Fist attempt issues - leaking extruder, jamming feeder

Fist attempt issues - leaking extruder, jamming feeder, bed not level etc.

  1. Leaking extruder - only a small leak and manual says unrepairable - this is silly.
    I tightened it and went OK luckily.
    But root cause of problem is the lack of imagination. The extruder head needs a thin copper
    washer to prevent 3.5Nm tigtening instruction. If applying such huge force, it is likely
    a simple mistake will end in leaking extruder or sheared thread.

  2. The gap between the extruder stepper motor and the big gear needs adjustable control.
    If the gap is too large, then the play in the gears will cause plastic to leak leading to bumps
    and parts being knocked off when the head bumps into it.
    If the gap is too small, the gears can jam every now and then.
    This crucial fixed ajustment mechanism needs to changed to
    allow a wing nut on similar adjustable mechanism to finely control that gap between
    the two gears so that there is no play in the gears. Or change to a belt drive.

  3. The bed is not flat on first assembly. The screw holes for the PCB make it bulge out.
    Bending the aluminium frame sorted it out. A bit sill again. Need a screw in the middle to prevent
    the bulging.

  4. The 0.25m mm bed height adjustment mechansim is a silly screw. That needs to change
    to something with a substantial handle to allow other flat surfaces to be added and the
    adjustment to be made rapidly.

  5. The Linux software on Ubuntu 12.04 ran into difficulties with command waiting messages.
    No clear error messages and wasted several days hunting for fix.
    Stumbled on to an article to download, patch and recompile the mono so that it can handle
    the Arduino running on non-standard 250000 baud.
    Some big messages need to be put up somewhere about that.
    Again silly - no point in hiding this information from Linux users.
    And surely it would benefit everyone if the software could output meaningful error messages
    when the serial comms failed to set the non-standard 250000 baud?
    This gives a clue to developers to go look in the right places for a repair.

Happy enough for the moment and thinking of order up a few more printers :slight_smile:

I builded mine a few weeks ago and I too used copper washers in the extruder.

I agree about the bed flatness too. It is a “stupid” construction, as are the endstops. My bed was higher on the edges so opposite of yours I mounted a micrometer next to the extruder and used epoxy to glue a glassplate to the heatbed. Then I made it level but that takes a lot of time. I improved that somewhat by using epoxy to glue the nuts in the tumbwheel and then a soft copper washer on top. Now they are easy to mount and adjust better. This way I managed to get it flat within 0.01mm.

I agree to about the adjustable control for the gear. I now spend a lot of time to get that as good as possible. The manual is great but I miss instructrions about how to set te gear freeplay and how tight you need to set the springloaded screw. It seems to make a difference in result but I’m not sure in what way. It can be I’m trying to compensate an other error.

I first used Linux (ubuntu). It had some graphical problems so I could not see how the model was orientated. Now I’m fighting with the 8201

Fred

The manual says spring loaded screw set to lowest setting with minimal compression.
I tried it with more compression when the extruder gears were not optimally spaced, and it jammed more frequently with higher spring compression.
After more adjusting, seem to have found optimal point for the spacing and the gears run failrly silent and smooth.

Haven’t yet tried glas bed, but I did notice the first layer peels off from PCB regularly as soon as the job starts.
If I stop the job after 1st layer, peel that layer off and start the job again, then it seems not to peel on second attempt.

When printing a cone shape, I notice the extruder head spends too much time at the top of the cone and not allow it
to cool before the next layer is laid down. I guess print more than 1 copy to get the head moved to elsewhere.

The graphics card I got for Linux PC is the Nvidia Geforce 210 - a very cheap and very powerful card for this kind of job.
(There are newer versions out which look identical except for fewer RAM chips
for which proprietary drivers for Linux are not yet out - but the open source drivers seem to work the graphics card
but slow when running applications like KiCAD 3D rendering.)

I managed to print 2 items yesterday each lasting an hour and they seem to print well.
I guess the next item to print is the 4 hour cover for the PCB :slight_smile:

Yesterday I had luck too. It printed two 1 hour jobs. Not faultless but that is not the printer to blame but the settings or the thingyverse object (I used that to limmit the risc on faults in the design)
Two toplayers where missing (one close to the base, the other the second top around 10 cm high) But there was enough filling to be usable.

