1. Got a question or need help troubleshooting? Post to the troubleshooting forum or Search the forums!

What...the... (E3D print head)

Discussion in 'Mods and Upgrades' started by Randseed, Apr 8, 2014.

  1. Randseed

    Randseed New Member

    Joined:
    Apr 2, 2014
    Messages:
    9
    Likes Received:
    0
    So I get my new E3D and am happy. I spend most of the evening installing this thing because the grub bolt fell off never to be seen again, so I had to replace that with an M3. . Finally, it's all wired up. The thermsistor is appropriately placed, is hooked up, and is not shorted. The fan is installed. Filament feeds down into the head (measuring it) while cold, so I know it's unobstructed. The fan is functioning properly (I wired it into AUX1). I modify the software for a maxtemp of 300C and a temperature sensor type 5 as the documentation instructs. It compiled. I upload it to the printer. Looking at version messages, I verify that the new version is running.

    Oh, I'm supposed to set it to 290C to start. That's what the assembly instructions say. It seems to heat up slower than I would expect, but I attribute this to the fan. I do verify twice that the fan is running. I wait. I fuddle with the filament screws (seriously, who came up with that design?) and chase the things as they fly off into the wall a couple of times.

    All of a sudden, the bottom falls out. What this thing has done is melt itself and melt part of the X-carriage.

    So clearly, a few things happened here. The thermsistor didn't work, because it read 280C (!) when all this happened, and it's physically on the heat block. Their heat dissipation system obviously didn't work, and their own part melted.

    1.1.JPG

    2:
    2.1.JPG

    3:
    3.1.JPG

    Any bright ideas? And I should have printed a spare X-carriage before doing this, but obviously I didn't foresee this happening. It was going to (ironically) be one of the first prints I did with the new head.
     
  2. 1d1

    1d1 Active Member

    Joined:
    Feb 21, 2013
    Messages:
    418
    Likes Received:
    216
    Bummer. I had a runaway once with no serious consequences. Here's what I would do: Wherever you bought it, send them the pics. Sanjay, Filastruder and Printed Solid are all excellent at following up on warranty issues. Clean up the heat sink and replace the thermistor. Check it out before installing the hot end in the ROBO. Cool it off, then find a way to secure it to your extruder carriage. As long as it is firm and relatively vertical, you will be able to print with it. If you do not wish to wait for the new fan/mount, etc. print the new extruder carriage anyway. Remember, this what I said I would do. The e3d is pretty efficient and will likely handle printing without cooling, especially if all you need is for it to stay rigid for the course of the print. OTOH, I make a lot of mistakes as I bumble through learning how to do all this. Also, e3d says 290 but even 250 will do. What you want is a final check of the tightness of the heat sink/stainless feed/nozzle at temp to eliminate any joint ooze that could lead to a feed block. Then you get to do much of the process over again when re-installing the new extruder mount, but look at how efficient you are since you know how it all goes now.
    Bummer... but good luck!
     
  3. Printed Solid

    Printed Solid Volunteer Admin
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    1,605
    Likes Received:
    1,003
    Whoah! Yeah, you must have gotten hotter than 290 at the thermistor for that to happen with the fan on! This setup is rated at up to 400C...

    If you bought from me, I can send you a new fan mount and thermistor, but I'm guessing the problem was not actually due to a failure of the thermistor itself, (although odds are it's toast now). Some troubleshooting is definitely needed before you power up again.

    Maybe one of our members who is a little more inclined to shooting electrical / wiring issues can help weigh in with troubleshooting tips?
     
  4. Randseed

    Randseed New Member

    Joined:
    Apr 2, 2014
    Messages:
    9
    Likes Received:
    0
    Yeah, it was interesting. Since a thermistor's resistance decreases as temperature rises, if there was a short it would have never started heating in the first place. That means that the circuit had to be wide open open. But there was clearly temperature change, it just wasn't very fast. It would stand to reason that by the time the thing got *that* hot, there would have been enough heat to overcome any problem with thermal conductance being sub-optimal. When the thing fell apart, it was reading 218C.

