Temp fail causing print fail

[bananapanic]

I upgraded to the HE280 hotend right when it came out. I've run into about every problem I can figure in trying to get it to perform consistently. I thought I was through with the last of problems-- (having to autotune the PID functions because I couldn't get the hotend above 240). I got a couple prints in on Urethane 235/70, but came in this morning to last night's print barely starting. The temps were at def. I'm doing the same print this morning. The hotend and bed got up to temp, but at some point I noticed the hotend at 115-120 and the bed is around 52. I was thinking it was the power supply, so I turned the bed off with no difference in hotend temp. I'm leaning more towards the hotend resister failing? It seems like a power supply would show at least some difference when depending on what it's powering.

[timskloss]

Temps at 'def' mean the thermistor is open circuit. Do you have a probe thermometer you can use to take the nozzle temperature? If the hot end is constantly heating it can get VERY hot and do damage, so this is not something to take lightly.

Also, do you have a voltmeter to check the power supply? I recently found my power supply to be 12.0 VDC when the system is idle, but 11.7 V when under a load heating the bed and hot end.

[IMBoring25]

It will go def for any thermistor fault...Open, short, or, on the newer firmwares, a watchdog failure where power is being applied but the temperature is not going up.

[bananapanic]

The thermister seems to be reacting normally in every instance. What's worse is I can set the hotend temp to 240 before printing and have it heat without issue. My PID autotune values seem erratic. The last revealed kp of 27.05,172.11, 59.83 with other values being similarly varied. I can check the power supply voltage, but will have to wait til tomorrow when I can grab a multimeter. Also, the morning's print I referenced earlier never "def"ed. It just stayed down at 217 even though it was programmed to 240. Eventually the print quality was compromised so I cancelled the print, but it was almost finished.

Edit 1: I swear it looks like just having filament pushed through the extruder is enough to cool it 15 degrees. I just did a preheat and the extruder was heating slower than I ever remember it. When it got above 200 it would go from .3 to .8 to 1... I remember it jumping degrees before.

Edit 2: Doing a calibration cube and the heat stayed decent until it started printing- dropped to about 231 and stayed pretty constant until layer 3 when the cooling fans kicked on (HE 280). The temp plummeted and finally "def"ed. It got to 217 before the last time I looked away to type this and now it's over. This issue seemed to occur spontaneously, as I didn't change any settings for this to occur.

Edit 3: The PID calibrations... They do nothing!

Edit 4: I ran home and grabbed my multimeter. The voltages across the extruder is 11.8 and the bed is 11.4. I don't think there's anything wrong with my power supply. Something must have burnt out on the HE 280. With all the trouble I've had with this thing I'm about ready to revert my printer. I haven't been able to use it consistently for weeks now because of the "upgrade." ugh.

[Xenocrates]

The thermister seems to be reacting normally in every instance. What's worse is I can set the hotend temp to 240 before printing and have it heat without issue. My PID autotune values seem erratic. The last revealed kp of 27.05,172.11, 59.83 with other values being similarly varied. I can check the power supply voltage, but will have to wait til tomorrow when I can grab a multimeter. Also, the morning's print I referenced earlier never "def"ed. It just stayed down at 217 even though it was programmed to 240. Eventually the print quality was compromised so I cancelled the print, but it was almost finished.

Edit 1: I swear it looks like just having filament pushed through the extruder is enough to cool it 15 degrees. I just did a preheat and the extruder was heating slower than I ever remember it. When it got above 200 it would go from .3 to .8 to 1... I remember it jumping degrees before.

Edit 2: Doing a calibration cube and the heat stayed decent until it started printing- dropped to about 231 and stayed pretty constant until layer 3 when the cooling fans kicked on (HE 280). The temp plummeted and finally "def"ed. It got to 217 before the last time I looked away to type this and now it's over. This issue seemed to occur spontaneously, as I didn't change any settings for this to occur.

Edit 3: The PID calibrations... They do nothing!

Edit 4: I ran home and grabbed my multimeter. The voltages across the extruder is 11.8 and the bed is 11.4. I don't think there's anything wrong with my power supply. Something must have burnt out on the HE 280. With all the trouble I've had with this thing I'm about ready to revert my printer. I haven't been able to use it consistently for weeks now because of the "upgrade." ugh.

Check your PID_Drive_Max, as well as the resistance of the heater cartridge. It should be ~3.6 ohms. If it's more, then it is not delivering sufficient power (Voltage drop is irrelevant here). If the PID drive max is less than 255, bump it up. The HE280 can handle in case there is an overshoot when tuning much better than the old stock hotend. In addition, once you've done that and retuned the PID, run a tune with the fans running at a medium speed, or manually adjust it to hold temp while the fans are running. It likely wants more P, to allow it to regain heat (Which will also make it heat faster), and this will require retuning I and D somewhat to compensate at idle. Repetier's Autotune is honestly pretty hit or miss.

[bananapanic]

I was able to test the heating element this morning. It walked decimal places both sides of 5 ohms, but would also erratically jump and fall as well. Is that representative of faulty hardware? Either way, based off of what you said it's too high?

[Polygonhell]

If it hits 240 when heating, it should be able to retain it when printing 20+C drop is a lot.
Is it just the cooling fan overcoming the heater?
If it is make sure you run the autotune with the Fan On (M106 S255)
Your I and D seem very high from the auto tune, I personally always end up tweaking them manually.

