Which is more accurate a TCR number or CSV file?

[mactavish]

FROM: http://www.djlsbvapes.com/dna200/ This reviewer seems quite knowledgeable, watched a bunch of his videos, he does use an external temperature probe, I have no idea of its accuracy or his results. Just thought I would post this and see if there are any opinions on this topic, or a response from Evolv, as while both methods can be used in Escribe software, it seems that even Evolv recommends the downloading of appropriate CSV files from SteamEngine,com and importing them, and using a manual entry of TCR value when CSV is not available, or you prefer to use your own customized value. A paste from the reviewer link above: "Regarding the temperature, in my opinion if you setup your Temperature Control profiles with the TCR of each type of wire, you will have an more accurate temperature and more stable vape as you can see on my testing below:"

[mactavish]

30 views, no opinions, guess I stumped the band!

[Spector NS5 RD]

i dunno i was wondering the same thing. i played around using both. both work fine for whatever wire you're using. as long as you have the correct tcr number, the vape will be just as good as a csv file, IMO. i know with a csv file it differs from a tcr number in the fact that it's not linear. a csv file can have slopes etc. for a wire material that is not a linear tcr. i think nickel has a non linear curve.

i'd say in theory, the csv would have the greatest accuracy. however, the tcr number input seems easier to do, in terms of convenience.

[mactavish]

Some newer mods, like my EVic VTC Mini, with firmware 3.0, allows for manual TCR input, but not CSV. If one or the other is more accurate temp wise, Evolv could boost their chip has an advantage. I believe temp accuracy is important, so let's see who weighs in on this subject, now that all the forum spam is gone for now.

[Viruk]

Please someone correct me if I'm wrong - this is my interpretation...

A TCR is a linear approximation, whereas a csv is a set of points on a curve - so I'd assume that the csv is a better option as it better represents the change in resistance due to temperature change

I also thought that Ni200 was a pretty linear correlation - I'm pretty certain its more linear than most materials anyway

[watcher64]

Ummm ...

 

The csv is just the TCR values put into a "comma separated value" file, you can have either linear or a curve. 


This makes them easy to load into escribe ...

 

Open one up with notepad, you will see temps and ohm values.

[Viruk]

apologies, my mistake - that'll teach me to post at 4am without reading properly!

I was thinking about other devices that use a single value for the TCR rather than a curve...

a csv will contain something like this..

"Temperature (degF)","Electrical Resistivity"
-100,0.670714795589447
70,0.999953627586365
200.170745849609,1.28299999237061
299.544677734375,1.50100004673004
400,1.70200002193451
500.341491699219,1.92499995231628
600,2.09100008010864
800,2.375

Generally, unless you have the equipment and expertise to make your own file you should rely on wire manufacturers or other users files (including steam engine). There are plenty of people on the forums here sharing their experiences.

[mactavish]

watcher64 said:

Ummm ...

 

The csv is just the TCR values put into a "comma separated value" file, you can have either linear or a curve. 


This makes them easy to load into escribe ...

 

Open one up with notepad, you will see temps and ohm values.

Doesn't explain the differences in his temperture results between the two, did you watch the video?

[jazzvaper]

@mactavish I did watched the entire video. And had watched his video on the Hotcig, which after mine died I returned for a COMPLETE refund.

[mactavish]

jazzvaper said:

@mactavish I did watched the entire video. And had watched his video on the Hotcig, which after mine died I returned for a COMPLETE refund.

I was asking "watcher64". Assuming he tests exactly the same with his temperature probe, he does publish different temperature results with TCR Versus CSV inputs. Guess I'm going to have to write and bother the SteamEngine author again, maybe it's the holidays, but surprised no official response from Evolv, they really need to either hire more help, or get some product experienced moderators here to help out. This entire forum is getting like the Apple site, just end users helping each other if they can, and little input from the source!

[watcher64]

mactavish said:

[QUOTE=jazzvaper]@mactavish I did watched the entire video. And had watched his video on the Hotcig, which after mine died I returned for a COMPLETE refund.

I was asking "watcher64". Assuming he tests exactly the same with his temperature probe, he does publish different temperature results with TCR Versus CSV inputs. Guess I'm going to have to write and bother the SteamEngine author again, maybe it's the holidays, but surprised no official response from Evolv, they really need to either hire more help, or get some product experienced moderators here to help out. This entire forum is getting like the Apple site, just end users helping each other if they can, and little input from the source![/QUOTE]

 

My thoughts would be that the CSV is of course not as accurate for some reason, and that could be steam engine, or it could be how escribe plots the points.  If you look when using the TCR, I checked with TI only BTW, the base resistance and the top end are different, which could throw the whole thing out of wack.

But honestly, I could care less about "exact" numbers, if it vapes well, is not burning, not melting my face off, then then I'm not overly worried about it.

