Replies

              • Developer

                Hi Peter,

                Standard logs are fine for any auto-tune attempt that didn't work.

            • T3

              Leonard,

              Any chance you can take a look at the logs from my autotune from Friday? I've tried multiple times and couldn't get it working at all and am stumped. I have it tuned alright but it would be nice to compare values to what autotune finds.

  • T3

    I have the same issue as Papayou.

    I'm flying a TBS Disco on 4S, DJI 920kV motors, DJI 30A Opto ESC, DJI 10x4.5 props, Pixhawk, GoPro Hero 2, FatShark 250mW 5.8GHz and a FatShark FPV camera. Using an X6R for the receiver and also have a Teensy 3.1 set up with my 3DR telemetry to get telemetry back to a ground station and my Taranis.

    I've done my manual tuning and things seemed to be in a good state so I saved those settings. Then I tried auto-tune while in AltHold and it starts off fine moving in the roll direction, granted it's moving off in one direction, and then after a while, I bring it back and let it keep going. Then it just stops without going through to the pitch sequence. I flip the switch back to LOW to get original PIDs and then back to HIGH and land, arm, and read the parameters but they're the same, so I assume it didn't complete as Randy indicated.

    I don't have logs right now but can upload them later tonight but is what Randy says true that if it can't find a good PID setting for roll, it stops? That would be useful to have in the Wiki at the very least. I have the GoPro video of it trying as well.

    I'll get logs up tonight and will also be sure to check the CG to make sure it's alright but if I remember correctly, it's pretty close and if anything, is only off in pitch if it's off at all.

    Looking at vibration as well, I found I'm within +/- 1 for Y but it's more like +/- 1.5 for X and am not sure what else to do there. It's still in the +/- 3 that I saw in the manual so I haven't worried about it.

    • T3

      Okay, here are the logs.

      Log 1

      Log 2

      I also just checked the CG and it looks like it wasn't balanced front to back properly but it's pretty close side to side so I'm not sure why it wouldn't get the PIDs properly for roll and move on to pitch.

      Dropbox - Error
      Dropbox is a free service that lets you bring your photos, docs, and videos anywhere and share them easily. Never email yourself a file again!
      • Based on what I can see:

        http://imgur.com/a/Tcgwi

        yt0T9vn.png?width=500ScL4Om6.png?width=500VS1FZjf.png?width=500

        Step 0 of AUtotune lowers your P value until 0.01 would be hit, Autotune fails therefore, try lowering D by /2 and retry.

        • T3

          Thanks again!

          I decided to go back to stock PID settings and start from there. I had it tuned to what I thought was pretty good and save a backup of all the parameters and then reset the PID values. Once I did that, Auto Tune worked just fine and it went through the whole process and was great.

          I went out earlier when it looked calm and tried there. Big mistake. I had a decent headwind and went ahead just to see what would happen. It took an entire pack for the process to go through but it went through both roll then pitch and finished. The PIDs were HORRIBLE but it was in the wind so what could I expect?

          Then this evening, I was doing a Save Trim procedure and also trying to get another parameter worked out and once that was done, I took another fresh battery and tried Auto Tune again in no wind. It did its magic and once it finished, I found it was EXPONENTIALLY BETTER than what I thought I had done with my settings. My stabilize roll/pitch P value jumped from 4.5/3.5 to 10.5/10.5, which was a surprise, and then it adjusted my rate PID values for roll and pitch after that and man, it's awesome. I wasn't sure what the phrase "flying on rails" really meant until after this but now I know. It's very impressive and I'm very happy I did it right.

          Thanks again for your suggestion. I'm sure lowering the D value by half would have helped but I just figured I would start at the default first since it was an easy enough change and went from there.

          • Developer

            Great to hear healthyfatboy!!

            And thanks  Apfel Korn for all the help you are giving the community!!!

        • T3

          Excellent! Thank you for answering and taking a look.

          Do you think it would be worth just resetting the PID values to default just to see what it comes up with? I already have the current PIDs saved so it wouldn't be difficult to do.

          Either that or I'll also try lowering my D term as well. I assume it's just the Roll D term that needs to be cut in half?

  • Just finished a perfect autotune and my copter is flying on wires! I would like to say a major thanks to the Arducopter Devs!

  • Hi, 

    I try AutoTune on a Tricopter x900 from HK and APM 2.6 runing on FW 3.1.5

    setup is:

    RcTimer 2812-880KV Brushless Motor

    ESC 30A Simonk Firmware

    10 x 4.5 Gem Props

    Dynam 2600mA 4C

    Ublox Neo 6m with compass

    The problem is that the copter increse the throttlr from 532 to 800 just a few seconds after I engage autotune.

    in steps:

    Take off

    Switch to AltHold

    Engage AutoTune

    now the copter rolls a few times left and right and all a sudden it raises up in the air.

    switched to Stabilize mode, gained controll

    back to about 6m alridude

    switch to AltHold

    engage AutoTune

    same thing, rolls a few times than raises up again.

    in the log file you can see this starting with line 1359 CTU thrin=532 out is 535

    after that Throut rises to exactly 800 and the copter starts to gain altitude until I switch to stabilize mode to cancel autotune.

    Things I did so far:

    re flashed Firmware (with erase and reset) 

    changed Flight controller from other quad I have

    I have no idea what I'm doing wrong or is wrong. Please any ideas are welcome. attached one log for reference.

    The copter fly's well in Althold and Loiter

    cheers

    2014-07-02 13-40-03.log

    https://storage.ning.com/topology/rest/1.0/file/get/3701772439?profile=original
This reply was deleted.

Activity