Adjustment to compensate for deck resistance

[GaryParr]

Does anyone know how to compensate for deck resistance?

I've got two different rta decks that consistently show different ohms readings with the same coil builds. One deck always hovers around .180 and the other around .246.

is there any way to correct for the deck resistance? My hunch is that the TCR can be adjusted to compensate, but it would require someone smarter than me to work out the math.

 

[ShowerHead]

What kind of mod are you using? Some have the ability to adjust resistance, some don't.
With that large a difference though, I likely would not do it. Are these the same tank, moving the deck from tank to tank makes no difference?

 

[KenD]

Well, the tcr should not be adjusted. That is based on qualities of the wire material and does not change due to external factors. Adjusting the initial resistance could do it, but in that case you would need to have a way to very accurately measure the "added resistance" (something like the DNA/escribe read mod resistance function). Honestly, I've never come across an atomizer that adds that much resistance. There's something wrong with it, and it would probably be best to fix what's wrong or stop using it rather than trying to compensate for it.

Sent from my Thor E using Tapatalk

 

[Frenchfry1942]

I slapped my decks around until they stopped showing resistance.

Worked for me.

 

[GaryParr]

What kind of mod are you using? Some have the ability to adjust resistance, some don't.
With that large a difference though, I likely would not do it. Are these the same tank, moving the deck from tank to tank makes no difference?

The mod is a Smok G-Priv, so I've got decent adjustment capabilities. Good question about the tank... it's two different tanks (Smok TF-V4 and a Vandy Kylin) but I'm using the same pre-wrapped 430 SS coils in both builds. There is minor difference in leg-length when all is said and done, but I've swapped these coils out numerous times and, while there is some fluctuation in resistance when swapping coils, the end result is always somewhere around a .065 Ohm difference. I've also borrowed other mods to check resistance with and they always match (within reason) the G-Priv.

Well, the tcr should not be adjusted. That is based on qualities of the wire material and does not change due to external factors. Adjusting the initial resistance could do it, but in that case you would need to have a way to very accurately measure the "added resistance" (something like the DNA/escribe read mod resistance function). Honestly, I've never come across an atomizer that adds that much resistance. There's something wrong with it, and it would probably be best to fix what's wrong or stop using it rather than trying to compensate for it.

I agree there appears to be something "off" about the deck with the high resistance and yes, I probably should stop using it for TC. But... it got me thinking about other decks that might be doing this same thing (to some degree or another) and how people compensate.

I'm not sure I agree that adjusting the initial Ohm reading is the right way to go. The mod measures the change in resistance, right? In a hypothetical .5 ohm build with .25 ohms coming from the deck and .25 from the coil, heating up to 50% change in total resistance (0.75) would actually mean you have a 100% change in the coil (.5) plus the original un-changed 0.25 from the deck. Changing initial ohm reading won't affect that. Seems more like lowering the TCR so the mod expects a 25% change would be the right way to go, no?

I slapped my decks around until they stopped showing resistance.

Worked for me.

This. This is just brilliant.

 

[KenD]

I agree there appears to be something "off" about the deck with the high resistance and yes, I probably should stop using it for TC. But... it got me thinking about other decks that might be doing this same thing (to some degree or another) and how people compensate.

I'm not sure I agree that adjusting the initial Ohm reading is the right way to go. The mod measures the change in resistance, right? In a hypothetical .5 ohm build with .25 ohms coming from the deck and .25 from the coil, heating up to 50% change in total resistance (0.75) would actually mean you have a 100% change in the coil (.5) plus the original un-changed 0.25 from the deck. Changing initial ohm reading won't affect that. Seems more like lowering the TCR so the mod expects a 25% change would be the right way to go, no?

The right way to go about it is to not use a faulty deck and expect good results. Any way you try to compensate will be a band aid that will, at most, only be accurate within a very small temperature/resistance range. By changing the tcr you're telling the mod that you're using a different material than you are.

Sent from my Thor E using Tapatalk

 

[papergoblin]

Does the deck show resistance with no coil(s) on it? If the deck shows resistance when blank, it's bad and needs insulators replaced or stop using, cause eventually you're gonna get a dead short.

