Stainless steel-what do you find works best, having trouble getting it to work correctly (316L and 317L)

[Voku]

I have the impression that the SS behaves like Kanthal . The Ti can even observe how much the resistance of even relative to ambient temperature .


sorry for my English

[Bobby]

Well, the erratic behavior occurs on different atomizers, DIY coils and factory coil heads, on SS304 and SS316L, when using different CSV files and when using the TCR setting, with and without preheat. Other users have confirmed that it happens also on different brands of DNA200 mods, and using different profile settings and that different TC wire (e.g. Ni200) with coils with a lower ?R per degree C will produce a smooth graph
I think that that rules out: bad connections of the atomizer, the coil build/head, a possible wrong composition of SS wire, wrong TFR, wrong settings, a specific brand of DNA200 mod, usererror and measurement inaccuracy. I think the only thing left is a bug in the TC firmware when using SS.

Can somebody from Evolv comment? Is this a know issue?

[HugeEgo]

Bobby said:

Well, the erratic behavior occurs on different atomizers, DIY coils and factory coil heads, on SS304 and SS316L, when using different CSV files and when using the TCR setting, with and without preheat. Other users have confirmed that it happens also on different brands of DNA200 mods, and using different profile settings and that different TC wire (e.g. Ni200) with coils with a lower ?R per degree C will produce a smooth graph
I think that that rules out: bad connections of the atomizer, the coil build/head, a possible wrong composition of SS wire, wrong TFR, wrong settings, a specific brand of DNA200 mod, usererror and measurement inaccuracy. I think the only thing left is a bug in the TC firmware when using SS.

Can somebody from Evolv comment? Is this a know issue?

It's really not Evolv's problem.  It is one of those "it is what it is" type of things.  The TCR of 316/317 SS is what it is.  Evolv can't change the laws of chemistry.

Is it possible they could make improvements to the board in the future to compensate?  I suppose it's possible, but I doubt this is something that could be improved upon with firmware updates.  It would take an even more accurate ohm reader (that can read beyond the milliohm range -- say in microohm range).  I don't know if this is physically possible to do with current technology, at least not accurately for our application where we are screwing in atomizers with 510 connections, etc.  It's already hard enough to get an accurate milliohm reading.  But I could be wrong, I am not an EE.

[Bobby]

HugeEgo said:

It's really not Evolv's problem.  It is one of those "it is what it is" type of things.  The TCR of 316/317 SS is what it is.  Evolv can't change the laws of chemistry.

For the measurement system the only thing that really matters is the change of resistance per degree C of the coil that it measures.
- With a 0.049 ohm Ni200 build (TCR = 0.00641) the ?R per degree C is ~0.314 milliohm.
- With a 0.47 ohm SS304 build (TCR = 0.00105) the ?R per degree C is ~0.5 milliohm.
- With a 0.35 ohm SS316L coil (TCR = 0.00092) the ?R per degree C of ~0.3 milliohm.
The TC system regulates the first one without problems and shows a smooth curve, so why the erratic regulation with the other 2?

Simplified, what I think happens is:
Once the fire button is pressed, the system sets the power and continuously measures the live resistance. Besides the switch from punch power to setpoint power, it does not change the power as long as the temperature setpoint is not reached.
Once the temperature setpoint is reached, the TC system uses a regulation algorithm (programmed in the firmware) to set the power to a level that will increase or decrease the temperature. The use of a damping factor prevents too large adjustments. This is done in a loop until the fire button is released.
The damping factor is probably calculated using the TCR value of the wire type.
With Ni and Ti the damping factor is correct, so the size of the power adjustments are appropriate, preventing the temperature from wildly overshooting.
With SS the damping factor is wrong and the size of the power adjustments are too large, causing the temperature to overshoot every time.
According to my logic; if a proper damping factor calculation is used with SS, this can be prevented.

[HugeEgo]

Bobby said:

For the measurement system the only thing that really matters is the change of resistance per degree C of the coil that it measures.
- With a 0.049 ohm Ni200 build (TCR = 0.00641) the ?R per degree C is ~0.314 milliohm.
- With a 0.47 ohm SS304 build (TCR = 0.00105) the ?R per degree C is ~0.5 milliohm.
- With a 0.35 ohm SS316L coil (TCR = 0.00092) the ?R per degree C of ~0.3 milliohm.
The TC system regulates the first one without problems and shows a smooth curve, so why the erratic regulation with the other 2?

Yes, all of that is correct.  Another way of looking at it for a given coil is to use the steam-engine formula for "TC Precision" (TCR * R * 1,000,000). 

However, consider the following: Let's say we vape at 232°C (450°F) and assume the room temperature is always 20°C.  Now, let's calculate a few different scenarios for various wires.  I chose these at random using a 3mm bit with roughly the same wraps for each.

