1. Interested in a popular vape? Short on time? Be sure to check out the Best Of threads for Plug-Ins and Portables.
    Dismiss Notice
  2. What does SSTB mean? See our glossary of acronyms.
    Dismiss Notice

TC Box Mods Firmware Discussions and Customizing

Discussion in 'DIY' started by Pipes, Feb 5, 2017.

  1. Pipes

    Pipes Addicted DIY Enthusiast Accessory Maker

    Messages:
    1,836
    Location:
    Southern Ontario
    Hey Guys/Gals.
    Opening this thread to have a spot to discuss and pool our knowledge regarding best firmware options to alter existing box mods to better suit our vaping needs. There seems to be more and more 510 based vaporizers coming to market in our world here. The TC technology is still new for our application but has being close to perfected for the needs of ECig users.

    So far, the "myevic" open source firmware seems to be a big topic as has very complicated and powerful upgrades for user accessibility to setups and alterations. We need to take advantage of the firmware to make these devices do our bidding.

    [​IMG]

    The reason is they all have the same chip set and the code has been leaked out and now open source.
    This chip set has proven to work well for our purposes. The newest firmware even allows for PID alterations of the heat up algorithm. The customized firmware is in a file called myevic.bin and is being changed regularly. Advanced options like 15 second time out is a nice option. The same file works on all the above compatible mods.

    For information and menu instructions go here.

    For the actual file and some history go here.
    The way this firmware works well for the Project is:

    Relevant Firmware Settings:

    [​IMG][​IMG][​IMG][​IMG][​IMG]

    For realistic numbers up .56 ohm to .600 ohm for starting point and calibrate further later. Raising this number for hotter and lower for cooler. Turn Check to “off”. Zero all will force a re-read of actual ohms.

    Note that manual entered ohms must be within 10% of actual read ohms or settings will change back when used. Including the check to “on”.

    Don’t forget to turn the temp down to around 380F after you manually upped the ohms.


    [​IMG][​IMG]

    After you alter the ohms by adding .04 ohms, the TCR should be changed to 170 to help the temperature hold linearly true.

    [​IMG][​IMG][​IMG][​IMG][​IMG]

    Set up 15 second timeout. To avoid hot preheat turn Algo to PID and change numbers as shown. This setting prevents preheat and ensures minimal charring. If Algo is off, expect 1-2 seconds of preheat. Changing these numbers will also affect the overall temperature and re-doing ohms calibration will be necessary.

    [​IMG][​IMG]

    Access this setting if going to use NFE software.

    I also now own a couple DNA units and really, really love them.
    The DNA chipset seems to handle a broader range of builds in a more stable manor. Maybe it's the sample rate or preset PID might have a more intelligent algorithm..? Not clear but the monitor graphing is generally spot on when using the software to see whats happening.
    What makes it really cool is the way you alter the TCR number. Not restricted to just altering the slope from room temperature to the end of the scale. But can change two spots along the curve to change the linearity of the curve. In fact the change is done in a dragging the points in a visual graph. The preheat is adjustable by stating how much, and for how long in drop down boxes. All done within Escribe software. Check out the demo.

    Now the "myevic" can be adjusted to be very close in performance, it's the PID to match your specific build is the key. Playing with the PID (Proportional, Integral, Derivative) can be a bitch but fruitful when you get it right.
    I do believe the EVic stock had the preheat a little high but also remember it seemed different with each firmware release. As with TCR presets.

    Here is a screen shot from inside the Escribe software. I set up a material called Project and adjusted the points to get the result I wanted. Altering the first two point adjusted the offset for scaling. Adjusting the third number is the critical one. It's how the temp changes from room temperature to the upper temp you ever plan to use. It this case 450F. The last point is set a minimal difference from the max usable to prevent the unit from running the temp a way too hot. Gotta love that part.

    [​IMG]

    The DNA devices have come down to reasonable prices now so don't be shy. Also, the material files are of course transferable.

    Fun Stuff, let get the discussion going...

    :sherlock:
     
    Last edited: Feb 6, 2017
  2. herbivore21

    herbivore21 Well-Known Member

    Messages:
    3,936
    @Pipes
    Great info brother, but if I may ask - will this work with any and every oil cart we might use? Are there any exceptions? I'd love to try this with my Puffco+ atomizer but I've never heard of anyone having success with these for TC use so far. I'd love to give this a try if it will work :D
     
    Copacetic likes this.
  3. Pipes

    Pipes Addicted DIY Enthusiast Accessory Maker

    Messages:
    1,836
    Location:
    Southern Ontario
    It would depend on the heating element material. The material needs to have a reasonable temperature coefficient. Stainless steel is very popular it seems these days. Materials like Kanthal do not have enough resistance change to work with.
     
