Temperature Control A Different Way?

[TheJester]

So here's two crack-pot ideas I just had. I haven't crunched the numbers or considered how to do a full implementation of it either. Now as we all know the changes in resistance of nichrome or kanthal wire are very small even at high temperatures.

So run with me, come up with a way to refute these, or come up with other ones?

Both operate under the assumption that we know the resistance of the coil at rest. Which is easy to obtain and almost all regulated devices are doing this already.

Here's the real crack pot idea...
1) Say we put a hall effect sensor underneath (or as a ring around) the atomizer to measure the magnetic field of the coil. Keep in mind, these coils we're using are basically solenoids. The magnetic flux would be proportional (probably not linearly hahaha) to the temperature of the coil and could be used to cross of the variables relating to how much power is being sent to the coil verses how much is flying through (IE: creating a magnetic field). Knowing the power being supplied to the coil would help a good deal as well.

I used a little app on my cellphone which serves as a crude magnetometer, and when I fire my atomizer at 15 watts I get a shift of about 4 microtesla being about half a cm from the atomizer. I imagine the field change could be as drastic as 10microtesla and read to two decimals of accuracy under better conditions.

2) Say the user selects what gauge wire to use from a menu for Kanthal (standard material right?). The chip can quickly calculate approximately how many wraps there are at rest, or even ask if it's unsure "2.5mm 7 wraps?" (IE Steam calculator). Why couldn't a mod interpolate the temperature of the coil from the power applied over time? Sure it wouldn't be perfect and values would have to be stored under certain conditions. Imperfect coils with super long leads could create hot-spots as well. Even if it didn't give the supposed +/-1 degree accuracy which nickel wire can, isn't +/-5 degrees "good enough" to prevent whick from charring when dry?

I'm not a fan of drilling through atomizers to put IR thermometers in place or whatever, seems finicky.

 

[rhelton]

Here is the data sheet from the manufacture, I see the possibility of what your brainstorming. My only concern is it would end up being something like the Provari and a Zatty, not very good when split up if that makes sense. But hey this is how things happen, everything starts with an idea and Its a good one.

Kanthal A-1

 

[TheJester]

Thanks for the data .

So I re-ran the magnetic field experiment last night on a temp control vapor shark and my iStick 20w at the same wattage. Yes I know the iStick is overpowered/tuned to voltage mean but whatever. It can easily be seen that there is a correlation between applied power, and magnetic field from the coil. Same atomizer 1.4ohm kanthal, and 0.15ohm nickel build. I did 3 "short" pulses, then 2-3 longer ones after to make sure it was just random peaks magically appearing on the magnetometer hehehe. The nickel wire gives a much bigger jump (less resistance) but that was expected.

Measurements were taken with a cell-phone (not ideal) and on the outside of a closed Marquis RDA atomizer, with juice and cotton, and everything.

Notice how the vapor shark LEAPS it's magnetic field, then really tones it down to a plateua to maintain temperature while the iStick just makes a plateua and holds it. This was the kind of graph we were expecting to see if there was a detectable magnetic field coming from the coil.

Now I'm wondering if a lower resistance coil (say 0.40-0.70ohm) of kanthal would be more preferable to get a larger field. I think the resistance/heat eats up a lot of the energy.

Here's the graphs (time on the X-axis and microtesla on the Y axis):



The gap for the vapor shark is about 58-78 microT so a change of 20 microT. The iStick is a lot lower... More like 4.8-7.8 so a change of 3 microT.

I'm kind of tempted to explore this a little further mathematically.

 

[Zanderist]

I think one issue with Kanthal is that it doesn't give up its heat right away, I suspect the nichrome wire needed for a snack doesn't have this problem.

You would have factor in a lag factor which change with material and guage.

You have time the coil till it was glowing hot, which would be different for each material and gauge.

Mayne to add into your idea, interpolate based off surface temperature of the atomizer, it resting a temp sensor on the top cap of an rda.

This just is a hypothesis.

 

[Flavored]

1. Isn't magnetic flux a function of current? Amp-turns? That wouldn't be a measure of temperature.
2. Interpolation of temperature in this manner wouldn't work because the device would know nothing about airflow or juice flow/wicking (both of which cool the coil).