Many of my attys show a different resistance with the same coils but that's due primarily to leg length. If you're set on using both, then build the one reading higher with a lower coil(s). If you usually use say 5 wraps, do 4 wraps. Resistance is resistance, doesn't matter to the mod if it's the coil or deck. The mod is going to calculate everything from resistance read. If you could fake the deck resistance to 0, it would still not run the same as the resistance is still actually higher, so all the info on the display will be wrong.

 

[papergoblin]

Sorry for back 2 back posts.

Check to make sure some wire isn't inside your posts/built up into the threads, that with long leads could be some of the problem.

 

[GaryParr]

The right way to go about it is to not use a faulty deck and expect good results. Any way you try to compensate will be a band aid that will, at most, only be accurate within a very small temperature/resistance range. By changing the tcr you're telling the mod that you're using a different material than you are.

Bah, who wants to do things the right way? Where is the fun in that?

Honestly though, I do agree the deck should probably not be used for TC. That said, I'm curious to find a reasonable method to work around this type of issue in less extreme instances.

Does the deck show resistance with no coil(s) on it? If the deck shows resistance when blank, it's bad and needs insulators replaced or stop using, cause eventually you're gonna get a dead short.

Many of my attys show a different resistance with the same coils but that's due primarily to leg length. If you're set on using both, then build the one reading higher with a lower coil(s). If you usually use say 5 wraps, do 4 wraps. Resistance is resistance, doesn't matter to the mod if it's the coil or deck. The mod is going to calculate everything from resistance read. If you could fake the deck resistance to 0, it would still not run the same as the resistance is still actually higher, so all the info on the display will be wrong.

Sorry for back 2 back posts.

Check to make sure some wire isn't inside your posts/built up into the threads, that with long leads could be some of the problem.

It doesn't register an atty when the coils are off, so not worried about a short. Good thing to ask though.

As for dropping resistance in the problem deck, even if I wrapped less and got the TFV4 down to the 0.180 of the Kylin, 0.065 of that would be coming from the deck itself, meaning the coil was only providing 0.115 Ohms. I'd still have the same issue... the mod would be looking for an x% change in resistance but since the deck resistance is static, the coil itself has to change more than it should for the mod to register the expected reading. This is why I'm thinking the TCR could be adjusted to compensate.

Good point about looking for buildup in the post themselves. I assume you are talking about where the coil gets mashed against the post by the set screw. I'm honestly not sure how carefully I've looked at those between coil changes. I'll be sure to examine them closely on the next cleaning.

 

[ShowerHead]

Are the tanks stainless or black/colored?
I had a few tanks that were black where the finish was thoughtfully applied to the deck area.
The finish was enough to throw resistance off by varying amounts at varying times.
I no longer have the decks.

Which is my advice to you.
TC is nice, all I use. It is also extremely sensitive to resistance. I've had lousy results when the resistance is off by 0.02 ohms. Oddly enough, the 'worse' a mod is at performing TC well, the better it will do with shaky tanks.

If you just HAVE to use the tank, then change the resistance manually and lock it. At least you will have to do it every time you change tanks and it won't carry over to a good tank like it would if you change TCR values to compensate.

 

[GaryParr]

Are the tanks stainless or black/colored?
I had a few tanks that were black where the finish was thoughtfully applied to the deck area.
The finish was enough to throw resistance off by varying amounts at varying times.
I no longer have the decks.

Which is my advice to you.
TC is nice, all I use. It is also extremely sensitive to resistance. I've had lousy results when the resistance is off by 0.02 ohms. Oddly enough, the 'worse' a mod is at performing TC well, the better it will do with shaky tanks.

If you just HAVE to use the tank, then change the resistance manually and lock it. At least you will have to do it every time you change tanks and it won't carry over to a good tank like it would if you change TCR values to compensate.