(on another note, is it possible that, when using a csv to plot the points , and adding this "checking" the chip just does not keep up.)

[mstave]

Vapes well AND not melting face off - My God, dude ... you have high standards!

[Viruk]

So - having found the relevant section in the video for him playing with the TCR for Stainless Steel, I actually stand by my initial comment...

The guy in the video appears to refer to a TCR as a single value he enters (a linear approximation as I initially stated) whereas the CSV is a set of data points to on a curve.

The CSV should be more accurate....

I'm referring to the section of video around the 47 minute mark - if he goes back and does this again later in the video please let me know a time reference so I can revisit it...

The TCR is only a single value he sets - so less data points than the csv (which is basically several data points on the temperature curve). He says "its actually a better way and better experience if you set the TCR" - but in the section I watched he doesn't explain why he thinks this is the case.

I don't actually think the way the guy in the video uses terminology is the best for clarity.

Having found his results section (which seems to be lacking in methodology btw... he just has some numbers listed in graph form) - he does say that there could be margins of error due to variance in the alloys used.

In terms of his lack of methodology related to my own (limited) experience with a csv, is for 316L stainless steel and I'm actually using a 0.4ohm Aspire Triton coil. I'm seeing a lot of bouncing around of the temperature in the device monitor. The issue of SS in temp control is discussed in other threads and the bouncing is not my focus here, however what number would I record for temperature if it is bouncing around like that?
I'm just too uncertain about his process to rely on the data he presents as an absolute value.
Are his coils really hitting and staying exactly at those temperatures? I doubt it - so whats the process for recording a temperature data point? how is this average value reached/calculated?
Without some context to how this figure was arrived at I think the value of stating the number is limited...

Rather than the question is CSV or TCR better; I think the question should be - are the data points in the csv more accurate for my coil material than a single approximation of the TCR value?
Are you familiar with the term "Garbage In Garbage Out"? your results will only be as good as the values set in the CSV or TCR. (This is of course assuming that the chip handles everything as expected with no bugs or errors - but I'm not opening that question up here)

So, in the video he claims that when you don't have the figures from the wire manufacturer the TCR is better; however this seems to be a fairly arbitrary claim. I guess YMMV - but without testing this out yourself with your own coils/materials you can't know which is better.

In my mind (and I could be wrong here!), by using a single value for TCR you lose some of the benefit that the DNA200 chip can give you from using a more accurate set of data - but that assumes the data points are correct...

Apologies if some of that is a bit rambling - I've tried to skim through the video and note my thoughts as I go.

[mactavish]

Agreed, a CSV file should be more accurate, but that's the contention here, that in this reviewers findings, they are not. I do know the difference between both types. I've watched a few of his review videos, he seems quite knowledgeable, and handy with his mod breakdowns. On his website,, he lists his temperture results (I'll paste them below for easy access), and while I do not know what temp sensor he uses, or its accuracy, if one assumes he does each test exactly the same, with the only difference being a CSV or TCR input, that's what makes the theory interesting. Sure, the atomizers and builds would have to be stable etc. But this guy is no dummy in my opinion. Without a sensor, I can't duplicate his tests, one could only try making two presets in Escribe, and then try and feel if one can detect a temperature difference in the vape, switching between the two. I mostly use Titanium wire, he claims a 26 degree higher temperature in the CSV file over the TCR, that's quite a bit higher, should be noticeable by feel, if true. When I get some free time I'll attempt it myself, on my Vaporshark DNA200. But I don't think this guy is just making things up, would serve no purpose, he seems pretty serious about his reviews, which why not quite as detailed as PBusardo's, they are certainly more technical then most other reviewers. ------------------------------------------------------------- PASTED FROM REVIWERS WEB PAGE: Here are my findings on the DNA200: Regarding the temperature, in my opinion if you setup your Temperature Control profiles with the TCR of each type of wire, you will have an more accurate temperature and more stable vape as you can see on my testing below: DNA 200 TC Tests: (TCR is using each wire TCR added on the "Special" Button under each profile on Escribe Software) (CSV is using the CSV files downloaded from Steam Engine) All tests where conducted with the temperature of 400ºF NI200: TCR: 397ºF CSV: 389ºF Titanium: TCR: 403ºF CSV: 429ºF 304SS: TCR: 401ºF CSV: 395ºF 316LSS: TCR: 404ºF (0.00100 instead of 0.00092 DNA200 doesn’t allow that TCR) CSV: 387ºF

[Viruk]

I'm not saying he's wrong - but I'd like to know his methodology to get to the results he found.

There are many reasons why his findings don't match our assumptions, but there are just too many possibilities to sensibly guess why that's the case.

Is he using cheap wire? Did he download the wrong csv?
In the video he repeatedly states the SS file as "316L" but selects SS 316 not SS 316L. I'm not saying he got it wrong, but its one possibility of the results not being quite right.

