A few beginner questions - mainly about insulation and thermistor mounting

Hello all,

I'm now building my first reprap (mendel prusa), but I have a few questions that I haven't been able to find clear and decisive answers to.

1. Should/can thermal paste be used to help mount the thermistor and resistor on the hotend, or is kapton tape enough?

2. Is thermal paste needed when mounting the hotbed thermistor?

3. Should there be insulation under the hotbed? (The hotbed is mounted 1cm above an aluminium plate)

4. Is a cooling fan needed for the J-Head MK IV-B hotend? ( It seems so ) But how and where should it be mounted?



If it helps you can see the current state of my build here:



I hope a kind soul can help me with some good advice here - thanks a lot :-)

Edited 1 time(s). Last edit at 09/15/2012 10:05AM by marlaur.

1. Thermal paste is helpful when mounting the resistor and thermistor in the hot end. If the resistor fits loose at all, wrap some aluminum foil around it before inserting.

2. Thermal paste is not needed when mounting the heated bed thermistor, but will probably help you get better temperature readings.

3. Insulation will help your heated bed get up to temperature a bit faster, but is not required.

4. A cooling fan is definitely needed if you are printing with PLA or if your carriage and extruder are made out of PLA. It can be mounted on the right side of the carriage as seen in your image. There should be a couple nut traps there that you can use to attach a 40 mm fan.

Final note: if you have problems with Z wobble (apparent as vertical banding on your prints), you can try removing the stabilizers at the top and bottom of your Z threaded rods.

---

Help improve the RepRap wiki! Just click "Edit" in the top-right corner of the page and start typing. Anyone can edit the wiki!

Thank you!

I've been soldering all weekend and hopefully I can already start printing next weekend.

I do think I need to find some kind of insulation to put between the hot bed and the aluminium plate. The PCB is warped so it doesn't fit my glass plate.


NewPerfection Wrote:
-------------------------------------------------------
>
> Final note: if you have problems with Z wobble
> (apparent as vertical banding on your prints), you
> can try removing the stabilizers at the top and
> bottom of your Z threaded rods.

If possible at all I would like to keep the stabilizers at the bottom - they hold all the weight. I don't like the steppers to take any force other than rotary, for people making CNC's that seems to be a big nono. Of course their steppers experience a lot more force than a 3d-printer-stepper would ever do... Also I could imagine that changing the printed couplings with some of a better quality could help if wobble turns out to be a problem.

Thanks again :-)

I'd place insulation around the periphery of the sandwiched pair. When a print is finished, you can remove a section and blow into it (or position a fan to blow into it) to cool the heated bed down more quickly, leading to faster release of the printed part.

you can pick up "real" 5mm-to-8mm shaft couplers on ebay for $5-6 (US) ea, although unless you get lucky and find them domestically, they take quite a bit of time to arrive from china. Personally, I think they're worth it and chose them over the printed versions for my impending build..

xiando Wrote:
-------------------------------------------------------
> I'd place insulation around the periphery of the
> sandwiched pair. When a print is finished, you can
> remove a section and blow into it (or position a
> fan to blow into it) to cool the heated bed down
> more quickly, leading to faster release of the
> printed part.


Interesting Idea. Thoug, I would have to make small islands in the middle as well, since that is where the pcb is warping the most.

Is much time saved by doing this?

Is the chance of cooling the glass plate unevenly not increased?

> you can pick up "real" 5mm-to-8mm shaft couplers
> on ebay for $5-6 (US) ea, although unless you get
> lucky and find them domestically, they take quite
> a bit of time to arrive from china. Personally, I
> think they're worth it and chose them over the
> printed versions for my impending build..

Are you using the ones that looks like springs?

Are your couplings and steppers alone holding the weight from the entire x-axis?

If your warping is so severe that you can't bring it back to flat by applying a glass plate over it, then if it were me, I'd contact the seller and ask for a replacement. They should not be warped so much that they have to be sandwiched on both sides to bring them to flat. Mine has some modest warp, along the diagonal, but not *that much that a glass plate, the corner mounts, and a bit of reverse warping (basically flexing the board - carefully - across the warp) won't clear it up.

I suppose you can put some form of interior buttress in to shore up the center of the PCB, but I wouldn't bother using insulation for the purpose, you'll likely need something with a bit more rigidity like an aluminum hex standoff. Insulation is meant to be "fluffy", not compressed to the point it can hold weight...

The time saved? A few minutes by manually blowing into the area from testing the theory out on a friends' machine (My printer is still a work in progress). With a properly mounted fan and tweaks to the firmware to control it (turn it on following print), I'm guessing (guessing) that the ~10 minute natural cooldown *with insulation in place can be cut nearly in half.

