Two steps forward, then subtract one

Well first off I made a mileston on Tuesday night at Tangleball, I fitted the stainless steel rods and mounted the cross beab attaching both linear bearings, plus came up with a novel apprach to afixing the belt to the bearing, more on that later.



Above is when I mounted the rods, and below is when I attached the belt to the linear bearing.



Did you notice how I did it?

Here is the cross beam


Compared to my sketch in Blender.


That was last night, and it was a hot night too, and I left Tangleball late, so I left the 3D printer in my car. What I did not think about was that I also had to go to the mall after work to do my meager Christmas shopping, as I had to get some bits for my friends' Christmas present which I will drop off tomorrow night, yay, another 2 hour drive there and 2 hour drive back, hope she will like it all the same.



Anywho, back on track, I had to leave my 3D printer at work to save it from being damaged or stolen while I went shopping.
I'm not popular with my boss about doing that, because people are gandering over it, trying to make sense of cocaphoney. Though I did iinvite one guy over to give some feedback, but moments before he did, I peered at it as I was heading from the bathroom, and I noticed something, something real bad.

The rods sag under the weight of the Z axis, greeeat, just what I need!

This is on top of the Y axis needing an extra 608 bearing, larger holes for the other linear bearing so that there is slack enogh to attach springs from one side to the other to pull both linear bearings into full contact with the rod.
Sorry no pic or video because the sag is so timy that the camera won't pick it up, but its a sag of 2mm.

When the guy came over, he suggested making L brackets to mount more 608 bearings on the top aluminium L angle, to offset the weight of the Z axis off the rod.

On all the other ideas, this one is the most doable, and I cam make it with some of the tools I will have available outside of work over Christmas, damn Christmas, ah well can't complain too much, as I need the rest.

current way of making this L bracket, which I will work on sketching up tonight, again, sorry, no photo, is a length of flat aluminium with holes at strategic locations, then bending it in a vice, obiously the 608 bearing will need to be in mounted in a slot for the screw to pass through and the whole thing to move up and down, though I feel I will need more 608 bushings, and if I want any more of those to be made up, then this Friday is my last chance before Christmas.
Once that has been done, I will need to attach yet more springs to equalize the load pressure off the rods.

I still need to mount the Z axis to the bracket and constrain t down with a bearing also, either one of the spare 625 bearings or a 608 bearing, again i've not thought this out completely so it will have to done on the fly.

Y is it taking so long?

So I wanted to give an update to my 3D printer.
Last time I said I wanted to have the Y axis completed this week, which may not happen as I have set time aside to do Christmas cards, so the intial plan of of heading down to Tangeball, and drilling holes for the stainless steel rod into the wood and making the cross beam will have to wait till next week, Tuesday night.


Here is most of the parts except the cross beam and the other stainless rod, which is cut to the same length.

 

It may not seem like much, but to redesign the the linear bearing in such a way it's under engineered to allow for my many mistakes.

Still plan to use 608 bearings with a busing inside so that I can use an M6 screw, the reason for using an M6 screw is that the thread pitch is enough to support the weight of the Y and Z axis  withouth worry of  thread stripping, as if I had used 8mm thread in the 3mm wall L bracket, it would just have 1 and a bit turnes in the thread, with 6 mm I have 3 turns of thread in the 3mm wall L bracket.

I took the oppotunity to do a short video explaining my crazy ideas, I can't help it, I love making robots, this project is no exception.

So i'll mount most of this this weekend, and glue on the pulley to the motor with LocTite 680 at work on Monday, and drill holes on Tuesday night.

Then it will be onto establishing the hight of the extruder to the bed, personal time frame to do that is mid to late next week.

I'm also going to have to take the half day from work on Friday to take up to my friend in Whanagarei her Christmas present.

Is newer better?

Well I decided to download and try out the latest and greatest host software and firmware.

