When I built my TOM I insulated the thermocouple with kapton tape and though all was well, and maybe it was. Yesterday I thought of grounding the thermal core (without coming anywhere near the thermocouple as you can ground one of the supporting bolts). When I started the bot again my temperature readings were off, and different when the heater was heating versus when it was not. I had only one explanation: the thermocouple was not really insulated from the thermal core any more. I removed it, added more kapton tape, put it back and sure enough temperatures are back to normal.
So not grounding the heater core is really a disaster waiting to happen because when one of the power (heater) resistors craps out you may well blow the extruder board through the thermocouple.
I'm very glad I grounded the heater core and strongly advise everyone to do so! Makerbot should really add this to the build instructions.
(And once you ground the core it's easy to tell whether the thermocouple is well insulated: if it isn't the temperature reading will not correspond to the real value (and if you ground it to the ground of the extruder board like I did the reading will differ depending on whether the extruder is heating or not).
insulated the thermocouple with kapton tape
Although that's the Officially Blessed Technique, it can fail without warning. The thermocouple bead will eventually punch through the tape under pressure from the clamping screw & washer. That's happened on my TOM and, from what I've read, to about half a dozen other folks.
There are very few good solutions to that problem, but the best suggestion I've seen so far is to epoxy the thermocouple bead into a solderless wire lug, then attach the lug to the Thermal Core using the screw and washer, under an insulation blanket.
I cast an epoxy tombstone around the bead on my TOM and it's holding up surprisingly well, but even JB Industro Weld epoxy is barely adequate for the task (it's rated for 260 C). One of the comments on that post has a link to a picture of the solderless lug trick; I like it!
In both cases, I think the thermocouple will read slightly lower than the actual Thermal Core temperature, but after you find a temperature / speed / flow combination that works for you, run with it. There's no need for an absolutely accurate thermometer, just one that's repeatable.
How would one ground the heater core? Clamp a piece of copper to the bolts and then run that wire to like… the case of the power supply or what?
I used a length of husky stranded hookup wire, connected with a solderless lug and a second nut to a Thermal Core support bolt:
That also shows the thermocouple in its epoxy tombstone.
On the other end, I kludged up a Molex connector with all four pins as ground, connected to (you got it in one!) the ATX power supply case. Some chunks of 12 AWG copper wire served as pins:
That wire won't carry current until the resistor fails, at which point it dumps the current to a safe spot without passing through the power supply's normal wiring. The connection also serves as a static discharge drain, which you definitely don't want to route to the circuit common through any wires going anywhere else.
My static control post has details on some other stuff to ground while you're at it, although I'll admit those are off-scale low probabilities compared to everything else that can cause problems…
Thanks Ed! That's a good idea with the molex connector, I will have to do that, but first I am going to check work and see if we have a smallish grounding bar maybe I can use.
Now, I used that Omega cement to seal my Thermocouple, but it didn't really work out well, so right now, all I have between the core and the Thermocouple is a piece of kapton. This worries me about it getting eaten through because of the washer, I had a nightmare trying to get it to stay under the washer too. Now I have no idea what the condition of the tape is any more. Grounding the heater core should protect my extruder controller in the even of another resistor failure? Or should I make certain that the thermocouple is electrically isolated as well as having the ground? I definitely want to use one of those rectangular speed nuts.
Ed's picture shows nicely how to attach a ground wire to the core. (It's exactly where I hooked up a wire as well and I'm glad to see I'm not the only one who uses whatever is available even if it's not a black wire for ground.)
When the power resistors are fine they are insulated so all is well from that end.
When the thermocouple touches the core (because contact was made through the kapton tape) you will notice because the temperature readings are off.
I noticed that the temperature reading jumped each time the heater went on or off. That's because the ground isn't exactly 0.000V everywhere. The slight offset that changes when the heater is on or off is also picked up by the chip measuring the output of the thermocouple and results in the jumps.
For instance: I am heating and read 60 degrees, then set desired temp to 0 (heater goes off) and suddenly measure 80 degrees. That's easily as much of a jump that can be caused by the tiny fluctuation of ground level. I insulated the thermocouple again and readings returned to normal: no jump when the heater goes on or off.
So the control panel's temperature readout is an easy way to tell whether the thermocouple is still insulated or not.
Now I have no idea what the condition of the tape is any more.
Most likely, it's been punched through. Just guessing, though. [grin]
Grounding the heater core should protect my extruder controller in the even of another resistor failure?
That's the belt in a belt-and-suspenders operation. The thermocouple is also grounded, so any protection depends on how the current splits between the ground lead you added and the thermocouple's ground connection. Given an essentially unlimited current source (the ATX supply) directly connected to the Core, even a small fraction of that current directed into the EC could be tragic.
There's also the ground-loop problem that causes erratic temperature readings. That's not a failure, exactly, but does show what happens when ampere-level heater currents travel next to microvolt-level signals.
Or should I make certain that the thermocouple is electrically isolated as well as having the ground?
That's the suspenders part of the operation [grin]. That way, you ensure the thermocouple won't get clobbered by anything else that happens to the Core.
I found that the round washer cuts through the kapton tape, so I used a small (440) rectangular speed nut from the hardware store instead. It's arched and has some spring to it so it captures the kapton taped thermocouple nicely without digging in. I've also grounded my core which is a good insurance policy.
This 440 rectangular speed nut?… It will thread onto a metric bolt that normally gets attached to the Makerbot heater core (I forgot the size, but it's tiny like a M1 x 3 mm)? I really want a better solution for securing the thermocouple, but
I also want to use the correct hardware that fits.
That screw is a M2- also the same size found in many laptop computers hold the case together.
I bought a Mk5 in pieces from MakerBot and had trouble finding those screws (same size for the resistors). My little cup of laptop screws became quite usefull again.
Hmm…. 2mm then…. I bought the 4-40, it doesn't thread on, but the socket cap holds it down so it's all good. Now, just have to wait for the weekend for big maintenance day. Going to ground heater core. Re-isolate thermocouple. Put this rectangular speed nut on. Change out power supply. Replace Conveyor beltless ABP with HBP. Test out Aluminum plate build surface. Test out Glass plate build surface. Do same wire management. Install some fancy LEDs. And finally, replace my Wade's style tensioner (which functions but looks ugly) with a more precisely printed Wade's style tensioner. Setup my new printer cart. Do some oiling and screw tightening.
I think that's it…. this post was for more my benefit than anyone elses. Now I have a list written down ;P
I would add a fan to both my x and y axis motors-after my recent long print failures which I attribute to heat building up in the motors even after setting and double checking the currents. It can't hurt.
Like a Heatsink fan combo? I have some Heatsinks laying around. I've been working up a way to mount one to the Z and Y. But every bracket I print ends up being slightly wrong in some way. I don't know about a fan on X stepper…. how can you fit anything in that tiny spot?
@Renosis
I grounded my heater core as you suggested, running a stranded wire and a solderless lug from the extruder plate down to the a screw in the outside case of the PSU. Oddly, when I connect the ground wire to the plate, the temperature shoots down to 0.
At first I thought there was a break in the kapton tape around the thermometer and a short somewhere. I removed the thermometer completely from the extruder and reattached the ground and sure enough, the temperature shot down to zero again! I applied more tape to the thermometer just to be on the safe side and slid it back into place. Just for kicks, I attached the ground again to the same result.
Any ideas what might be causing this?