  4. nosmoking

    nosmoking Fogmonger

    Messages:
    965
    Location:
    DC Area
    Try reading up here...also keep reading out that thread to learn a little more about your Puffco Plus atty. Feel free to reach out to me if you need any help setting up your Plus atty in TC mode.
    http://fuckcombustion.com/threads/h...-you-actually-loved.23182/page-4#post-1076778
     
    Pipes and herbivore21 like this.
  5. herbivore21

    herbivore21 Well-Known Member

    Messages:
    3,936
    That info doesn't seem to be forthcoming... I wonder if I can hit up someone at Puffco? :)

    IMO medical stainless steel makes the most sense, being a metal that's going to cope well with repeated heat cycles without dangerous decomposition.

    EDIT: Just saw the post above which seems to have the details required! Ok I need to try this out! Thanks @nosmoking I stopped reading that thread a while back and wouldn't have noticed that post otherwise!
     
    Last edited: Feb 6, 2017
    nosmoking likes this.
  6. ander

    ander highfromthemorning

    Messages:
    328
    Location:
    mitteleuropa
    I was already using the last version of myevic firmware (@m0qu4, also thanks for the "back" menu button! Small things but...) and never dealt yet with the PID... until today. Am I wrong or it wonderfully stabilize the coil glowing..?! My temperature oscillations were of 50 - 80°C, now are about 5-10 degrees.
    I'm doing some glowing tests at around 240°C, and it works very well but I've noticed that if the coil starts from cold, the temp first goes up to 270°C and then it stabilizes around 240-245 °C. If possible, how to deal with this peak..? (edit-> seems a constant peak of 30°C)
     
    Last edited: Feb 9, 2017
    KeroZen and Pipes like this.
  7. Pipes

    Pipes Addicted DIY Enthusiast Accessory Maker

    Messages:
    1,836
    Location:
    Southern Ontario
    Your into the fun stuff.
    First, ensure you have the boost turned off.
    To pretty much kill the preheat, try the P value up to 1600. Slows down the initial heat up but minimized the oscillations.
    However, I like the little burst to get the herb warmed up quicker. This is where preference comes into play and gives us choices.
    Gotta love it!
     
    WakeAndVape likes this.
  8. ander

    ander highfromthemorning

    Messages:
    328
    Location:
    mitteleuropa
    And on the path to perfect convection..?
    Good point!
    :bowdown:


    edit-> I can't see much difference in raising the P value over 1350. Low wattage seems to increase the peak duration so from 45w now I'm at 100w. Seems more stable this way.
    But enough with glowing tests... now it's the time to some more consistent tests... :nod:


    edit2-> Ok the higher P works, but with a lower I
    Perfect.
    The PID setting really changes everything. In the last days I was precisely wondering why... "why my girl's hairdryer has a stable resistance and my coil doesn't..?"
    Thanks guys. I can now finally start the fine tuning...:tup:
     
    Last edited: Feb 9, 2017
  9. KeroZen

    KeroZen Chronic vapaholic

    Messages:
    1,664
    Location:
    On Air
    Ha @Pipes, I wanted to create this thread the other day before going to bed but I felt asleep like a turd! :zzz:

    Anyways we don't care who did it, what matters is that it's finally there! Now the fun can begin! :rockon:

    So I would like to ask a question to our mighty readers: who's already been in touch with the developer? (I could use plural here, but last time I looked ClockSelect was pretty much the only committer in the repository)

    Do you think he would be willing to implement specific changes and features for our special application or not? Some e-cig users want to have nothing to do with our kind, for obvious reasons, so he might be against that.

    If the answer to that second question is no, who would be willing to participate in a fork of the codebase? We could have a special build specially crafted for our "deviant" use case maybe. Not that I have that many ideas for now though... We could at least have a build with preconfigured settings for the Project and/or with the stuff we don't need removed (some kind of idiot-proof working out of the box version if you will?)

    As I said in the DT thread, I gave a quick glance at the code and was a bit worried by the almost total lack of comments and the frequent use of magic numbers (ie numerical constants that are not #define or named const) The whole code is surely clear for its creator but not really meant to be shared or worked-on easily by others. I'm not saying it's outrageously complex either, but some parts don't seem necessarily obvious at first glance...

    At first I thought we could try to improve on that refresh rate front, reading the coil resistance more frequently for instance, but it turns out the PID algo is already running at 50Hz (unsure if the ADC reading is at that same rate but it should be) With PID's there's an upper limit for the refresh rate past which there's not much gain or worse it can make them unstable.

