Project Night on Wednesdays // 2 Dec 2015

[Marek Kuziel]

Hi everyone,

Here is a summary from yesterday's `Project Night` session.


3D Printing

Winston came along with disassembled extruder part of the `i3` printer. The new part from ABS has a smaller nut-locker so it'll require a bit of fiddling to get the nut in properly. It's easier than printing a new bracket with adjusted sizing. (See `i3.jpg`)

I have re-done PID tuning on `DMvII #71`. Values were slightly different than last week. The only thing that changed since last week was room and initial hot end temperature. This time the hot end was colder which resulted in better end values for the PID. The hot end seems to operate correctly with the new values.

After that I tried to re-level the base and stop position for the hot end. What a mission. I did manage to get two partial prints out of it. In both cases the extrusion either stopped in the middle of the print or the leveling was just plain wrong. The conclusion is to throw out the bed completely and replace it with something that can be actually leveled.


Analogue Computer

Andrew's long term project to create purely analogue computer is coming along really nicely. It's great to see the progress over the time. I was a sneaky bugger and took a short video of one of the panels that Andrew had on the table. Enjoy here: https://www.youtube.com/watch?v=tLoKNxKPXbo


Arduino vs Chromebook

Neil (the other one) came along and worked on connecting Chromebook and Arduino. News from "Today I Learned" department: Chromebook is based/done via Gentoo. Who knew? (See `chromebook-arduino.jpg`)


Website Building

Simon came along for the first time. He worked on a website for his club. Being the "you work with them computers" guy, he was tasked by his club to make a website for them. He's using Wordpress.


Committee Meeting

It was the first Wednesday of the month and so we had our regular (*brief*) committee meeting. Nate, Quentin, Andrew and myself attended. Kevin, Callum, Rob and Pete were missing. The main points discussed and agreed on:

* The Foundation will reimburse everyone who has been putting money into the Foundation before we buy new tools etc.
* Paid membership will be offered in January and valid for 12 months.
* Paid members will have right to vote and borrow equipment and not pay gold coin donations for sessions.
* Members will need to sign a waiver ie. you are responsible, not the Foundation if you want to borrow any tools/equipment home.
* From January, gold coin donations will be mandatory for non-paid members. First visit free.
* Inventory of tools/equipment needs to get up-to-date.
* Last Wednesday `Project Night` will be on December 16 and first `Project Night in 2016 will be on January 13.

Marek

[Neil Glasson]

Looks like progress is being made.  Sorry I couldn't come last night and next week is looking unlikely too.

When I find nuts are too tight in 3D printed parts I usually file down the flats on the nuts.  3D printing often produces undersized holes - particularly when holes are very small.  If holes are circular I fix them with a drill.  If they are hexagonal I either fix them by designing the model accordingly or filing the flats down a bit.  You can also insert nuts with a soldering iron.  It might be possible to adjust calibration of the 3D printer to achieve dimensional perfection but I prefer to run a bit hotter than recommended so that I get good layer to layer adhesion and stronger parts.  I also generally prefer holes to be too tight rather than too loose.

Regarding bed leveling - I don't think it is necessary to throw out the bed.  It might be worth talking with Diamond Age and finding out what the design intent was with the Y axis support parts.  There are zig zag slots in the plastic bits under the bed.  I suspect these may allow some form of leveling adjustment.  Also the bed is just velcro'ed on.  If we just drilled a hole and fitted a small screw into each corner of the support plastic, these screws could push up the bed in each corner giving a rigid vertical constraint in each corner while not separating the velcro.  I think the thickness that the velcro grips will be a bit variable which will be causing your issue.  Because the MDF frame is very rigid and precision CNC routed, unless the Y axis rods are bent, the Y carriage will run true.  The only thing that can disturb the bed level is variation in the velcro grip thickness.  Variation from side to side (in the X direction) is corrected by adjusting one Z stepper relative to the other Z stepper.  Extruder clogging will hopefully only be a case of finding the right settings and assembly technique.  When all is sweet, you should easily be able to push filament thru by hand.  Hope this all makes sense.

Cheers

Neil

[Marek Kuziel]

On Thu, Dec 3, 2015 at 4:59 PM, Neil Glasson <boy...@gmail.com> wrote:

Looks like progress is being made.  Sorry I couldn't come last night and next week is looking unlikely too.

No worries. I also cannot make it next Wednesday, because I am off to Wellington to attend Kiwicon, so Andrew will be hosting the session.

I feel quite happy about the PID tuning progress. The rest will follow if we keep doing baby steps and improve everything that needs to be improved step by step. Overall, for me personally, this is a great experience in learning the alchemy of 3D printing. The amount of variables that all work together to deliver a *good* print is staggering.

 

When I find nuts are too tight in 3D printed parts I usually file down the flats on the nuts.  3D printing often produces undersized holes - particularly when holes are very small.  If holes are circular I fix them with a drill.  If they are hexagonal I either fix them by designing the model accordingly or filing the flats down a bit.  You can also insert nuts with a soldering iron.  It might be possible to adjust calibration of the 3D printer to achieve dimensional perfection but I prefer to run a bit hotter than recommended so that I get good layer to layer adhesion and stronger parts.  I also generally prefer holes to be too tight rather than too loose.

Yep, that's what Winston is planning to do: file down the flats on the brass nut.

I suggested heating the nut with a soldering iron, but the idea didn't get enough support. I guess because it's a "one way street" when you do so :-)

The big gear on DMvII #71's extruder has its nut locker quite loose and you can tell why it is better to have them as tight as possible.

 

Regarding bed leveling - I don't think it is necessary to throw out the bed.  It might be worth talking with Diamond Age and finding out what the design intent was with the Y axis support parts.  There are zig zag slots in the plastic bits under the bed.  I suspect these may allow some form of leveling adjustment.  Also the bed is just velcro'ed on.  If we just drilled a hole and fitted a small screw into each corner of the support plastic, these screws could push up the bed in each corner giving a rigid vertical constraint in each corner while not separating the velcro.  I think the thickness that the velcro grips will be a bit variable which will be causing your issue.  Because the MDF frame is very rigid and precision CNC routed, unless the Y axis rods are bent, the Y carriage will run true.  The only thing that can disturb the bed level is variation in the velcro grip thickness.  Variation from side to side (in the X direction) is corrected by adjusting one Z stepper relative to the other Z stepper.  Extruder clogging will hopefully only be a case of finding the right settings and assembly technique.  When all is sweet, you should easily be able to push filament thru by hand.  Hope this all makes sense.

All makes sense. Thank you. I guess my suggestion for "throwing it away" was a bit exaggerated.

You are right about the velcros. Those are the biggest source of errors when it comes to the bed levelling. They need to go first.

Also the M12 rod that sets the point for where the hot end stops by pressing down the Z end switch is something I'd replace with M6 or something that will give us more precision. M12 is a shocker when it comes to fine tuning how much above the bed the hot ends sits. It has a huge impact on quality of prints.

I really like how Wilson (http://reprap.org/wiki/Category:Wilson) does its bed. It's MDF with four screws+springs in each corner with a hot bed on top. See http://reprap.org/mediawiki/images/archive/f/f8/20140223190151!Wilson.jpg

That way the level can be really fine tuned with a very little error.

--Marek

 

--
You received this message because you are subscribed to the Google Groups "The Foundation" group.
To unsubscribe from this group and stop receiving emails from it, send an email to chchspace+...@googlegroups.com.
To post to this group, send email to chch...@googlegroups.com.
Visit this group at http://groups.google.com/group/chchspace.
For more options, visit https://groups.google.com/d/optout.