Hotted not maintaining temperature

I'm having problems with maintaining my hotend temperature. If I heat the hotend to a set temperature and let it sit there, it will reach and maintain the temperature. If I run autotune, no issues. When I start to print after the first layer the temperature starts to drop. On the first layer, no issues but starting on the second layer the temp starts to slowly drop. Looking at the temperature graph, It obvious that the heater cartridge is trying to recover the temperature. To see if might have a loose or broken wire, I'd run the gantry around using the "jog" buttons in Repitier but the temperature held. Only when I print do I have the problem,

I replaced the thermistor because I broke it trying to figure out the problem. No change. Today I replaced the heater cartridge. No change. The power supply voltage holds steady at 12.3 volts. Putting a meter across the D10 terminals, I can see the ramps controlling the hotend. When the hotend temp is dropping I can see 11+ volts on the heater cartridge, so it's trying to make it hotter. Looking at the temperature graph it shows some temperature recovery but not enough to trip a thermal fault. I've never see the voltage to the heater greater than 11.7 volts, is that normal or should it reach 12 vdc?

My next thought is something is wrong with the ramps board. This problem started a week back - the print before it happened came out perfect. No hardware or software changes made. E3D hotend, ramps 1.4, Marlin RC3.

Any ideas what to check nexrt?

You have 40W 12V heater cartridges? Resistance should be about 3.6 Ohms. What temperature are you trying to maintain? Assuming it's not something ridiculous, if you have 3.6 Ohms for the heater and 12V at the ramps the only logical explanation is a broken wire. I would suggest the crimps where the wires are connected to the solid cores coming out of the cartridge - I had a loose one that behaved very much as you describe. But if you changed the cartridge you'd have to be very unlucky to have both of them be loose. You could check the screw terminals at the ramps end and any connectors between them and the heater. If nothing is obvious, run a new cable between ramps and heater and see if that solves the problem.

Quote
JamesK
You have 40W 12V heater cartridges? Trying to maintain a 210 degree temp. It'll drop down to 200 or so recover to about 205 and then usually starts dropping below 200. I've temporaly disable the thermal_protection_hotends so it wouldn't fault out on Thermal runawy alarms. At least I can see it seems to recover a bit.

Resistance should be about 3.6 Ohms. What temperature are you trying to maintain? Assuming it's not something ridiculous, if you have 3.6 Ohms for the heater and 12V at the ramps the only logical explanation is a broken wire. I would suggest the crimps where the wires are connected to the solid cores coming out of the cartridge - I had a loose one that behaved very much as you describe. But if you changed the cartridge you'd have to be very unlucky to have both of them be loose. You could check the screw terminals at the ramps end and any connectors between them and the heater. If nothing is obvious, run a new cable between ramps and heater and see if that solves the problem.

Yes, 40W 12v cartridge. I measured the resistance of the original cartridge (came with e3d) as 5.2 Ohms, the replacement measures 4.7 Ohms and a third one from the batch of heaters ordered measured 3.9 Ohms. The replacement heater is wired it directly to the Ramps terminals - I wanted to verify that it was the fix before I spent any effort installing it permanently. The original does have solder joints which is one of the reasons I suspected it was part of the problem. I did recheck terminals to make sure they were tighted down after your response - good and tight.

Since the third cartridge had a resitance closer to 3.6 ohms I installed that one to test it. The same problem as before, if anything the temperature fell off faster than the previous heaters. So it seems I can rule out the catridge heater being the problem. This has been the most frustrating problem I've had with my printer.

Stating to draw at straws here and replaced the Ramps board. Not sure I expected to make a difference. So the only electrical components not replaced are the Mega board and the power supply. Printer electronics are not that complicated...

Yes, the straw supply is starting to get a bit short. When you measure the voltage, where exactly are you measuring between? The mosfet heater circuits are switched on the negative side, so people occasionally get confused by measuring between supply 0V and the heater + terminal on the ramps, but that will always read the supply +V value (unless the poly fuse is cutting out). If you measure between supply 0 and the heater - terminal on the ramps it should show 12V when off and close to 0V when on.

How is your setup wired up? Are you using the 'normal' connections for heatbed, extruder and part fan or have you swapped anything around? And you have both sets of 12V inputs connected, and those connections are solid and not overheating?

I've had problems with poor thermal conductivity between the heater cartridge and the heater block, using a cheap one with an oversized hole and the cartridge held in place with a set-screw. That was improved hugely with some Z5 thermal grease, but it also showed up as an inability to autotune the pid, and you've said that works ok.

Do you use a heated bed? If so, your power supply must be capable of way more than the hotend needs, but if not, maybe the psu is struggling. Definitely clutching at straws now...

