OK, so it is very easy to set an overheated temperature. Now the problem is clear! The solution is not quite so easy now! Still, one option is a small thermostat set to open the heater power connection if overheat is approached. But a thermistor and an electronic control would be better.

 

OK, so it is very easy to set an overheated temperature. Now the problem is clear! The solution is not quite so easy now! Still, one option is a small thermostat set to open the heater power connection if overheat is approached. But a thermistor and an electronic control would be better.

Yes! as you said before in the above post that is needed a thermistor. Can I use a non inverting comparator with thermistor to sense the temperature? And what about thermocouple will that be possible ?

 

Yes! As you said before in the above post that is needed a thermistor. Can I use a non inverting comparator with thermistor to sense the temperature? And what about thermocouple will that be possible?

I want you to tell me more about the below sentence that you have posted:

"But a thermistor and an electronic control would be better"

 

When all of this began you claimed in post #16 you had a Type K thermocouple which at this point I seriously doubt. In the interest of keeping this simple and easy to do I suggest you measure the resistance of the temp sensor wires. If there is a thermistor out there you will get a resistance typical of ambient temp. Do this with the wires disconnected. Next. reconnect the wires and measure the voltage across them with the unit heating up. What do you see? Assuming a thermistor is being used you should see a changing voltage as it heats up. At set point the voltage should remain a constant.

What this will do is help determine what is out there sensing temperature and if what is there can be incorporated into an overtemperature solution as to what your goal here is. That will also tell us if the existing temperature sensor can be paralleled for an overtemperature solution.

Ron

 

When all of this began you claimed in post #16 you had a Type K thermocouple which at this point I seriously doubt. In the interest of keeping this simple and easy to do, I suggest you measure the resistance of the temp sensor wires. If there is a thermistor out there, you will get a resistance typical of ambient temp. Do this with the wires disconnected. Next. Reconnect the wires and measure the voltage across them with the unit heating up. What do you see? Assuming a thermistor is being used, you should see a changing voltage as it heats up. At set point the voltage should remain a constant.

What this will do is help determine what is out there sensing temperature and if what is there can be incorporated into an overtemperature solution as to what your goal here is. That will also tell us if the existing temperature sensor can be paralleled for an overtemperature solution.

Ron

Good! This was already answered in my previous post!

I will repost it below

 

Good! This was already answered in my previous post!

I will repost it below

What is did was I removed the heating core from the handle and connected the two yellow wires to mains and connected the black wire with the black probe of the meter and the red wire to red probe of the meter and set the meter to 200k. After that I switch the mains to heat the heating core and after some minutes the meter start reading from 0.0,0.1 to 9, 1.0 to 9,2.0,2.0 to 9. I waited and it reads up to 3.9 almost getting to 4.0 that where the metal case around the core started overheating and turn reddish

 

Those readings confuse me a bit. I am guessing that The thermocoule is egerating a voltage that confuses the resistance readings, especiallyon thehigher resistance ranges. Do a resistance check on a lower resistance range without heating. and then reverse the poarity and take the same reading again.

 

Also, keep in mind do not power the heating element less the blower running. Less any airflow over the element it can burn up pretty quick. Next part of what I suggested is with the sensor wires connected look for a voltage across the sensor wires with it running. At this point I question if the sensor is actually a thermocouple because if a thermocouple the board it connects to would have to have CJC (Cold Junction Compensation). If the sensor is a thermistor you will see a varying voltage which will be much higher than the millivolt output of a thermocouple. Devices like this seldom use a thermocouple since a thermistor is much less expensive and easier to read, while not as accurate as a thermocouple they are normally good enough. Thermistors also do not require CJC that I mentioned. I also suggested using a magnet on the wires, a Type K or J will have one magnetic lead.

Ron

 

Those readings confuse me a bit. I am guessing that The thermocouple is operating a voltage that confuses the resistance readings, especially on the higher resistance ranges. Do a resistance check on a lower resistance range without heating. And then reverse the polarity and take the same reading again.

