Hi, Christian!!! Have you built ur heater? I made thermostat for gas analyzer long ago and have some experience whit this. what is the problem you solve, may be I could help))) p.s. Sorry for my english

Hi Dmitriy,

I do have the heather running (Nichrome wire connected directly to a power source). Right now the only way to increase or decrease the heat is by manually increasing or decreasing the voltage and current directly on the power source. I also have a thermistor sensing my temperature every few seconds.

However, I would like to have a much better automatic control on the temperature using a microcontroller. I would like to be able to code a routine to increase or decrease the heat generated by the Nichrome wire. In order to accomplish this I am thinking in placing a Mosfet in between the power source and the Nichrome wire and control it using PWM from the microcontroller. Does this makes sense?

I can already measure temperature using the thermistor and microcontroller, so the only other thing to accomplish is the control of the increase or decrease of temperature.

Thanks for your help and do not worry for the English, mine is also not perfect ;)

Best,

Christian

Hi Christian

it seems that the PMW mode is the best way to regulate temperature. I used uC PIC16F877 in my design (it was long ago) for converting analog signal to digital form and for sending it to PC port. PC's software (delphi) was intended to calculate time of duty cycle which then was sent back to the uC.

But almost every modern uC is programmed using C language with elementary math support so you don't need a PC

Sorry, it is too late. Tomorrow morning I'll attach the pictures with the formulas and schemes.

Night

Thank you very much Dmitriy.

Regards,

Christian

Hi Christian

in the attached file I tried to depict the implementation of the PWM regulator on the microcontroller (uC)

There is a formula in the green frame to calculate the duty cycle. The smaller delta t, the higher temperature stability!!!

if you're interested in the details, please share what type of thermal sensor and microcontroller you use. What type of MOSFET is available?

See u later

Hi Dmitriy,

Thank you very much for your detailed implementation response. It suits very well to what I had in mind. I specially like the formula for duty cycle, I think I will give it a try and start coding it in my uC.

I am using a thermistor of the type 204GT (high sensitive thermistor); a basic stamp 2p2 as uC which I am already using to sense resistance change from the thermistor; I also have some spare MOSFET like FQP45N03L but I can order almost any other online. I will start working on the implementation and will upload results as they come.

Best regards,

Christian

Best wishes and success

Hello Dmitriy,

The PIC16F877 does have a PWM channel output. Reading ADC values (for example based on the Thermistor output) and using it to calculate the required PWM count would be another option to make the system autonomous - no need for a PC or software for just to calculate the PWM duty cycle.

Just a suggestion!

Hope it will be useful! Best luck!

Hi Prasanna

Thank you for suggestion. Using embedded PWM is a realy good way.

In my case there were some limitations. Even slowest PWM mode was too fast and noisy switching low-impedance load. PC synchronised heater control and ADC requests via that PIC16.

Hello Dmitriy and Christian,

Thanks for wonderful discussions!

Best Luck!

Hi Christian Mendoza-Buenrostro,

I also want to make such device. What is the voltage/Current required to heat the Ni-Cr wire to maintain 300 degrees C. And what is the length and diameter of the Ni-Cr wire?

Hi Anand Thati

Power depends on temperature and heat capacity of the object. Do you have a soldering iron? copper piece temperature is maintained at 300 degrees, and the input power is 20 - 100 W. NiCr electrical resistivity is about 1,2 ohm-mm ² / m, so wire with 1m lenght and D=0.2mm has electrical resistance of 40 Ohm.

P = U^2/R or P = I^2 * R

Good Luck!

Thanks #Dmitriy,

is it possible with 2.5W input power?

of course!!! Try to solder smd resistor 56 ohm (size 2512, 1W) to power source of 12 V. This resistor will dissipate 2.5 watts. If tin is melted, you have goals! Esle use smaller one. Resistor's heat insulation will accelerate its self-destruction)))

Take care of your eyes

Hello Dmitriy,

Would that be a good solution to use those 2512 cakes to be used as heaters, particularly for a prolonged use?

of couse NOT! It was just an example of experiment to estimate a mass of object which can achive the needed temperature (300 degrees @ 2.5W). I beg your pardon for my english(((

Thanks Dmitriy,

But i want to make it with Ni-Cr wire.