The tops of the decks, connections, and posts are free from any anodizing so it is something else causing the resistance. And yes, I completely agree that this "problem" deck should not be used for TC. I'm more curious to find a reasonable method to handle a situation like this... where there is some resistance built into the system which throws off the temp control.

Changing the resistance manually appears to be the common suggestion, but it seems the wrong approach. I would have to set the resistance to some unknown value so that when the coil reached the right temp, the % change in resistance of the coil + the static resistance of the deck = the % change in resistance the mod was expecting. So, if the coil was really only .180 ohms at room temp and supposed to be .237 @ 460F then the expected change to reach that temp was a roughly 32% rise in resistance. Since the deck adds 0.065 to every reading the mod takes, I'm shooting for the mod to recognize that .302 is the correct ohm reading @ 460F which, working backwards, means I have to set a starting resistance of roughly 0.229. Unless, of course, I want to drop the target temp to 420F at which point I have to re-work the calculations again to get the new starting reading I'd need to lock in.

That's why I think setting the TCR is a better option. Especially because on this mod, changing the TCR is no more complicated than changing the resistance. By tweaking the TCR, you alter the % change in resistance the mod is expecting so that the static resistance from the deck is accounted for. If the room temp reading is .245 (0.180 from coil, 0.065 from the deck) and the target reading @460F is .302 (0.237 from coil, still 0.065 from the deck) then that works out to a 23.26% change in resistance. I just have no clue how to work out what that difference translates to in the required adjustment to the SS 430 TCR of 1380.

 

[ShowerHead]

That's why I think setting the TCR is a better option. Especially because on this mod, changing the TCR is no more complicated than changing the resistance. By tweaking the TCR, you alter the % change in resistance the mod is expecting so that the static resistance from the deck is accounted for. If the room temp reading is .245 (0.180 from coil, 0.065 from the deck) and the target reading @460F is .302 (0.237 from coil, still 0.065 from the deck) then that works out to a 23.26% change in resistance. I just have no clue how to work out what that difference translates to in the required adjustment to the SS 430 TCR of 1380.

Keeping in mind that TCR is a curve that starts at room temperature, I still think it won't be a good experience.
But, if I were to experiment, I'd start with taking 23% off 138 and add a zero to the end of the result. So, you'd have about 106 then add the zero for 1060.

Here's what I had to do for Crazy Wire SS430 before I binned it.
"Temperature (degF)","Electrical Resistivity"
-100,0.815199971199036
68,1
800,1.80519998073578

Here's what actual SS430 from zivipf looks like.
"Temperature (degF)","Electrical Resistivity"
-58.0000,0.9034000000
68.0000,1.0000000000
212.0000,1.1104000000
302.0000,1.1794000000
392.0000,1.2484000000
482.0000,1.3174000000
572.0000,1.3864000000
800.0000,1.5612000000

For the heck of it, here's SS316L
"Temperature (degF)","Electrical Resistivity"
-58.0000,0.9230640000
68.0000,1.0000000000
212.0000,1.0800000000
302.0000,1.1260000000
392.0000,1.1680000000
482.0000,1.2070000000
572.0000,1.2460000000
800.0000,1.3369666667

 

[Coastal Cowboy]

The temperature control algorithm in your chip is just a model of a real world relationship between a dependent variable (temperature) and and independent variable (resistance). You have to calibrate your model in order to make it behave like the real world relationship it simulates.

There are two ways of calibrating a simple model with one dependent variable. One way is to change the intercept with the Y axis. The other is to change the slope of the curve.

For our purposes, the intercept point is the cold resistance of the build and the slope is determined by the coefficient--i.e., temperature coefficient of resistance.

If you're confident that you're using the correct TCR for the material in your build, and that the material will perform as predicted, then you should calibrate the model by changing the intercept. Yeah... fudge on the cold resistance to account for whtever garbage the deck is contributing.

Why? Because the slope of the curve isn't going to change. It doesn't matter if you start at 50 or start at 100 if the rate of change between 1 and 500 is the same from start to finish.