The images below are the reading I got without heating and reverse polarity

 

The images below are the reading I got without heating and reverse polarity

I think the sensor in the heater core is type k thermocouple and its use to sense the temperature because of their hardness. And also so don't think thermostat will be embedded in the heater core. Do you have anything to say about this post?

 

Also, keep in mind, do not power the heating element less the blower running. Less any airflow over the element it can burn up pretty quickly. Next part of what I suggested is with the sensor wires connected look for a voltage across the sensor wires with it running. At this point I question if the sensor is actually a thermocouple because if a thermocouple the board it connects to would have to have CJC (Cold Junction Compensation). If the sensor is a thermistor you will see a varying voltage which will be much higher than the millivolt output of a thermocouple. Devices like this seldom use a thermocouple since a thermistor is much less expensive and easier to read, while not as accurate as a thermocouple they are normally good enough. Thermistors also do not require CJC that I mentioned. I also suggested using a magnet on the wires, a Type K or J will have one magnetic lead.

Ron

Do you have a circuit diagram on how to use thermocouple to sense the temperature and give feedback to the non inverting comparator to be able to get higher voltage to shut down overheating with the help of relay?

 

Do you have a circuit diagram on how to use thermocouple to sense the temperature and give feedback to the non inverting comparator to be able to get higher voltage to shut down overheating with the help of relay?

There are chips on small breakout boards. The AD4895 comes to mind and early in the thread another member mentioned one. Not knowing exactly where you are a Google of "Analog Output K-Type Thermocouple Amplifier ad8495" should get you some hits. The inexpensive versions come on a breakout board ready for a thermocouple input, they give you the CJC I mentioned and give you a voltage output proportional to the temperature measured. All you need for your application is a very small bead junction K thermocouple. This gets you from the mV out of a thermocouple to a usable voltage. From there, as Bill mentioned just feed it into a simple comparator circuit. Since you really do not want to read temp you only need to keep an element from going over a predetermined point. Here is an example on Amazon. There are other chips on other breakout boards which will do the same thing. I have noticed most newer stuff uses an I2C serial output which you do not want.

It was way back in post #2 and he suggested what I just linked to.

Ron

 

The images below are the reading I got without heating and reverse polarity

That low resistance value and the fact that the readings are not identical says to me that the sensor is indeed a thermocouple. So that does require a much higher gain amplifier. I have not used a high gain amplifier as a comparator, so it seems that you will probably need both an amplifier and a separate comparator, unless you can find an IC that has two sections that would work. Probably a dual opamp will be able to provide the gain in one section and the comparator function in the second section, but you may need an external transistor to drive the relay.
Then there is a bit of decision making has to be done, because it might be more reliable to have the protection system keep a relay energized to allow the heater to work, as one choice, or it might be better to have the protection system operate the relay to switch off the power when it starts to overheat. That scheme could also operate some alarm to warn about the overheating, so that the user could reduce the temperature setting. So there needs to be a bit of deciding done.
There are two more considerations, which is where will this additional protection system be installed, and what will power it? That needs to be decided before work starts. Once you have a circuit you can see how big it will be and how much power it will require.

 

Ok, nice to hear that. Anyway, I built non inverting comparator and I connect the inverting input + of lm358 to the negative of thermocouple and two 100k resistors in series as a voltage divider connected to Vcc of lm358. After that I added the non inverting input - of lm358 and the positive of thermocouple connected to the middle of the two join resistors in series. I then connect the positive
Of led to the output of lm358 and 100 ohm resistor connect to the negative of the led to join the gnd. After all the above mentioned, I connect the Vcc and the gnd of lm358 to 8vdc, I then connect the heating core to the mains to heat, when it start heating at the normal temperature range nothing happens so I waited for the heating core to overheat, and then I saw the metal case around the heating core turns red, within some few minutes the led start glowing, note at the normal ranges the led did not glow. But I read the output voltage the meter and it was not enough to switch 12v relay. Do I have to use darlinton pair to get high gain to switch the 12v relay or how should I do that?

