Asymmetrical prints

[Tincho85]

Hi all, I've recently installed the ball-cup kit and re-flashed the printer with the lastest firmware by SeeMeCNC.
After following 626Pilot's guide to get it dialed in perfectly, I noticed my prints are not symmetrical.
But I don't think this is because of the recent update. I might have this issue from before.

Here are the measurements:
https://c1.staticflickr.com/9/8147/2892 ... 40fc_o.jpg
[img]https://c1.staticflickr.com/9/8147/2892 ... 1d04_c.jpg[/img]

70x70 cube: 70.76mm (101.08%) x 70.15mm (100.21%)
50x50 cube: 50.50mm (101%) x 50.10mm (100.2%)
30x30 cube: 30.30mm (101%) x 30.00mm (100%)
15x15 cube: 15.05mm (100.33%) x 14.95mm (99.66%)


Things to consider:
* This is a RostockMAX V1, with 20 tooth timing pulleys.
* The melamine frame is custom made, but has the same geometry as the RostockMAX V2 (equal T-slot towers distance to the center of the bed), so it's essentially an aesthetic change.
* I didn't buy the new injection carriages (I'm already regretting it), still using the old ones with a printed adapter for the ball joints. This adapter was printed one at a time, in the exact same position, to be sure all 3 of them have the same dimensions.

To figure this out, I've tried:
* Rotated the effector platform.
* Rotated the arms.
* Downgrade to 0.91 firmware version.
* Different filament.
* Different slicer (cura, slic3r and S3D)


The problem could be related to the melamine frame, but I can't understand why the 15x15mm cube is undersized in one side and the 70x70 cube is oversized in the same location.

Any thoughts?

[nebbian]

If one belt is slightly more stretched than the others (ie the Z belt) then it will cause this effect. Sadly Repetier doesn't allow different steps per mm for each tower (unlike Marlin).

Try moving the belts around and see if the problem follows the belt. Or try measuring the distance each carriage moves (measuring this precisely is difficult, at least for me) when you move the effector down 100mm.

Yes I have seen and measured this on my own deltas.

[Tincho85]

Hi nebbian. Thanks for your insight. I like your thinking.
Still at work, but I will play with Z belt tension later today.

[Tincho85]

So, I've released some tension from the Z axis. It got better, still not perfect but helped.

50x50x10mm tall cube: 50.34mm x 50.01mm

I'm going to loosen it a bit more and see if it improves.
Will come back with updates.

[Tincho85]

In the meantime... will order some new belts. They might have worn out.
The last time I changed them was 1 year ago.

[Tincho85]

Uhmmm, I guess it was just luck that it got better by loosening the tension.

50x50x10 cube: 50.45mm x 50.01mm


Any other ideas?

[Nylocke]

Its unlikely the belts are stretching that much to be an issue. They have fiberglass or kevlar strands in them that negate stretching. Its more likely to be one of the issues detailed by others in the same post you linked. Not sure if you checked how square your towers are, I looked through your build log but I didn't see you using a wixey or square, or what calibration method you're using but those are two places to check. Also someone had an issue with an oblong nozzle. Not sure if you've had any nozzle crashes or printed with any abrasive filaments but that might be an issue too.

[Tincho85]

Its unlikely the belts are stretching that much to be an issue. They have fiberglass or kevlar strands in them that negate stretching. Its more likely to be one of the issues detailed by others in the same post you linked. Not sure if you checked how square your towers are, I looked through your build log but I didn't see you using a wixey or square, or what calibration method you're using but those are two places to check. Also someone had an issue with an oblong nozzle. Not sure if you've had any nozzle crashes or printed with any abrasive filaments but that might be an issue too.

The effector platform was rotated and therefore the nozzle, but the issue persisted. So I don't think that I have an oblong nozzle.


Tower lean was removed from the equation because I fixed it last year, but it seems that you are right. The last time I checked my Z was off by 0.6 degrees and corrected it to 90.1º
I checked again today and it's back to 90.6º

The only problem is that I don't have the proper tool for this job (± 0,5º accuracy), but that's what I have at the moment. Maybe it's worst than 0.6º. Using this digital angle finder

If you have a good tool to reccomend, a link would be appreciated.
I was considering this one: https://www.amazon.com/Calculated-Indus ... 223&sr=8-3


Thanks for the help


EDIT:
Right now it's like this:
X: 89.9º
Y: 90.0º
Z: 90.6º

[Nylocke]

I have one of these http://www.wixey.com/anglegauge/ the type two I think. Its accurate +/- .2 degrees, but the Type 1 is +/-.1 and its not much more. Pretty similar to what you linked. You can find the wixey on amazon too.

[Tincho85]

Thanks for the help Nylocke. In Amazon, I can't find any international buyer willing to ship the one I linked or the Wixey.
I will be buying this one, it looks good enough for this application: http://www.ebay.com/itm/New-Dual-Axis-0 ... Swn8FXRrJB

Accuracy:
0 to 20°: ± (0.1°) 
20 to 70°: ± (0.2°) 
70 to 90°: ± (0.1°) 

But I'm too lazy to re-assemble the machine right now and correct it. If the V3 upgrade is released soon, I might wait and do it altogether.

[Nylocke]

Okay, thats unfortunate. Your custom MAX is one of the coolest things I've seen on this forum, its truly unique.

[Tincho85]

Thanks, I really appreciate the kind words.
Still haven't really decided to do the V3 upgrade. First I want to see those nylon filed parts and what they are used for. The idea of a much stiffer machine is very appealing, but maybe I can merge it into my build, time will tell.
I know for sure that I wont buy the melamine parts, as they are too heavy and big for international shipping.
When work slows down a bit, I will try cutting some 6061 aluminum with my MPCNC, maybe a custom metal Max? who knows.