Wednesday, November 16, 2011

Extruder drive-screw

It seems there is only one way to mount the dremeled transport screw. That is with the nut in the gear. The gear it self imparts a slight twist to the filament, if the head of the screw is in the gear it's counter-clockwise. If the nut is in the large gear it is clockwise. This may seem trivial, but I found with changing the spin direction I got a more consistent extrusion flow. With counter-clockwise spin it was really inconsistent. I have to assume this is because the hole the filament goes down is slightly fluted from using a drill bit to clean the print... or from the drilling of the PEEK barrier. This still needs more testing though, I have another duplicate extruder almost setup. Maybe I can replicate the results with the other extruder.

Monday, October 24, 2011

Transport screw testing

This transport screw has been working well, much better than the other hobbed bolts I had made. This way of making the drive screw is a lot more consistent and quicker to make.  I'm making a cut area of about 22mm, this way the bolt can be put in either direction, with the head in the gear, or the nut in the screw. The head in the gear seems like it makes more sense to me... I tried a cut area of 20mm, this works, but when I added two washers just behind the large gear a small amount of uncut threading could be seen in the idler wheel hole, 22mm seems to be just right.

Sunday, July 24, 2011

July Fixes

So I started thinking about the endstop problem I was having while using the teacup firmware. I had wired my endstops to be normally open, when everyone else... on the planet... wired theirs to be normally closed.  a G28 would send whatever axis away from the minimum (the second part of homing), and when I closed the switch would stop because it looked like it finally got away from the minimum. Then I found this mechanical endstop schematic, and I wondered if maybe I had a tiny stroke last year... at 28. It seems I couldn't connect anything properly. 

And I found the speed issue I had yesterday was because MAXIMUM_FEEDRATE_E was set too low. Here is the now more... sensible diff.

Saturday, July 23, 2011

July 2011 Updates

       I found that the firmware I had been using would require too much work going forward and I started a switch to using the Teacup Firmware.  Here's a diff of the changes I made.  I mostly made some small changes to my hardware to match closer to the firmware standard.  It seems to work great, except for a few middling things I still have to figure out, speeds are off.  This also means that I can switch to other host software and I'm presently using the 0025 release of ReplicatorgReprap Host worked well enough, but I can debug a lot easier with Replicatorg and it has included Skeinforge for a while.  Skeinforge is the one thing I don't think I'm changing anytime soon.  I also found in the course of this updating that I'm terrible at electronic design... well I suppose I'm better than I was... because I found the problems. I was terrible, I'm somewhat better now.  I had wired my heated bed in silly way, sharing the 5v- with the 12v+, the resistors were wired in two parallel sets of series resistors, which is now completely in series 12+ and 12-.  I also had added a 5v line to each of the minimum limit sensors... unnecessary with physical switches.  I cleaned up some wiring and fixed the pwm board to add some diodes and change out a transistor.

       Earlier in the month I bought a groove mount thermal barrier from, which I found is based just about 30 minutes north of here... cool shit, on both points. I don't see any reason why to use any other design for a thermal barrier for a while. It's a great design and having tried it with PLA and ABS.  It works, and works well... I'm buying 10 once I have some money.

Thursday, April 21, 2011

April 21, 2011

I replaced one of the broken TIP120s on the pwm heater boad with a MOSFET, testing with a multimeter looked like it was working properly, so I connected it to the heated on my extruder. Testing that proved, or created, a short in the nichrome heater somewhere. Instantly smoked and fizzled. So I killed it and left it to today.  I have it dissected so far, and I did not find the short, but there was a blob of black PLA where the PWM leads connect to the nichrome wire, so it could have been in there... or something else made a terrible smoke, acrid smell, AND killed the heater... maybe a tiny Genie... for my First wish, a thousand more wishes and my Second; a simpler to disassemble hot end with a resistor heater block, and Thirdly; a peanut butter, banana, cream cheese sandwich.