 

[GaryParr]

Keeping in mind that TCR is a curve that starts at room temperature, I still think it won't be a good experience.
But, if I were to experiment, I'd start with taking 23% off 138 and add a zero to the end of the result. So, you'd have about 106 then add the zero for 1060.

Here's what I had to do for Crazy Wire SS430 before I binned it.
"Temperature (degF)","Electrical Resistivity"
-100,0.815199971199036
68,1
800,1.80519998073578

Here's what actual SS430 from zivipf looks like.
"Temperature (degF)","Electrical Resistivity"
-58.0000,0.9034000000
68.0000,1.0000000000
212.0000,1.1104000000
302.0000,1.1794000000
392.0000,1.2484000000
482.0000,1.3174000000
572.0000,1.3864000000
800.0000,1.5612000000

For the heck of it, here's SS316L
"Temperature (degF)","Electrical Resistivity"
-58.0000,0.9230640000
68.0000,1.0000000000
212.0000,1.0800000000
302.0000,1.1260000000
392.0000,1.1680000000
482.0000,1.2070000000
572.0000,1.2460000000
800.0000,1.3369666667

Can I ask how you arrived at 23% to take off 138?

The temperature control algorithm in your chip is just a model of a real world relationship between a dependent variable (temperature) and and independent variable (resistance). You have to calibrate your model in order to make it behave like the real world relationship it simulates.

There are two ways of calibrating a simple model with one dependent variable. One way is to change the intercept with the Y axis. The other is to change the slope of the curve.

For our purposes, the intercept point is the cold resistance of the build and the slope is determined by the coefficient--i.e., temperature coefficient of resistance.

If you're confident that you're using the correct TCR for the material in your build, and that the material will perform as predicted, then you should calibrate the model by changing the intercept. Yeah... fudge on the cold resistance to account for whtever garbage the deck is contributing.

Why? Because the slope of the curve isn't going to change. It doesn't matter if you start at 50 or start at 100 if the rate of change between 1 and 500 is the same from start to finish.

Correct, BUT... the percent change is different. 50 to 100 is a 100% increase for 50 units of value, whereas 75 to 125 is a 66% change for the same 50 units of value. I guess it all depends on a mod determines change... is it based on amount of value or percent of value?

 

[GaryParr]

I suppose this would all make more sense if we can visualize what I think is going on. In the table below, I took the temp and resistance factor values for 480 SS off steam engine. I then used my coil Ohm as a starting point to calculate the actual Ohms that the mod would see at the given temps. I then calculated the Ohm difference from step to step and the % change from base (same as the resistance factor). I then took the calculated coil ohm readings and added in the constant resistance from the deck to get the "system" values.

The graph on the left is what, I think, everyone is telling me is going on. The mod sees a higher resistance because of the deck, but at the end of the day it doesn't matter because the resistance sill rises the same... just starting from a different point.

The graph on the right is what, I believe, really matters and why the graph on the left is irrelevant. The mod uses TCR to calculate the % change of resistance needed to hit a target temperature. And when you add a static, non-changing value to your reading, it throws off the percent change.

The mod uses TCR to figure out what % change is needed, starting from the base resistance at 68F, to hit the target temp. So, if I set my mod (as it is now) to 392F, it multiples the starting resistance of the system .245 by 1.2484 and ends up with 0.305858. Using the table above, that puts the actual temperature of the system at somewhere around 490F because the mod will not read 0.305858 Ohms until the coil is contributing 0.240858 of that value.

 

[Coastal Cowboy]

I suppose this would all make more sense if we can visualize what I think is going on. In the table below, I took the temp and resistance factor values for 480 SS off steam engine. I then used my coil Ohm as a starting point to calculate the actual Ohms that the mod would see at the given temps. I then calculated the Ohm difference from step to step and the % change from base (same as the resistance factor). I then took the calculated coil ohm readings and added in the constant resistance from the deck to get the "system" values.

The graph on the left is what, I think, everyone is telling me is going on. The mod sees a higher resistance because of the deck, but at the end of the day it doesn't matter because the resistance sill rises the same... just starting from a different point.