    That being said, a PID is not necessarily the most optimal solution in our case. I mean it's convenient, well documented and ubiquitous and al. but it's possible that other algos taking the asymmetrical aspect of our system into account (it heats up way faster than it cools down at first) could be better. Maybe at least something distinguishing the two parts (a- getting to temp fast without overshoot, b- holding temp during the draw) I don't know, so let's discuss!
     
  10. ander

    ander highfromthemorning

    Messages:
    328
    Location:
    mitteleuropa
    My scientific approach can be resumed in "Ouch! Very hot!"... so I can only like your proposals. And benefit of your discoveries..!

    Here my latest and really, really stable settings:
    Coil readed and locked at .575 ohms. 100w. TCR 190. Algo settings: P 1600, I 40, D 0.
    I can barely see the incandescence at 230°C and seems good right? Need tuning for sure, but the coil behaviour is so stable now!
    I don't have the same stability with a TCR 170 and a locked resistance of .600 ohms... that's strange or maybe... I don't know what I'm talking about..?
     
    Last edited: Feb 9, 2017
  11. funkyjunky

    funkyjunky moooo Manufacturer

    Messages:
    246
    hey!

    nice thread!

    just wanted to deposit the latest tubo PID settings, since they seem a bit different from others:
    P: 4000
    I: 600
    D: 10

    no real reasoning apart from feeling right. mind you the tubo has three glass tubes that absorb some heat and likely delay the system much more than with the project. i think that you be corrected with less I. i dont think the D part does a lot...wanted to have nonzero there :D
    the way these settings act are very fast temperature rise, overshoot of about 10degrees from cold, 5 or less when already warm. after the initial overshoot they stabilize extremely quickly. i think its because of high P, this makes the hole PID act very fast and aggressive.

    for the project id suggest high P but maybe not so high I as in my case.

    also concerning "calibration" of temperatures. im still using the ohms and tcr numbers to calibrate. i found that TCR 185 seems to be it. and interestingly i settle with ohms a bit lower than the mod reads initially, like 0.015ohm lower.
     
  12. Pipes

    Pipes Addicted DIY Enthusiast Accessory Maker

    Messages:
    1,836
    Location:
    Southern Ontario
    This is interesting indeed...? Maybe this is "before" a few power bursts to break in the coil?
     
    funkyjunky likes this.
  13. funkyjunky

    funkyjunky moooo Manufacturer

    Messages:
    246
    no its after its fully burned in and i calibrated it.
    i havent had time to measure temps with the project to compare. i just found that i dont hit the correct lower temps if i go higher ohms but lower tcr (which also matched high temps). i calibrate with two values, 150degC and 210degC.
     
    Pipes likes this.
  14. Hippie Dickie

    Hippie Dickie The Herbal Cube Manufacturer

    Messages:
    2,418
    Location:
    where the Cube rules!
    i use two sets of PID values - one set for over setpoint temp, one set for below setpoint temp ... more stable in my case.

    but i have a k-type thermocouple for temp reading, so, kind of like a fair witness, i don't have to assume the temp - i think trying to convert coil resistance changes to temp reading would be really hard to do, but maybe that's not the case.
     
  15. funkyjunky

    funkyjunky moooo Manufacturer

    Messages:
    246
    yeah its more or less the case. but its more like calibrating than guessing. since many conductors change resistance with temperature you can try to hold them at a certain resistance which results them being a certain temperature hot. then when you calibrate them you can correlate temps to those ohms.

    i bet you could do the same with another material probably (i dont know what you use) and a shunt resistor in series with your coil.

    in our system (project,tubo) the thing is we want to trick the software to display the temperatures not at the coil itself but at the load within a small distance. thats why we want to shift it, there is already an approach to just offset the value, which admittedly does make the most sense or you can trick the software by setting higher/lower resistance and higher/lower temperature coefficient of resistance(resistivity?)
     
    KeroZen likes this.
  16. Hippie Dickie

    Hippie Dickie The Herbal Cube Manufacturer

    Messages:
    2,418
    Location:
    where the Cube rules!
    i found (using BBQ thermometer) that the herb temperature rises to the coil temperature within about 5 seconds, so knowing the coil temperature is sufficient.

    also, from my readings, the coil temperature is affected by the coil geometry, rather than just the resistance ... while the k-type has a fixed geometry and well characterized temperature response.
     
    funkyjunky and KeroZen like this.
  17. ander

    ander highfromthemorning

    Messages:
    328
    Location:
    mitteleuropa
    exactly the same feeling...

    P 1750
    I 80
    D 10

    I tried to push the P higher but I feel it more stable around these numbers...
     