I found out my hot air rework station is perfect for correcting small mistakes.

I printed a few small things on the pcb before I had the glass. But those were small footprints.
I found out it is very important to have an absolute clean printbed. If I clean it with 70% alcohol the object is good fixed I was afraid to break it.

I noticed a big footprint seems to give more wrapping problems. My methode at this methode is setting the nozzle at 0.2. Then it makes a brim (is that the word ? An extra layer under the object after printing twice a circle around the object. I use those to check the Z hight and amound of filament.
During that I monitor the result and adjust the Z ax stepper so the result is a nice flat shape. I think I’m now there. I will measure the distance and then set an offset value in slic3er.
I first changed some parameters around the layer thickness and amound of filamen. Then I focused on the retraction and Z proces. Played with values for both and that seemed to give an improvement.

I use a laptop that runs W7, Ubuntu 12.x and since a few weeks most times Mint because it is in base ubuntu but then with gnome. My laptop is a cheap 3 year old compac 620. I run Freecad under mint, repetier too and tried 123D under wine but that does not work smooth. Sketchup under playonlinux/wine works very well. I like 123D cad the most but my graphical station only runs on Linux. My second favourite is Freecad. It can do allmost everything you can imagin but it is a lot more complicated to learn.

I have the electronics not covered.
Next thing I’m going to check is the PSU. I’m now checking, repairing and modding a powerfull ultraclean linair powersupply made by Delta that would be more than capable for the printer. (not the chinese crap but the famous Dutch Delta).

Hi pa4tim,

the extra layer under the object is called ‘raft’. ‘Brim’ are the extra outlines added to the first layer to improve adhesion to the heatbed. You will have to remove the ‘brim’ after printing has finished.

Do not expect to make your first print ‘perfect’. It takes some time to get to know your machine. Time will tell you how to improve your prints.
Just ‘try and learn’, I am sure you will get satisfactory results.

Enjoy,
BigDee

[quote=“gplsquared”]Fist attempt issues - leaking extruder, jamming feeder, bed not level etc.

  1. Leaking extruder - only a small leak and manual says unrepairable - this is silly.
    I tightened it and went OK luckily.
    But root cause of problem is the lack of imagination. The extruder head needs a thin copper
    washer to prevent 3.5Nm tigtening instruction. If applying such huge force, it is likely
    a simple mistake will end in leaking extruder or sheared thread.

  2. The gap between the extruder stepper motor and the big gear needs adjustable control.
    If the gap is too large, then the play in the gears will cause plastic to leak leading to bumps
    and parts being knocked off when the head bumps into it.
    If the gap is too small, the gears can jam every now and then.
    This crucial fixed ajustment mechanism needs to changed to
    allow a wing nut on similar adjustable mechanism to finely control that gap between
    the two gears so that there is no play in the gears. Or change to a belt drive.

  3. The bed is not flat on first assembly. The screw holes for the PCB make it bulge out.
    Bending the aluminium frame sorted it out. A bit sill again. Need a screw in the middle to prevent
    the bulging.

  4. The 0.25m mm bed height adjustment mechansim is a silly screw. That needs to change
    to something with a substantial handle to allow other flat surfaces to be added and the
    adjustment to be made rapidly.

  5. The Linux software on Ubuntu 12.04 ran into difficulties with command waiting messages.
    No clear error messages and wasted several days hunting for fix.
    Stumbled on to an article to download, patch and recompile the mono so that it can handle
    the Arduino running on non-standard 250000 baud.
    Some big messages need to be put up somewhere about that.
    Again silly - no point in hiding this information from Linux users.
    And surely it would benefit everyone if the software could output meaningful error messages
    when the serial comms failed to set the non-standard 250000 baud?
    This gives a clue to developers to go look in the right places for a repair.

Happy enough for the moment and thinking of order up a few more printers :)[/quote]

These are exactly the same points we thought about when building and calibrating our printer these past two weeks. The leaking extruder we solved by heating it up and tightening it slightly, how ever a washer would probably be better. Our doesn’t leak any more.

The gears and jamming problem we also had a bit of in the trial runs. A bit of fine adjusting with filament in the extruder solved that and it seems to be running acceptable. A bit noisy but nothing I can’t accept. A belt would probably run a lot smoother, I agree.