 

[TheJester]

Flavored - you are correct in that it is a function of current. Current is also a measure of resistance at a given potential which I know that you know. The interesting thing about this approach is it doesn't say how much energy is lost to resistance, but actually the opposite. It says how much energy is being sent through the coil and not lost to resistance/heat. So if one knows that value, and the power being sent too the coil, one can figure out how much heat is being dumped and tune accordingly.

DNA40's temperature control doesn't know anything about airflow or juice directly either. However, if the coil cools from say juice or introduction of air, the resistance changes (the magnetic field changes as well). Do you see what I am getting at?

Zanderist - You're right, going purely theoretically requires a lot of known values about lag-time and things of that nature. That's why I mentioned it last. That thermometer/thermocouple idea is a good idea as well. I just wish there was a way to do this without modifying any RDAs, or sticking wires into a dripper. Drippers like the hobo don't have too much room between coils and the cieling .

 

[Zanderist]

The issue is with the design in relation to modified rda. People hold on the idea that it will be a primitive twist connection at 510 thread forever.

I think a 4th generation device will have a quick connect connection with built in thermistor that will be coated in an interview coating.

 

[Flavored]

Flavored - you are correct in that it is a function of current. Current is also a measure of resistance at a given potential which I know that you know. The interesting thing about this approach is it doesn't say how much energy is lost to resistance, but actually the opposite. It says how much energy is being sent through the coil and not lost to resistance/heat. So if one knows that value, and the power being sent too the coil, one can figure out how much heat is being dumped and tune accordingly.

DNA40's temperature control doesn't know anything about airflow or juice directly either. However, if the coil cools from say juice or introduction of air, the resistance changes (the magnetic field changes as well). Do you see what I am getting at?

Zanderist - You're right, going purely theoretically requires a lot of known values about lag-time and things of that nature. That's why I mentioned it last. That thermometer/thermocouple idea is a good idea as well. I just wish there was a way to do this without modifying any RDAs, or sticking wires into a dripper. Drippers like the hobo don't have too much room between coils and the cieling .

In a heating coil, you don't lose energy to resistance, you get heat. That is the objective. If you wish to measure current, there are less complicated and more accurate ways.

What the DNA 40 does know is the temperature of the coil, directly, presuming you've used the specified coil material.

 

[TheJester]

Flavored - Losing energy to resistance is HEAT. Aka Joule Heating. Joule heating - Wikipedia, the free encyclopedia , if you've never heard of the effect read the first paragraph and the section on "Power loss and Noise". On spec sheets, when people talk about power dissipation on a resistor, they are talking about heat. IE, why people put heat-sinks on components such as the PCU on you're computer. Basic rules of thermodynamics, (thermal) energy can't come from nowere, it's coming from the battery losses on the coil.

Well then what is the problem, if we have an accurate way to measure voltage and current simultaneously across the wire why can't we get temperature control out of nichrome? I have to admit, I haven't looked up hall effect sensors or crunched the numbers yet so I can't speak of it's efficacy, but I still think it's possible to go this route. Thinking about it though, how do people measure current typically? Isn't it hall effect sensors or inductors? I mean when I google Current Sensor that is all I see available.

The DNA40 doesn't know the temperature. It measures the resistance (probably measures that indirectly too through voltage and current). The resistance is proportional to temperature with-in a certain range for Nickel. It probably is for nichrome too, but they lack the accuracy to measure it, I can't say for sure, but I think that is the problem. Anyways, the temperature is interpolated, not measured directly. I realize that might sound like I'm saying the same thing, but it's really not the same. I'm trying to come up with a way to do something similar for a coil with very low resistive changes when heated.

I just figured I'd rather deal with a gradient of 10-20 micro tesla with two (maybe 3-4 depending on the component) additional decimals rather then be stuck with the 0.00V. The beauty of the idea is that magnetic fields permeate pretty much all common materials with low losses unless there is induction, so it is non-invasive to any RDA. 5-6 decimals of a magnetic field measurement sounds great to me, I wonder how many digits the DNA40 is using for their resistance sensing?

