hot end temperature oscillations

This is on a kossel variant using Johann's modified marlin for deltabots. Im using a authentic j-head hotend with a .5 nozzle with the resistor that came with it. When it shifts to the second layer and speeds up the temperature starts oscillating back and forth 10-15C. This happens even when I dont drop the temp on the sequential layers as well. I have tried numerous things and still the issue is there so im not sure if anyone has any pointers as to where to check. Also my heater power does not show up in repetier unless im auto tuning the PID as shown below. I think thats just a setting in the firmware but im unsure where to look. I attached my config.h file so maybe someone can look it over and see if I have some strange setting in it. .

I have tried the following:
-auto tuned PID numerous times
-cartridge heater
-several thermistors
-new power supply
-ran shielded wire all the way back to the board
-routed all power wires/stepper wires/ etc away from thermistor inputs
-tried the other spare thermistor input on my azteeg x3
-insulating the hotend

Well I had a problem very similar to this about a year back. I ended up having to replace the hotend because I was getting some cross-talk between the heater and the temp.sensor. I sure hope you dont have the same issue.

Look at this and see if it sounds similar to your issue.

-Nick

It does sound somewhat similar but mine seems to be speed related. If I run it real slow its fine but I shouldn't be going too fast. I was only printing 65mm/s with abs. That shouldn't be a issue on a jhead but I may just order a e3d and try that.

If you get oscillation at steady state your P value is too high. Also, most firmware abandons PID control when you are outside of 10 deg C band of target.

Basic rules for PID controllers
1. Adjust the proportional term to get fast response, but you will get oscillation at steady state.
2. Adjust the integral term to get close to zero steady state error (getting rid of oscillation), but you will get overshoot.
3. Adjust the derivative term to lower overshoot.

This is an iterative process. For auto tune to work you'd have to run it while extruding at the target speed.

Is there any way to run a autotune at the desired speed besides pushing filament in manually? I have just been using the m303 command at the printing temp. This is how my solidoodle was calibrated and it has no trouble with oscillations.

You'd have to modify the firmware to use autotune while extruding. Try using the Ziegler Nichols method for PID parameter selection.

[www.ni.com]

In short, set I and D to 0. Increase P to get the system to oscillate around the target value. (Done - Your system oscillates already). Note the P value and the frequency of oscillation. Then set the P, I and D parameters as specified in the Z-N table in the link above.

Edited 1 time(s). Last edit at 11/12/2013 10:41PM by Threegreens.

Try lowering your PID MAX. In my experience, 265 defualt is too high for the much more efficient cartridge heaters, resulting in ridiclous PID values. Try 150 and re-run autotune. If it still overshoots lower it more.

Edit: suasage fingers

Edited 1 time(s). Last edit at 11/13/2013 06:35AM by foshon.

I had temperature occillation problems on my Mk V Jhead. Typically with set point on 235 deg C, it would sometimes shoot up as high as 247 deg C and cut out when the max limit was set on 245 deg C.
I retuned it and widened the 10 deg C band in Marlin to 15 deg C which helped stop it overshooting as much. It did not completely eliminate the problem though. Final thing which solved my problem was fitting a small enclosed fan duct around the peek bit of the Jhead. My previous fan was was switching in and out with the block heater and causing temperature occillations. It is very stable now.

foshon I forgot to mention it but I have done testing running the PID_MAX down to 100 but it still oscillates. It just seemed to slow the initial heatup of the hot end.

dave3d I ran a fan on it but to no avail since it still oscillated. I ordered a new e3d to try as im thinking it was something faulty with the jhead I have even though it is an authentic one.

Also thinking more into it why does the heater power not display when normally printing but if I run the PID autotune it displays properly? Is there something in the settings that maybe I missed?

Ok hooked up the e3d hotend and the fluctuations are still there.....

Ive tried all the different thermistor and hot end inputs and still there as well. It seems it is a speed thing because the issue starts when the steppers start moving fast. I know I have enough power and I dont get any voltage drops at the PS. Is it possible that the steppers are somehow backfeeding into the 5v line and throwing the readings off? Or maybe the USB backfeeding? I do have the jumper set to power the board from the power supply. Should I try to use the usb 5v power instead? Also when my machine is not plugged in the steppers are hard to turn but when I plug it in they can be moves by hand with ease uness they are jogged with the software.

Just want to update everyone who was curious or helping out. I solved the problem which was something wrong in the firmware. I transferred my settings to the latest fork of marlin and can hold stable hotend temp now.

Looks like your PID loop is going into an unstable region. Try reading up on PID loop tuning...