Heater Decoupled since updating to latest firmware

This is for general discussions. Also FAQ can be found here.

Moderator: speedake95

Post Reply
User avatar
gfeliksdal
Site Admin
Posts: 405
Joined: Sat Feb 25, 2012 10:40 pm
Location: Netherlands
Contact:

Re: Heater Decoupled since updating to latest firmware

Post by gfeliksdal »

Hi DDME-Marc,

So with the D value of 80 you have experienced a lot of errors? Did you by any chance see in the log area of repetier, what the error message was?

My observation with the 0.91 vs 0.92 firmware is that the controlled temperature with the 0.92 seemed a lot less noisy also temperature fluctuations are a lot better also for the whole termperature range. The D value for 0.91 was almost factor 10 higher. That seems to confirm your theory, that noise gets amplified by the large d-values.

I've tried the value 46, but imho it gave a too large overshoot when there was a big termperature change. From cold start or when with dual head going from idle temp to operating temp. But if that will be the solution, we have to tune the parameters in another way I guess.

The temperature signal is at the moment already filtered. Not sure how it exactly takes place. But it takes an average of a certain amount of measurement. Repetier also suggested that increased filtering might be the solution or also for the heater decoupled loop that there should be a certain amount of outliers before the error should be triggered. He would look into it if the current PID tuning wouldn't work.

I was hoping that the new values would be the solution, but it somehow still pops-up with some users.

User avatar
DDME-Marc
Posts: 71
Joined: Sun Nov 24, 2013 11:04 pm
Location: New Zealand

Post by DDME-Marc »

Hi Guillaume,
gfeliksdal wrote:So with the D value of 80 you have experienced a lot of errors? Did you by any chance see in the log area of repetier, what the error message was?
Firstly I never attempted printing parts with a D value of 80, although I did evaluate values up to 120 during simple filament purge tests. Ultimately my goal to keep P, I and D values close to the auto test values were based on developer discussions followed (such as in GitHub). I was aware that we already filtering signals, however I also got the understanding that minor over shoot was a trade off in avoiding noise. Concerns surrounding noise from D values for some developers even pushed them to try running PI controllers only (eg no derivative action to dampen temperature oscillations).

Once I got close to auto tune values I then proceeded to print 12 successive parts, over which I progressively push values even closer to auto tune results. To further complement tests I then complete a different 3 hour print with the settings below complete with decoupling time of only 8 seconds (now set back at 12).

http://forum.felixprinters.com/viewtopi ... 2&start=70

I also never went as far as to measure for noise with the 0.91 firmware because we simply do not have the facilities in the company. Fortunately all previous problems have been overcome, such as the opto sensors which were resolved by simple re-routing.

Although discussion of noise is theoretical, this is where I was hoping other users may be in a position to provide additional data/testing.

On the plus side I am thankfully still running fault free.

Cheers,

Marc
Felix 3.0 Dual Head * E3D Titan V6 * Repetier-Host V2.1.3 * Repetier-Firmware 0.92.9 (01/08/17 - Modified) * KISSlicer Pro - 1.6.3 * Arduino 1.8.5 *

olce
Posts: 9
Joined: Sun Aug 31, 2014 9:29 am

Post by olce »

Hi,

I'm having similar problems.
I upgraded to 20141107 - FIRMWARE - Repetier 0.92 - FELIX_3_0_DUAL_extrusion last week or so.
I was able to print e.g. 4 of those labyrint giftboxes (so this means 8 parts, each taking about 2 hours)
without much issues. Once or twice I had to restart because of the decoupling issue...
Yesterday I had to test something for another issue I had (posted elsewhere on this forum)

And then it started : it couldn't even keep the temperature for 10 minutes, most of the times, I got the feared decoupling error after 5 to 10 seconds after printing the purge line

I read through this thread and used the posted values for P, I, D to test with (I'm still a 3D Printing novice/newbie). I tried :
(P, I, D)
20, 0.2, 75 : worked fine for a few days (had to restart a 2 hour print 2 or 3 times), since yesterday, doesn't work at all anymore
41, 0.1, 260 : can keep it going for more then a few minutes, never past 10 minutes
20, 0.4, 46 : no success, decouple after a few seconds
125, 0.5, 650 : (from the original firmware) : no success, decouples after a few seconds

additional observations :
- during the trials, I can see the temperature is very jumpy while the printer is printing... variations 180-199 degrees and larger
- when I manually set the temp to 195 (via RHost) and leave it like that without doing anything else: the temp remains very stable (almost a flat line in the graph)
- when I manually move the heads over the X-Axis (via RHost) sometimes the temp remains stable, sometimes I can see it jump (up or down)
- I also went back to firmwares dated 20141002 or even 20140910 hoping I could reverse the situation, but no luck there either.

It is a bit frustrating (and worrying) that last week I almost had no issues with this, and now I even cannot get the smallest print finished

is this a software issue, or am I hit by an hardware issue?