    Beats the hell out of me.

    Do you have an STL file for the fan mount? Presumably it's printed in ABS?
     
  5. Mike Kelly

    Mike Kelly Volunteer

    Joined:
    Mar 11, 2013
    Messages:
    6,967
    Likes Received:
    2,276
    Well this is certainly a new one. So you said it had a short? Or was your thermistor loose?

    I don't know why they advise people to get it up to 300C. Thermistors can't read reliably at that temperature and will very drastically overshoot.
     
  6. SteveC

    SteveC Well-Known Member

    Joined:
    Jun 12, 2013
    Messages:
    894
    Likes Received:
    316
    Randseed, I'm really sorry that this happened. What a mess. The slow heatup as read by the themistor was definitely a warning sign. I find that the E3D's ceramic heater heats up much much faster than the stock power resistor. The strange thing is that the thermistor read at about your target temperature, right? Following is some background on the thermistor that might help you think about the problem but it does not explain it reading at the target temperature.

    If the NTC (negative temp coefficient) thermistor was open circuit (infinite resistance) the temperature read should be zero and the firmware fail safe should have cut off the heating. If the thermistor was shorted (zero resistance) the read temperature should be maxed out and the heater should have cut off. The main problem that I have heard of is if the thermistor just falls off and looses thermal contact with the hot end. Then the read value will be near room temperature and the heating feedback loop will cause runaway heating which is what you had.

    There is a firmware change suggested by Thor that was discussed in one of the firmware update threads http://forums.robo3dprinter.com/index.php?threads/marlin-firmware-update-2-6-14.1725/ that looks for and cuts off heating if a fast 5C drop in temperature is detected. This would occur if the thermistor fell off or lost thermal contact with the hot end. The only problem is that the 5C threshold is probably to small and situations like the print fan suddenly going at high speed during bridging could cause it to trigger. It is a great idea that needs some testing and tweaking.

    So other than a strange software crash or a failed ADC in the Atmel micro I have no idea what could have caused the thermistor to maintain a resistance that reads near the target value.

    I fully expect someone to chime in with an obvious solution that I missed here;). For encouragement here are some photos of the Robo and E3D in the middle of a 10+ hour print of a voronoi structure. Its fine structure is probably near the limit of what the printer can do.

    IMG_1855.JPG IMG_1859.JPG
     
    2 people like this.
  7. SteveC

    SteveC Well-Known Member

    Joined:
    Jun 12, 2013
    Messages:
    894
    Likes Received:
    316
    Mike, I agree that it really looks like the thermistor lost contact somehow.

    Randseed, One thing I just noticed is that you mention connecting the fan to AUX1. I think that has Vcc on it which should be only 5V passed from the Arduino Mega's regulator. If you are using the stock E3D 12Volt fan this may barely turn it on. It probably is unrelated but it really should be changed to the PSU's 12V.

    http://reprap.org/mediawiki/images/f/f6/RAMPS1.4schematic.png
     
  8. Printed Solid

    Printed Solid Volunteer Admin
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    1,605
    Likes Received:
    1,003

    I don't have the STL, but I can either get it or make one up quickly. If you need it, I'd probably just pull one from stock and print a new copy for myself when I have some spare time.

    I agree that Mike and SteveC's hypothesis of losing thermal contact is probably the most likely culprit.

    PS. You definitely need to upload this to the epic 3d printing fail website!
     
  9. Randseed

    Randseed New Member

    Joined:
    Apr 2, 2014
    Messages:
    9
    Likes Received:
    0
    I'd appreciate it, Solid. I drilled two holes in what's left of the existing one and got a half-ass assembly out of it.0

    SteveC: That's what makes no sense. The thermistor was, at worst, right next to the hole. I haven't run the math, but it seems like there should have been even enough radiation if the thermistor was somehow sitting in the hole (or right next to it) and not in contact to have triggered out before it hit a true temperature of in excess of 400C. The idea of a near 200C temperature delta is a bit...odd. Particularly since it had to be in contact even if that contact was bad. I will definitely have to install that failsafe.

