EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: NiHaoMike on July 13, 2013, 04:26:20 pm
-
Are there any reliability issues with permanently shorting output to ground of a 7805 and using it as a heater? The application I have in mind is a crankcase heater for a refrigeration compressor. A TO220 package nicely thermally couples to the underside of the compressor, whereas a standard power resistor does not.
-
I don't think the thermal overload protection will play nicely. IIRC the hysteresis is large, so you'll just temperature-cycle. Use a transistor (or LM317, but you waste more power in the sense resistor) set up to a rough constant-current mode.
-
You can get thick film power resistors packaged in a two pin TO-220 package (http://www.ohmite.com/search.php?function=results), intended for low inductance/RF applications. Search for TCH35 and TBH25 on that page. Tend to be expensive, though...
-
We used to manufacture a diesel heater (diesel is a thick syrup when cold) that used metal clad wire wound resistors as the heaters, they are perfect for this kind of thing.
Ie ..
http://www.ebay.co.uk/itm/1R92-Ohm-25W-Metal-Clad-Resistor-Wire-Wound-Axial-EC28-/160930841963 (http://www.ebay.co.uk/itm/1R92-Ohm-25W-Metal-Clad-Resistor-Wire-Wound-Axial-EC28-/160930841963)
http://www.ebay.com/itm/METALCLAD-RESISTORS-50WATT-WIRE-WOUND-/251187175012? (http://www.ebay.com/itm/METALCLAD-RESISTORS-50WATT-WIRE-WOUND-/251187175012?)
The TO-220 are also perfect little heaters.
You just have to pick the correct value and wattage.
-
hey
As mentioned above . you can buy TO-220 package resistors .
But I think you really should try my idea :
ATX power supplies are ridiculously cheap ,
You may want to use IRFP250N's with small Nichrome wier wound resistor as proper shunt :
http://www.ebay.com/itm/Nichrome-resistance-wire-11-AWG-gage-5-ft-for-foam-cutting-kilns-ceramics-/321009587312?pt=LH_DefaultDomain_0&hash=item4abda99470 (http://www.ebay.com/itm/Nichrome-resistance-wire-11-AWG-gage-5-ft-for-foam-cutting-kilns-ceramics-/321009587312?pt=LH_DefaultDomain_0&hash=item4abda99470)
using Op07 as fine negative feedback op amp :
http://www.ebay.com/itm/4-pc-OP07CP-OP07-DIP-8-LOW-OFFSET-VOLTAGE-Op-Amp-/180842619352?pt=LH_DefaultDomain_0&hash=item2a1b0f61d8 (http://www.ebay.com/itm/4-pc-OP07CP-OP07-DIP-8-LOW-OFFSET-VOLTAGE-Op-Amp-/180842619352?pt=LH_DefaultDomain_0&hash=item2a1b0f61d8)
At 12V I could ran that Nmos at 120W . 24/7 !
but my room became hot . LOL ! ( that was 5 units . so 5*120W is 600W ... so yes ....... :palm:
and the bills for the electricity was way higher , but that was in the name of science ! ^____^
I didn't ren that config for more then two weeks , so I can't tell you long term specs , But BK8500 uses 8 of these for dissipating 300W , so 37.5W per package looks like overkill , but 80W per package seem to be the limit for long term performance .
How big is that device ? , can you tell us what material is it made of ? , because aluminum has 910(J/C) heat capacity , and water got 4200 , and so on . so that we can tell you what power to use in order to reach (X)temp at (X) amount of time
good luck ! :)
-
Typical crankcase heaters for the compressors of the sizes I'm dealing with (less than one ton) are rated on the order of 40W. I'm actually looking at some setup that can go up to the order of 100W (I'm already using a modified PC power supply to run some stuff) so that it can leave the heater off most of the time but quickly boil off any refrigerant that migrated inside just prior to starting the compressor. I'll definitely explore the option of using transistors as both the heaters and control devices, but I also just had an idea to wrap the bottom of the compressor with some copper wire and then connect that coil to the unrectified side of the PSU to make an induction heater.
-
Just go and buy a band heater for it and use it. Simple and works well. You can get a PTC one that is self regulating as well, high heat when cold and dropping to almost zero when it reaches around 30C.
-
It's one of those "tall and skinny" rotary compressors, so a cheap band heater won't fit.
-
They make them for those.....
http://www.ries-gmbh.de/index.php?module=products&index (http://www.ries-gmbh.de/index.php?module=products&index)[products][category]=127&lang=en
I rarely use them as here heat pumps are not exactly going to spend a lot of time in heat mode, though if the unit has a heater on the compressor I always connect it to power so that the compressor is happier with no liquid condensing in the suction line by the unit.
-
I think the linear reg heater would work well as long as you don't exceed the max operating junction temp of the device. Doesn't sound like you want to heat up that case to those high of temps anyway.
I think it would be a cool project because it uses one of those devices in a way it wasn't original intended for. There will likely be some challenges though in that for the initial temp ramp you will likely exceed the max current of the device before it reaches steady state so a max current hold plus temp regulation would have to be used.
-
I actually tried the induction idea and it worked very well. I simply wrapped 5 turns of AWG18 around the bottom of the compressor and connected that to a modified PC power supply I usually use for exciting flyback transformers and stuff. (About 20Vpp at 30kHz.) It only took a few minutes for the bottom (sump) of the compressor to get warm to the touch. Power usage was about 70W with the PSU turned all the way up. Efficiency should be better than a heater operating from the output since it bypasses the rectifier. A heater operating directly from mains, of course, would be more efficient, but then I'll need a high voltage relay or SSR to control it. A Peltier could possibly be even more efficient, but the bottom of the compressor doesn't have a flat spot big enough for a typical Peltier.
There were also a few designs that used the start winding of the compressor motor and its run cap as the crankcase heater. But that seems to only work well on some compressors since on mine, the top part got warm but the bottom stayed cold. Which means it won't be effective at vaporizing the liquid refrigerant out of the oil.
Theoretically, it could also be possible to use some phase change medium (wax or something) to store heat coming off the discharge line and then transfer that to the compressor sump as needed. I might explore that option if I end up with enough leftover room in the case.