Controlling temperature of heater wire.

[asimpson]

I wish to run a Nichrome heater wire nearest to its maximum temperature and with a high watt density. I am concerned that when heated body which is a steel block attains maximum temperature traditional temperature sensing may be too slow.

I am currently using thermocouple with on/off control. I know a PID would be better.

Is there a control method to directly sence heater temperature fast enough to prevent heater burn-out at 900C ?

 

[MacGyverS2000]

Wasn't this same question posted just a day or two ago?

You have to define what the ragged edge actually is. Once you have that, then you need a sensor that can give you an accurate value. Then, you need to act on that sensor value quickly. And so on down the line. When it comes down to it, a PID (or similar) is your best bet... but if you're running at the ragged edge, you'll need to do some real fine tuning to make sure you don't go beyond that edge.

Why the need to run right at the edge?

Dan - Owner

http://www.Hi-TecDesigns.com

 

[benta]

At that temperature I'd expect the wire to be glowing.
Perhaps a colour sensor?

Benta.

 

[asimpson]

Yes I did post before however the thread vanished.

Cant use optic sensor. Device will be enclosed.

 

[IRstuff]

You were already answered in your duplicate posting in the Heat Transfer forum; you can measure the resistance.

http://wiretron.com/nicrdat.html

Nonetheless, speed of temperature sensors should not be an issue:

http://www.omega.com/prodinfo/TC_Graphic1.HTML

TTFN
faq731-376

 

[asimpson]

I understand I can use resistance measurement however I have not found a comercially available controller that can do this and I do not have the electronics skill to design my own controller from scratch. Perhaps I should have mentioned that.

On the temperature probe issue. The speed of the probe is not in question. I am concerned that given the temperature of surrounding material (+900C) if heater is given full power the probe will have to be incontact with heater wire to regiser its temperature rise in time and this would be fairly difficult in a compact system.

In a cartridge heater the hot junction is usually at the tip. Heat might have to travel from heater wire through ceramic and MgO powder with only a temperature gradient of only 50C. I would like to reduce this responce time and get finer control.

Thanks

 

[MacGyverS2000]

If you cannot make your own controller, then why are you asking for info on how to control it? It seems that no matter what method we offer (and measuring resistance is about the most simple method I can think of), you'll still be in the same boat of not being able to use that info.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[IRstuff]

" Heat might have to travel from heater wire through ceramic and MgO powder with only a temperature gradient of only 50C."

So, one can easily imagine using a 50°C lower setpoint until the temperature stabilizes, and then slowly ramp the temperature up. Since all the behavior is deterministic, it's a matter of characterizing the "plant" that the PID is driving and including the response profile into the PID programming. This is something that needs to be done, regardless, although other systems might not need such complex startup. Given that you know you have lag, your PID has to be tuned accordingly.

I should point out that 50°C is probably as close as I would want to get to the spec melting point anyway, since the melting point is dependent on the exact composition of the nichrome, and nothing is ever made identically the same.

TTFN
faq731-376

 

[itsmoked]

I agree with IR. Do it like everyone else does. They just limit the on time during warm up. Look at electric kilns. They cycle the acres of Nichrome on and off until the kiln gets quite hot then they transition to temp control. Continue to use a T/C with PID once it's near setpoint.

Keith Cress
kcress -

http://www.flaminsystems.com