, then you can take your rack & I/O and what-not PLC controls with you. Your house buyer can be blissfully ignorant of the great control scheme you had, because the thermostat does its thing and ignorance is bliss, isn't it?


Dan

Thats a good idea, I tend to see this being so cool, that others would be impressed. Where potential home buyers could see it as being a head ache to have deal with.

I have already considered still using the original thermostat to control the system, and use the PLC to monitor room temperature, and control the dampers.

I figure that as rooms come to set temperature, closes the damper. Causing more air flow into the rooms not to set temperature. Where the original thermostat still controls the main part of the house.

Then monitor air flow, make sure not over loading the system, and also warn you its time to change the air filter.

not sure this will work, I'm basing my theory on my first choice of career as a HVAC technician, took 2 years in technical school, then 4 years working for a contractor started as a installer, then service work. did this then went onto working maintenance in a cold storage since then became a industrial electrician. Then onto doing PLC work for the last 15 years.

You might also consider the dampers to fail open. If there is a total failure (more for future owners) they might not notice it not working efficiently and not cuss out that damn designer. Also consider if the PLC is shutdown that it will revert to simple controls incase the PLC looses it's brains. If you are gone who will be able to reload the program?

plan to use ice cube relays to feed the dampers, so when the PLC sends an output to close the damper, will turn on the relay, in turn the N/O contacts of the relay will power the damper to close.

if the the PLC fails all ice cube relays become deengerized then any N/O contact would be open, and in turn dampers will all be open.

will also put in a switch to go from auto to manual. basically take the PLC out of the loop.

plus any problem with the dampers, any HVAC technician could handle

Hi Joe and others,

I am an HVAC Tech specialising in controls for the last 10 years, mostly BMS/DDC of course. I really like your idea of a PLC to control a home HVAC system. I contract myself out to the oil & gas sector for HVAC commissioning, repairs, maintenance etc and most systems offshore are controlled by a PLC (usually AB)

Have you considered the need for a VSD on the supply air fan? If you are going to start closing dampers you are going to need one.

I would also seriously consider using analog dampers, with a feedback, I normally use 0-10V stuff from Belimo.

Just a few ideas for you,

cheers

Charlie

It is not hard to use two relays to feed the existing thermostat into the PLC, and then let the PLC decide when and if the system actually runs. You could use an infrared sensor to detect if anyone is actually at home and have the PLC setback the On time, or delay the run. Use 3 NC relay contacts for the Fan Relay, Heating Contactor, & Cooling Contactor, so if PLC is disabled, or removed, the HVAC runs normally.

That is a good suggestion about using infrared sensors. Would you need one in every room?

also this is probably dumb question but it sense someone if they were sitting watching TV?


the thought has crossed my mind, but then get the problem of low air flow across the evaporator coil. then I thought if to many dampers close will still have the problem of low air flow. Figured I could measure the air flow if it gets to low start opining dampers until I get an propitiate air flow


this could be a possibility. I will have to look into this further

Joe, I installed an infrared motion-sensor light switch in my bathroom. It has two adjustments, Sensitivity, and Time Delay. The time delay can be set to keep the light on between "motions" in the room. It fits in a standard wall box and works very well. I have been using it for years. We older guys that get up frequently in the night really need one of these! It saves fumbling for the switch at night in the dark. You just walk in and the light turns on. Walk out, and you have enough time to get back to bed, and 3 minutes later it turns off. It avoids having to leave the light on all the time.

For HVAC control, you would need to get a little more sophisticated. There are "occupancy" sensors available to use in controlling various fire and burglar alarm systems. You COULD put one in each high-use room, with inputs back to the PLC. If any one is ON, the PLC assumes the home is "occupied". I would put in a time delay, so that if any sensor is triggered in a 60-minute period, the "occupied" relay stays on.

If you have an alarm system, you could use a contact the alarm-set keyswitch to also set the "occupied" relay to occupied or unoccupied.

You could just mount a simple On/Off selector switch near the normal entrance/exit door.

Joe

Off the top of my head there are two occupancy sensors
One is motion - these are not recommended for "sitting areas" ie live room bathroom etc. The most common are wall mounted. The other is infrared these I believe can come in wall mount or ceiling mount. There is a third which combines the two these also are in wall mount and ceiling mount and are I believe recommended for bathrooms.

Have not dealt much with the infrared version so cannot answer your question regarding sitting ie how long would it take for sensor to decide you are part of background IR level?? Lithonia may be good source for info OR you can do as I did steal the directions from a box at Home Depot be careful though that is where they hide the little magnetic theft detector - believe me I know.

Dan Bentler

Dan,

It depends on the Sensitivity and Time settings on the infrared sensor. If you set it too sensitive, it will trigger from air movement from the vents, or a cat in the room. If the time is 3 minutes, then if it sees any little flick of the hand in a 3 minute period, it stays on, resets the timer and starts the 3 minutes over again. It is hard to stand in a room and not trigger it. If you make an effort to stay still you can fool it, but it is not easy.

Yeah your right, People are surprisingly sensitive to heat variations in a home... the modern digital home thermostats adapt to anticipate the heat and don't need adjustment...



_________________
Programmable Thermostats