Given that we're running out of hardware possibilities, you might want to consider throwing Repetier firmware on the board for a quick test. One of the things I feel Repetier does better is thermal control, but the main point would be to try a completely different firmware and see if it behaves the same way or not. There's a relatively easy to use online configuration tool which provides a download link as the final step: [www.repetier.com]

If you can get past the ugly home screen and unfamiliar menu structure you might even find you like it

Quote
JamesK
Yes, the straw supply is starting to get a bit short. When you measure the voltage, where exactly are you measuring between? The mosfet heater circuits are switched on the negative side, so people occasionally get confused by measuring between supply 0V and the heater + terminal on the ramps, but that will always read the supply +V value (unless the poly fuse is cutting out). If you measure between supply 0 and the heater - terminal on the ramps it should show 12V when off and close to 0V when on.

How is your setup wired up? Are you using the 'normal' connections for heatbed, extruder and part fan or have you swapped anything around? And you have both sets of 12V inputs connected, and those connections are solid and not overheating?

I've had problems with poor thermal conductivity between the heater cartridge and the heater block, using a cheap one with an oversized hole and the cartridge held in place with a set-screw. That was improved hugely with some Z5 thermal grease, but it also showed up as an inability to autotune the pid, and you've said that works ok.

Do you use a heated bed? If so, your power supply must be capable of way more than the hotend needs, but if not, maybe the psu is struggling. Definitely clutching at straws now...

I'm measuring across the +/- terminals where the heater is connected at D10 I get 0vdc when it's idle up to 11.7vdc when it's driving the hot end hard, I've never seen a full 12v measuring at that point. I'm only using the 5

Yes, basic setup hot end on D10, parts fan on D9, hot end fan is wired across the power supply terminals and I don't have a heat bed. Running dual Z motors. The printer has a 14x14" bed so I'd need a separate power supply if I add a bed in the future. I'd have to verify but I think my PS is 30A.

Quote
JamesK
Given that we're running out of hardware possibilities, you might want to consider throwing Repetier firmware on the board for a quick test. One of the things I feel Repetier does better is thermal control, but the main point would be to try a completely different firmware and see if it behaves the same way or not. There's a relatively easy to use online configuration tool which provides a download link as the final step: [www.repetier.com]

If you can get past the ugly home screen and unfamiliar menu structure you might even find you like it

I looked at Repitier early on on was turned off by the questions to pre-configure it.. At the time, I didn't know enough about firmware vs just editing the configuration.h in Marlin. Does Repitier support servo bed leveling? I have BLTouch (highly recommended) auto bed leveler on my printer. I'll have to check it out.

Thanks for supply more "staws"

Edited 1 time(s). Last edit at 04/02/2016 01:17PM by gmartell.

Quote

I'm measuring across the +/- terminals where the heater is connected at D10 I get 0vdc when it's idle up to 11.7vdc when it's driving the hot end hard, I've never seen a full 12v measuring at that point.

I haven't measured mine recently, but 11.7 is probably reasonable after some voltage is dropped across the mosfet. If you can get the clips on the wires at the cartridge end it would be interesting to compare, just in case you're dropping voltage in the wiring for some reason (I know, it's very unlikely).

Quote

I'm only using the 5

Lost me there, could you clarify?

One thing you might try is reconfiguring to use the heatbed mosfet/pin for the hotend, just in case there's something odd going on with the pin on the mega (I was going to say the mosfet, but you already swapped out the ramps).

I think Repetier supports servo activated Z probes for ABL, but I haven't tried it yet - on the wish list. BLTouch looks very nice, but I'm going to give LeadingLights piezo sensors a try first because I often change hotends so a probe-less design would be a plus.

Sorry, meant the 5A terminals, not the 11A terminals. I have another Mega around here so I may just swap it.

Thought about it when I swapped the Ramps but I learned a long time ago to only do one thing at a time when troubleshooting.

I hadn't seen LeadingLights piezo before. Saw the original concept posted but didn't realize someone had developed the concept this far. Pretty cool but wouldn't work on my printer without major gantry redesign. Definitely something I will share with the folks at are local makerspace. They'll just rip a printer apart on a whim just to try something new!

Lower the blow of the fan a little bit if it happens after the first layer and you have your slicer configured to start the fan on layer 2.

While I don't think I know the cause I upgraded to Marlin RC4 today and the problem went away. Couldn't print with RC3, upgraded and it worked! I did figure out the the hot end was heating because the fins were warm on the hotend and it takes a lot of heat to overcome the cooling fan on an E3D. So I have to believe that there was something screwed up with the thermistor reading. Maybe it was in Marlin maybe I do have an issue with the Mega 2600 board. Your suggestion about trying Repiteir firmware made me try Marlin RC4.

