Shop OBEX P1 Docs P2 Docs Learn Events
Modulating Heater Project(Anyone Interested?) — Parallax Forums

Modulating Heater Project(Anyone Interested?)

mctriviamctrivia Posts: 3,772
edited 2013-01-17 08:56 in General Discussion
I wish to bury a heat tape under neither each of my garage doors inside the concrete slab. near these inside a 3/4" rigid EMT conduit I would like to place a thermistor to measure the slab temperature.

Placing this in the slab is the easy part. The hard part is if I want the slab to last and not use a lot of electricity I need to PWM the power to the wire to very the amount of heat being generated by the wires to keep the slab at 1deg Celsius. Everything attached to the mains must be CSA or cUL approved. A quick search finds http://www.digikey.ca/product-detail/en/DRA1-MCX240D5/CC1737-ND/2197276 as a suitable SSR and a wall wart designed to run on 240V should work nice to power the controller.

I was wondering if anyone had any suggestions on the easiest way to get an accurate temperature(keeping in mind leads may be 20' long) and weather or not I need any special circuitry to drive the SSR or if I can just run a wire from a propeller IO line directly to it.

Any comments especially about potential problems that may come up are very appreciated.
«1

Comments

  • Duane DegnDuane Degn Posts: 10,588
    edited 2013-01-13 11:45
    I'm pretty sure SSR can be driven directly from a Prop. I've use a SSR with a Prop directly and more importantly, I've seen Beau drive a SSR directly from a Prop I/O pin.

    By PWM, I assume you'll use a very low frequency? I use a SSR with my drying oven and I think the frequency is 0.1Hz (turns on once every ten seconds).
  • mctriviamctrivia Posts: 3,772
    edited 2013-01-13 13:41
    I was thinking along the lines of 4hz with duty cycle in half cycle increments.
  • mctriviamctrivia Posts: 3,772
    edited 2013-01-13 21:09
    here is a much cheaper SSR. Uses 5V signal but I can probably apply a 5V power rail through pull up resistor and use prop to sink to gnd to turn relay off. Am I current that if it does not have zero crossing feature it will act as regular relay cutting in and out at any point in the curve where as if it has the zero crossing it will cut in any time but out only at zero crossing?
  • kwinnkwinn Posts: 8,697
    edited 2013-01-13 22:14
    I would suggest using a zero crossing SSR to minimize any chance of electrical noise. Driving an SSR directly from the prop is not a problem for most SSR's. One (or several in series) temperature dependent resistors could be used to measure the temperature although LM35,s might be more accurate.

    Do keep in mind that the slab temperature may vary from one side of the garage to the other so you may need more than one sensor.
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2013-01-13 23:41
    Heating by electricity has huge power demands. Circuits are often 20 or 30 amps at 120VAC or 240VAC.

    Frankly, I think PWM is micro-management as the concrete will provide a thermal lag due to it storing a great deal of heat. There are existing in slab heating solutions that are electrical or hot water. So I have my doubts that you need to develop a new one. But if you are trying to create a more efficient system, it should start with a complete look at the whole thermal context, not just the electrical control.

    That means you have to figure out how much heat the slab will actually store (it is the thermal equivalent of a capcitor), the adequacy on insulation to the underside of the slab (ground water and no under-slab insulation are significant factors), and ambient temperature.

    If you can figure some sort of time constant, you may need to only cycle the heat in very slow sub-hertz intervals.

    In Taiwan, I live in an all concrete building. Nearly all the tenets only turn their A/C when needed to save money. But I changed my approach to keeping it one 24/7 at a higher temperature because the concrete stays colds once it is cooled down. It is expensive to recool. The bill is about the same as if I just had it on when needed, but I have eliminated a huge humidity problem.

