This relates, at least in part, to why Ni200 is such a "bad" wire for TC.
BRANDON of Evolv was, in an early interview, cited countless times, by the 'spaced-coil-crowd' (read, fanboi's and fangurls), for saying (paraphrase) if you blow on the (contact) coil the temperature accuracy is affected, therefore only build spaced coiled, or spaced
coils are "recommended".
The TCR of other wires, mostly, eliminates this bugaboo.
Interesting, I had not heard of that quote (I wasn't around until a while after the 40's launch and not into TC until later still.)
I hadn't associated spaced/contact particularly with accuracy. I had thought Ni200 couldn't be contact because it doesn't properly oxidise and will short/fluctuate resistance.
I don't really understand what Brandon could mean by that, because any real coil has airflow against it - sucking not blowing, but same deal pretty much. A contact coil would vary how the air flows around/between the wraps, but I don't see how it would be all that different overall. Especially for the DNA 40 which seems capable of coping fairly well with both no airflow and lots of airflow.
Do you have any references/inks to Evolv's original quotes on this - or people discussing it?
So far all of my temp testing has been on spaced coils - not only because I assumed that was required with Ni200, but also for the practical reason that it's far easier to get the tiny bobble at the end of my probes held tightly against a single wrap of a spaced coil: I can sort of hook it on, then secure the wire in such a way as it is being gently gently pulled against the coil, locking it in place against the wrap. With a contact coil, I would need to
press it against the wrap, which is harder with the probes I have. Even using a wick to sandwich the probe between wick and coil often doesn't seem to apply quite enough pressure to ensure fully accurate readings.
This will hopefully change when I get my new flexible-rod probe, so I will try contact coils as well then - it will be interesting to see if it makes any practical difference for accuracy with Titanium, SS, and NiFe.
My explanation for potential differences in practical accuracy with wet+airflow is as follows:
The tc mods just measure res for temp control so why would airflow matter, other than requiring more power to achieve target res/temp?
Granted airflow and wicking affect rise/fall time of temp but shouldn't effect tcr values/ temp accuracy......yes?
When doing "real vaping" tests - adding juice and airflow - we're moving on from only testing the mod's ability to measure the resistance of the coil and apply power to heat it to a certain temperature and keep it at that temperature.
Or rather, we're expanding our look at that to consider a more complicated case. We're now looking not only at the raw resistance readings and TCR calculations, but also the sophistication of its power management algorithm and the frequency with which it performs its calculations.
The dry coil case is, generally speaking, the ideal case for the mod. It is heating a piece of metal which has (almost) no external influences upon it. It is in almost complete control of the temperature of the coil. Not
absolutely complete control, because the background temperature is still much lower than the coil so in the absence of power it will naturally cool down. But the background temperature is known, and the rate at which it cools the coil is relatively mild and relatively constant.
Wet coil + airflow is the opposite. Now there is a varying, unpredictable force cooling the coil, acting against the power the mod applies.
I think of it with the following analogy: Imagine lying on your back and elevating a ping pong ball by blowing on it steadily, such that it's floating above your mouth. Imagine there being a line drawn on the wall, against which you are trying to maintain the height of the ping pong ball. This is fairly straightforward (for as long as you have breath, anyway
)
Now try to do it when there is a fan in the ceiling blowing downwards onto the ball, and further there is water dripping from the ceiling onto it. You don't know the strength of the fan nor the volume of the water, and they are changing moment by moment.
That's why I consider wet coil + airflow to be more of a challenge for the mod. In particular, it will highlight deficiencies in the mod's frequency of resistance reading and calculation. With the dry coil case, it could theoretically check the resistance only, let's say, four times a second, and still do a reasonable job of maintaining the temperature. In a real vaping situation, with juice boiling off at varying rates, and airflow hitting the coil in fluctuating amounts, four times a second would not be nearly enough to accurately maintain temperature. If it really was four times a second, I'd expect to see big lurches in the power graph - a rapid decline in temperature for 250 milliseconds, following by a lurch upwards as the mod re-measures and realises the coil is under-temp and quickly applies a large burst, then another decline.