The cheapie board I bought for another project has a CH3400 serial chip on it and that driver doesn't work on a Mac so I couldn't swap it out. I think I'm going to order a Sainsmart one to replace it. Now I need to go back and try to switch back to RC3 to see if the problem comes back. Great that it works but still don't know the cause...

I've had a problem or two like that. It's unsettling to say the least. I must admit, when sigxcpu suggested the fan I thought that was it, and I'd over-looked the obvious. Not sure if it's good or bad that it turned out not to be the problem!

Quote
JamesK
I've had a problem or two like that. It's unsettling to say the least. I must admit, when sigxcpu suggested the fan I thought that was it, and I'd over-looked the obvious. Not sure if it's good or bad that it turned out not to be the problem!

'

I wish it was so obvious - still bothers me that I really haven't solved the problem and I somehow got lucky.

Ok since you are still struggling maybe some of this might help to find the cause of your problems:

The ramps board needs a stable power supply with enough to spare for pulsed operation (bang bang).
Normal ATX power supplies in the lower range often suffer from massive breakdowns with rapid changing loads, a good load on the 5V rail helps to stabilise the 12V output!
You can overcome a weak power supply by adding a booster capacitor as used for high end car audio systems - the big cans with a voltage display.
Depending on the cable length and type you will get wrong readings from the thermistor!
This is especially true if the wires for hotend and thermistor are not twisted like litz wire.
This does not happen in all configurations and seems to depend on the frequency set for the PWM control.
In some cases adding a small 10nF capacitor parallel to the thermistor connect can help to stabilise the readings.
The hotend itself also need wires that are thick enough to handle the current without adding too much resistance.
I had a few pre-wired hotends that came with just 0.5mm cables but the center of these wires was filled with a fibre string almost like headphone wires.
Replacing them with standard speaker wires gave me a very fast heating hotend.
Last but not least there is the general thermistor problem where you get totally wrong temp readings.
I replace all thermistors with a selection I ordered in one batch and confirmed to be close to each other at room temp and in boiling water.
I also only use types with 1% or less tolerance.

If you have the chance test the output to the hotend with an oscilloscope.
It will show the voltage, how the signal rises and how clean the signal is.
If you check the supply voltage on the second channel at the same time you can also see if the heating has any effect on how stable your supply voltage is.

As a last resort clean up the Arduino.
Sometimes the EEPROM values become corrupted and cause all sorts of problems.
Take the EEPROM Clear sketch from the examples, adjust your memory size to whatever you have, the Mega2560 has 4096kb for example.
Then put the firmware back on and check if there is any difference.

Well I spoke too soon on fixing my hotend heating problem and learned a hard and stupid lesson. I had moved to Marlin RC4/5 and it seemed that it fixed my issues and for the most part it did. I was still having the same issue on some parts where I would lose heating control and the printer would shut down. There were improvements in RC4 from a thermal control standpoint, so I think it solved part of my problem but not all of it.

I disabled the thermal protection on the hot ends which prevented the shutdowns and it was obvious that the firmware was controlling the hotend temperature - just 10 or so degrees below setup. It get's stranger.

As part the process of debugging, I rolled back to RC2 and the problem was still occurring even though I never had the issue with this version. While playing around I sliced the part in a different slicer (Slic3r) and the problem went away - completely and consistently. Tried again with my normal slicer - Matter Control - problem came back. More mystery - how can a slicer make a difference?

As I was physically looking over the printer for the hundredth time I noticed my parts fan had gotten knocked aside, was pointing directly at the heater block and it was RUNNING! My LCD display is located on the side of the printer so it wasn't obvious that the fan was on.

I have a very large printer (320x320x320 build volume) and have to physically move it to get behind it to even see if the parts fan is running. My fan is setup to only run when doing bridges, so I don't pay any attention to where it's pointed. Unless I'm going to do the rare bridge part, I just ignore it. Matter Control (normal slicer) had an update a week or so back and now it was running the fan starting on the second layer where it hadn't been before! Same parts, different behaviour. I'm not blaming the updated software but something changed when I installed the update. Maybe I changed a parameter or something. Who knows for sure. Fan is now disabled in the slicer.

I can't believe I missed this. So I am back to my original heater and the upgraded thermistor. All is working well on Marlin RC3. I would like to go back to RC5 but I'm having issues with auto bed leveling on that branch. That's an issue now on GitHub.

So thanks to everyone who helped me with this. Many of you were closer to the answer than I realized and I really got deep into the weeds when it was rather a simple. Hard lessons are the best lessons.

Edited 1 time(s). Last edit at 04/08/2016 01:30PM by gmartell.

Well there you go - sgxcpu for the win after all

I guess that's one benefit of my part fans being so noisy, no way I can miss when they are on, although my extruder fans are giving them a run for their money at the moment.