    In other words, this is first a thermal problem. It needs a thermal analysis and appropriate thermal objectives. You don't really need a tight steady-state.
  • mctriviamctrivia Posts: 3,772
    edited 2013-01-14 04:51
    The concrete will provide a thermal lag as you say so at the surface pwm at 4hz or .1 hz will not be noticeable. The problem is the thermal wires put in concrete get very hot(40 deg Celsius) so the traditional method of turning them on and leaving them on until at temperature tends to make the concrete break down quicker then it should since it gets really hot around the wire but stays cold where the wires aren't and then slowly heats up throughout the slab. Since I am not trying to heat the room with the wire I plan to put a lot less in then would normally be used.(1/10th). The garage is well insulated(R9 under floor, R29 on all walls, R50 on roof). I will have a forced air heater for control of air temperature(will likely be off all the time but when I need to work in garage) I just need a small amount of heat to keep the slab above freezing so the garage doors do not freeze to the slab like it does in my current garage and for longevity of the slab I want to keep it as close to a stable temperature the best.
  • kwinnkwinn Posts: 8,697
    edited 2013-01-14 07:51
    Do you want to keep the whole slab above freezing or just the section under the garage doors?
  • mctriviamctrivia Posts: 3,772
    edited 2013-01-14 07:58
    All I care about is under the door but I am told that would be bad for the slab. So my plan was to make 4 zones. By each door and each half of remaining floor. To save money will do by door at normal 4 inch spacing and rest of floor at 3 foot spacing.
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2013-01-14 08:34
    Actually, having a very short run may be a bigger problem than long runs as the overall resistance is less -- thus requiring some sort of limit to voltage/current. With a big floor, the wire can just be made longer to bring the resistance up to a nice limit. So I suppose you are right about using PWM for this situation.

    Another big problem with concrete in sub-zero conditions is spalling due to moisture in the surface of the concrete freezing and expanding when it turns to ice. The results are the surface of the concrete falls off. So with constant melting and thawing, it may be best to apply a moisture seal to the concrete as well -- such as Thompson's water seal. This may be a bigger source of having the concrete fall apart than overheating. It is hard to say.

    I can appreciate that you want to heat around the doorways as there are drafts and the areas may actually become icy. If you can install copper tubing and pull your heating wire inside it, it may be more forgiving for future repairs. And if worse come to worse, you might shift to using a hot water system to heat the concrete. This may sound a bit too much to engineer, but there are a lot of electric baseboard heaters that have an internal hot water system. One might be cannibalized and provide hot water that circulates by natural convention.

    There are concrete slab heating systems that can simply attach the wire to an existing slab and then pour a strong thin layer of concrete to cover. Others can do the same with tubing and a water heater. I doubt if these will be bad for an existing slab, but it certainly helps to have a slab that already has some wire mesh reinforcement inside and is on stable well drained soil. Heated concrete slabs are actually very nice as the heating doesn't vary as quickly as forced air and solves problems of having a room that is warm with a floor that is cold.

    http://www.warmup.com/us/in-floor-heating.phtml

    http://www.concretenetwork.com/radiant-floor-heating/

    http://www.youtube.com/watch?v=vt_Thcs_jvU
  • PliersPliers Posts: 280
    edited 2013-01-14 08:51
    Perhaps it would be better to heat the bottom of the door?
  • kwinnkwinn Posts: 8,697
    edited 2013-01-14 09:06
    If the objective is only to keep the door from freezing to the floor it may be more economical to place the heater on the bottom of the door. That does create a problem with providing power to the heater but a sliding contact at the top of the door would take care of that. Another option would be an eavestrough heating cable under a protective metal extrusion on the floor.

    Are you pouring a new slab for an existing garage or putting up a new garage?
  • mctriviamctrivia Posts: 3,772
    edited 2013-01-14 09:24
    Building whole new house nothing there yet. Heating the slab would be better since if I do the door ice can still for around on ground and moving wires don't last as long.
  • kwinnkwinn Posts: 8,697
    edited 2013-01-14 10:19
    mctrivia wrote: »
    Building whole new house nothing there yet. Heating the slab would be better since if I do the door ice can still for around on ground and moving wires don't last as long.

    In that case you may want to look at something like IPEX tubing embedded in the concrete with a pump recirculating a heated antifreeze solution. A bit more complicated than a heating cable but has some advantages:

    Lower temperature gradients so less likely to harm the slab.

    Temperature sensors are accessible, not embedded in the slab.

    Antifreeze solution can be heated by electric, gas, or solar, whichever is most economical.
  • mctriviamctrivia Posts: 3,772
    edited 2013-01-14 10:22
    I had thought about that but thought the wire may be more control able and also less maintenance. Also no need for constant running pump.
  • mctriviamctrivia Posts: 3,772
    edited 2013-01-14 10:27
    Oh and electric is the only option where it is located. Plan to put pv panels on roof eventually though.
  • kwinnkwinn Posts: 8,697
    edited 2013-01-14 10:27
    I plan to build a new house in the not too distant future and it will have in floor radiant heating with zone controls for the individual areas. It is a bit more expensive than the standard forced air systems but provides a much greater level of comfort and lower operating cost.
  • kwinnkwinn Posts: 8,697
    edited 2013-01-14 10:46
    mctrivia wrote: »
    Oh and electric is the only option where it is located. Plan to put pv panels on roof eventually though.