Actually I'm not completely sure how the DNA40 measures the resistance of the atomizer? voltage divider and power loss at the DC-DC regulator? Current sensor?

Keep in mind those measurements I took were with a onboard magnetic field sensor on a smart-phone... A lot more distance (lower magnitude field measurement), and the IC used on smart-phones probably isn't a super nice one. Phones aren't exactly designed for that sort of application...

 

[Flavored]

You're looking at it backwards . . . heat losses in the type circuits you are speaking of are a bad thing. They don't wish to lost energy to heat. In our coils, we wish to "lose" it all to heat. Plus, since we use DC, our coils are resistors, and resistors heat up with current flow.

And, yes, you are "saying the same thing." If you know the temperature coefficient of a conductor and measure it's resistance, comparing it to a baseline value, you do, indeed, know its temperature.

 

[Dampmaskin]

Temperature and heat are two different concepts. When you fire a mod at 20 W, you already know the heat - it is 20 W. If I understand you right, the temperature is what is interesting here. And the temperature can not be measured via, or inferred from, the magnetic field from the coil. It may be possible to measure the heat that way, but assuming that we're using a reasonably accurate VW mod, we already know the heat.

 

[TheJester]

Flavored - I don't see how I am looking at it backwards... What I mean to say is a measured value is not the same as an interpolated value, that's all. For example, say

I take a coil and do a resistance measurement on it at 15 watts, cool the temperature is with-in 1% of it's suspected value. Horray good engineering was done for us.

Now let's say I am building a device(the intent of this thread), and I have an error measuring voltage say +/- 50mv error, and current with absolutely none. Say those 15 watts are 3v (+/-0.05), and 5 perfect amperes (what a miracle!). What is my power? Well, 15w (+/-0.25). Still a small error, but I don't want to go through an entire example including how resistance/heat and thus temperature changes with coil uniformity/spacing. For example, It would be interesting to test how the PID on the DNA40 fires a straight wire.

I mean, I've had my DNA40 tell me it hit 210*C when I had a short or hot-spot of some sort and no vapor was coming off that thing. I've changed atomizers on it, and it didn't recognize that I did(didn't ask me if I put in a new coil). It read the right resistance but it's PID was way off... That's what I mean when I say it's different to measure temperature DIRECTLY verses measure resistance changes. PID doesn't go by the absolute temperature of my coil as if a thermistor would. It's tuned from a cold coil for a reason.

A prime example would be the battery life indicator on the DNA40. You think that's a direct measurement of the battery? No way, but the idea it was made probably was relatively sound. This is totally off-point though, like, what a tangent...

Dampmaskin - You are correct temperature and heat are two different things. Temperature can be calculated from heat depending on the environment.

Is that really how mods are designed though dampmaskin? Is it 20watts of heat being generated at the resistive heating element, or is it 20 watts of power being shot through the coil inwhich a good deal of it goes into generating heat? I'm asking a lot of questions because I legitimately do not know but want to know. Something tells me they set their DC-DC regulator to 20 watts and whatever happens, happens. I really doubt any chip on the market today is doing that much checking, beyond calculating a maximum and minimum battery limit for safety based on measured resistance.

 

[BlueridgeDog]

You have an excellent idea. My only question is: How is this better than the method that the market is currently going with i.e. nickel?

Personally, I think the future will have some temp. sensing thermocouple integrated to the atty.

 

[Flavored]

What I meant by backwards was you seemed to be quoting information on electrical and electronic circuits that spoke of heat losses in circuits. Nearly every electrical and electronic device gets warm/hot as it works, this laptop is currently warming me up right now. That is heat loss, serves no useful purpose (unless I'm cold). With our resistive heating coils, the purpose is to heat the liquid, there is no loss. Coils used for other purposes (transformers, inductors, etc., applied largely in AC circuits) have resistive heat losses.

The DNA chip does not predict a coil temperature for a give wattage, it measures the temperature and, if you have reached the temperature setpoint, will vary the wattage to maintain the temperature. Draw harder/create more airflow to cool the coil, and the chip will increase the wattage to keep the temperature up. Lose juice flow to reduce cooling on the coil, and the chip will back the wattage down.