 

First, you really do want a LM358 Operational Amplifier. Since you want a comparator I suggest using a comparator like the LM397 or similar comparator such as the LM111-N/LM211-N/LM311-N Voltage Comparator . Since you only need a single channel comparator. Next, you really do not want two resistors in series to get your Vref (Voltage Reference). I would suggest a small ten turn 10K potentiometer. This allows you to set a range of reference voltages. If you need a schematic I can likely post one. I suggest that you avoid using an operational amplifier as a comparator. Your pot wiper (Voltage Reference) and Voltage In can go to either the inverting or non-inverting inputs depending on what you want the output to do. Here is an example using an LM339 comparator.


Vin is shown as an increasing and decreasing voltage. Vref is the reference voltage set by the potentiometer, in this case about 6 volts. Vout goes low when Vin exceeds Vref. This is just an example of a comparator circuit.

Ron

 

First, you really do want an LM358 Operational Amplifier. Since you want a comparator I suggest using a comparator like the LM397 or similar comparator such as the LM111-N/LM211-N/LM311-N Voltage Comparator . Since you only need a single channel comparator. Next, you really do not want two resistors in series to get your Vref (Voltage Reference). I would suggest a small ten turn 10K potentiometer. This allows you to set a range of reference voltages. If you need a schematic I can likely post one. I suggest that you avoid using an operational amplifier as a comparator. Your pot wiper (Voltage Reference) And Voltage In can go to either the inverting or non-inverting inputs depending on what you want the output to do. Here is an example using an LM339 comparator.

View attachment 254821
Vin is shown as an increasing and decreasing voltage. Vref is the reference voltage set by the potentiometer, in this case about 6 volts. Vout goes low when Vin exceeds Vref. This is just an example of a comparator circuit.

Ron

If you wouldn't mind do you have any reasons why you suggest that I should avoid using an operational amplifier as a comparator?

Note: I'm still a learner.

 

Certainly you will need a transistor to drive the relay. An MPSA13 will work well, I have used them in a standard product.
AND you will need an adjustable setpoint voltage so that the setpoint will be much below the heater case glowing. Really, set it just above the hottest temperature that you use. And while an actual comparator is better for high performance operation, an opamp avoids needing a pull-up resistor.

 

If you wouldn't mind do you have any reasons why you suggest that I should avoid using an operational amplifier as a comparator?

Note: I'm still a learner.

Sure. Give this a read with a focus on:
"The difference between an op-amp comparator and a voltage comparator is in the output stage as a standard op-amp has an output stage that is optimized for linear operation, while the output stage of a voltage comparator is optimized for continuous saturated operation as it is always intended to be close to one supply rail or the other and not in between".

Something else to consider. Most comparator chips use an "open collector" output., this is why you see R1 in the example circuit I posted. Next as Bill points out in a real example the output of the comparator will drive a transistor to drive a relay. Since you just want switching to drive your relay I would consider a common switching NPN transistor like the 2N2222 which can easily drive a 0.5 amp relay coil. Once a relay is chosen the rest can come together. I also am not saying an operational amplifier used as a comparator won't work or is a poor choice, I am saying I see a comparator as a better choice.

Ron

 

If you wouldn't mind do you have any reasons why you suggest that I should avoid using an operational amplifier as a comparator?

Note: I'm still a learner.

I used a ready-made chip the AD595 for Ktype thermocouple gives out 10mV per degC.

https://www.analog.com/en/products/ad594.html#product-overview

 

W

Sure. Give this a read with a focus on:
"The difference between an op-amp comparator and a voltage comparator is in the output stage as a standard op-amp has an output stage that is optimized for linear operation, while the output stage of a voltage comparator is optimized for continuous saturated operation as it is always intended to be close to one supply rail or the other and not in between".

Something else to consider. Most comparator chips use an "open collector" output., this is why you see R1 in the example circuit I posted. Next as Bill points out in a real example the output of the comparator will drive a transistor to drive a relay. Since you just want to switch to drive your relay I would consider a common switching NPN transistor like the 2N2222 which can easily drive a 0.5 amp relay coil. Once a relay is chosen the rest can come together. I also am not saying an operational amplifier used as a comparator won't work or is a poor choice, I am saying I see a comparator as a better choice.

Ron

Well understood and very informative