Newby and a simple temperature controller

[dethaspagan]

Hey guys, i am trying to build a simple temperature controller. That will turn on a heater when it is less than a set temperature and turn on cooling device when it is to hot and does nothing at the set temperature. To do that i did it a bit of research and i put together the below. Which i hope you can see.



I built it and it does not work. Which is kinda what i was expecting. If you could give me a hand on how to go about fixing it. i think it may have to do with my resistors but i am kinda stumped and sick of going around in circle's.

oh and this is my first circuit.

I have had a good look around the forum and it looks all pretty cool if not complex.
so any help would be appreciated.

thanks,
Tim

 

[betwixt]

What happens at the moment?

Adjusting the variable resistor should make the output pin of the IC go from +12 to -12V as long as the relays are only drawing a small current. Try disconnecting both of them and use a voltmeter between pin 7 and ground to see what voltage is there.

Also check the following:

The schematic needs two power supply rails, +12V and -12V with the middle connected to ground, is this what you are using?
Like this, -12V----PSU----0V - LINK TO GROUND - 0V----PSU----+12V

There is very little current available to operate the relay coils, are you using suitable relays?

Brian.

 

[FvM]

I see, that your circuit is derived from an LM335 datasheet application. Your variant has at least one error and one bad property.
- The LM311 has an open collector rather than an push-pull output, so it can't drive two relays as you intended. A standard OP could, if the relay current isn't too high.
- The circuit has no hysteresis. This is acceptable when directly driven a heating resistor as in the original circuit but very bad with relays. Near the setpoint you get continuous relay clattering and a rather short relay lifetime.

More generally, I doubt if a heat/cool controller should be designed as on-off-controller, because it must be expected to cycle permanently between heating an cooling.

 

[dethaspagan]

ok the relays have a resistance of 400 ohms and the diodes have a 1vdrop so the current is 11/400 = 27.5mA.

I think i have figured my original problem the lm317 needs 3mA of current load currently it has closer to 1mA. I will make another 5k branch parallel to the lm335.

any idea on how to implement the hysteresis?

 

[KMoffett]

As Fvm said?

Near the setpoint you get continuous relay clattering and a rather short relay lifetime. ....because it must be expected to cycle permanently between heating an cooling.

In order to get the HEAT<> OFF<>COOL operation you will need to replace the single comparator with two in a "window comparator" configuration with two setpoints. If the temp is below the lower setpoint one comparator turns on the heat. If the temp is above the upper setpoint the other comparator turns on the cooling. If the temp is between the two setpoints both comparators are off and you neither heat or cool.

Ken.

 

[dethaspagan]

ok that makes sense would i be able to use 2 * lm311 to do this or would i be better of using a TL074 op amp. i have both.

 

[KMoffett]

The TL074 is an opamp. The LM311 is a comparator...use the 311s.

Ken

 

[dethaspagan]

with the lm335n temperature sensor. i am getting rather odd voltage drops approximately 1.20v and i am at room temperature. it is uncalibrated. I have tried several different lm335n's but they are all giving me the same value so it cant be a fault with the component. does it need to be calibrated to have just kelvin values or could an outside factor be affecting it.

i am really stumped.

 

[gordon55y]

dethaspagan,
You are getting 1.2V on lm335
Check polarity.
See your schematic for polarity.
Is this a 3 pin device?
Or is it the 8 pin dip?