Since I still have an amount of nichrome wire I'll just have to rebuild the same design.  I might as well get some pictures of what I've been doing to wrap my hot end.


I'm rethinking some things I thought were obvious decisions previously.  Well decisions is probably not quite accurate, since I didn't really think it through, which why I am now.  I've been making SAE parts... because I live in the US, one of the last foolish countries using imperial measurements. I've thought for years that continuing to use imperial measurements is rediculous, and backwards as a country.  To be the change that I want to see I'm moving personally to metric as much as possible, trying to think, and understand the world around me in metric terms instead of imperial.  I'm going to stop making SAE repraps.  Again, it's backwards and silly to not use metric.

This is also part of the reason that I posted parts online earlier in the week,  I had a whole set, and my printer fully collapsed... I'm still fixing it presently... and I had some older spare parts that matched what was missing.  So I sold it all, now I can get on to printing Metric parts, and doing the little I can to push the United States to metric.

Monday, April 18, 2011

April 18, 2011

I'm posting this set of PLA Prusa Mendel parts to Ebay in a moment... (here it is) it's been a long to trip to finishing this set.  There was a power brown-out here about 2 weeks ago, I've never found a need to buy a UPS until then. I was in the middle of a long print, 1 of 2 last parts for this set of parts, the computer shutdown, my printer's power supply developed an odd tick to it, which I found later... So after hours of trying to find out why I couldn't hold at a decent heat after the power outage, and the printer kept stopping for no reason in the middle of prints I discovered the PSU was dying, then a transistor for the heater blew up with what seemed like a thunderous clap.
I switched the PSU out, switched to another pwm channel on the heater board... everything seemed to be coming up roses. I still couldn't get the heater to work properly... anyway lots of desoldering, repairing, and resoldering... still no good solution. So the last two parts to the set were the extruder block, and 39 tooth gear, I had a few spares sitting around... They aren't as high quality as I'd make today, but it completes the set. And I can move on to rebuilding electronics.

I have a couple MOSFETs I got from digikey that I think I can use to repair the PWM heater board.  And I'm starting on a new arduino board similar to the sanguinololu in principle, but I have an atmega328p.  I'm thinking a simple Arduino, bare-bones laid out on perfboard, one heater circuit, one thermistor circuit, connections out to sensors, and stepper boards.  If I can repair the PWM heater board, this will be a start on a new 'bot, but if not; I have a replacement for the rat's nest that is running my printer now. Point to point wiring has worked, but it is certainly not intended to last as long I've been using it.

*exhale* Time to go for a run.

Friday, March 25, 2011

March 25, 2011

Using a Diskette power connector for power and thermistor

Thermistor on top, yet to be cleaned up.

Thursday, March 24, 2011

New Heated Bed

My previous heated bed was nichrome wire based, I counldn't get it above 60°C, so it wasn't usable for ABS, but worked great for PLA.  It started out as an experiment, and through various hacks I kept it working.  Finally, coming up to another necessary repair I decided to build a new bed. I bought some Ceramic wirewound resistors from Digikey, fireproof, I'm paranoid. I'm first trying to attach them to the same aluminum plate with JB-Weld. I needed a small putty knife to apply the jb-weld to small ceramic blocks, so a couple lines of openscad later:

cube([50,10,1], center = true);
translate([15,0,0])cube([30,15,1], center = true);

I have a putty knife the exact size. For now, I'm just waiting for the jb-weld to dry. If the jb-weld doesn't hold for the comparatively low temperatures needed for a heated bed... well I'll have to figure something else out...

the putty knife is on thingiverse now

Thursday, March 17, 2011

Happy St. Patricks Day

Printing a Mendel Pen Holder, hopefully I can start making my own circuit boards with this pen holder and this post about using vinegar, Hydrogen Peroxide, and salt to etch. I found the etching post through Hackaday
Shit's gettin real.

