I am planing on making a little PC controlled thermostat . Maybe I will use this to control a combustion furnace or possibly conduct an automated
distillation...who knows. Anyways...
I know 'normal' relays wear down because of moving parts so it would not be a good idea to have them open and close frequently and unnecessarily.
However what about solid state relays? Would it be foolish to use a solid state relay to effectively deliver X% power; that is for 7 seconds have a
circuit closed and for 3 seconds open to deliver 70% power or would I quickly destroy the relay? Just to make sure you understand the meat of the
question, is the life of a solid state relay determined by the number of times it opens and closes or by the number of hours of use it gets?
Thank you in advance.Magpie - 7-11-2007 at 22:39
I think as long as you stay within the voltage and wattage limits of the relay its life would be virtually unlimited. I believe they also are rated
for an upper limit on switching frequency, but it would be very high and not a concern for your application.
I built such a homemade thermostat using a solid state relay:
As you can see in the picture it is installed against a heat sink, as recommended.smuv - 7-11-2007 at 22:55
Thanks Magpie,
I have a question regarding your thermostat (I guess PID is a more correct term), let's say you set your furnace to 200c, does it turn heat off
exactly at 200c and turn heat on right when the temp falls to 199 or does it hold the temperature in a certain range (I am just trying to
understanding the logic it uses so I can program accordingly)?chromium - 7-11-2007 at 23:35
Solid state relays are very reliable when switching common heating elements. They are problematic if you want to switch inductive loads or devices
with high startup currents (such as transformers or some incadescent lamps). 1/1000th second of overcurrent or overvoltage is often fatal for them.
Solid state relays must be choosen for maximum possible transients not for average current/voltage level.
Usally thermostats try do hold temperature on certain range. If you want temperature not to exceed 200C you need to switch off before 200C because
there always is overshot as heating elements and its surroundigs deliver heat quite some time after power-off.
[Edited on 8-11-2007 by chromium]Magpie - 8-11-2007 at 11:00
The PID controller responds to the "error signal" which is the difference between the observed temperature and the setpoint temperature. The PID
controller is tuned to optimize this response so that the error signal, or delta T, is minimized. This tuning can be done either manually or through
the "auto tuning" feature.
As chromium has indicated every load (furnace, heater, etc) will have different response characteristics. The tuning of the PID controller optimizes
the response of the heater.
In practice, the observed temperature will oscillate around the setpoint. Good tuning minimizes this oscillation.Eclectic - 8-11-2007 at 11:05
Very good industrial PID controllers can be obtained off Ebay for under $50, many with a communication channel for direct computer control. I'm not
saying you shouldn't make your own, but you might want to take a look on Ebay.smuv - 8-11-2007 at 12:36
Thanks for the pointer Eclectic, but for now I am doing this more because I want to see if I can interface something simple with a computer (lets see
what 1 semester of programing has done for me). It is something I know i will find an application for later on. Also, the thermostat will be almost
free to make thanks to some free samples sent to me from maxim (semiconductors).
Thank you Magpie and chromium, the information is greatly appreciated.
[Edited on 8-11-2007 by smuv]
[Edited on 8-11-2007 by smuv]trilobite - 8-11-2007 at 14:21
Read up on triacs/thyristors and triac driver optocouplers. Triacs do need some cooling as they will have some voltage drop, but the heatsinks
shouldn't be a big problem. Inductive loads can be controlled when you provide the triac/thyristors with a proper snubbing circuit and choose the
driver correctly for your application. Here are some Fairchild application notes for you:
Commercial solid state relays designed for turning motors on and off are fairly bullet proof. The requirements for driving resistive heating elements
are not so demanding.Twospoons - 8-11-2007 at 17:01
Put in appropriate transient protection, make sure you use an over-rated ( by about 20%) solid-state relay and you should be fine.
That said, I've had one fail spectacularly - it jumped about a metre into the air, trailing a 30cm blue flame. I wasn't being very nice to it at the
time chemrox - 8-11-2007 at 22:32
I hope this is related enough:- I'd like a pointer or two on using the (thermocouple ?) leads on my heating mantles to control heating so as to
control flask temperature. I want to specify a T and have it cut out when the temp gets above that.smuv - 8-11-2007 at 23:20
Quote:
id like a pointer or two on using the (thermocouple ?) leads on my heating mantles
Do you mean that your heating mantle has an integrated temperature sensor (its just surprising to me)? If so it is very likely a thermistor due to
the much lower cost of such devices; if the leads leaving your heating mantle are composed of the same material it further indicates that a thermistor
is what you have.
After that...I can't help you much...that is unless you want to interface the thermistor with your computer Magpie - 9-11-2007 at 10:09
chemrox:
I have some old Glas-Col heating mantles that have the embedded thermocouple. I have never used this feature.
Somone else on this board advised me that that is what they are.
A thermocouple provides a voltage at the lead ends (in millivolts) that can be converted to a temperature. You just need to know what type
thermocouple it is plus a lookup table to convert the mv to deg C. See the Omega Instrument site for tables, etc.
