Yes this is a very good idea. That or real temperature measuring - like pbusardo now does with a temperature probe.
Fortunately, I also now have a temperature probe device, that goes to 300°C and says it's accurate +/- 1°C. Unfortunately, it cost $5 from FastTech, so I am not sure quite how good it will actually be. I'm hoping maybe "better than nothing", and I plan trying wrapping a coil around it soon.
No, but that's the whole point - we adjust NP to counteract the known/expected imperfections.
The best example is Static Resistance. We know for certain that when we see an ohm reading of x.yzΩ, it is not the true reading of the wire. There has to be
some Static Resistance: the mod 510, the atty 510, the connection between atty and coil and the atty itself all contribute resistance.
We hope that that static resistance (SR) is low, and we make our best attempts to minimise it - the easiest way being to ensure the build is tight, and with some high-SR RTAs (eg KF4, SR measured at 0.03Ω) tweaks have been found to reduce it. Even thoroughly cleaning the screws has been found to make a difference of 0.005Ω-0.01Ω.
The beauty of NP, and also Dicodes' Coefficient Scale (CS), is that we can make calculations - we see what our real resistance is, then we estimate (or even measure - see below) the SR, and we calculate a revised NP/CS value that takes into account the estimated real resistance.
In a perfect/future world, TC mod makers would allow us to enter an offset - "This is what I read the resistance as, but you, oh vaping master, infinitely superior to me, a mere machine, know that that is wrong - please tell me how much to adjust this by?" It might use slightly different language. Though on Evolv devices, this feature will not be added; instead Evolv have an alternative they are implementing after receiving suggestions from dr g, whereby if you think the temperature is wrong a small hammer flies out of the top of the mod and hits your head until you agree that it's Completely Accurate. I for one welcome our Evolv Overlords.
But in the absence of such resistance offset adjustment, we can use coefficient adjustments.
The whole SXK NP thing has been unfortunately clouded by the fact we're having to throw it into unexpected use to counteract SXK's chronic low resistance. But that is far from its only use. I will get my Dicodes 2380 tomorrow or the day after, which I paid All The Moneys for, intending it as my benchmark TC device; I will no longer have to adjust CS for poor resistance readings, rather I hope to use it to adjust for estimated/measured SR.
As for estimating/measuring SR:
- Firstly, estimating - just knowing that a big-metal RTA will have more than an RDA, and complicated RTAs (KF4, Squape, etc) even more.
- Eg 0.03Ω for a KF4 and similar atty is a good estimate - 0.02Ω at the least.
- Secondly, indirect measurement - that's what I'd call the Water Test or any other kind of actual temperature test.
- Do the temperature test, adjust NP/CS until the temperature configured is actually achieved, and your new NP is the right one for that target temp and build and wire
- From that, work backwards to work out the SR of the atty/build (I can add this feature to my calculator spreadsheet.)
- You might say, why work backwards when one already has a good NP?
- Because the NP/CS value is specific to a target temp and wire type and starting resistance: the SR is not (or much less so)
- Once the correct NP/CS value is known, the SR can be worked out from it, and then that SR can be plugged into future calculations of NP/CS values for other target temps and other wire types, without having to repeat the water/temp sensing test each time.
- Also, shared with others to build up a knowledgebase of atty:SR pairings.
- Third, direct measurement - shorting the atty and measuring the resistance
- The most extreme, and requires accurate resistance measurement right at least down to 0.01Ω
- Short the posts of your atty with a tiny piece of low-resistance wire, then take a measurement
- Ideally do this off a mod, on a separate resistance reader - but probably not a $10 ohms reader because they're not usually accurate that low.
- If doing it on a mod, be sure not to press fire
Or else be very confident your mod will say Low Res and not try to actually fire To be fair, all should, and some have been tested in this exact way.
- This method was pioneered by @ndb70 and @funkyrudi has also been doing it - I have yet to try myself.
- Once I get my Dicodes 2380 I plan to try it and then use the resulting measurements to adjust the CS for maximum accuracy.
Oh and to answer the question "what reads accurately to 0.001Ω" - the Dicodes claims to be accurate to mOhm resolution, 0.001Ω. But annoyingly, and baffingly, it only shows to 0.01Ω granularity, even on the dedicated Resistance Reading screen. So it will round the last digit.
The Yihi chips display three decimals (during Set Resistance) so we can hope they're reading it accurately, but they haven't stated an actual resistance accuracy. And the fact that they truncate the last digit, not round it, for the main display, is also baffling. But we can hope they at least use the last digit in their TC calcs - hope, but not assume I think.