Tuesday, March 15, 2011


So, where was I... I got through all the PLA tests I wanted to do, and actually used a whole pound of Black PLA. In the end I printed a full Prusa Mendel, and better quality than most of the parts on my printer... I've posted those Black PLA parts to eBay, since I still don't have all the electronics yet for a second printer.  I decided in the course of my Prusa upgrades and switching to PLA that it's probably best to have two extruders for each printer.  I don't mean a two headed printer, but and extra extruder sitting by in case something goes wrong.  Then I can just swap in the second one and make repairs on the first.  PLA Doesn't jam as much as I was led to believe, maybe older designs did, but Wade's extruder design works great.  When PLA does jam, which it did on my old extruder, it... takes a while to fix... days.  I did have more blobbing and lots, and lots of strings from PLA, it makes a thread like filament when the extruder moves.  By the end of printing the Black Prusa I had the extrusion reversing far enough and quick enough that it mostly wasn't an issue (15mm at 100mm/s).  I think I've gotten as good with printing as I can without more exact measuring devices.  I'd like to try using a small digital scale to compare the grams output in Skeinforge to the actual weight of printed objects, and a caliper to get some measurements on plastic output vs. input. 

Thursday, February 10, 2011

Jan 21, 2011

It seems the problems I've been having with my new Z axis is because I was trying to use my old threaded rods. These original threaded rods are 1/4", the Prusa parts I'm using are 5/16"... whoops... so I'm reworking this, in addition to installing kFreeBSD/Debian on another system.  (I don't know why this didn't publish when I wrote it)

PLA tests

I've read a few places that PLA is difficult to use in the warmer places of a reprap, that it can warp, and deform.  I haven't seen anything definitive, just some anecdotes, so I have some testing to do.  I'm printing up a prusa X-axis and extruder to replace the ABS bits on my reprap.  I'm making them as a direct snap in replacement. I should be able to keep the same Skeinforge settings, print for a while and see how it comes out.
In other news I'm using my heated bed again, it makes PLA prints almost perfect.  I still get little the "character" of PLA issues; blobs, hair-like strings, but distinctly much better than anything I've done before.  Unfortunately this bed, as it is, will not work for ABS, I can only get it up to about 60C.

Wednesday, February 2, 2011

Prusa/Catalyst PRINTS!

On the left is the best print of my printer when it was fully a Catalyst. This took about an hour. On the right, in black PLA, is what just printed from the same printer modded to half Prusa. As good if not better quality in about 22mins... and it could still go faster, this was at about 10mm/s. I'll find the upper limit of the feedrate and flowrate in the coming weeks. WHOO HOO!!

Prusa/Catalyst works

So it's up and working. I was able to start with one design of repstrap, and by replacing parts and rebuilding bit by bit I now have another machine, while it functioned pretty much the whole way.  I have added my heated bed back on, that is the last thing I'm dealing with. When trying to heat the bed the reported temperature dropped... it seems I reversed the vcc and gnd to the thermistor circuit, whoops.  I haven't printed anything on it yet, but all the bits and pieces work correctly so I should just have to fix the bed thermistor and adjust some settings in Skeinforge.

Thursday, January 20, 2011

Prusa upgrades installed

it took some tinkering after installed but everything is together and moving again. It looks like I have some backlash on the new x-axis, how unexpected... so I have to turn lash on, and I'm turning the feedrate up to 10 from 5.7, trying a proportional flowrate of 8.7. trying that out in Skeinforge, oh and lash set at default of 0.2mm.

A problem to fix first, for some reason the z axis is stopping about 5mm up, makes a grinding like noise... both motors stop.

I changed it to 1/16 stepping, whisper silent, works, but still stopping in the same place. I'm going to check the captive springs, they might be jamming somewhere. It appears that the hexagonal holes the 'captive' nuts on the threaded rods are too big.

17.(optional) Insert a spring into the top of the hexagonal channel
oh, apparently it's optional I guess I could try without the springs.  In taking it apart I found the ends of the springs were jammed into the top nut on one side, and the threaded rod hole on the other.

