Heating water quickly and fixed temp

[AyaanS]

Hi

i am 14 years old and need help on a science project for which I need to heat water immediately in a test tube at a temperature as close to 50 degrees centigrade as possible but not more than that and keep the water heated at this temperature only.

please advice

thanks and regards

 

 

 

[Sensei]

25 minutes ago, AyaanS said:

i am 14 years old

Two thousands years ago, you would have wife and awaiting children at such age..

25 minutes ago, AyaanS said:

i am 14 years old and need help on a science project for which I need to heat water immediately in a test tube at a temperature as close to 50 degrees centigrade as possible but not more than that and keep the water heated at this temperature only.

Surround test tube by aluminum foil, place precise thermometer inside. The easiest would be to use Arduino, with temperature sensor, with custom made program which will read temperature and keep steady temperature..

 

Edited June 27, 2018 by Sensei

[Bender]

You also need some heating resistor to put inside (if it is water resistant) or wrap around the tube. Arduino can control relay to switch heater on and off. Usually you use a dead band: switch on at eg 49.5 and switch off at 50.5. (That is called a hysteresis controller or a bang-bang controller.)

[John Cuthber]

On 6/27/2018 at 1:31 AM, Sensei said:

Two thousands years ago, you would have wife and awaiting children at such age..

Surround test tube by aluminum foil, place precise thermometer inside. The easiest would be to use Arduino, with temperature sensor, with custom made program which will read temperature and keep steady temperature..

 

Without a heater that's not going to get very far (and I'm not sure what you think the history lesson achieves).

 

53 minutes ago, Bender said:

You also need some heating resistor to put inside (if it is water resistant) or wrap around the tube. Arduino can control relay to switch heater on and off. Usually you use a dead band: switch on at eg 49.5 and switch off at 50.5. (That is called a hysteresis controller or a bang-bang controller.)

A bang bang temperature controller is (usually) easy to implement, but a proportional one will give  much better control.

If you know how much water is in the tube then an interesting variation would be to measure the temperature of the water, calculate the energy needed to increase the temperature , charge a suitable capacitor the right voltage and then discharge it directly through the water itself.
That gives you the possibility of heating  the water in microseconds, at the expense of being much less practical.

 

[Bender]

1 hour ago, John Cuthber said:

A bang bang temperature controller is (usually) easy to implement, but a proportional one will give  much better control.

Proportional control requires implementing PWM, and I doubt the possible gain would matter for a school project. The noise can also be annoying.

[John Cuthber]

12 minutes ago, Bender said:

Proportional control requires implementing PWM

PID controllers have been widely used for a long time; rather longer than PWM has been common.

If you use an arduino, doesn't it have PWM built in?

[Bender]

1 minute ago, John Cuthber said:

PID controllers have been widely used for a long time; rather longer than PWM has been common.

If you use an arduino, doesn't it have PWM built in?

PID is horrible to control temperature which usually has a large time constant and no means to actively cool after an overshoot. Just P will work but requires a voltage regulator. An Arduino does that with PWM, which is supported on only some of the outputs. It is not that difficult to implement, but for what is likely to be a first-time user, I'd stick to bang-bang for quick success. 

In this case I'd use bang-bang myself, and only look for alternatives if necessary.

[CharonY]

On 6/26/2018 at 6:31 PM, Sensei said:

Two thousands years ago, you would have wife and awaiting children at such age..

This is totally off-topic, but I was at a talk a while back that discussed child bearing in different societies and through time. Interestingly, menarche (onset of menstruation) is highly dependent on nutritional and other life factors. Thus, in modern society menarche (and thus the possibility of pregnancy) is earlier in modern times than before the industrial revolution (though at the beginning there could have been a dip as early urbanization was associated with higher morbidity). In Roman society it is estimated that menarche was expected around age 14 with pregnancy typically 2-4 years later. 