Any advice is greatly appreciated

Olivier


P.S. : all prints are PLA, target temperature 195°
P.P.S : (I bought the printer in September, did few prints since then without issue, but not yet a large amount, so I would be surprised if there would be mechanical damage already?)
Felix 3D 3.0, DUAL extruder setup
Firmware: 20141107 - FIRMWARE - Repetier 0.92 - FELIX_3_0_DUAL_extrusion
RepetierHost : FELIX Printers 1.0.6

Dreide
Posts: 176
Joined: Sun Sep 29, 2013 12:32 am
Location: Lausanne, Switzerland

Post by Dreide »

Try with tightening the set screw that holds the thermistor in place.
Felix 2.0->3.0dual * Repetier (host+firmware) * KISSlicer Pro, Simplify3D * Cubify Design

User avatar
DDME-Marc
Posts: 71
Joined: Sun Nov 24, 2013 11:04 pm
Location: New Zealand

Post by DDME-Marc »

Hi Olivier,

Full credit to Guillaume, as his latest release of firmware (11-11-14) using PID values of 20 0.2 75 provides a very stable temperature across a multitude of print environments. For me personally I tried a range of PID settings over a period of time. In the end I ended up with very similar values as Guillaume, settling for the same P and I values, and a slightly lower D (derivative) value of 72 instead.

I would suggest its less likely to be a firmware issue, and more likely a hardware issue such as the thermocouple, thermistor. Also getting a clean/strong signal from the thermistor could also be an issue. Over and above Dreide's suggestions, I would consider checking through the thermistor/thermocouple connections to make sure they are suitably secure, and if possible using an electronic contact cleaner at each termination point.

Cheers,

Marc
Felix 3.0 Dual Head * E3D Titan V6 * Repetier-Host V2.1.3 * Repetier-Firmware 0.92.9 (01/08/17 - Modified) * KISSlicer Pro - 1.6.3 * Arduino 1.8.5 *

NikolaiD
Posts: 16
Joined: Wed Sep 03, 2014 6:38 pm

Post by NikolaiD »

My printer has been working solid since I got it in September, a few days ago I started getting the decoupled errors. I haven't updated the software, I haven't updated the firmware, nothing is different.
And it got exponentially worse. It got to the point where it would decouple on the first layer all the time.
I uploaded the new firmware, no change. I changed #define EXT0_DECOUPLE_TEST_PERIOD to 20000, uploaded again and now it prints fine again... Perfect, steady temperatures, I did several two hour prints and it hasn't decoupled yet.
It seems like some numbers in there don't like to play well with each other...

olce
Posts: 9
Joined: Sun Aug 31, 2014 9:29 am

Post by olce »

Hi All,

just wanted to give a quick update to my decouple-situation
after a conversation with support (Mark) from Felixprinters and some testing, we suspected the termistors to be defective. I recieved new termistors (a newer/updated/improved version as the ones shipped with my Felix 3D 3.0 when I bought it)

I replaced the ones in my first extruder.

Since then, I've been printing without any decoupling issue, for a week or 2 (about 10 prints)

So I think my decoupling issue has been solved.

Thank you Mark from Felixprinters for the help :)
Thank you all here for trying to help :)

Olivier
Felix 3D 3.0, DUAL extruder setup
Firmware: 20141107 - FIRMWARE - Repetier 0.92 - FELIX_3_0_DUAL_extrusion
RepetierHost : FELIX Printers 1.0.6

NikolaiD
Posts: 16
Joined: Wed Sep 03, 2014 6:38 pm

Post by NikolaiD »

Very interesting. I do have an extra thermistor, I might try to replace it next if I keep having issues.

An update on my situation is that since I changed EXT0_DECOUPLE_TEST_PERIOD to 20000, I was able to print for two days and then started decoupling again.
So, when the printer has been sitting overnight, It starts printing well, then after 10-20 minutes, the temperatures start jumping around and it decouples (No difference if I use 0.91 or 0.92).

I was watching the temperature curve tab when it decoupled. The screen stopped updating for about 10 seconds, the next click showed a very low value, the heater went full on, then it overshot, then full off and like that until it decoupled.

My thought today was that the microcontroller is overheating. The power supply gets hot right under the ATmega chip and the chip gets pretty hot itself.

This morning I put a small heatsink that I had laying around on it and a fan blowing on the circuit board:
Image

Since that, my temperatures got SOLID. I printed all day with zero fluctuation! For the first time in weeks...
Image

I really hope it keeps this up. If it does, I'll pick up a little proper Raspberry Pi heatsink and see how that does. I'm excited to have it working again, so I wanted to share.
Thoughts?
Attachments
DSC_0734.JPG
DSC_0735.JPG

Dreide
Posts: 176
Joined: Sun Sep 29, 2013 12:32 am
Location: Lausanne, Switzerland

Post by Dreide »

NikolaiD wrote:So, when the printer has been sitting overnight, It starts printing well, then after 10-20 minutes, the temperatures start jumping around and it decouples (No difference if I use 0.91 or 0.92).
But the decouple feature has only been introduced with 0.92, so I wonder what makes 0.91 detect a decoupling event? Maybe yours is just a different problem.
Anyway, interesting that you could fix it by just cooling the µ-controler.
Felix 2.0->3.0dual * Repetier (host+firmware) * KISSlicer Pro, Simplify3D * Cubify Design

NikolaiD
Posts: 16
Joined: Wed Sep 03, 2014 6:38 pm

Post by NikolaiD »

Interesting. Mine decoupled with 0.91 as well, so I switched back to 0.92 because I get better prints. I'm not sure if it's fixed yet. I'll do some more printing today and see how it is. Fingers crossed. :)

One thing that has me curious (I haven't looked into it yet) is, is the microcontroller clocked differently in 0.91 vs 0.92?
Printing with the same PLA, same file, same settings to get the same result in 0.91 I have to print at 190C and with 0.92 I have to print at 220C.

Post Reply