• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Silicon Chip Magazine Thermostat circuit.

Hello. First post as I am needing some advice.

My old and faithful modded laminator died recently. It used an household iron thermostat to control the temperature and worked very well for the toner transfer method of PCB fabrication and creating sheet metal decorative items using the toner and etch method.. But having got a new unit to modify I decided to use a digital circuit to control the heater elements. After some browsing I came across this unit in the Silicon Chip magazine Jan 2007.

I made the PCB myself and have built it. The fun started when testing. To set the unit to a given threshold you measure the voltage at the wiper of VR1. This equates to a given temperature (though testing the actual temp will be done to ensure accuracy). It can be set to sense High to low or Low to high and has adjustable hysteresis.

Set up as is in the diagrams it seemed to be working backwards. Though I may be confused. So the set up and use of his circuit will be...

Using the normally closed relay contacts to switch the heater elements on and off. The thermistor will be fitted to the aluminium extrusion, where the heaters, are located to detect the temp of course. Temp setting to affect transfer of the toner will be 170° - 180°. So initial state when powering up Laminator and the thermostat circuit. Heaters live. Relay not activated. When temperature reached Heaters off, Relay activated.

My problem is as set up it initially wouldn't trip the relay. The LED indicator was on at power up so the relay should have tripped. Initial test state of pots was full anti clockwise on hysteresis and turn set point to full anti clockwise until the LED and relay activated as per the article text. This would be 0°. Initial state in H/L was relay activated (once fixed). So at this point the initial state was power on relay activated. Not what I wanted.

I solved the relay not tripping by changing the indicated resistor to 1K. I should point out I used a BC558 as I had one to hand and not the specified BC337. Is this why it wouldn't work initially?

To get the correct initial state of relay in the NC and switching at higher temperature I had to change link1 to the H/L setting and reverse the diode D3 as described in the text. So it works, subject to calibration and fitting to the laminator. But I would like a second opinion on the accuracy of my correction and understanding of the H/L - L/H setting.

I am just a hobbyist with electronics. I know enough to be dangerous but successful enough with most projects and fault finding where I can.

Thanks in advance
T&M
 

Attachments

Les Jones

Well-Known Member
The reason the circuit did not work at first is because you used a PNP transistor instead of the NPN device specified. I think changing the base drive resistor to 1 K enabled the PNP transistor to work as an emitter follower. This would also mean that the switching sense would be reversed. (I.E when it should be on it would be off and vice versa.)

Les.
 
Thank you for your reply........ How embarrassing is that! OMG! I can't believe I missed that. Right best correct that. I have a BC548 I can use.

regards

T&M
 
Could be worse. You could have thought you were using a BC548 when you were using a BC558.
Quite. Been a while since I have done any projects and I let an assumption on my part make this mistake. Lesson learned though.

Put a 10K resistor back in place and fitted the diode back as per the text and changed the H/L - L/H jumper back to H/L and replaced the transistor with a 2N3904. All working perfectly now.

Cheers

T&M
 

dknguyen

Well-Known Member
Most Helpful Member
Yes. Hysteresis is positive feedback and zero positive feedback is no hysteresis (aka open-circuit in the positive feedback loop).

So increasing the resistance there in some way, either by adding a fixed resistor in series with whatever is there or replacing the pot with another pot with a higher resistance should work.

Those resistance values are influenced by other resistors in in the circuit (i.e. the other two resistors on the non-inverting input of that op-amp) so you might run into a wall with high high you can make the hysteresis resistor and still have it work properly. If you do, then you should be able to decrease those two resistors somewhat (namely the 10K) so that the threshold voltage is driven harder relative to the positive feedback voltage.
 
Last edited:

EE World Online Articles

Loading

 
Top