Heat output element

Question

I'm building a little quiz game in Arduino and I already have it displaying a multiple choice question and people can answer by pressing pushbuttons.

But anyways now I wanted to know if it's possible for there to be an element (maybe some sort of resistor or diode?) that heats up depending on whether you get the question right. So you would answer the question with one finger and put your another finger on the element to tell if you got it right. Obviously I don't want this heating to the point of actually burning someone, just enough so that they can tell if they got it right.

Also it would be cool if this same element would also be able to cool down (below it's initial temperature) if the answer is wrong. And obviously the heating up (or cooling down) has to be relatively quick, you can't be waiting a couple minutes. And as an added challenge I don't want this to be sucking up all my batteries :)

Any ideas about what this element / circuit can be?

Answer 1

A resistor will heat when power is applied.
A physically small resistor can be heated in under a second with a finger applied.

You could get the sensation of cooling by having a cold source that is thermally attached to the rear of the resistor.
Turn off heat and you get cold.
Balance cold and you get "neutral".
Exceed cold input to get hot.

Heating only to a small (physically and power rating) resistor you can use under a Watt to heat. You could use two resistors spatially separated to give Yes/No indications.

A Peltier TEC ("Thermoelectric cooler") will heat or cool. A small one will usually still need in the "several Watts" range to cool.

A vibrator motor as used in cellphones and pagers to give tactile / haptic feedback or alarms may be easier to use. Vibration level and vibration insulation will need experimenting to get non audible result.

---

How large a resistor?

One approach to resistor sizing is to suck it and see - but you may burn your lips :-)

Smaller resistors tend to have lower power ratings.
They get hotter at lower power inputs BUT as the total heat energy is lower they are less liable to be able to be easily and quickly sensed.

I'd guesstimate that 1/8th of a Watt would be ample and it may be that significantly less is acceptable.

Here is a Digikey page for a Panasonic 1/8th Watt 10k resistor - by itself its not muh use, but after you've seen the following material you can use this as a starting page for searching for related resistors.

From that page you can find the specification page for Panasonic ERJ6ENF family resistors and on the bottom right of that page is a treasure trove of links - many useful.

One is this spec sheet for a family of SMD resistors](http://industrial.panasonic.com/www-data/pdf/AOA0000/AOA0000CE2.pdf)

An extract from page 6 of the above data sheet is shown below.

Larger version of table below is here
As can be seen, power ratings by size are shown in the red bordered table at left.
Power ratings of from 0.03 Watt to 1 Watt are available in this range.

If you want to try an 0.05 Watt (50 milliwatt) resistor its type code is 1GN and it is an "0201" resistor. That's small!In the dimensions table at right you can see that an 0201 resistor is about 0.6mm x 0.3mm (!!!).

An 1/8th Watt or nearly so is available in 0402 0603 & 0805. The table tells you that an 0805 is about 2mm x 1.25mm =~ 0.080" x 0.050" - you can solder one of those between stripboard tracks or see it well enough to solder wires to it (with care).

The table below from page 7 of the datasheet tells you about temperature and power rating. While exact temperatures are inexact and dependant on circumstance, the table says that eg a 6E resistor at 70 C can tolerate full power dissipation and at 150C it can tolerate none. That implies that it will BE at 155C in 70C ambient at full power - or Trise = 155C-70C = 85C rise. ie the 1/8th Watt resistor will have a body temperature rise of about 85C above ambient at 1/8th Watt. Very roughly you get about 0.85 C rise above ambient per percent of power rating. That's probably horizontal in free air mounted on an FR4 PCB with xxx tracks and YYY copper and .... . ie this is approximate.
Max water temperature that an average person can stand to you can keep their hands in is 65C. At around 25C ambient (very warm room) you'd get to 65C at 40/.85 = 47% of power rating or 1/8W x ~= 50% = 1/16W or about 60 mW.
The best you can say is that the real value will be different but that this gives a guide.

Power in a resistor = V^2/R, so for a given R power required and thus voltage required can be calculated to get a given &age of rated power. If you have a 3V3 supply and want 3V3 applied to give you 100% power in a 1/8th Watt resistor the R = V^2/Power = 3.3V^2/0.125 = 87 Ohms.
Use eg 82 Ohms.

If you can get through hole 1/8 W resistors they may be easier to work with.

Nichrome wire (toaster element or electrical supplier) is an option.

Answer 2

The fundamental problem you have is that to make a detectable level of heat you require a similar amount of power, heat conduction acts slowly and a cooling response takes even more power.

However, there is a solution that will allow you to "gain" up the response and as such automatically reduce the response time and the reduce the power consumption.

You need to evoke a known physiological/sensory response. When finger tips feel alternating strips of hot and cold that are close together it confuses the sensory system and is perceived as being a lot hotter than it actually is. With only a few degrees difference you can convince people that they are being burned, you need a toned down version of this.

I initially thought that it would be great to have a Peltier cooler and have an array of hot and cold metal stripes (with insulators between) that the cooler would pump heat from one set of strips to another. But I suspect that if you have alternating heating and ambient stripes (with a large heat sink) and if made small enough then the effect becomes very pronounced. This means that all the benefits I mentioned above become apparent.

I am now speculating, but perhaps a field of 0402 resistors with areas of ground plane behind such that the finger tip contacts the resistors and the background at the same time would evoke this response. But you will need to experiment. When I experienced this it was a sandwich like structure with small stripes.