    The heating cable should work well as long as the control system keeps the cable from getting too hot in relation to the surrounding slab. To that end you may want to put one or more sensors next to the heating cable and some in the slab to monitor and control the temperature difference.

    I wonder if you could use the heating cable itself as its own temperature sensor. Most heating elements do change resistance with temperature so it may be possible to monitor the heater cable temperature by measuring the current draw or the resistance when the power is off.
  • mctriviamctrivia Posts: 3,772
    edited 2013-01-14 14:33
    Resistance can't be used because no legal way to measure it. I could use a ct around the wire and an ad converter but could we not just use the thermistor in the slab assuming trying to heat as slow as possible? If only increase duty cycle when temperature is dropping should keep tight tolerance with minimal heat deviation throughout the slab.

    May be worth running ipex through main slab to allow for future heating of entire slab and use the heat wire by the door only
  • kwinnkwinn Posts: 8,697
    edited 2013-01-14 18:33
    mctrivia wrote: »
    Resistance can't be used because no legal way to measure it. I could use a ct around the wire and an ad converter but could we not just use the thermistor in the slab assuming trying to heat as slow as possible? If only increase duty cycle when temperature is dropping should keep tight tolerance with minimal heat deviation throughout the slab.

    If I were to do something like this I would do the following:

    - Use 10K thermistors so the wiring resistance will have minimal effect on the reading.
    - Place the thermistors in pairs with one close to the heating cable (cable temp.)and the other as far from it as possible (slab temp. - centered between runs of heater cable?).
    - Use the slab temp. to control the power to the heating cable so the temperature difference between the slab and cable is kept constant (say 10 - 20 degrees C) until the slab reaches the preset temperature of 1 degree. At this point the controller would maintain the slab temperature.
    May be worth running ipex through main slab to allow for future heating of entire slab and use the heat wire by the door only
    Certainly a lot cheaper to do during construction and it would allow you to heat the garage for use as a work shop in future.
  • mctriviamctrivia Posts: 3,772
    edited 2013-01-14 19:14
    Sounds like a good plan. Any trick to read these without using an ad chip?
  • kwinnkwinn Posts: 8,697
    edited 2013-01-14 20:03
    Use the thermistors in rc time constant circuits on 2 prop pins would be the simplest. To get accurate readings near the setpoint you will need to calibrate at least one of the thermistors around that temperature. It might be simpler to add one of the digital temperature chips to do that.
  • CircuitsoftCircuitsoft Posts: 1,166
    edited 2013-01-14 21:01
    You could probably cut the heat tape into shorter (1/20 length) sections so it'll run at 12V. From there, run romex to parallel them, then you can find a CE certified power supply, and control it yourself.

    If you want to do it yourself, you can find a bunch of Ohmite OY series resistors, which have a -.09%/Deg. C rating curve. It's not hard to build a circuit that will modulate power into it while simultaneously measuring its resistance. Is that something you're willing to do?
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2013-01-14 23:22
    Well, if it is construction of a new home, I'd use a proven system. My personal preference is hot water/or other fluid as it can be augmented with solar panels.

    I don't see any reason that a proven in slab electrical system that is properly installed should damage the concrete. And you have opportunities to provide adequate drainage, good insulation, and steel reinforcement.

    Using a custom designed gadget is likely to never perform the way you want and installation of a proper system after completion will likely cost more, not easily included in the financing, and not add value to the home.

    The main issue here is the stability of your soil. If you are on clay soil or perma-frost, you have some rather special design issues. The clay soil will 'heave' with large changes in moisture content. And perma-frost needs to stay frozen if at all possible, as it turns to mush and muck.

    Your biggest cost savings is not going to be in active control of a limited system, but in proper insulation and proper moisture barriers which minimize unnecessary heat loss.
  • mctriviamctrivia Posts: 3,772
    edited 2013-01-15 04:48
    You could probably cut the heat tape into shorter (1/20 length) sections so it'll run at 12V. From there, run romex to parallel them, then you can find a CE certified power supply, and control it yourself.

    The heat wire is a parallel heater. There are 2 copper wires that run the length of the wire with a resistive material between them. Cutting the wire only reduces the heating capacity it does not lower the voltage.
    Well, if it is construction of a new home, I'd use a proven system. My personal preference is hot water/or other fluid as it can be augmented with solar panels.

    I don't see any reason that a proven in slab electrical system that is properly installed should damage the concrete. And you have opportunities to provide adequate drainage, good insulation, and steel reinforcement.

    Using a custom designed gadget is likely to never perform the way you want and installation of a proper system after completion will likely cost more, not easily included in the financing, and not add value to the home.