Yes, the couplers I bought are the simple cut cylinder design. Far better than printed in my book and worth the few extras dollars and wait time, but maybe I'm just a bit of a tech snob...idk idc. . Technically, the screws hold the weight of the extruder translation stage The couplers simply merge the motor shaft to the screw, although I see your implied point, since the screw literally hangs from the motor shaft.

The hot bed PCB warp is not too bad. I've only tried to hold the PCB and glass together using my fingers around the edges, and as far as I can see the middle of the PCB is not touching the glass when I do so. I might get better results using those paper clamps that are so popular.


> Yes, the couplers I bought are the simple cut
> cylinder design. Far better than printed in my
> book and worth the few extras dollars and wait
> time, but maybe I'm just a bit of a tech snob...

Actually I've thought about using couplings like that from day one of my build, but in all the stores where I bought my other parts they were sold out. So I ended up just using the printed ones - for now. It seems like I might have to import from the US or China in the end.

> since the screw literally hangs from the motor shaft.

And this works fine? Interesting. Couplings are often a bit springy so I imagined it would be like hanging the whole thing from rubber bands, but perhaps the weight is enough to keep it steady?

Well, they're not that springy! They do have a bit of flex (since that's their purpose), but axially they seem to be quite stable overall, so since the weight they carry is constant, and their motion is pretty well constrained to a momentary action of only about 60 degrees for a 0.2mm layer height during layer transition, I don't think it will be an issue. The most I could see is a having to insert a brief pause into the firmware after each layer transition to stabilize, presuming of course that I encounter issues with mechanical ringing from using them. I see them in use for the z-axis quite a bit in photos of printers, so I suspect that aside from academic concerns, it's not an issue. But nonetheless good to know so as to consider if any issues do arise...

As far as "bonding" the PCB to the glass, you might also consider using kapton if the clips are a nuisance. I wish I could machine the glass plate so as to apply the thermistor to it instead of the MK board (and still mount flush against the heater board). Also thought about adding another layer between the MK-1 and glass of thin lexan through which I could feed the thermistor onto the glass, but I think it's too poor a conductor.

BTW, after writing last night, I put that one suggestion about warp to action and reduced mine considerably just by gently flexing the board a few times against the warp. It's very close to flat now. ~maybe 1-2mm off flat. I'd recommend it, since you apparently have warp across both axes. And afaik, central bow will not be solved by perimeter mounting methods of any sort. I've heard too much about the plate breaking, so I wouldn't recommend using adhesive to mate the two either, (ie, don't get clever and use high temp RTV to glue the glass plate to the heater board) since you may need to replace it on occasion.

Quote

since the weight they carry is constant

When the extruder moves from one end of the axis to the other the weight distribution between the two shifts dramatically. I would expect the axis to see-saw slightly.

---

[www.hydraraptor.blogspot.com]

Perhaps I should have said "near constant". Personally, I think deviations in weight distribution from side to side will be inconsequential wrt the couplers. I could be wrong though, and if it becomes an issue, I'll print some of the plastic couplers. No biggy. The couplers can be re-purposed if they don't work out.

Edited 1 time(s). Last edit at 09/18/2012 08:45PM by xiando.

If printing with ABS, bed insulation will help keep the bed temp stable, not necessaraly for speeding up the initial heat up stage. I noticed that before I insulated, the temp would wonder around 95-105C, now that it's insulated it's about 109-111C during the print.

I think the z axis mechanism works better with the motor mounted on the bottom as in the Mendel90 and the Prusa i3. In my Prusa i2, I have the motor in the bottom using Nophead's coupler. With the z threaded rods resting on the motor shafts, I don't have to worry about the x-carriage dropping when the rods slide inside the coupler.

Those aluminum couplers are not the best idea... the z lead screw is what holds the X axis up, and those aluminum coupler are springy enought to influence the z height and you will get random banding because of the linear rails binding just ever so slightly, I'm looking at an alternative, but will more than liking go with a "solid" solution as there needs to be ZERO angular change at the motor coupling, only the X axis nut, which is 75% of the time 75% away from the the top of Z, also if your properly (physically) calibrating things with your fingers before any lead screws are installed you will be suprised at how much things bind, if you bought a printer or built one, and simply "measured" things... that isn't good enough, I would tear out your lead screws or belts and move things with your hands to make sure everything is smooth before you move forward.

You go for it BAM. No one's stopping you from doing whatever you like. I will be using my motor couplers, but thanks.