Yup, the springs were the problem.
Pictures to come...

guh, more jamming PLA

I've decided I might have been wrong before, I'm trying 190 as the temperature again, maybe I can get the extruder idler block to work.

Out of frustration and impatience I started the print, and pushed the unattached bowden cable into the extruder after it was up to temperature... strangely that is working...

It came out, but there's a lot of strings, and seemingly too much plastic.  I'm increasing feedrate to 5.7, trying to print 39 tooth gear. I'm doing the same thing to start the print, hopefully this will not be a longer term solution... well it's jammed again... another unfinished print.

PLA working... ish

I'm trying out a difference in settings I found here. I'm hoping by also lowering my "Infill Perimeter Overlap" ratio while raising the "Infill Interior Density over Exterior Density" ratio to 1.0 I can get slightly better prints. I've had a few successful prints in 4043D PLA, and I'm continuing with that.

It appears after further attempting to print, that at present I cannot reliably print with PLA. I either need be able to print with a higher flow rate or not have a bowden cable in the middle of the extruder assembly. the heat from the hot end is travelling up the pla filament into the bowden cable where it softens and creates blockages. quicker prints would forgo this, but best would be quick prints without the cable assembly.

Catalyst meet PLA

Trying to get PLA to print, I'm working at printing wade's extruder, it's jamming on occasion. I'm turning up the the temperature 5 degrees from 190 hoping this will prevent jamming.  I was considering an oily sponge connected through my bowden cable But I'm trying out just turning up the heat a bit.  I found there was a blockage, a large blob above the thermal barrier and below the nut holding the bowden cable to the hot end.  This might be a reason to remove the bowden cable.
*disassemble extruder*

That was the jam that stopped the extruder... how to fix this...
trying to print again...

I think the problem is a result of metal conducting heat up to the nut holding the hot end of the bowden cable in place, PLA expands or simply melts enough into a blob and bob's your uncle.

I don't think it's likely that using a bowden cable with PLA is workable.

Monday, January 10, 2011

posted on eBay

I posted the ABS parts from a Wade's Extruder. Here's the listing...
 I finally have some really decent prints coming out of my repstrap. I managed to make a couple of wade extruders, I wanted to replace some parts on my extruder. Instead of saving them until I can assemble my first child reprap I'm selling them. These are the parts that turned out the best, I took the second best ones for my use.

As you can see in the pictures the motor block warped furthest from the bearing block.  There a couple dark spots in the block also. These parts are printed in uncolored ABS.  I started cleaning these up, but some of the holes will still have to be reamed or drilled.  Only what's pictured is included.  If you have any questions, please ask or check my blog.

Wade's Geared NEMA extruder
Beefed up idle bracket for Wade's extruder by nophead
Gear with Setscrew for Wade's extruder by nophead

Friday, January 7, 2011

Mendel Axis Pulleys

Telling the Truth / Lying to machines

It's likely that most of my thought processes, no matter what I'm doing, are colored by my conception that I'm an artist.  As such I'm most often interested in how to use a reprap.  To me this new tool isn't just something where I can find the best settings in Skeinforge, and click print with each new object. There are varying scales and best ways, little hacks to get best prints. I think at this point I understand enough about the process and settings in Skeinforge that I can start exploring the many ways to print. An artist is not defined by what medium they use, but how the medium is used.

I'm starting small, while within the scope of printing other repraps and reprap parts. Yesterday I was fully focused on an Axis Pulley. I have printed about 10 of these in the last couple months, never happy with the results. They come out, they're usable but it always seemed It was possible to make them better, but I couldn't articulate to the computer how.  Starting with the assumption that the hardware machine is limited, but nothing could be done to get about these limitations, I started looking to firmware and software.

top usable, but ... bottom aesthetically pleasing
From hardware up; the firmware is still a work in progress of hacking from an older standard reprap code. There were a few things I thought I could get working that would make my prints higher quality. Namely; 1. a working heated bed, 2. better extruder control 3. fewer pauses 4. quicker prints.

