Another UK build

Re: Another UK build

Postby NeoTheFox » Sat Apr 05, 2014 6:59 pm

Let me suggest RAMPS-FD instead, I am using it, and it is very powerfull and usefull. Also if we could ever make all the kinematics in firmware (I really hope I could make it to auto-calibrate points for segmentize with z-probe) 98Mhz will be way better. And it cost almost like RAMPS 1.4

You just keep outrunning me with the progress :D This squrrel looks awesome, btw.
NeoTheFox
 
Posts: 92
Joined: Tue Jan 28, 2014 5:54 pm
Location: Moscow

Re: Another UK build

Postby Nicholas Seward » Sat Apr 05, 2014 7:02 pm

+1 for the squirrel print. I have tons of strange looking squirrels . My love of squirrels has translated to systematic squirrel torture.
Nicholas Seward
 
Posts: 738
Joined: Mon Nov 25, 2013 10:41 pm

Re: Another UK build

Postby NeoTheFox » Sat Apr 05, 2014 7:03 pm

Nicholas Seward wrote:+1 for the squirrel print. I have tons of strange looking squirrels . My love of squirrels has translated to systematic squirrel torture.


You should collect them all and expose them as modern art.
NeoTheFox
 
Posts: 92
Joined: Tue Jan 28, 2014 5:54 pm
Location: Moscow

Re: Another UK build

Postby timrwilliams » Sat Apr 05, 2014 7:38 pm

Neo, I hadn't seen RAMPS-FD. It looks a powerful way ahead. I'll put it on the 'wish list'

Nicholas, your squirrel comment reminded me of the Tom Lehrer lyrics -

'And maybe we'll do
In a squirrel or two
While we're poisoning pigeons in the park'

(Sorry! that should probably be another thread)
timrwilliams
 
Posts: 87
Joined: Fri Feb 14, 2014 10:32 pm

Re: Another UK build

Postby imrahil » Sat Apr 05, 2014 8:09 pm

next week I will be testing new board from PJD Automatyka (it's the same very smart guy from metal spools offer :mrgreen:) - it's controller based on teensylu/printrboard:
Image

personally I don't like sandwich-like constructions like RAMPS - now I have Sanguinololu, next one will be single-board one too.

ps. sorry if I posted too much ads in last posts :roll:
imrahil
 
Posts: 68
Joined: Thu Dec 12, 2013 10:19 am

Re: Another UK build

Postby timrwilliams » Thu Apr 17, 2014 8:07 pm

A moon gazing Easter rabbit to join the woodland gang (squirrel/owl/rabbit) - My family just don't treat this as a serious project!
I still havn't got the Z calibration right - the rabbit is too short.
I have checked each arm calibration. The linearity seems good but there is certainly some hysteresis, between 0.1 and 0.4 mm, and it seems randomly scattered over the arms operating lengths.
The Segmentizer base values I measure also seem rather variable in reproducibility.
From other comments on the forum, I guess that my springs are too weak. This is fine for running the motors cool but I have ordered some stonger ones - the present ones were just out of the spares box.
I have also upgraded the Sanguino board to an Atmega 1284 so shall spend the weekend switching from Sprinter to Marlin - with all the extras that will allow. I'm thinking bed probe.
Attachments
GUS8.JPG
GUS8.JPG (232.23 KiB) Viewed 12007 times
GUS7.JPG
GUS7.JPG (222.62 KiB) Viewed 12007 times
timrwilliams
 
Posts: 87
Joined: Fri Feb 14, 2014 10:32 pm

Re: Another UK build

Postby NeoTheFox » Fri Apr 18, 2014 5:33 pm

What slicing software are you using? Slic3r used to case same hysterisis for me, because of how it treats Z movements. Problem gone with Cura.
NeoTheFox
 
Posts: 92
Joined: Tue Jan 28, 2014 5:54 pm
Location: Moscow

Re: Another UK build

Postby timrwilliams » Fri Apr 18, 2014 6:30 pm

Thanks for the suggestion.
I AM using Slic3r. I started 3D printing with ReplicatorG several years ago but never upgraded to Cura. - Another learning curve to climb!
I discovered that the Marlin versions I have in use, do not have Bed Leveling implemented, and I have just downloaded a more recent version. So more fine tuning to come.
timrwilliams
 
Posts: 87
Joined: Fri Feb 14, 2014 10:32 pm

Re: Another UK build

Postby NeoTheFox » Fri Apr 18, 2014 6:46 pm

There is no curve, most of slicing software has the same options, and cura actually got fewer than slic3r. Good luck :D
NeoTheFox
 
Posts: 92
Joined: Tue Jan 28, 2014 5:54 pm
Location: Moscow

Re: Another UK build

Postby timrwilliams » Wed May 14, 2014 4:56 pm

I have fitted new springs - about x 5 stronger than original ones. The new ones are 22 x 5 x0.7mm
Rough bench tests indicate that they lengthen 3mm for a force of 10N
I have set them to about 6-8 N. At this setting I can see no visible change in spring length for a full range of arm movement. (unlike the previous springs which were 0.5mm wire) . I did have to up the current drive to eliminate any skipped pulses.
This raises the interesting point (Nicholas has mentioned this in other posts) as to the need for springs at all. Would the natural compliance of the system be enough to keep the system tuned.?
Personally I like the idea of a bit of compliance for those difficult times when things go wrong and some 'give' may prevent a broken part. Thats just me being pessimistic!
Result - improved but not perfect.
The picture (GUS20) shows some results. The timeline goes L -> R and you can see the height problems in the early prints. The calibration arms should be 30 x 10 x 5mm
To start I was getting up to 10% variation in height. Now its about 2-4%.
Attachments
GUS20.JPG
GUS20.JPG (277.45 KiB) Viewed 11909 times
timrwilliams
 
Posts: 87
Joined: Fri Feb 14, 2014 10:32 pm

PreviousNext

Return to GUS Simpson

Who is online

Users browsing this forum: No registered users and 1 guest

cron