Noob UK build

Re: Noob UK build

Postby timrwilliams » Mon May 26, 2014 9:51 pm

I can't say I find Marlin better. Apart from its predecessor Sprinter, I have never used anything else.
I am sure others are working with Repetier, and I seem to remember that Nicholas posted his config file somewhere on this forum.
If you really do want to try Marlin, I can certainly let you have a copy. But there are several versions downloadable. I am using a version which has separate Cartesian/Delta config files and also (if my memory serves me) has an advanced config file which carried some variables.
If you do go down that road, its best if I PM the whole Marlin to you.

I have just noticed - the third thread on the first GUS index page may help.
timrwilliams
 
Posts: 87
Joined: Fri Feb 14, 2014 10:32 pm

Re: Noob UK build

Postby Niggle » Tue May 27, 2014 5:33 am

I'm also using Marlin at this point.

The config file is a lot messier than the Repetier equivalent.

I switched because I was getting confused by Repetier's handling of the EEPROM. Also, I have an LCD panel for my controller and find that Marlin makes it more useful.

However, your problems look to be caused by your config file not matching your wiring. Copying someone else's config file is not going to help you solve that problem. In particular, the z motion speed is related to setting for Z that are different from the equivalent settings for X and Y, but direction and limit of movement settings tend to be unique to each build.

If you create a config file for a delta printer and compare it with the one you have, using a 'diff' tool, the Z changes should be highlighted. The rest is likely to be more along the lines of trial and error.
Niggle
 
Posts: 84
Joined: Tue Feb 11, 2014 7:27 pm

Re: Noob UK build

Postby ChrisK » Tue Jun 03, 2014 8:46 am

I decided to have a go with Marlin and it looks like I have got the basic settings OK my GUS is beginning to act like a printer now and the LCD controller is pretty cool.
I got some ballpark steps/mm readings with a ruler as my caliper was too small.
On the subject I was wondering if this lo-tech alternative could work. If you were to tie a string around the protuding nut and pass it through a stick on cable tidy stuck to the other nut as a guide? You could use the guide to mark the string before and after move and then measure up. As i am only measuring the moves I could use my calipers here. If it would work it would allow to mark off a series of moves insucession to check across the full range or allow averaging.

My printer seems to be starting the setup fine, lifting to all the endstops but looks like it's then moving down too far and seems to think it's drilling for oil
http://vimeo.com/97176499

Am I right in thinking my next step is to set the positions at the begining of sevmentise??(once i have a hot end that is)
Do you manually controll the head from the top, maximum positions to points microns above the bed and make notew to type into segmentise??
ChrisK
 
Posts: 63
Joined: Mon Dec 30, 2013 6:14 pm

Re: Noob UK build

Postby timrwilliams » Tue Jun 03, 2014 9:45 am

Thats looking really smart, Chris.
I am sure that you are right - 'Segmentize' bed values is the next step, as you describe, using manual controls after zeroing in the home position. Its a bit tedious at first, but you soon get the hang of it.
I started off with 6 sets of values plus bed center, working over a small central area (squirrel sized!). Ideally one should go for lots more values spread over the full print area.
What hot end will you use?
The Cal method you suggest sounds quite practical and easy to fit on.

Tim
timrwilliams
 
Posts: 87
Joined: Fri Feb 14, 2014 10:32 pm

Re: Noob UK build

Postby ChrisK » Tue Jun 03, 2014 10:28 am

I'm probably going to go for the widely available J head.
Thinking of making up the difference in length by adding another wooden deck under the basalt.
Until that is in place I'm going to try out my calibration technique and better dial in the steps/mm values
ChrisK
 
Posts: 63
Joined: Mon Dec 30, 2013 6:14 pm

Re: Noob UK build

Postby timrwilliams » Tue Jun 03, 2014 1:05 pm

A J head works well and is easy to buy here in the UK, kit or ready made. I did fit one directly into the GUS hub as designed, but although it was a good mechanical fit, I thought it was too deep and was obstructing some of the J head cooling fins. I didn't even try it in that position and used a stand off adaptor. This works well without a fan. If you do mount the hot end directly, I would get a fan fitted right from the start. I think @brandonh 's experience with heat distortion in a hot car is a lesson for us all.
timrwilliams
 
Posts: 87
Joined: Fri Feb 14, 2014 10:32 pm

Re: Noob UK build

Postby Niggle » Tue Jun 03, 2014 6:00 pm

The point of the calibration part of segmentize is to sort out how to compensate for the thickness of the bed, variations in the height of the hotend and the hub components and the placement of the limit stops on the arms. Adding material to the bed will not help.