The graph on the right is what, I believe, really matters and why the graph on the left is irrelevant. The mod uses TCR to calculate the % change of resistance needed to hit a target temperature. And when you add a static, non-changing value to your reading, it throws off the percent change.

The mod uses TCR to figure out what % change is needed, starting from the base resistance at 68F, to hit the target temp. So, if I set my mod (as it is now) to 392F, it multiples the starting resistance of the system .245 by 1.2484 and ends up with 0.305858. Using the table above, that puts the actual temperature of the system at somewhere around 490F because the mod will not read 0.305858 Ohms until the coil is contributing 0.240858 of that value.

The graph on the left is what the algorithm in your chip is doing. TCR is a fixed value that describes the slope of a curve. Once you lock a starting resistance and set the TCR, the chart on the right doesn't happen.

 

[GaryParr]

The graph on the left is what the algorithm in your chip is doing. TCR is a fixed value that describes the slope of a curve. Once you lock a starting resistance and set the TCR, the chart on the right doesn't happen.

I have to disagree with you on this. At least, I think I have to.

First, can we agree that the values in my table are correct? If so, then tell me where my logic breaks down in these steps.

1. Mod reads starting ohm of 0.245 (0.18 from coil, 0.065 from deck)
2. I set my mod to a temp of 392F
3. Mod uses TCR to calculate target resistance of 0.305858 (0.245 X 1.2484)
4. Mod applies power until resistance reaches 0.305858

If all of that happens as I describe, then I'm hitting a temperature well over 392F because if we take that target resistance and subtract the constant from the deck (0.065) then we see the coil is actually at a resistance of 0.240858 which it doesn't hit until a temperature around 490F.

I'm quite open to the possibility that I'm wrong, but I'd really appreciate if you can tell me how.

 

[GaryParr]

Let me add something to that first chart.

When I created it originally, I used the coil ohms of 0.18 as the starting point. Each point on that graph is the calculated resistance needed to hit the required temp as represented by the grey line. The green line is just each one of those points with the constant deck resistance added to it. That's why the curve matches. It is what everyone seems to think is happening.

I've added a new line to this, the black one. This is the same calculation used to generate the grey line, except I start with the system resistance of 0.245. Now, the difference in slope becomes evident.

 

[GaryParr]

BAZINGA!

Putting all of this down in tables and graphs led me to the answer.

The Coil Ohm column is what the coil needs to hit in resistance for the given temp. The Mod Calc Ohm column is what the mod thinks the resistance needs to be based on the starting system (system means 0.18 coil + 0.065 deck) resistance. What I want the mod to do is calculate the System Ohm column since this just coil resistance plus the deck constant.

Google tells me that TCR is calculated using temp in C as:

TCR = ( R2 - R1 ) / ( ( R1 * (T2 - T1 )
​

So, I added Temp in C to the chart and then just did the math.

( 0.264872 - 0.0245 ) / ( ( 0.0245 * (100 - 20) ) = 0.001014
​

This seems to work out exactly for all of the temps. I can verify the TCR calc by using the Coil Ohm column instead of System Ohm and it gives me 0.00138, which is the known TCR for SS 430.

 

[Coastal Cowboy]

Let me add something to that first chart.

When I created it originally, I used the coil ohms of 0.18 as the starting point. Each point on that graph is the calculated resistance needed to hit the required temp as represented by the grey line. The green line is just each one of those points with the constant deck resistance added to it. That's why the curve matches. It is what everyone seems to think is happening.

I've added a new line to this, the black one. This is the same calculation used to generate the grey line, except I start with the system resistance of 0.245. Now, the difference in slope becomes evident.

Deck resistance is not dependent on temperature. It's constant at every point across the curve. In fact, there should be ZERO or minimal resistance contributed by the deck. If there is more than minimal resistance from the deck, the way to compensate is to increase the starting resistance.

The only way to generate the black line is to make the deck resistance plus coil resistance temperature dependent as a whole system and develop your own TCR. If you think that's what's happening, then you've proven that the RTA deck is faulty and you should follow @KenD's advice. Either fix what's wrong with it, or get rid of it.