[bananapanic]

Thanks for the response. Just to be clear, the numbers I listed before were only my Kp values to demonstrate how varied my autotune was. I just ran it again with the fans on to VERY different results. I ultimately set the values to Kp=20, Ki= 4, and Kd=27. While these helped the printer get to temp much quicker and hold it better, at about layer 3 it started a very slow decent. So with my temp set at 235, printing a 1cm cube with 30% infill, it just finished with the temp bottoming out at 220. I imagine if it was a longer print I was on the cusp of it failing again.

Edit 1: Autotuned it again with the fans as low to the print bed as possible to maximize their cooling effect. New PID values are 17.42, 2.39, and 32.37. Still dropped down to 220 before the end of the 1cm^3 print. I think it's the heating element.

[nebbian]

BananaPanic, try setting the nozzle heat manager to 1, and trying again with another autotune, followed by manual tuning. (do this in the eeprom settings) I've no idea why 3 is the default on Repetier, but I've found it particularly unreliable. Heat manager 1 is the classic PID loop, where changing the PID values actually does what you might expect. In your case, if the temperature drifts like that, I'd bump up the I value. But first you need to change to heat manager 1.

[bananapanic]

looks like mine is set to 1 in the configuration file. It says #define EXT0_HEAT_MANAGER 1. I'll try messing around with increasing the I then.

Update 1: I upped the I from 2.37 to 4.37. During the 1cm^3 print the temp dropped down to about 223 at the second layer, but near the end of the print it crawled it's way up to about 230. Should I keep increasing the I until I can get it to muster 235 consistently? That's what I'm trying next. What I don't get is why this problem would arise from seemingly nowhere?

Update 2: Yeah, this isn't going to cut it... Trying to print 235 it's still dropping down to 218

[nebbian]

Yeah what you've got sounds to me like some other problem, not tuning. Perhaps a dry joint (hot connection) somewhere? If it were me, I'd replace the heater cartridge and go from there.

[rurwin]

There are two values that limit the maximum drive: DRIVE_MAX and PID_MAX. Both of these are overridden by the values in the EEPROM. In vanilla Repetier (I'm looking at github) PID_MAX defaults to 255, but it's worth checking your EEPROM to be sure. DRIVE_MAX defaults to 140, which a comment in configuration.h suggests may not be enough for ABS. Someone did mention that though so I assume you've checked it.

Note that EXT0_HEAT_MANAGER is also overriden by the EEPROM so check the EEPROM value is also 1 (That is the default in vanilla Repetier.)

Deadtime control uses only DRIVE_MAX and not PID_MAX, so it might be worth an experiment to set EXT0_HEAT_MANAGER to 3.

Also PID_CONTROL_RANGE (in configuration.h, not the EEPROM) sets the area outside which the heater will definitely be on full (PID_MAX). The default value is 20C. Do you see a difference in behaviour within 20 degrees of the setpoint? If it drifts down further than 20C then it's definitely a hardware fault, probably in the heater circuit.

Both Deadtime control and the I term of PID control use Infinite Impulse Response (IIR) filters on the current temperature. If you are getting intermittant large spikes to high temperatures (low resistance) it could cause both of these algorithms to see a constant high temperature and behave accordingly.

It may be worth setting I_GAIN and D_GAIN to zero. You will be able to get close to the setpoint using P_GAIN alone, albeit not quite and with large over-shoots. I don't know if I'd want to print like that, but you should be able to see the behaviour by switching the fan on and so forth. With only P_GAIN the algorithm has no memory of past events -- it's only reacting to the current temperature. So if you still see downward drift it is highly indicative of a hardware failure.

[Doug68]

I've also recently switch to the HE280 and I've also had odd losses of temp and 'def' come up a few times mid print.
Now this is going to sound goofy, but I believe the issue is the cable connector working itself loose out of the board.
I'm sure I have enough slack in the cable - bowden, but sure enough after the last 'def' I pushed the thing down properly and it fixed it, similarly I was seeing the hot end losing temp so pushed it back in a bit and it was magically cured.

I've now bodged on a cable tie on there to try and hold it in place.

[Mac The Knife]

I've also recently switch to the HE280 and I've also had odd losses of temp and 'def' come up a few times mid print.
Now this is going to sound goofy, but I believe the issue is the cable connector working itself loose out of the board.
I'm sure I have enough slack in the cable - bowden, but sure enough after the last 'def' I pushed the thing down properly and it fixed it, similarly I was seeing the hot end losing temp so pushed it back in a bit and it was magically cured.

I've now bodged on a cable tie on there to try and hold it in place.

Here's a clip PartDaddy made up for the cable connector http://repables.com/r/808

[Doug68]

Here's a clip PartDaddy made up for the cable connector http://repables.com/r/808[/quote]

Excellent, that'll be next in the queue to print! Thanks.

[sfuelberth]

So I had also started to experience some issues like being described here, some fluctuation of temp during prints and now I am not getting any heat at the hot end. I have measured a resistance of 6.0 ohms over the heater cartridge, I ordered a new cartridge but I am curious if I need to look at anything else? Also I have a newbie question with the PID values if setting them manually how exactly would you determine these values?

[Doug68]

See page 269 of the V2 build manual and follow the instructions there to do the heat calibration.
http://download.seemecnc.com/rostockmax ... dition.pdf

Have you made the clip to hold the connector in?
Have you check the connector is properly seated after the issues start?

[geneb]

The stock PID values are good for the HE280. I strongly suspect you've got a wiring problem - likely a bad crimp on the heater cartridge wires.

g.

[sfuelberth]

I'll check out the crimp on the ferules and I will look at the V2 documentation. Thanks.