The steepness of each decline would vary greatly according to the strength of the external forces at that time, and therefore the lurches upwards will each need to be of different magnitude. It's quite likely the mod would never be at or even near the required temp, because that would require that it could, in one shot, apply just the right amount of power to take it perfectly at temp, including factoring in the forces acting upon as it applies that power which are potentially of different magnitude than those before. Errors will compound.
Four times a second is an exaggeration to show the point - I am sure most mods do it hundreds of times a second. The principle is that the more often the mod can do that measurement, the less it will have to correct for, the more chances it will have, and so the more accurate it will be. And a deficiency of frequency will be highlighted far more on the wet coil case; it could be perfectly accurate on dry and hopelessly inaccurate on wet, depending on the frequency.
Brandon of Evolv has also made the point that the DNA chips can measure resistance while also applying power; it was his (admittedly biased) impression that other mods may not. It's certainly possible he's right about some, and that would definitely be another impact - one which, once again, would cause much larger fluctuations and inaccuracies on a wet+aired coil.
The (early) SXK chip is one we associate with "rough" or "pulse-y" TC, and that's almost certainly a symptom of what I, and Brandon, describe: insufficient frequency of resistance readings, and/or being unable to read resistance while also applying power. Hence the feeling that it's suddenly applying a large "pulse" of power, to correct the coil temperature.
Then there's the smartness of the power algorithm: how does the mod know how much power to apply to achieve a given temperature increase? Does it have a linear scale, 1°C = X watts? Or does it vary as the temperature gets higher? Does it always apply pre-heat, or only in certain circumstances?
And then, how well did the mod maker study this? Did they test only with Ni200, or with Titanium too? Did they study it in only a dry coil situation, or only a wet/real vape, or both? How good are their numbers? How sophisticated is their algorithm - it might be the simple case of 1°C=X W of power, or they might try and be much smarter, trying to predict what the airflow/juice drain on the coil will be in the next millisecond, so as to apply the power for it as it happens, rather than after. Did they analyse the rate and effects of juice evaporation? I'm betting most did not, but maybe some did? Either way, will it make a difference to the accuracy if the juice is 50/50 vs 100% VG?
Which then gives rise to yet one more factor: the assumptions of the mod makers in their planning for their power algorithm. Although I've described above that I consider dry coil to be the easier case, that doesn't necessarily mean that I expect to see in my testing that dry coil is most accurate. Because mod makers are planning for real vapes, not thermocouple tests. They
expect juice and air, and perhaps they found they can be more accurate if they assume in their calculations it's always going to be there. Therefore making it accurate only with a "real" wet+air vape, and inaccurate with a dry test.
There's one way that can definitely be true: if the mod simply cuts power completely and stops vaping when it detects a "dry" coil, as the Yihi SX Mini M does for example. If after 4 seconds of vaping it's stuck at maximum temp, it flashes that "DRY COIL" message and simply stops firing altogether. When I tested the SXM on a dry coil, I had to release and re-press fire every three seconds: hardly representative of a real vape!
Another example of things possibly being different between dry/wet is shown on the Dicodes Titanium graph: we see it persistently over-heating at the start of each vape. For an apparently highly accurate, expensive TC mod, I find that surprising. But it could be that that only applies to a dry coil, because they have assumed there will be factors acting against the coil to cool it and planned their power delivery accordingly. (Though frankly I think it's still a black mark against the Dicodes that it does this, as the ultimately accurate/responsive mod would be able to cope with any scenario. I need to test it more carefully.)
Another example is the Joyetech graph - all over the place in my dry coil testing, but as we know providing a subjectively decent (if somewhat offset) TC vape.
I hope to study all this in a lot more detail starting now. I did do some tests months ago, both wet and dry, but mostly with Ni200 and not with all factors controlled (such as SR), and not with a nice test rig. I plan to do things better now and hopefully get a few more answers.
