3
\$\begingroup\$

I am building a smoke generator based on heating mineral oil with a resistance wire. The smoke will be used for automotive diagnostics. I've read on this site where calculating the temperature of the wire is complicated due to the large number of variables. The recommendation was to experiment. I don't want the oil to get so hot it bursts into flame but I want smoke. For this question assume the desired temperature is 350F. The voltage will be 12v (car battery).

My thought was to attached one end of a 3 foot piece of 0.5Ohm/ft. wire to one side of the battery and attach the other lead to various points along the 3 feet. At each point I could use a thermocouple to measure the wire temperature between the two battery leads. Would this work? I am concerned about the effect of the extra wire that is outside of the two battery leads.

The determined length, using this method, will be a good starting point as the straight wire configuration does not represent the final design. The final design will have the wire wrapped around a 2-3 inch length of Tiki Torch wick. The ends of the wick will be in the oil and the wick will pull the oil into contact with the wire.

My question is will my method work for determining the wire temperature. I could cut wires of various lengths but would like to avoid doing that if I can use a long wire and put current through just a portion of it. And is there a better way?

\$\endgroup\$
10
  • 1
    \$\begingroup\$ Is mineral oil really the best liquid to use? There's fog machine solution which I assume is safer. \$\endgroup\$
    – DKNguyen
    Commented Feb 28, 2023 at 20:14
  • 2
    \$\begingroup\$ In short, don't assume!! Verify. What I did was to go visit a university library and studied the topic carefully. Another approach, perhaps, is through experimental testing -- though that can be expensive, time-consuming and otherwise just duplicate what others have already done for you. But you cannot know -- there is no possible way to know -- about your safety situation until after you've worked out the difference in temperature between smoke and flash. If they are close, then you have a very different problem ahead. You can't just blindly guess about this. Be safe. \$\endgroup\$ Commented Feb 28, 2023 at 20:18
  • 3
    \$\begingroup\$ If you experiment to find the temperature of the wire with it straight out, wrapping it round a 2-3 inch length of Tiki Torch is guarranteed to change its temperature radically. Do the experiment with the final physical form that you are going to use. \$\endgroup\$
    – Neil_UK
    Commented Feb 28, 2023 at 20:21
  • 1
    \$\begingroup\$ Engineering this is easier if you work in SI \$\endgroup\$
    – D Duck
    Commented Feb 28, 2023 at 21:41
  • 1
    \$\begingroup\$ To be clear I'm not guessing or assuming. I used the word assume above simply to have the discussion. I'll use the actual value in the design. Fortunately this is not rocket science but only a smoke generator. Using mineral oil is used every day in this same application so while there is risk there is risk in everything. \$\endgroup\$
    – dbsoccer
    Commented Feb 28, 2023 at 22:37

3 Answers 3

6
\$\begingroup\$

Nichrome wire which is popular for heating applications has a rather flat temp co of resistance. This is not always desirable. If you get wire that has a good positive temp co then you are better off .When the oil sucks heat away the resistance falls driving more power from your 12 VDC tending to reduce the fall in temp .If you run the wire dry with no oil the temp will try to go up but so does the DCR which will reduce the temp rise. So this wire makes things more idiot proof .It is easy to measure the resistance of the wire by whatever means and infer wire temp from the manufacturers tables.

\$\endgroup\$
2
  • 3
    \$\begingroup\$ Excellent! That's actually where I was leading the questioner. Don't design systems to make something safe -- engineered systems fail. Use physics to make something safe. Nature doesn't fail. \$\endgroup\$ Commented Feb 28, 2023 at 22:59
  • 2
    \$\begingroup\$ In general this is called "PTC heating wire" \$\endgroup\$
    – jpa
    Commented Mar 1, 2023 at 14:41
4
\$\begingroup\$

The main problems with using a thermocouple are that the current going through the wire will affect reading the tiny voltage produced by the thermocouple (insulating it will be difficult) and the thermocouple wires themselves will greatly affect the temperature so that your reading, even if accurate(say you measure immediately after removing current), is not representative of the wire temperature sans thermocouple.

The oil will also draw heat from the wire (or possibly add to it under some conditions), so testing without it is not representative.

I suggest testing with a variable supply. You can glomp onto a piece of wire with screw terminals, for example.

In any case if you need a controlled temperature a piece of wire may not cut it. Perhaps you need a heater and a sensor with a closed-loop controller. In which case you might pick a small reservoir that is easy to both heat and sense the temperature of. Perhaps combined with a metering pump. This would be similar to commercial glycol based fog machines which produce a relatively safe “fog” for visual effects.

Note: you could use the wire itself as a sensor but that is non-trivial and I’m not sure the results will prove to be useful.

\$\endgroup\$
3
\$\begingroup\$

I would not cut wires of various lengths. If you are using a fixed voltage, I would start with one you know is too long and thus too cool, and and just keep cutting it shorter until it gets hot enough. The alternative is to use a fixed length with a current source and turn it up until it gets hot enough and then record that current. Then, knowing wire resistance per length you can calculate the voltage required to achieve that current for any given length.

Actually, instead of cutting it shorter each time you could just move the electrical contact down the wire so current is flowing through a shorter length. Saves wire.

I think it would be tricky to make good contact between the end of a thermocouple and the wire and I don't think it's necessary. Immerse it in the oil bath when doing this since the temperature will not be the same as the air (one of those complications in the calculation method). Then slowly turn it up until the oil smokes. That's the current you need.

You can do guesstimates for the initial length:

\$Watts = \frac{V^2}{Lr}\$ where \$L\$ is the length and \$r\$ is the resistance per unit length

And you should also calculate the current to make sure your battery can actually supply that current and do so for the length of time you need.

If you use enough heating wire its resistance will swamp out the resistance of the normal copper wire and you can neglect it.

\$\endgroup\$
10
  • \$\begingroup\$ I don't have a value for the required watts to make the oil smoke. What I am trying to avoid doing is making several prototypes of the final design with various lengths of wire. The power source, hence the voltage, is fixed. While an automobile's 12 volts it is what it is. Again, this is not rocket science. The difference between smoke point and ignition point is a known concern which is one of primary reasons for my asking the question on an engineering forum. Many people have built similar devices by simply using random lengths of a resistance wire of unknown properties. \$\endgroup\$
    – dbsoccer
    Commented Feb 28, 2023 at 22:37
  • \$\begingroup\$ My attempt is to be a bit more precise. Based on the input I've received, I feel my attempt to ballpark the wire length will still include too many unknowns. The best way is to create prototypes of various resistance values until I narrow in on one that both works (produces smoke) and gives me a comfortable safety margin. Thanks. \$\endgroup\$
    – dbsoccer
    Commented Feb 28, 2023 at 22:37
  • 1
    \$\begingroup\$ @dbsoccer Keep in mind that the automotive "fixed" 12 volts is not really 12V. Could easily be a steady 14.5V with the engine running. This could mean the difference between smoke and fire... \$\endgroup\$
    – Mr47
    Commented Mar 1, 2023 at 8:30
  • 1
    \$\begingroup\$ @dbsoccer I assume you are going to be bracing or winding the heating wire in some way and in that case you can make one of the electrodes moveable along the length and that would let you experiment as well as maintain adjustability. \$\endgroup\$
    – DKNguyen
    Commented Mar 1, 2023 at 22:36
  • 1
    \$\begingroup\$ @dbsoccer Yes, the same kind of idea that large wirewound variable resistors use where you unscrew a conductive clamp, move it to a new spot, and rescrew the clamp. \$\endgroup\$
    – DKNguyen
    Commented Mar 2, 2023 at 0:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.