1. a working heated bed, I know how this works and had a low temperature heated bed working previously, but not cost effective for finishing in a day.

2.The extruder works, but for some reason... likely because I'm using old code intended for a DC motor, the stepper motor will not reverse.

3. fewer pauses... This could be fixable in just code, but the code possibilities are limited by the hardware, so this is out.

4. quicker prints. This is really dependant on hardware and the two previous points working

SOoo... the extruder firmware code; I managed to get some of the functions from a more recent reprap firmware intended for an Arduino Mega with a stepper extruder.  I cannot run all the code intended for a Mega on the much smaller and weaker Arduino 328p, but the algorithms and a couple functions could be adapted and in the end I managed to get the extruder to go in both directions. WOO HOO!

But still my pulleys looked unfortunate. So software. What can i do in Skeinforge? I played with this all yesterday, really 90% of the day went to adjusting things in skeinforge, trying different infill ratios and patterns. No matter what I did the pulleys still came out about the same, with gaps in the gears. So then I tried lying, which comes just before sledgehammer solutions. I told skeinforge the plastic coming out of the nozzle was less than it actually was. the setting originally was .48mm for a .50 nozzle, I dropped this to .45 then .42 and I started lowering the flowrate in speed from 5.6 to 5.4 and 5.3 I finally end up at a layer height of .36 and a flowrate of 5.0 an 11% reduction in flow and 25% lie in layer height. But it worked you can see the difference in the pictures above.

Sate of the Machine 2011

State of my repstrap:
It works, it works consistently, but there was much changed and adapted to get to this point. For myself and for anyone else on so perilous a path I thought I should document what finally worked to get this PITA to work.  I'm going to spend 2011 working to produce as many repraps as I can, many of the problems I've had over the last year could have been solved much quicker by having a second machine.  I have to figure out where I'll be selling the ones I don't keep, for the time being it's likely to be eBay.
the heap that finally worked

Things added;

heater board (PWM Driver 1.1)

Dc driver board

Plastruder Mk4

arduino proto shield, sorry don't remember from where

Thermistor board

Old computer Power Supply, to replace the 18v PS

washer flywheels, I needed a fly wheel on either end of the X axis threaded rod. This though I'm finding is making the X-axis inaccurate because the threaded rod overturns.

Blue Painter's tape on wood base, the original plexiglass build plate warped terribly.  But presently I'm printing on a plexiglas plate on top of my heated bed, which is off... as it doesn't get warm enough for ABS

IDE Cable; I found splitting some old ide cables where needed made for cleaner wiring

5W-30 motor oil, I needed oil... I had it on hand, probably not the BEST idea

Things lost:

2 - 3 thermistors, yea I broke a few when making and remaking extruder nozzles

Feet of nichrome wire, well I kept it and ended up using it to make a heated bed

many feet of Kapton tape, but I still have more

patience... I totally lost it a few times

pounds of ABS... it's all in a box, hopefully I can recycle it some how in the future. Or I'll throw it out when the box is full

I managed to print Wade's Extruder and have been using that with much MUCH success.  The Makerbot MK4 / Dc Driver board have been relegated to backup status.

The path forward:

I've been working on printing a Prusa Mendel out of ABS, most of the printed parts are done.  Some of the parts are being adapted right now to replace the Z and X axis on my present machine.  I want to replace the threaded rod X-axis, which means I have to replace the Z-axis. That is coming in the next couple days and weeks.  I bought some black PLA from Ultimachine a week or so ago, I'm going to try switching to PLA for a while starting later today.

I've been thinking about a modified Makerbot I might make make in the course of the year... more about this at another time.

I had been keeping extensive notes on my computer before Christmas, I'll post those as soon as I edit them for readability.  Going foward I'll be keeping my daily notes on this blog, seems like a good use of this space.