I set forth to do the normal changes, but couldn't for the life of me figure out why the host would crash each time I wanted to jog an axis.  It was rather stupid of me, I cleaned the 5D gecode -> Interpreter and reinstalled the files, again, and again.  Finally after posting on the forum, I tried something else out, delete the applet directory, as this is where a fresh compile of the code goes.

Presto!


It worked, powered up the machine, then the host software.  Doing it this way instead of the other way around which I had been doing just fine for the old firmware worked.  I hate to think wat will happen when I use the bluetooth at the end of the build cycle, as I did have it going before quite reliably.

I also upgraded my Java version, I hadn't done it because of frequent host crashes, however another recent post suggested editing the repRap.bat and change "Java" to "C:\Program files\Java\bin\java.exe" or where ever it is located.

That fixed that long outstanding issue.

Still have to restart my computer after exiting from a crash or even a normal exit.
It is still unknown if this is a problem with Java, or its implementation in the host, my feeling its the host, but what I just don't know.

Lastly I found that with the atest firmware it wants to home too quickly for my flatbed scanner, this I believe is due to the acceleration in the firmware, sadly because it is desgned for equal stepper motors all round, which I don't, I want this turned off anyways, which I have done now in the firmware configuration file.



Forwards and nwards to the next problem, redesigning the Y axis, as folding the internal flap didn't work out, I did this at work when I wasn't even thinking, I was just too tired and not with it.



Which was ok, as I had to wait about 15 minutes between charge cycles when fixing units anyway, which had charging faults.

Using this method for the Y axis, adds strength and alows for an easy way to attach a bar so that it can reach an endstop if I need to, plus affixing the belt onto it as well.
http://forums.reprap.org/read.php?12,59159

Been too busy to blog

Although I haven't said anything for a while, its not to say i've been idle, rather the opposite, between real life, work, a robotics competition etc, i've only just got back into it again.


So I mounted the partly built Z axis, and found a major issue, forcing a redesign of this.



So I sketched up an idea, using 608 bearings


And made a mockup of it.


Then I went back to working on finishing the Z axis.  I made up two blocks, 1 threaded drive block, the other a guide block, and no matter what I did, I got binding, so I am still working on this, even today.
The photo shows the severe misalignment contributing to the bidinding issue, also I have to get new blocks made.

To the right of the photo is another view of the mockup of the linear bearing for the Y axis, working on this at the same time as the Z, and using the suggestions people have made, I am now making  them.



I am still not happy with the bending I had done in the mockup, so before I do this for real, I reckon aother test is in order.

Tweaking the extruder

Interesting developments this week, managed to get the extruder to extrude PLA at a consistant rate without stalling, but only through the host test extrude, still this is big step forwards.

I've also begun work on redoing the Z axis.

Before I get too far there is still one little thing that needs doing, setting the value of the extrusion rate in firmware for the motherboard, E_STEPS_PER_MM.
For the standard Mendel with a splined shaft this value is 0.706, while Adrian B. has set this to 2.0 for his geared extruder, though I found out last night that this number is wrong, its too low.
First I had read this blog post
http://www.brokentoaster.com/blog/?p=358

After which I had to ask Adrian if this is equation he used, which it isn't, instead Adrian says that its

1 / (PI . Dpw / Stepper steps) x ((ratio of filament in to nozle dia or Fd / Nd)squared)

If 

PI = 3.14159265

Dpw = 5mm

Stepper steps = 400

Fd = 3mm

Nd = 0.5mm

Comes to 0.707 which is wrong, so I fiddled the equation a bit to come to

((PI x Dpw / Stepper steps) x ((ratio of filament in to nozle dia or Fd / Nd)squared)) / 2

I get 0.706858, and the note in the firmware says its emperically adjusted, so by taking 3sf past the dp you get 0.706.

I don't know how correct this is yet, so i'll update the wiki later

Some Progress on the extruder

After more than 50 iterations on my "JunkStruder" as I like to call it, a major delay in acquiring fibreglass coated Nichrome to overcome the intense heat needed to extrude PLA, ~275 degrees Celsius seems to be working, but not by the stepper yet.