[John Cuthber]

20 hours ago, Bender said:

PID is horrible to control temperature which usually has a large time constant

It is precisely because of the large time constant that people use PID controllers.
I'd use bang bang control if I was just trying to keep my home-brew warm.

But if I was trying to "Win" a school project, I'd get a PID

I look forward to Sensei's explanation of how one might heat the water without a heater.

[Bender]

25 minutes ago, John Cuthber said:

It is precisely because of the large time constant that people use PID controllers.

The large time constant will cause the D-action to react almost exclusively to sensor and digitisation noise.

Any significant I-action is likely to cause an overshoot, which is exactly what he doesn't want, especially since he can't actively cool it down again.

So a proportional control: sure. It will be slower, noisy and take more effort, but could be slightly more accurate.

[John Cuthber]

OK, why did anyone ever go to the trouble of inventing the PID controller?

The whole point is that you can set the P, I and D terms independently and avoid the issues you mention.

(Still a  less interesting question that how does Sensei's "heater free heater" work? but I guess that's a lost cause).

 

 

[Bender]

For linear applications where you can control in both directions, the sensor are sufficiently reliable (in case of the D-action) and the potentially added accuracy is actually worth it.

Position and implementation of sensor and heating element will have a bigger influence here than the choice between controllers.

To reverse the question : if PID controllers are so great at controlling temperature, why don't room thermostats use them?

[John Cuthber]

All the deceit furnaces I have seen have PID controllers- but (obviously) no active cooling.

"if PID controllers are so great at controlling temperature, why don't room thermostats use them?"

Because bang bang is good enough and a lot cheaper. It has, of course, also been going since long before PID controllers were invented.

As an amusing aside, you might even get away with feed-forward control in this case.

If you know the temperature from which the water is starting and you know how much of it you have, you can calculate the energy needed.

And you can deliver that very quickly indeed.

it fails on the 

On 6/27/2018 at 1:11 AM, AyaanS said:

keep the water heated at this temperature

requirement.

 

[Bender]

1 hour ago, John Cuthber said:

Because bang bang is good enough and a lot cheaper.

Exactly.

Although I must admit that if I had to judge a science project, I would be impressed with a 14 year old successfully implementing the PID, even if using one didn't really make sense.

1 hour ago, John Cuthber said:

All the deceit furnaces I have seen have PID controllers- but (obviously) no active cooling.

How are they heated and where are the sensors positioned?  How important are the I a D actions?

I have seen a PID in an industrial machine as well, but only for the primary control with hot or cold water. The water itself was still heated with bang-bang.

(What is a "deceit furnace"? Google didn't provide a useful translation.)

Edit: another application of PID in temperature control would be for controlling the temperature of a continuous flow.

Edited July 1, 2018 by Bender

[HB of CJ]

A test tube may not work that well.  The problem will be control while introducing lots of heat to raise the temp quickly.  You might end up with a locally non controlled heat increase which could result in a small steam explosion.  Apron.  Hull body protection.  Full head and eye protection.

It almost sounds like you are trying to make a flash steam boiler.  Be careful.  Have fun.  Document everything.

[John Cuthber]

13 hours ago, Bender said:

What is a "deceit furnace"? Google didn't provide a useful translation.)

Doh!
All the deceit decent furnaces I have seen have PID controllers.
The cheap ones use bang bang, but if you want good temperature control- and control of rate of change etc- you need PID

13 hours ago, Bender said:

How important are the I a D actions?

Very- for exactly the reasons that
(1) you gave earlier and

(2) people use PID controllers..

 

13 hours ago, Bender said:

another application of PID in temperature control would be for controlling the temperature of a continuous flow.

In that case I'd definitely try to include some sort of feed-forward if I wanted good control (OK- at the expense of a more complex system but that's always an issue)

If I have a flow of x litres per minute and I want it at 50 C, but the incoming temperature is y then I can calculate the power requirement to a good precision. I'd use that to supply the bulk of the temperature rise, and use a PID to fine tune it.

 

Edited July 2, 2018 by John Cuthber