Dear colleagues

I've just tried to estimate a size or mass of warming object (somethg about 7mm*3mm*1mm). Also temperature depends on heat exchange with environment.

Anand Thati wrote:

is it possible with 2.5W input power?

Try this:

NiCr wire d=0.2mm, l=1500mm (R ~= 57 Ohm)

Use power source 12V (>300mA)

P = U^2 / R = 144 / 57 = 2,5 W

wire tightly put on temperature sensor

to connect properly use refractory alloys or screw connections

Hello Anand,

Sorry for the late replay. Dmitriy has given a lot of useful tips for your question, try following them. As for my case, I am using a Ni-chrome wire embedded in a commercial extruder tip from the Rap Man project. I do not remember the particular length and diameter for the Ni-chrome wire being used but you could download the Internet manuals to get that info, look for extruder manual rapman on Internet. The Ni-chrome wire is embedded in a fire cement at the tip of the extruder, the temperature sensor (thermistor) is also inside the fire cement in very close proximity to the Ni-chrome wire. I am not sure if you could obtain 300 C using this approach but you can give it a try and let us know.

In my setup, I am using the commercial extruder I mentioned plus a power supply set to 12 V. I have experimentally found that a current of around 1.3 A is just enough to heat the tip containing the Ni-chrome wire up to around 190C. Again, this will depend on the length and diameter of your Ni-chrome wire as Dmitriy has suggested you. I also recommend to not increase your voltage more than 12 V, there is an upper limit for the Ni-chrome wire (if you go pass it the Ni-chrome wire will simply burn at the middle). Maybe you can set your power supply to 12 V and adjust the length of your Ni-chrome wire to get the temperature you need.

I also recommend that you try to keep the heat as much as possible by isolating the Ni-chrome wire from the environment (or else the heat would be rapidly dissipated to the ambient). In my case I created some sort of insulating cover to the fire cement containing the Ni-chrome wire by using heat resistant tape. This will make it more efficient in terms of dissipated power.

I hope we have provided some help for your questions.

Christian

Hello Prasanna Waichal,

I have tried to use resistors as heat elements before as in the theoretical example Dmitriy mentioned but it is not a good idea as he also pointed out. The resistors will heat really fast and of course everything will be hotter around them for some period of time but it is not the most efficient way to generate heat. I remember one of the problems in my particular application was that the solder used was melting away so the resistor will just disconnect and move away from the position they were at.

Thank you christian, you have given good information.

Just what I had in mind!!! Thanks a lot!

Share thoughts in English is not my strong point :(

Hello Christian,

Yes indeed, the resistors get fried up and also loosens up the solder joint. And more important is that the wastage of heat (dissipation) on the PCB.

I am just giving a suggestion - I remember there was one article published long time back in Electronics For You, New Delhi. It used our writing pencil (or rather the graphite lead there) as a heating element to burn a solder wire. Although it is fragile,

I would suggest go for some industrial grade graphite slab or cake, simply pass current through it and see it gets heated up. Another advantage you can get is that since this is a soft material, you can EMBED your temperature sensor such as a thermistor or more precisely the RTD element into the HEATER body itself.

So let's try this one also. A little engineering will be needed to fit the graphite to the perfect shape and then to fix the sensor BUT IT WOULD BE WORTH TRYING!

Best Luck!

For your comments colleagues!

Thanks,

-Prasanna

I made such a controller years ago for heating a Nichrome wire to red hot temperature. It was then used to pull out a glass tube and make it into a micro tube with small diameter. Nichrome  has a well known resistance to temperature curve so if your circuit applies a heating current and measures the voltage across the wire you can control the temperature with FB. I actually used a SCR in a bridge so it was fairly complicated but a DC controller would be a lot easier. I had inherited a non-working SCR in a bridge design and was stuck with this.

Maybe you should use an ON OFF controller with PID.

It is extremely simple and works perfectly.

Arduino mega inside 3d printer Prusa I3 uses it too.