I notice in the video that for SS he has a CSV file from DJaquith as well as files from steam engine - which one was used to compare against the TCR value he uses? I've seem some posts from Jaquith regarding different CSVs for SS and I'd be very interested in his opinion on this topic.

In the video, he mentions using a wire manufacturers csv being better than a TCR value - but he provides no rationale for that statement. He doesn't say he tested it - is he just assuming this is correct?

There just seem to be too many assertions without rationale or evidence of testing for me to accept it at face value. I'd like to repeat - I'm not saying he is wrong or mistaken, he might be absolutely correct in every way; but to me it just sounds like there are too many unknowns/assumptions in this process.

What were the pre-heat and punch settings? From what I've read, SS doesn't seem to do very well with pre-heat set too high.
What were the coils used in the test? Was the same coil used? If so, did it adequately cool to ambient temperature before a new test was performed?

Anyway - I'm curious about the result and any verification we can get. I don't mean to just bash this review.

Mactavish - if you do test this yourself with Titanium I'd be interested in your result.
I have some SSV Titanium but have yet to use it as I've only had my DNA200 a few days. I have the manufacturers CSV and I'm curious how it will perform - but as the SSV titanium is some alloy I cant use a single TCR to compare that myself.
I could try the SS Triton coil with a TCR but I can't measure the temperature as I don't have the equipment to do it - I would only be able to provide (completely unscientific) feedback about whether I noticed any difference and possibly a screenshot from device monitor about the stability of the temperature...

[VapingBad]

Both are a method of expressing the relationship between temperature and resistance, both can be set to different degrees of accuracy, a TCR is limited to a linear-relationship and the csv is not.  If the material has a perfectly linear-relationship/flat-curve they are as good as each other, but for materials that don't the csv can be more accurate.

So a csv is definitely a superior method.

[HugeEgo]

Steam engine's TCR values are listed for the "vaping range."  If you look up the TCR of titanium at 20°C, you will usually find .0035.  However, that is for room temperature.  In the vaping range the TCR changes from .0036 to .0037.  This is where Steam-engine gets the value of .00366.

So, Daniel plugged in .00366 as his TCR number, which IS the same thing as a .csv curve.  The .csv file is simply a file that provides a TCR number for a certain temperature (since most metals have varying TCR's depending on temp).  This allows the user to put in the exact TCR number for, say, 450F.  However, 0.0036 is going to be very close throughout the whole vaping temp range.

As for why his .csv provided a less accurate temp, I don't know.  It should be exactly the same as inputting the TCR manually.  Perhaps he used the wrong file.

If you're really concerned about it, go buy a thermocouple probe and test yourself.  You can get cheap Chinese digital thermometers for around $20.

[Vaportron]

How would you test the actual live temp? Shoot a coil with an infrared thermometer? Hitting the fire button to do a dry burn test isn't really accurate. During a vape (that is a dry hit), you are inhaling and providing air Cooling to the coils which just blowing a fan (even at high speeds) just doesn't really replicate the event.

[Spector NS5 RD]

^^^^^if you have an accurate temp probe, one that plugs into a good multi meter, that should work. stick it in between the coil, maybe wrap a tiny bit of cotton around so it doesn't short the coil out. 

[Vaportron]

ChunkyButt200 said:

^^^^^if you have an accurate temp probe, one that plugs into a good multi meter, that should work. stick it in between the coil, maybe wrap a tiny bit of cotton around so it doesn't short the coil out. 

AH very good. Thanks for that info.

As far as which SS TCR "code" is best i like 0.00100 as one reviewer had put it.
Has anyone else been using this "code" ?

[Vaportron]

ChunkyButt200 said:

^^^^^if you have an accurate temp probe, one that plugs into a good multi meter, that should work. stick it in between the coil, maybe wrap a tiny bit of cotton around so it doesn't short the coil out. 

AH very good. Thanks for that info.

As far as which SS TCR "code" is best i like 0.00100 as one reviewer had put it.
Has anyone else been using this "code" ?

[Spector NS5 RD]

i'm using the csv file from steam engine's wire wizard (304 ss). a little high on the tcr value but works good enough for me. i'm hearing 0.00092 is the correct/most accurate when it comes to 316.

[Vaportron]

ChunkyButt200 said:

i'm hearing 0.00092 is the correct/most accurate when it comes to 316.

thats what i heard about 316L but i cannot enter that into escribe
 

[Spector NS5 RD]

do you have an older version of escribe? make sure youre running the latest. i can enter that value on my escribe. so my guess is you need to update.

[Vaportron]

Yes vey good. I have updated escribe (i thought it was suppose to keep itself up to date) and have changed the tcr to 0.00092! Thanks Chunky. However what is everyone using? Now I'm vaping at 350F compared to 390F (0.00100). Now if this thing is reading the temperature why do i have to turn it down? Because it is just guessing?