hysteresis problem

[snoop_electro]

hello all

I got a problem with one simple circuit . Problem is about hysteresis ..

To make things clear : when the temperature is at the desire level, the relay is cut... but 10 sec after the cut, relay is one again for 40 sec... another cut.... and 10 sec later on again.. and so on.

How do I can add hysteresis control to this circuit... by adding a capacitance ? Or I just have the choice to add a timer delay at the end of the cycle.. to be sure that the off-on-off-on etc.. has a minimum of time between the on-off segment

thanks for help

 

[Sebi]

You can play with R2 and R4, parallel cap with thermistor, but - i think - need a small heatsink for thermistor, and/or a new place in room.
Try out this modification.

 

[snoop_electro]

thanks, i will try it !!

 

[0x34]

I'd say about 4.7 to 6.3 uF rather than 47uF.... But the cap should resolve the prob.

 

[john1]

I think that the activity of R2 is upsetting the
normal hysteresis associated with most temp sensors.

It might be better if the output did not alter the
conditions at the input.

Maybe try it again without R2 connected?

 

[john1]

Hmm ...
yes the output is supposed to alter conditions at the input,
but not within the unit,
its supposed to alter the temperature,
and the sensor operates,
then the conditions alter at the input.

the hysteresis timing is thereby determined
by external factors.

(hope thats a bit clearer ...)

 

[Roff]

John, as far as I know, thermistors don't have hysteresis.
The feedback through R2 yields about 100mv hysteresis at the bottom of the pot, but only about 10mv hysteresis at the top of the pot. The amount of hysteresis depends a lot on the temperature threshold setting. Also, the circuit probably won't even work at the bottom end of the pot, because the +input will be biased outside the common mode range of the 741.

 

[0x34]

To avoide the feedback loop's voltage from affecting the input, you need to get away from the Bipolar OpAmp and use a FET input OpAmp. The FET version has no input conditioning resistors and therefor a change in voltage level on one input will not affect the other input. 8)

 

[john1]

Hi Ron,
You're quite right thermistors don't have hysteresis.
I was thinking of the interval between the heating
and the response from the sensor.
This gives a temperature range whilst the heating rises
and the sensor operates, and of course a similar range
whilst the temp drops and the sensor operates again.

So yes a thermistor does not have hysteresis, but the
result is that a(n) hysteresis effect still occurs.

The disadvantage with systems based on a single temp
sensor is that the system cannot rest, it will always
be either rising or falling.

Now that i have looked carefully i can see that the R2
does as you say provide hysteresis, by effectively
moving the input point a little, so this system could
if conditions allowed, come to rest between the two
settings.

I didn't really follow the bit about the relay engaging
and disengaging at intervals after the required temp
was reached.

I have re-read it a few times but i still don't get it.
My monitor is small, i am at 640 x 480, and i am fed
up with panning left and right.

I think sebi has the right approach to make the hysteresis
into a variable setting, but i would not have changed
the 100 res. as much as that, maybe to 200.

Changing that for a 1000 would push its voltage up quite
a bit, i think i would rather leave that value as it is,
or maybe a bit more like double its value.

Still, see how it goes.

 

[snoop_electro]

hello all !!

I'm working on the problem.. and maybe add a bit of delay with other circuit...

I will post the final on this web

Ric