thinking of building my own electronic controls for radiant

Aidan (UK)

I’m a mechanical engineer (heating, chilled water, ventilation, etc.) & I started getting involved with some BMS/DDC building control systems about 10 years back. The BMS was needed to control the mechanical plant that I was designing, specifying or installing.

BMS systems can monitor or control anything; gas, water, oil and electric meters; switch pumps, fans or heaters; control temperature, humidity, de-hum, room pressure, etc.. The cost was virtually the only limitation.

The thermostats you mentioned are technical dinosaurs; why would anyone want an electro-mechanical switch assembly with a +/- 1 degF accuracy when you can control to +/- 0.1 deg with a 40c bead thermistor?. Why does the UK have thermostats with mains (230 V) ac when you can run better control wiring on 10 V dc? This is assuming you have a compatible electronic controller on the other end of the wiring.

Having invested the effort to learn about BMS systems ( to the extent that I was reprogramming some of the low-bid control software) I thought I’d like to instal a similar system in my home. There’s nothing available, excluding commercial stuff at £300 for a controller and £500 for the supervisor software, which doesn’t allow you to program the outstations. Why not? I can buy a £5 calculator in the supermaket, why can’t I buy a £100 electronic heating controller? Why can’t I monitor my heating systems performance on my PC or download a suitable control strategy from the internet?

The problem is in getting PID controller with analogue inputs and outputs. My electronic abilities are nowhere close to enable me to assemble something from semi-conductors. You would probably have to get a PLC with analogue inputs and outputs from somewhere. You can sometimes find them on e-bay, but non-obsolete ones with the right modules and software are rare. I’ve bought an old BMS outstation, but don’t know if it works yet. The biggest investment is in the time needed to learn how to make any such system workable.

I think you’ll find that the lack of accuracy is a culture shock. The older generation of BMS outstations had a cycle time (review all inputs, reprocess all outputs) of 15 seconds and this was/is perfectly adequate for most building control applications.

I think this is very relevant to The Wall and hope you’ll keep us posted. Please let me know how you get on or if you find a good source of information. The voice control sounds a bit ambitious to me; get a working system first and then add the bells and whistles later

jerry_2

a tech guy thinking about rolling his own controls

I am a tech guy who is not quite knowlwdgeable enough to be considered dangerous yet.
I have read a bunch of stuff on this wall and the other parts of the web site. This is great stuff and you are certainly wise and fun people to listen to.
I have some background in controls (systems, aircraft) and like tech projects. I am going to remodel my house, and for a bunch of reasons want to use radiant (something about using forced air to heat a 19' area just seems dumb.)

It bothers me that the average japanese rice cooker has a good deal more smarts than the average house temperature control system. If they built an aircraft control system with the oscillations of even the fancy thermostats, you'd have quite a number or people using those air sickness bags.

Sooo, I am thinking I would like to take a stab at building designing my own software to control the heat and cooling of the house. I'm not too worried about the cooling side. I have a compressor in now and I understand how to move damper solenoids and how to detect/avoid icing. It's the heating side that seems like it's a bit trickier.

The basic system design is zone per room/area with a total of ~2250sqft of radiant floor surface.

First, if you assume the time is free since it's a project for me, is this a sane thing to try to do? Clearly no one else will ever be able to deal with this, so I'll need to have wiring in place to allow simple set point thermostats to be used instead.

The thermostats will be direct read thermisters (maxim/dallas semi one wire, $3 each) and possibly a cheap radiant heat sensor with A/D, along with a few buttons and a display. I might even try to work in a microphone that would allow you to push a talk button and say "warmer" or "I'm cold". (That's all the other members of the house really want to do.)

If this is bordering on sane, would any of the wet and wise people here like to contribute their wisdom and wishes for an ideal heater control system? I'm game to see what I can cook up and happy to share my ideas and results.

While I'm bothering you folks, the warmboard product has caught my eye. It's very expensive subfloor but they claim is makes for super system efficiency and lower total install costs. The efficiency stuff matters to me for environmental reasons, but the numbers have to make sense as well. Have any of you folks done much with warmboard for floor radiant and what have been your cost and customer satisfaction results?

thanks so much,
jerry

PS. I can safely state that the desire to oversize boilers isn't limited to heating professionals. When I did the rough heal calcs for the remodeled house, the numbers came out to under half of what I hadexpeced and a third of what I would have guessed for the size of the heater.

jerry_2

using the wall well

I don't want to clutter this with things that aren't appropriate to the wall, so let me know when the thread should move of this forum.

thanks again,
jerry

Dave_22

Well...

