Connect with us

SSR Heat Sinks

Discussion in 'General Electronics Discussion' started by Fish4Fun, Dec 17, 2013.

Scroll to continue with content
  1. Fish4Fun

    Fish4Fun So long, and Thanks for all the Fish!

    464
    105
    Aug 27, 2013
    In addition to my electronics hobby, I have a home brewing hobby. I have a fairly large system for a hobbyist, and do 20-25 gallon "batches". Regardless of scale, brewing requires a fair amount of heating water (and then cooling it back down, hehe). I started with Propane, but for various reasons have migrated to "All Electric".

    The heating elements I use are 3kW-6kW @ 240Vac, so after more than a year of "flipping breakers" to control the power, I have decided to give SSR's a try. I just received some "Ruike SSR40DA" SSRs. This PDF http://www.google.com/url?sa=t&rct=...=MhYNiYC97YH9t6jSNR6DYQ&bvm=bv.57967247,d.cWc from Crydom absolutely confirms my low marks in thermodynamics, and it appears to be as "dummed down" as one can make Thermodynamics, lol.

    While IN THEORY the fairly large heat sinks provided with the SSRs **should** handle the 45W of SSR heat gain, but I have little faith. My plan from the get-go was to mount the SSRs (five of them) to some 4" square aluminum tube stock (I happen to have on hand) with the SSRs on the "outside" of the tube stock and the heat sinks on the "inside" of the tube stock. I then planned on making water tight "end caps" for the tube stock and using a thermal sensor to regulate water flow through the tube (and over the heat sinks inside). With my ground water temp being ~58F (15C), I feel 100% confident I can keep these Chinese SSRs cool.

    ....but then the stupid side of me thought, "hrmm, 45W * 4 (most that should ever be on at one time) is almost 200W! That's a lot of heat...hrmmm how could I use that? (YES, the elephant in the room is the 24kW being switched!) As luck would have it, my brew room is located in a room below a built-in hot tub. The hot-tub is generally set < 104F ( 40C ), and it would be fairly trivial to "dump that heat" into the hot tub....1500L @ 104F + 200W for ~6 hours = 1.2kWh would only raise the hot tub temp ~1.24F (0.7C).......

    ...BB's @ Tanks, LMAO! But, it did lead me to my actual question, "Knowing I can keep these SSRs darn near cold, what is a good target temp? While I haven't seen a real data sheet on these specific SSRs, 125C seems to be the "max" on most NON-Chinese SSRs. I know I should assume these should be run less than 75% of rated current, and I assume "cooler is better", but how cool is cool enough?

    A side-bar question, "If I wanted to implement PWM control for maintaining temperatures, should I go to the trouble of timing the PWM with zero-crossing?" My thought was to use a 4 second base frequency and have the "on time" go in 20 cycle steps (ie 4 seconds * 60Hz = 240 cycles per period, duty cycle would range between 1/12 to 1 or 500W to 6kW in 500W "steps". ) Maybe I am over-thinking this? I mean realistically I have been doing just fine flipping a breaker, lol (well, I have NEVER been happy with the breaker thing ;-) )

    This is my FIRST adventure into SSRs, so any more salient suggestions would be VERY MUCH appreciated!

    Fish
     
  2. (*steve*)

    (*steve*) ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd Moderator

    25,497
    2,839
    Jan 21, 2010
    You probably need to understand how heatsinks work.

    You know (I haven't checked, but I'll believe you) that the dissipation will be 45W.

    There are various heatsinks in that pdf, the important difference is their thermal resistance given in degrees C per W.

    Let's assume you choose a 2 degC/W heatsink. Let's also assume the temperature in your brewing room is 40C (because that's the temperature of your very large water bath).

    This means that the temperature of the device will rise by 90C (2 degC/Watt * 45 Watts). That means 130C, and that's too hot. (It also doesn't take into account other inefficiencies)

    If you can get a 1.5 degC/W heatsink, the temperature rise will be 68C, leading to a temperature of 108C -- and that's still very warm.

    Adding a fan would probably fix the problem, but your idea of attaching it to something carrying a cooling liquid would be even better.

    As for PWM, the thermal mass of your tank is so large that you could probably get away with a very slow switching, what you suggest would be way fast enough.

    Assuming these SSRs use mosfets internally (are they specified for DC switching?) then the heat dissipation is related to the square of the current. Operating them at 75% of their rated current will produce only a little over half the heat of running them at full power. It might be worth checking.

    If you can attach them to a cooler surface (say one that has beer on the other side) then the heat lost by the SSR could also be used to heat the beer.= and the junction temperature would stay far lower.
     
  3. Fish4Fun

    Fish4Fun So long, and Thanks for all the Fish!

    464
    105
    Aug 27, 2013
    Thanks Steve!

    Being a little bit less verbose, here is what I have: http://www.ebay.com/itm/271231725184?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1439.l2649

    If you can glean from the picture the efficacy of the heat sink then you are WAY better than I am. I know your feelings on Chinese parts, and to a large extent I agree, but I am hoping that by staying significantly under spec on current (rating 40A, largest load ~25A) AND force cooling them using water that I can make them work long enough to tweak the rest of the system and then replace them as they fail with higher quality parts. I just didn't want to drop $100+ per switch into this until I was sure I had everything right.

    So is attempting to keep the SSRs ~40C a "good target"?, or should I aim lower?

    Fish
     
  4. (*steve*)

    (*steve*) ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd Moderator

    25,497
    2,839
    Jan 21, 2010
    Sorry about that response. I didn't even notice who was asking. I probably covered stuff you're familiar with.

    Keeping these things as cool as possible is good, but as the allowable junction temps are way above 40 C, that would be fine. I prefer to keep things cool enough that I wont get a surprise if I touch it.

    And yeah, I'd try to keep within 50% of what might be very optimistic specifications.

    I've not had a SSR fail on me, but I suspect that it would fail short circuit. And that could be undesirable to say the least.
     
  5. Fish4Fun

    Fish4Fun So long, and Thanks for all the Fish!

    464
    105
    Aug 27, 2013
    Thanks Steve!

    I have used thyristors extensively in the past, just not configured as a "packaged SSR". I did spend a couple of days "mocking up" a thyristor SSR using 6 * BTA24 in parallel with a common trigger. I tested it on a single 1500W heating element and was pleased enough with the results to order these Chinese modules (5 for less than $60 seemed too cheap not to try them. I doubt I could buy/make the heat sinks for that, lol) The heat rise in my BTA24 mock-up (with NO heat sinks) @ only ~6A convinced me that I would need to be aggressive in heat removal. (Using Chinese Math, 6 * BTA24 = 150A rating LOL ;-) ) I was just looking (in a very long-winded way) for some advice on a target temp, and your response was the perfect match for my verbosity!

    Again, Thanks!

    Fish
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-