    And thanks for the AUX1 comment. I'll have to change that over. I don't think it had anything to do with this because all it really would have done was make it take a tad bit longer to melt down. The fan in question is the side-mount fan, which was spinning fairly well but didn't seem that effective.

    Also, interestingly, I stripped the entire thing, repositioned the thermistor, and turned it on. The same thing happened, but this time of course I turned it off way before it committed suicide.

    But God does have a sense of humor. Through some creative drilling and hardware usage, I managed to get the original extruder head installed again and it will actually print wel. Unfortunately, now the hobbed bolt isn't working and stops feeding the filament, so I have one of those on order. Ironically, even if I print a new assembly I still have to get another hobbed bolt. So I have it on order from printrbot.com.

    What's the thermistor part number so I can add it to my impending Mouser order? Also, where do you guys go online to order mechanical components (bolts, nuts, etc.)? I went to Lowe's and that was an annoying experience. For example, I couldn't find an M4 nut in the entire aisle. I'd rather just mail order everything anyway.

    And yeah, I can just hear the voice in the background: "EPIC....FAIL." ;)
     
  10. SoLongSidekick

    SoLongSidekick Active Member

    Joined:
    Mar 6, 2014
    Messages:
    220
    Likes Received:
    68
    I cooked a heating block to the point it was charred (thank God didn't quite get to what happened to you) by having the thermistor come loose and sit right next to the hole. That's what might have caused this.
     
  11. Mike Kelly

    Mike Kelly Volunteer

    Joined:
    Mar 11, 2013
    Messages:
    6,967
    Likes Received:
    2,276
    They really need to add the code into marlin to detect a loose thermistor. It can be seriously dangerous if it comes loose.

    For metric I recommend Ace Hardware. They tend to have a better stock of small metric stuff.

    The thermistor is a semitec http://e3d-online.com/100k-Semitec-Thermistor?search=thermistor

    on mouser: http://www.mouser.com/ProductDetail/Semitec/104GT-2/?qs=/ha2pyFaduhDub/UzqX5sAEhF7n126iQt/2y0r9Rzc3Z8Hkmeih3gA==

    You should also pick up some ptfe insulation for thermistors. http://www.mcmaster.com/#5335k9/=rg82tx or https://www.hotends.com/index.php?route=product/product&filter_name=ptfe insulatino&product_id=68
     
    2 people like this.
  12. SoLongSidekick

    SoLongSidekick Active Member

    Joined:
    Mar 6, 2014
    Messages:
    220
    Likes Received:
    68
    You're damn right, and is very prone with a new E3D installation. The kapton tape that is included with the hotend kit isn't enough to really secure the thermistor unless you really think it through. Without using any sort of fire cement, the only really safe way to secure it is with wrapping kapton tape around the heating block a few times; see below.

    [​IMG]
     
  13. SteveC

    SteveC Well-Known Member

    Joined:
    Jun 12, 2013
    Messages:
    894
    Likes Received:
    316
    That's great that you resurrected something the worked from that mess.

    The E3D stock thermistor is a Semitec 104GT2.

    That fail safe code works great but the temp delta value needs some work. A good test for it is that it triggered when I clamped onto the hot end with a crescent wrench to initially tighten the E3D nozzle. This dropped the temperature quickly and triggered the error message on the LCD.

    I found that my local Ace Hardware has a ton of metric hardware. Lowes/HD hardly has any.
     
  14. SteveC

    SteveC Well-Known Member

    Joined:
    Jun 12, 2013
    Messages:
    894
    Likes Received:
    316
  15. JohnStack

    JohnStack New Member

    Joined:
    Mar 9, 2014
    Messages:
    139
    Likes Received:
    65
    I know the directions "say" 290C; however, I haven't had this come loose after heating to 190.