I don't think your crazy. I'm a Controls Engineer and I plan on doing my controls myself. Most of the controls are just relay logic outside of the analog stuff. I just happen to have an Allen Bradley Programmable Logic Controller with analog and digital inputs and outputs. It is Ethernet compatible so I will have it hooked to my computer to trend and tweek during startup. This should leave me with endless possibilities . I'm thinking a touchscreen panel view to monitor or change system settings. I can't wait!!!!

Richard Miller_2

NO!

Stay around. I don't have a lot to offer except to mention that everything you want to do except voice is already available at suprisingly low cost. But if it is the challenge you are after then get after it!


Only thing I would ask. Since you already mentioned it here on the Wall you have a responsibility to keep us updated! Let us know if you decide to do it. If you do go ahead with the project we expect full details and pics in regular reports. That's in the rules around here somewhere. I think anyways.

Good luck and I am anxious to hear how it turns out for you.

Can you integrate it with Palm software?

Floyd_3

Sounds like you have really...

put alot of thought into this already.... by all means, please keep us informed of what you come up with....
some "outside the box" thinking can do nothing but make the rest of us think in different ways and of the many other possibilities that are out there....

As you may have gathered by now.... there are many in this field that are very slow to deviate from the "tried and true" methods that have proven themselves over the years...but there are also many here that wil go out on the limb and try things new, especially if they can do it in a safe enviornment like in there own house, where if it fails on Sat. night at 11:30 they can just jump outta bed and go hot wire the sucker till they get time to figger it out.

Hey, if you come up with sumptin that needs playin with let us know... we'll give the old once over and give you the straight scoop!!!!

Look forward tohearing more from you!!!!

hot_rod

More than one way to skin a cat!

Quite possibly the ultimate control system still does not exist. To be "perfect" I suppose every customer would have a different control. Since we all have different comfort needs, and live in a wide variaty of climates! Have fun with that. Make sure you have a way to retro fit a "commoin man's" system should you ever want to sell that home

As far as WB, I have found it to be amoung the best at providing near perfect radiant transfer and comfort. It provides the best across the board transfer, utilizing the lowest possible supply temperatures, period. At least currently

Takes a bit more planning, installation work, and some care to protect throughout the construction till final floor covering. The latest panel construction is a big improvement over the first "wafer board type" product. Get all the installers and framers "in your boat" before you have the product delivered, would be my best advise.

Keep us updated on your project.

hot rod

jerry_2

Analog controls aren't so bad

Those TCV on constant circulation systems are actually quite good at keeping temps stable if they are well located. It's actually quite a bit of work with control systems to reach this level. Their downfall is with things like low water temps, automatic setbacks and other things like that.

One big part of my goal is to both run the energy efficiency very high and also quantify the energy in use. This leads me to things like low water temperatures which make all sorts of things harder (that 20 degree drop is a bigger deal when it's from ambient +30 to ambient +10 rather than ambient +100 to ambient +80.) It's like new car engines, which get amazing power and efficiency, but with more computers and hoses than I can deal with.

I'm actually leaning away from a PID solution, I think there are going to be any number of things that make that approach more difficult that a digital solution. Also, I'm a digital kind of guy, Analog in, analog out, bits in the middle. I took an analog computing class (they still offered it when I went to college) and learned how hard it was to build stable circuits.

just my $.02

Aidan (UK)

analog controls

Yes. You have to skin your cat your way. Dave's brewing his PLC device, I'm working on a BMS system and you're doing your analogue computer thing. I don't know what an analogue computer is.

I think TRVs are fine, they'll control the room temperature to the extent that you can't notice any variations. I intend using a PID module to turn the heating water temperature as low as possible, whilst ensuring there's still sufficient heat emitted from the radiators with a temperature sensor. I think you need a P&I control device to get stable control of an automatically reset temperature.

bruce pirger

So this is were the nerds are, eh?

> put alot of thought into this already.... by all

> means, please keep us informed of what you come

> up with.... some "outside the box" thinking can

> do nothing but make the rest of us think in

> different ways and of the many other

> possibilities that are out there....