The really important thing is to get consistent steps/mm. This is easiest if you completely dismount an arm. Leave the motor attached so you can use the host software to extend the arm until the end stop triggers.

If you have nothing better, fit a bolt through the bearings at the hub end, wrap a piece of line around the shoulder bolt and the hub bolt and mark where the ends meet. If you have a long enough ruler, take two or three wraps. Divide the measured length by two (or four or six) and use the result as the maximum arm length. (The number will be longer than the actual distance between the bolt centers, but the math will compensate for that.)

Now use the host software to close the arm by 100mm (to make the math easy). Take another measurement the same way. The difference in the length of the line should be EXACTLY 200mm (or 400 or 600) shorter than the original measurement. Tweak the steps/mm for the arm and repeat the process until this is true. Now do a check, closing the arm by 200mm or 250mm. If the check gives different steps/mm, your have a non-linear arm which will cause long term problems. Pressing on, repeat the process for the other two arms.

Consider how much time you spent messing with pieces of string and how much faster it would have been to use a caliper. Was your time worth the cost of a caliper?

Once you have the steps/mm numbers correct, you can move to collecting data points. I have been using Marlin with an LCD and click encoder so I can use the encoder to hove the hotend and read the coordinates off the LCD.

Driving the extruder through the bed. The happened to me for two reasons.
1. Repetier firmware would move to 'home' and leave the internal coordinates at [-3000,-3000,-3000]. A move to [0,0,0] would try to drive the hotend through the bed.
2. Marlin firmware boots assuming that the hotend is at [0,0,0] regardless of where it actually is. I usually park the hotend on the bed, so any move will have unfortunate consequences. If the hotend is close to the bed, G28 will fully extend one arm and the next arm will not have the leverage to move. Otherwise, almost any G1 with positive coordinates will target a location below the surface.
Niggle
 
Posts: 84
Joined: Tue Feb 11, 2014 7:27 pm

Re: Noob UK build

Postby ChrisK » Wed Jun 04, 2014 1:37 pm

I'm considering increasing the bed height not to deal with calibration problems but to try and keep the mechanics away from the point where they lose strength or bind themselves (Does Nicholas describe this are as the singularity?) due to a slightly shorter hot-end being used.
What do you mean by non-linear arm? How does this happen and what would be a solution?
ChrisK
 
Posts: 63
Joined: Mon Dec 30, 2013 6:14 pm

Re: Noob UK build

Postby timrwilliams » Wed Jun 04, 2014 2:25 pm

A common cause of a different steps/mm value when you reverse direction of movement is weak springs leading to hysteresis. Do I see in your video that you have dispensed with springs completely. I know Nicholas has suggested this in previous posts but I have never tried that, just going for stronger springs. Does 'no springs' work for you?
I also see that you are using Aluminium spools, that should remove any cyclical non-linearity from eliptical printed spools.
I reckon your arms will be as linear in response as one can get.!
timrwilliams
 
Posts: 87
Joined: Fri Feb 14, 2014 10:32 pm

Re: Noob UK build

Postby Niggle » Wed Jun 04, 2014 2:58 pm

Raising the bed to compensate for a short hot-end is reasonable. Measure the height of the hub bolts above the bed with the hot end touching. The target height is about 80mm.

By non linear, I mean that the steps/mm changes as the arm moves. If an arm didn't print perfectly or got warm enough to warp slightly you could end up with this problem. If the fishing line is binding anywhere or if any of the joints is stiff enough to cause the motor to skip once in a while, you will have this problem.
Niggle
 
Posts: 84
Joined: Tue Feb 11, 2014 7:27 pm

PreviousNext

Return to GUS Simpson

Who is online

Users browsing this forum: No registered users and 3 guests

cron