    Last edited: Feb 14, 2017
    funkyjunky likes this.
  18. ander

    ander highfromthemorning

    Messages:
    328
    Location:
    mitteleuropa
  19. ander

    ander highfromthemorning

    Messages:
    328
    Location:
    mitteleuropa
    Yesterday I started monitoring the activity of my Reuleaux, and with success... but today I made an upgrade from NFirmwareEditor-v6.4 to v6.5 and something seems wrong: if I launch the Toolbox and his monitor, all readings are out of scale... I can not verify since I've made a files overwrite...
    Since I simply can not find the 6.4 version anymore and anywhere, someone still have it and could share..?

    Talking about PID... I've made a reset and a progressive increase for first only of the P and then for I. The behaviour of D is still a mistery for me, especially in our application (and from my humble experience maybe completely useless?)... I Ended up with a P:600 and I:60, with an oscillation of 5-10 °C: temperature oscillates mostly under my chosen temp and I like that. m0qu4 acquired an oscillation of only 2-3 degrees... What's your..?
     
    Joaon and KeroZen like this.
  20. Joaon

    Joaon Well-Known Member

    Messages:
    131
    Location:
    France
    [​IMG]

    Last settings for Project : P=1850 I=200 D=10 , 190°C - TCR 185 - 75w - .600Ω Locked ,


    2 secs heat up to 190°c, and it stabilize itself well, and around 30w - 10A max.

    Still doesn't monitored the VTC mini, but definitively interested in !
    (any suggestion for monitoring on Mac? )

    :ninja:
     
    Last edited: Feb 19, 2017
    hardboiledfrog, Pipes and KeroZen like this.
  21. Hippie Dickie

    Hippie Dickie The Herbal Cube Manufacturer

    Messages:
    2,418
    Location:
    where the Cube rules!
    i think of D as the "toke detector" - the derivative signals a change is taking place, so it will be the first factor to counteract a temp drop from cooling due to taking a toke. so it moderates (damps) the effect of P, and eventually I - as shown in the animated GIF above.
     
    hardboiledfrog, ander and KeroZen like this.
  22. ander

    ander highfromthemorning

    Messages:
    328
    Location:
    mitteleuropa
    Found NFirmwareEditor-v6.0, and readings are ok again.

    Some interesting info with equal settings (algo excepted):

    algo off
    [​IMG]

    algo PID: P 550 - I 50 - D 0
    [​IMG]

    algo PID: P 550 - I 50 - D 10
    [​IMG]

    algo PID: P 1850 - I 200 - D 10 (Joaon settings)
    [​IMG]

    Note in graphs 2 and 3 the results of a Derivate of 10...
     
    Last edited: Feb 20, 2017
  23. Pipes

    Pipes Addicted DIY Enthusiast Accessory Maker

    Messages:
    1,836
    Location:
    Southern Ontario
    You've changed the PID numbers pretty dramatically from:
    [​IMG] I being using your numbers and they feel just perfect! In fact, I've updated the Eraser FAQ sheet to reflect this improvement.
    Almost looks the same but the front end power is more on the new settings without overshooting the temp. Very nice. Going to give the new PID a go!! :rockon:
    [​IMG]
     
    Last edited: Feb 20, 2017
    ander and Joaon like this.
  24. ander

    ander highfromthemorning

    Messages:
    328
    Location:
    mitteleuropa
    You know what's funny then..? Despite of oscillations... it seems that I have more constant results (in terms of AVB, taste...) with Algo OFF... :hmm: Incomprehensible for the moment.

    Edit: bah... bullshit... I've found my sweet spot again around 205°C. This thing is pretty sensible!!!
     
    Last edited: Feb 20, 2017
  25. KeroZen

    KeroZen Chronic vapaholic

    Messages:
    1,664
    Location:
    On Air
    I'm cross-posting from the DT thread, sorry:

    This "kicking out of TC" has been my plague, with the Project and with custom RDA's. Yeah most if not all the time there was a rational explanation of course, but when you get a mouthful of thick and smelly smoke, that's not what you want to hear! Made me curse a lot!

    I wonder if we could maybe 1) disable this safety in code altogether... dangerous? or 2) replace it by an error message and cut power simply.

    My vtc mini is acting up once again. It's the second time it does that: whatever I do and whatever I plug into (swapped RDA's with wildly different resistances) it reverts to power mode. It's like I can't use SS-TC or TCR anymore. It's not asking me whether I want to use the new coil or not either, it doesn't complain, it just switches back to power mode.

    I can select TCR mode in the menu and as soon as I press the trigger to confirm and exit, it reverts! Anyone else had this issue? I can make it stick to TC if no RDA is present, then as soon as I plug one in and hit the trigger it switches.

    EDIT: Last but not least I can't change the "check" option in the "coils" menu anymore, it's always on. And this menu is no longer updating/reading the resistance values in real-time or when I click... really weird.
     

Support FC, visit our trusted friends and sponsors