A warning to eCalc users

Posted by DG on June 4, 2015 at 9:28am

eCalc estimates for throttle @hover have grossly changed in recent versions at least for my KDE 2814XF-515 motors.

According to eCalc throttle @hover should be 65%. In reality it averages ~45% with the exact same configuration as shown below. When first considering building this quadcopter, eCalc was close to reality when using 13x5.5 APC props and 12000 mah battery. I clearly remember plugging the numbers in and being satisfied my motor/prop/weight combination would work and were well below 50%. 

I have contacted eCalc and their position at first was I must not be interpreting the flight data correctly. This is nonsense, as the flight logs clearly show Throut and Thrin both firmly below 50% @hover in stab mode. If eCalc data were correct, the copter would barely have enough power after takeoff. The pwm agrees.

eCalc says current draw is 6.07A * 4= 24.28A total. Reality is ~20A. On my Y6 eCalc says 36A. Reality is ~26. 

I don't know how many scenarios these errors apply to. This is just a cautionary note to double check eCalc before using it.

6dy3s5z.png

You need to be a member of diydrones to add comments!

Join diydrones

Email me when people reply –
Follow

Replies

  • Antonie Kruger June 9, 2015 at 7:40am

    Ouch...I've just built a Talon 820 hex with 6 KDE3510's and KDE ESC's. I ran the calc around a month ago. I did maiden today...around 3 minutes checking that all seems sound. I'll put her through her paces over the weekend an compare to my original ecalc sheet. My design current draw has changes as well and the mixed flight time has reduced by 30%. Cant wait for the weekend.

    • DG > Antonie Kruger June 9, 2015 at 11:50am

      what is your actual current draw at hover compared to eCalc?

      I have Frsky Lipo sensor that gives me total and per cell v on the Taranis. It is much better than the PM which as voltage drops the error increases telling me the total is say 14.5 when the Lipo sensor says 15. Battery failsafe wants to land the copter, but I know there's a lot more flight time left. I'm to the point of disabling the PM and use just the Frsky Lipo sensor, but then if there's a problem (flyway cough cough), it should have battery FS functioning. I wonder if there's a way to integrate the more accurate Lipo sensor into the FC. 

  • This reply was deleted.
    • K V June 8, 2015 at 6:53am

      Values are better than a couple of days ago but still not what i used to see a few weeks ago.

    • DG > K V June 8, 2015 at 10:22am
      Agreed, the numbers are still quite a bit off compared to when I originally did it.
    • K V > DG June 9, 2015 at 2:58am

      Also, I forgot to mention this but ecalc doesn't take the weight of the propellers in the calculations. If you are using sub 12" props, its not a big deal but 13-18" CF props add about 50gms each and when you go to 26-29" CF props, each adds about 100gms.They really have to add this to their calcs.

    • DG > K V June 9, 2015 at 11:37am

      I think you're expected to take everything into account when entering the frame weight (not including drive). For example, my quad is 390g, but I estimated the total at 590g (FC, GPS etc.) before eCalc does its thing.  It worked out quite well when eCalc was giving good throttle @hover estimates. Their current estimates were off though.

  • This reply was deleted.
    • DG June 7, 2015 at 1:45pm

      Now that eCalc is supposedly fixed, I'll compare the two.

  • OwlPic June 4, 2015 at 7:13pm

    There are several problems, the biggest is the prop data is taken in a static condition there the prop is just throwing air and a vacuum forms over the front of the prop. In contrast in the real-word is the prop screws it's way through fresh air, and the faster you go the harder (denser) the air become (as the square of the velocity), so with denser air the prop is more efficient.

    Many other consideration such as the electric motor torque variations when the wires heat up due to the change is copper resistivity with heat.

    • DG > OwlPic June 4, 2015 at 7:30pm

      The point is eCalc was pretty darn accurate on throttle @hover just a short few weeks ago. Now it is completely bonkers.  They've acknowledged the problem and we'll just have to wait and see.

    • Gary McCray > DG June 4, 2015 at 8:02pm

      Sorry to see this,

      I have been using Xcopter calc (Ecalac based) for some time to set up my quad copters and it has always been remarkably accurate.

      I also have a set of KDE 2814XF 515KV and it was certainly within a better than 10 percent error in every facet.

      I don't think the T Motors are generally as good as the KDEs and the KDEs definitely have a superior bearing stack.

      Better bigger bearings and 3 of them in most motors.

      The KDEs are also much better designed to withstand high heat and over current conditions than the T Motors.

      I am now trying some of the new KDE 2206XF 2050 motors on a small quad and they also appear to be superior.

      The only problem with KDE is that they lack a motor size in the 8 to 10" propeller class where the majority of good quadcopters on the planet now exist.

      No Phantom or Iris or Flamewheel size motors.

      The 2814 is best for slightly larger copters in the 12" to 14" prop sizes.

      BTW T Motor props are the best I have found (at least in the 10" - 14" sizes).

      I have used APC and Gemfan carbon filled and carbon fiber and they can't compare to the equivalent size T Motor props.

      The T Motor props are quieter, lower vibration, smoother flying and deliver the longest battery life.

      They are also seriously not cheap, but you do get what you pay for.

      Best Regards,

      Gary

This reply was deleted.