    Second - and what they don't mention:

    It is definitely ok to bench test these before you print. In other words, I heated mine up without installing it. I then tightened everything down. I then checked temps, etc...

    I went ahead and submitted a suggestion to the E3D forum.

    http://forum.e3d-online.com/viewtopic.php?f=5&t=104
     
    5 people like this.
  16. CAMBO3D

    CAMBO3D New Member

    Joined:
    Feb 20, 2013
    Messages:
    642
    Likes Received:
    296
    the firmware should have taken care of the temp overshoot, if it doesn't reach said temperature or doesn't sense temperature rise (for instance if the thermistor fell out). it should have shut it self off. At least in repetier firmware it does this..... not so sure about the marlin firmware.. but then again its a setting in firmware..may not have been turned on.

    also I've had my ed3 up to 300c+ and its only warm to the touch..... not enough to burn or melt anything. so that being said your hot end had to be super hot in order for the heat sink to get that hot and melt the plastics.. (also another reason I don't use plastics on the hot end parts.. especially for an all metal hot end..

    kapton is not a good way to secure and insulate your thermistor wires.

    thermistor wires should be insulated individually with high temp tubing. Then secured in the hole via some kind of adhesive or epoxy.. (high temp copper rtv from automotive stores work great and is removable easily when need be) also....it doubles as insulation for your heating block (so fans don't interfere with thermistor readings)

    examples of my setup on another machine:
    CIMG2157.JPG custom all metal mount
    CIMG2167.JPG heating bock insulated and wrapped with kapton as extra measure
    CIMG2265.JPG notice i have rtv holding the thermistor, all thermistor wires insulated. kapton wrapped as extra measure.

    all this insures you have a reliable and safe to operate printer but the printer is only reliable if you make it reliable.. all dependent on the builder.


    ANOTHER CAUTIONARY NOTE...... ABOUT CARTRIDGE STYLE HEATERS. IS THAT ....IF ALLOWED TO CONTINUE TO HEAT THEY WILL GET RED HOT AND START A FIRE.

    OLD STYLE HEATING RESISTORS WOULD JUST BURN THEMSELVES OUT.. SO LESS RISK OF FIRE...
     
  17. JohnStack

    JohnStack New Member

    Joined:
    Mar 9, 2014
    Messages:
    139
    Likes Received:
    65
    Good feedback Cambo!
     
  18. SteveC

    SteveC Well-Known Member

    Joined:
    Jun 12, 2013
    Messages:
    894
    Likes Received:
    316
    Cambo, As far as I understand the Marlin firmware does not shut off unless the temperature goes to zero or maxes out. If the thermistor is still working but is detached from the hot end the firmware will have no idea and you will get runaway heating. The Thor firmware mod looks for a sudden drop. To be robust the drop trigger value needs some tweaking though.

    In my E3D setup I have the heater cartridge side toward the print fan duct so the thermistor is on the front left side. I only have a small ammount of the Permatex Ultra Copper RTV over the thermistor and PTFE covered leads. With this I see no change in the temperature readback even when the fan goes to max for bridging. Having more RTV around the heater could help keep the melt zone more stable but it sure makes a mess.
     
  19. Randseed

    Randseed New Member

    Joined:
    Apr 2, 2014
    Messages:
    9
    Likes Received:
    0
    This almost makes me want to engineer a setup where there are two temperature sensors that average each other. This is easy enough to do with the existing hardware.

    Where do you guys go for mail order hardware parts in the US?
     
  20. Mike Kelly

    Mike Kelly Volunteer

    Joined:
    Mar 11, 2013
    Messages:
    6,967
    Likes Received:
    2,276
    McMaster, Amazon, Mouser

    more 3d printer specific

    printedsolid, filastruder, hotends.com, e3d-online
     

Share This Page