    The main issue here is the stability of your soil. If you are on clay soil or perma-frost, you have some rather special design issues. The clay soil will 'heave' with large changes in moisture content. And perma-frost needs to stay frozen if at all possible, as it turns to mush and muck.

    Your biggest cost savings is not going to be in active control of a limited system, but in proper insulation and proper moisture barriers which minimize unnecessary heat loss.

    with 12kW of solar panels I should be able to supliment my entire household electrical bill to $0 a year. water based solar panels do not work well here because of the extreme weather conditions and they have been shown to reduce the life expectany of the shingles. In floor heating both electrical and water based have been shown to reduce the life expectancy of the slab unless kept at a constant temperature.
  • mctriviamctrivia Posts: 3,772
    edited 2013-01-15 04:53
    kwinn wrote: »
    Use the thermistors in rc time constant circuits on 2 prop pins would be the simplest. To get accurate readings near the setpoint you will need to calibrate at least one of the thermistors around that temperature. It might be simpler to add one of the digital temperature chips to do that.

    I am surprised the ds18s20 does not seem to be in the obex. 25 for $100 seems reasonable for all digital temperature monitoring.
  • BeanBean Posts: 8,129
    edited 2013-01-15 05:04
    To my way of thinking, if the goal is to prevent ice on the surface of the concrete, then that is where the temperature sensor should be.
    Not embedded in the concrete.

    How about using embedded metal pipes and circulate hot air through them ?

    Bean
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2013-01-15 19:31
    The 'damage to the concrete' has to come either from thermal cycles of expansion and contraction. This is usually overcome by the inclusion of a mesh of steel reinforcement. I do understand that you worry the electric heat elements are too extreme.

    Or by the expansion of water within the concrete when it freezes (concrete is normally porous). This can be controlled by a moisture barrier under the slab and by sealing the surface of the slab with a penetrating sealer (such at Thompson's water seal).

    One alternative is to heat the slab by direct solar, if the slab is properly design as a heat storage element. This is a passive solar approach that doesn't require any electrical or water pipe in the concrete. You just have to have enough available sunny days to accumulate heat and adequate thermal windows to warm the slab. Even in snow country and the far north, it is possible to get passive solar proving 90% or more of the heating when sunny days are available.
  • mctriviamctrivia Posts: 3,772
    edited 2013-01-15 19:41
    No windows in the garage so passive solar won't work. Garage doors are on south and to expensive to get insulate glass garage door. Not to mention more secure without easily broken glass.
  • mctriviamctrivia Posts: 3,772
    edited 2013-01-15 20:37
    if anyone is interested here is the house design.

    Insulation is R9 rigid under basement and garage floor, and around perimeter of house
    R20 spray foam on east,south, and west walls of main house, R33 spray foam on north wall
    R20 fiberglass insulation on garage walls.
    Roof is R10 spray foam with R40 blow in on top of it.

    Lots of insulation.
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2013-01-16 01:58
    Well, for the most part you are going about it in the right way.
    It seems the building codes have ample amounts of insulation required.

    The dilemma is that this is a garage and will often be as cold as the outside.

    If the doorways you want to heat are small, you could actually put your heater wire in an aluminum or bronze threshold atop the concrete. This might be a nice alternative as repair, maintenance, and replacement are not in the concrete. If you need to heat the bottom of the big garage doors, why not have the heating on the bottom door edge? If it is a design detail on the door, it again is more easy to repair, maintain, and replace.

    Maybe you could detail the slab to be better insulated and have the garage heated in another manner when you need to.

    I lived in Oregon where I was a general contractor and did some interesting slab on-grade construction. One method is to actually create a slab that has loose fill (such as un-compacted gravel or sand ) under it when poured, and then the gravel/sand eventually subsides to create an air space. You still provide insulation under the slab, and you provide a good moisture barrier - but the slab is engineered to actually support itself over a dead air space by the perimeter that has extra rebar. In some cases, with a large floor, you need an additional footing running down the middle of the slab with extra rebar. This results in much less problems with ground water below the slab and adds significant insulation as dead air space. Plus the floor is much better reinforced than the usual 6 x 6 wire mesh. It is engineered as a suspended concrete floor.

    For a moisture barrier under the slab, I used 50 pound roofing with glued seams on top of the loose fill and under the footings continous.

    Of course, you could just double that R9 under the garage slab if the insullation is cheaper than the modifications.

    With icing and wet/dry situations, sealing the top of the slab after it is poured with a water seal will get rid of spall and may actually eliminate the need to heat the slab unless you just want that for occasional use of the space.

    I did look at your plans and these solutions seem quite possible to adopt.
Sign In or Register to comment.