› ›

Dead Time parameter has no influence at all

icefireicefire
in Extruder & Bed
So I'm using a RAMPS 1.4 board with Repetier Firmware 0.91 and Repetier Host. I tried using dead time control for the hot end because PID ist more difficult to set up. However, the dead time parameter (Extr. 1 PID P/Gain) has no influence on the temp curve at all.

I tried with 0, 5, 7 and 11 seconds. The curve always looks like this:

image 
So what am I doing wrong?

Thanks :)

Comments

  • RepetierRepetier
    The curve until control range is always full power for all heat managers. Only in the last 20° it starts kicking in. From your graph I see your dead time is around 4 seconds. If you set this, you will see it disabling power a few seconds before you reach target temperature and then it only goes on in spikes for a few seconds. For testing the max pwm you need target temperature as 120°C does not cool down as fast as at 210°C.
  • icefireicefire
    edited November 2015
    Thanks.

    So I set the dead time to 3,75s (that would be the result of a precise graph analysis). Setting the target temp at 120°, waiting a few seconds and setting a new target of 240° results in a graph like this:

    image

    I have no idea what happens at 53:45 (maybe related to draught in the room??). However, it seems to work fine now.

    Is there any point in setting up PID control? I mean, the temp control is more sophisticated but the temp will inevitably overshoot before reaching the target. On the other hand, with dead time control I never seem to reach the exact target value (temp oscillates between -1° and -0,25° of the target temp, see graph). 

    image


    Is there any way to combine dead time control with PID control? That would be really great..

     
  • RepetierRepetier
    Ignore 53:45 it is from the algorithm where control range starts.

    PID can adjust a bit better to different loads. If you print faster with dead time control you will go lower a bit. With dead control you also need to set the pid max pwm so that cooling and heating are nearly as fast and there is still enough reserve. Then you get best results. Since 1-2°C more or less are no serious problem it is a easy solution with no big overshoots that adapts rapidly to new loads.
Sign In or Register to comment.