Strange behavior / Eleaf Invoke / AF 190624

Discussion in 'Issues' started by Tastenkrueppel, Apr 24, 2020.

  1. Tastenkrueppel

    Tastenkrueppel New Member

    Joined:
    Aug 16, 2018
    Messages:
    13
    Likes Received:
    1
    Greetings vapelings,

    i lately encountered some kind of strange behavior that led to bad vaping performance. The setup is an Eleaf Invoke on Arcticfox 190624 with a Wotofo Serpent Elevate equipped with an SS316L three-core fused clapton at about 0.2 ohms of resistance (which worked perfectly before i cleaned up the coil).

    While examining the device monitor i came up to the following:

    * As soon as the fire button is pressed, the cold resistance rises from 0.207 ohms to 0.218 ohms (maybe due the Shunt correction factor of 0.97?)
    * There is a sudden peak in temperature at about the moment when the temperature reaches its "target" value...
    * ... followed by a pretty massive drop in power (which seems logical after such a peak) and a resulting poor performance
    * when the fire button is released, the temperature is constant for about a second, drops suddenly and cools further down at a "normal" rate

    I recorded this finding in a .CSV and attached it (renamed) as Invoke_Pufflog_Elevate.txt

    Afterwards i reflashed AF three times ("my" version with different font and numbers, the original AF and back again to "my" version) -> no change in that behavior after any of the flashes.

    Afterwards i checked if one of my picos did anything similar with an Ammit 22 equipped with the same wire but with a higher resistance due to a bigger diameter and one more wrap -> nope, pico/Ammit works fine.

    Lastly i screwed the Ammit onto the Invoke. The Ammit has its own profile on both pico and Invoke so i switched to that profile beforehand. Result: the Ammit behaved similarly to the Serpent Elevate: Fire -> temperature rises -> peak -> power drop -> poor performance.

    I gave up, screwed the Ammit back to the pico, the Elevate onto the Invoke, switched back to the Elevate profile on the Invoke and decied to drop a report on here. Guess what happened next... now the Elevate behaves perfectly normal on the Invoke. Recorded this too (Invoke_Pufflog_Elevate2.txt, also a renamed .CSV).

    Well, my problem is solved now, but seriously...: WTF happened there? It came totally out of the blue (for me) and made me change the cotton in the Elevate because i thought it was done (it was actually, kind of). And it "repaired" itself by swithing the profile.

    Maybe you guys find something with the provided information while i hope i can "fix" strange things by switching profiles :-D

    Best regards

    Christian
     

    Attached Files:

  2. Tastenkrueppel

    Tastenkrueppel New Member

    Joined:
    Aug 16, 2018
    Messages:
    13
    Likes Received:
    1
    Maybe i should add that i'm using custom material curves downloaded from Steam Engine and a preheat curves that both worked perfectly "for ages" (on the mentioned Invoke, 3 classic picos, 2 pico RDTAs, 1 pico Squeeze, 1 RX mini and 1 Cuboid).
     
  3. maelstrom2001

    maelstrom2001 Developer
    NFE Team

    Joined:
    May 4, 2017
    Messages:
    1,288
    Likes Received:
    1,174
    Hello!
    It seems that the problem was in the saved "cold" resistance which was incorrect. Some boxes can not measure resistance of screwed coil precisely, which may lead to such TC behavior. The best way to fix cold resistance saved in profile is:
    • screw cold atomizer (coil temperature should be equal to ambient);
    • let it settle for a couple of seconds;
    • start Device Monitor and see Live Resistance chart;
    • set cold resistance in profile equal to Live Resistance measurements;
    • lock cold resistance.
    Another workaround without Device Monitor:
    • screw cold atomizer;
    • go to Main Menu -> Profile, select Coil item and hold Fire button to force measure coil resistance;
    • exit menu and lock resistance.
     
  4. Tastenkrueppel

    Tastenkrueppel New Member

    Joined:
    Aug 16, 2018
    Messages:
    13
    Likes Received:
    1
    Hi,

    thanks for your reply, but the second part of your answer is basically what i use to do. First of all, the resistance is always locked, because i know how utterly important it is, especially with SS-coils below 0.2 ohms. I build a new coil or rewick it, pour liquid on it and watch the live resistance on the device (which moves up and down in a range of about 0.003 ohms and if it's significantly more, i check back, if the coil is screwed correctly). I'll go to the coil menu then and keep the fire button pressed to fetch the resistance, especially when dealing with new coils. I reenter the menu to check back and mostly lower the fetched resistance by 0.001 or 0.002 ohms to what seems best according to the watched live resistance before. This procedure has always worked well. I also use to check the cold resistance after letting the device stand on the table over night to be fully sure that the coil is at room temperature. Sometimes i lower the resistance by another 0.001 or 0.002 ohms, but that's not always necessary.

    The point is: why does the cold resistance "change" as soon as the fire button is pressed (you can see this in the CSV-texts)? It's locked after all. In the meantime i checked a couple of devices (the Invoke, two of the picos, the RX Mini and the Cuboid) and i've seen this occur on every single one. Mostly the differences were not more than 0.003 ohms, but still present.

    Puzzled greetings
     
  5. maelstrom2001

    maelstrom2001 Developer
    NFE Team

    Joined:
    May 4, 2017
    Messages:
    1,288
    Likes Received:
    1,174
    Nevermind it. For some time it was implemented in debug purposes. For real coil temperature calculations, a blocked/cold resistance is still used.