The part I do not understand about your idea is how you intend to measure temperature. Unless I'm misreading, you have suggested a different way to measure current, a way tha tis more complicated than in use right now.

 

[TheJester]

BlueridgeDog - I don't think my idea would be better than nickel wire (if it works). Nickel wire works great, at best it would be on par. This is more so for fun, rather than changing the vaping world. It would be cool to make a "wire" invariant temperature control mod/chip. I am definitely with you all in thinking that a thermocouple atomizer makes the most sense. Instead of putting 12 coil posts for octa-tetrahedral-reimannian-scaffold coil builds, put a little thermowell? Makes too much sense hahaha. Someone needs to make the jump and have it catch on though.

Or why hasn't an RDA come out which bottom-feeds without involving passing eliquid through a 510 connection? Not sure, in my opinion it makes more sense than those negative pressure RDTA's and their poor seals and dripping control.

Flavored - I think we could talk in loops about something totally irrelevent. I'm not looking at it backwards, I am approaching the problem a different way. I guess I'll just do the math and see how the numbers look then post back. In the mean-time there isn't much else to really say .

 

[Dampmaskin]

Is that really how mods are designed though dampmaskin? Is it 20watts of heat being generated at the resistive heating element, or is it 20 watts of power being shot through the coil inwhich a good deal of it goes into generating heat?

An electrical heating element, unlike an electrical motor, actuator, transformer, conductor or semiconductor, has a theoretical efficiency of 100%.

 

[TheJester]

I have read in technical manuals that Kanthal was supposed to have efficiencies ranging from 60-70%. I forget the system they were referring too, I think they were using it for forging purposes. I guess it all comes down to what is efficiency. How insulated is the system and what are we focussing on heating. Even with a theoretical 100% efficiency, I was able to detect a magnetic field off of the coil via induction, and power still ran through the wire back to complete the circuit. So there is some amount of energy available to be used.

I'm trying to set up a finite element method simulation for this idea but it's going to take some time. Have a lot of work to do so I'll return to it soon enough.

 

[Dampmaskin]

When you run current through a coil to heat it up, the current doesn't end up in the coil and disappear, or disippate as heat. We're talking closed circuits here, and the current going out of the coil equals the current going into the coil, even though power is "lost" in the coil. What drops off is the voltage, not the current.

See Kirchhoff's circuit laws.

If you have read that a Kanthal heating element has an efficiency of less than 100%, I would very much like to see the source of that statement. I suspect you must have misread it.

Or maybe you're refering to the power factor in an AC circuit? That does not apply to most atomizers, as these run on DC.

 

[mcclintock]

The magnetism will be proportional for current for a given coil. It will not change with temperature unless the wire changes resistance. Resistance wire materials are chosen for approx. constant power flow regardless of temperature. Therefore you'd still need nickel, titanium or similar wire to sense temperature using this method, and compare voltage and current to determine temperature. I see no reason the result would be more or less accurate than with the DNA40 circuit. It wouldn't know the actual resistance but it doesn't have to, it just has to know when the reading is cold for a baseline and see how it changes relatively. However it requires and additional component compared to the DNA40, which surely just measures the current by the voltage drop across a low value resistor in series with the load, luike other ammeters (which does cause some power loss).

The efficiency of the coil circuit is decreased by resistance getting to the coil, and the part of coil leads that don't heat juice, or just heats the posts. The DNA40 (I don't have one) may allow setting the temperature within one degree but that's its Precision, its Accuracy is surely less. This is OK since, in the end, you're setting it for a good vape regardless of the numbers. There would be no reason to limit settings just because it doesn't act like a perfect calibrated thermometer.

 

[Dampmaskin]

The precision of the DNA 40 is 10°F or 5°C. Depending on the conductivity of the atomizer and 510, the accuracy can be ... well, in my experience, all over the place to be honest.

Exactly how it measures the resistance I don't know, but its efficiency seems to be very high. 92% according to Evolv's spec sheet.