I did have a big problem that I was very happy in my decision to use a BakeLite thermal break was a good one.
Essentially I broken the filament right at the top of the thermal break, gah!  But when I applied a butane torch flame, yeah that's right a butane torch right on the thermal break, I stuck down a piece of filament so that it welded together, then just turned on the extsruder heater to remove the stuck filament.

Thanks to Paul Marson, a mate of mine who suggested using a pot handle, turned on an a lathe to produce a thermal break, any other plastic like PTFE or PEEK would have melted.

On Friday night, cause I really don't know how to use a lathe, yet, Paul turned out a piece of stainless steel tube that I got from the hobby shop a couple days before, 1/4 inch OD 0.049 wall, OD turned down to 6.0mm and ID drilled to 5.0mm as i'm currently using a MindKits stepper motor rated at 23Kg/cm or 2.3Nm.
Why turn down such a tube, was because to make up the bulk of a package from the US, IE the fiberglass nichrome wire, got some nozzles and a new gear wheel that does indeed grip the filament much better, at the same time, and another stepper motor all from MakerBot.

Reason this post is so, so delayed is because the minor iterations i've been doing are so minor that its not worth blogging about, and also, the recent package too over a moth on a 10 day priority post from the US.
I'm so angry with USPS, that they can't even read the address label properly, which clearly states New Zealand not Zealand in the Netherlands, so yeah, it went to the Netherlands.

A chat with MakerGear also revealed that I can get more stuff I require for the hot end from them, including Wade's Exstruder, if anyone wants one, plus their heat core looks very cool.
Right now, i'm out of cash, and visa is quite high after paying the quarterly bills as well.

I hope to have the extruder running, and post a video of the iterations i've gone through to get it to work, though current experiments show that I have the RepRap host set to 300 degrees and 275 degrees read from the thermocouple just to extrude PLA, but at least the fumes from PLA are much more bearable than Kapton tape adhesive, smell like burnt sugar. :)

Little more done

Small update to where I am at.

About the dead extruder.
Fixed:  Due to frequent crashing of the RepRap host, the extruder goes way over temp, this eventually caused oxidation on the contacts, so I have cleaned it up, and made fresh connections again, and the hot end now works.
Reading on the forum someone mentioned about just hooking up the two lines from the motherboard that goes to the extruder board, instead just hook them onto the stepper driver 1.2 board that I have, and it works, what a relief that I don't have to fudge the silly Arduino code.  I forgot that the enable line is automatically held in an "on" state.

The opto end stop and the steppers only driving in one direction problem.
Fixed:  I am a douche, NopHead told me that the driver boards NEED the endstops, even in "dry" mode, so I made one endstop (opto), and the board its connected to goes both backwards and forwards, thanks to him that problem is solved.
The second opto endstop failed, I just couldn't get it to work, so instead I rummaged through my parts *ahem*junk*cough* bins and found a few tiny microswitches.  I bent the lever so that there is no chance of the top of the lever biting into the plastic of the flatbed scanner, wired most of it up, just got to solder on the ethernet cable tomorrow, and th x axis will be ready to go backwards and forwards.
I felt that with the x axis flatbed scanner having the opto endstop on the moving carriage was going to be a bad idea anyway, just gave me the final nudge to use a microswitch, and have it glued down on one fixed end.

One thing has gone wrong, and i'm out of cash, quite literally, on one of my V2.3 driver boards the chip is blown, however I can get one or a few from the local BIG electronics supplier RS components, but as I said I have no spare cash, so the Z axis has no driver for it at the moment.

Thats not a really big deal right now, as my attetion is getting the extruder completed and running, so I can bolt it to the Z axis, and finish the mechanical part of the Z axis.

Next week, I suspect that Tim Carr from MindKits.co.nz will have finished his Mendel, so i'll be oogling over that at the Arduino next week Thursday.