316L stainless:
Initial resistance at 20°C = .341?
Final resistance at 232°C = .409?
TCR = .00094
Total ?R = .068?
Total percent increase in R = 20%

Ti Grade 1 coil:
Initial resistance at 20°C = .247?
Final resistance at 232°C = .439?
TCR = .00366
Total ?R = .192?
Total percent increase in R = 78%

Ni-200 Coil:
Initial resistance at 20°C = .05?
Final resistance at 232°C = .114?
TCR = .006
Total ?R = .064?
Total percent increase in R = 128%

As you can see the total change in resistance is 128% in nickel, 78% for Ti, and 20% for 316L.  Even though the absolute rise in R (316L and the Ni-200 coils had almost the same total change in R) is not that much different, I think the overall change in resistance relative to the starting resistance of the coil is somehow affecting the accuracy.  Whether this can be overcome, I do not know.  

TL;DR --  Ask John.  I would be interested to know as well.

[Bobby]

@HugeEgo Thanks for the reply, mate. Although it did not provide the explanation, it made me rethink the case and I think I found the error in my logic:

The resistance (and temperature) cannot be measured directly. The voltage and current are measured and with these measurements, the resistance and temperature are calculated using the reference values at room temperature and the TCR.

The total ?R (and therefor the ?R per degree C) might be the same for the SS316L and Ni200 examples, but the total ?U and ?I (and therefor the ?U and ?I per degree C) are not. These are much smaller for the 316L coil compared to the Ni200 coil (and the Ti coil, see tabel below).
So the inaccuracy comes from the low rate of change in voltage and current as the temperature changes. A small error in these measurements - due to fluctuation and/or due to the fact that these values are close to the resolution of the measurement system - has a big impact on the calculated temperature. This might be why the regulation is erratic with as low TCR values as Stainless Steel.
I also wonder if this erratic regulation also occurs with other mods and Stainless steel, but nobody is aware of it since you cannot see the TC graph?!

These are just my thoughts so it would be nice, to get confirmation of this theory by Evolv or another expert.

			P=	15,00 W	?T=	212°C
Wire type	?20oC	?220oC	?T1-T2	U20oC	U232oC	UT1-T2	I20oC	I232oC	IT1-T2	?U per oC	?I per oC
SS316L (V4A)	0,341 ?	0,408 ?	0,067 ?	2,26 V	2,47 V	0,21 V	6,63 A	6,07 A	0,57 A	0,99 mV	2,67 mA
Ni200	0,050 ?	0,116 ?	0,066 ?	0,87 V	1,32 V	0,45 V	17,32 A	11,39 A	5,94 A	2,13 mV	28,00 mA
Ti Gr I	0,247 ?	0,430 ?	0,183 ?	1,92 V	2,54 V	0,62 V	7,79 A	5,90 A	1,89 A	2,90 mV	8,91 mA

[HugeEgo]

Yep, I think you are onto something.  The change in voltage and/or amperage needed to maintain the temperature are too small to do on the fly accurately (or at least this is the best explanation in my non-EE view).

[Johan]

"Temperature (degF)","Electrical Resistivity"
-58,0.915064
-0,0.95433
68,1
121,1.15
302,1.2
390,1.25
490,1.30
590,1.35
800,1.40

lordmage, or anybody - I wanted to try these numbers but having trouble hitting the exact number with the mouse. Can't figure out how to input with the keyboard. Also, how do you set electrical resistivity points. Managed to "split" one to add another, is that the only way? I have been loading one of Jaquiths .csv's and trying to tweak it, but I'm curious as to how to start from scratch. And yes, it does appear that I know just enough to probably get myself in trouble lol. Thanks

edit - I got them as close as I could and tried it with 316L on a Cthulhu V1 and it actually worked. Was happy, then it just quit temp control. Grrr....

[Bobby]

I would copy/paste the list you posted in a .txt file, save it, change the extension to .csv, load it in EScribe and upload it to the mod.

A .csv file can be edited manually as a plain text file by using notepad.exe (or as .CSV file with Excel off course). To tweak the y-value of the points by keyboard input, you can right-click the point.

If TC is not working, it might be that the build you are using has a big mass (e.g. Claptons, etc.) and is heating up too slow. I read somewhere that this should be bigger than 150 degree F in 1s or TC will not be activated and the set power will be applied without TC regulation. If this is the case you can increase the pre-heat/punch settings.

[Johan]

Ahh, yes (slaps palm against forehead). Thank you Bobby.

[Bobby]

Johan said:

"Temperature (degF)","Electrical Resistivity"
-58,0.915064
-0,0.95433
68,1
121,1.15
302,1.2
390,1.25
490,1.30
590,1.35
800,1.40

Btw: are you aware that these numbers are too high for SS316L? In all my 4 csv files (collected from different sources) the numbers are ~ 0.1 lower.

[Johan]

No I didn't. These were taken from a previous post and thought I would try them.

[Johan]

But I think I'm chasing my tail here anyways. Just tried some NiFe52 and getting similar results.   Nickel and titanium work fine with this mod/board. Gonna call it a day and use stainless steel on devices that work. Not inclined to get another dna200 at this point though...