>

> As you

> may have gathered by now.... there are many in

> this field that are very slow to deviate from the

> "tried and true" methods that have proven

> themselves over the years...but there are also

> many here that wil go out on the limb and try

> things new, especially if they can do it in a

> safe enviornment like in there own house, where

> if it fails on Sat. night at 11:30 they can just

> jump outta bed and go hot wire the sucker till

> they get time to figger it out.

>

> Hey, if you

> come up with sumptin that needs playin with let

> us know... we'll give the old once over and give

> you the straight scoop!!!!

>

> Look forward

> tohearing more from you!!!!

bruce pirger

So this is where the nerds are, eh?

LOL...I'm an astronomer in my day job...and I build data acquisition and control systems for our cameras and spectrographs.

I thought about building my own control system as well...it would indeed be a blast. But then I found Tekmar...and well, I think they've done a fine job...and not all that expensive. I'd look at their stuff...its rather wonderful...tekmarcontrols.com I believe.

On the other hand...I am a bit disappointed that I can't just plug in my Firewire line and have complete access to the control from the net...been thinking about adding that part to my control actually. But it might just be easier to place the various thermistors or whatnot on the pertinent locations and just monitor everything.

You know, it doesn't seem like Tekmar has much competition.....LOL.

chris smith

controls

ge fanuc makes a versa max controler,16 points exspandable to like a 250 or so, you can get the controler and software package for around $425, of coarse you will still have to get rtds for temp sensing, and other stuff but i found this product to be faily cheap,or if you want to totally build your own there are alot of small mirco computers that can be programed with basic. here are couple of links

http://www.geindustrial.com/cwc/products?pnlid=2&famid=12&catid=139&id=vmax

http://www.opto22.com/

chris smith
porter maine

Mike T., Swampeast MO

Some things to remember if you design your own controls...

Being massive, most radiant panel systems respond quite linearly to changes in supply temperature. Any slight deviation from true linearity is regularly exceeded with changes in outdoor conditions like wind, sun, precipitation, etc. If you try to achieve too tight of a reset "curve" without considering the atmosphere you can easily wind up with not enough heat--say a nighttime blowing snow. Trying to automatically compensate for ALL outdoor conditions (as opposed to temperature alone) can get EXTREMELY complex so it's generally best to "loosen up" the curve somewhat to compensate.

Again, being massive, most radiant systems "want" to maintain stasis and do not respond rapidly to requests for changes in temperature. Since a product like warmboard has its conduction surface on TOP, it is most certain to have relatively quick response--but remember, it's STILL relative. If you really want the ability to change temperature rapidly, supplemental heat in the form of panel or iron radiators will be MUCH better.

Proportional two-way flow devices are an extremely effective and efficient method to control heat. They are of course analog in nature and generally require some physical manipulation to change setting. Danfoss does offer some automatic setback devices for their TRVs--for panels/iron rads/etc ONLY--not on their radiant panel devices.

Zoning with valves/circulators certainly works BUT since these devices generally don't "communicate" with one another and are generally digital in nature, strange and unintended flow balance problems can occur. If digitally zoned to the extreme (micro-zoning) boiler short-cycling can be a serious concern.

Two-way proportional devices are able to self-compensate for nearly any condition. While they have no "hard-wired" communication with ane another, they have indirect communication via change in pressure. As any one device changes setting (thus delta-p) the rest of the devices automatically compensate. With some reasonable "headroom" in the reset curve they allow rises in temperature to be as quick as possible while reducing overshoot as much as possible--VERY important with high mass systems like panel radiant.

I'd sincerely suggest that you look carefully at the Danfoss line of valves/controls paying particular attention to those devices that act proportionally. You can get as fancy (and expensive) as you like or use devices that while unsuitable for electronic setpoint modification offer wonderful comfort/efficiency.

My personal experience is that once you have proportional control many of the circumstances that make occupants want to change things disappear and it becomes MUCH easier to be comfortable with "general" settings.

While it would be possible to achieve automated setpoint adjustment in individual spaces, it would be highly complex and extremely expensive--particularly if using proportional control. Whole-system modification would be reasonable by intentional "starvation"--just be VERY careful not to starve too much and offer FAIL-SAFE control to bypass the electronics should they fail.