circles not round

[jason]

Ok just about tamed the extruder/hot end.

Is it just me but does when printing circles ( yes I know no such thing..once your model has been through stl no circles all triangles)
I get say 9 oclock to 3 oclock as a single sweep but then 3 oclock to 9 oclock as another sweep..


Not withstanding blobs but theses seem to be real features..even scale with model..don't think its mechanical backlash since it scales..and are at the same position in each layer...
Using repetier and slicher....
Thought's anyone

[ratmonkey]

Can you post a picture?

[jason]

]The larger is 50 mm dia
The small one and lid is 13 mm

[ratmonkey]

Looks like your rack/pinion-pair is not tight enough.
Must be very noisy while printing... :-)

[jason]

Mmm...no measurable backlash with steppers powered up but will check..
How much mesh of the rack should there be...ie is the rack cut 100% or 110%? Deep..or just as tight as possible. ..


But might explain dissimilar step/mm for x and y that are required for square prints.....

..thanks

[ratmonkey]

X and Y should have similar steps/mm...
Rack and pinion should be just as tight as possible without blocking.
The axis has to move smooth.
If you have to push hard, then it is to tight.

BTW. can you post your stepps/mm for x and y?
is one of them 2-times the other?

[jason]

x 41.8 y 42.2.
Yeah a factor of two would be obvious..
Yes first two jumpers ON.
At 42 42 I get 24.8 by 25.2.. for a 25.00 cube
Z cal is un proven yet...top bottom fills crap at moment

[ratmonkey]

The racks cant be 100% identic. Yours have a difference of app. 1%.
That is ok.
If a square of 25x25 comes out as 25x25 with you x41.8/y42.2 and there is no backlash, then the problem is somewhere else.

Maybe your steppers miss some steps while printing circles. If this happens, you could try to increase/decrease the steppercurrent a bit.
If the current is to high the driver could overheat and breaks down until the temp is low enough.
This also leads to missing steps.

[jason]

Mm missed steps..that had occured to me.
Can't quite believe that with how consistent the print is..
Wrong but repeatable. .same artifact at different diameter. ..
On cubes there is "overshoot" at one corner (blobs)..but that is I think extrusion issues
....
As to tolerance on the rack or maybe the pinion. .typical. .
..
Bearing in mind people have printers with printed gears..and can print gears..it beggars belief to think that a printer as apparently solid should struggle..
Am I expecting too much...
Will 'scope the drivers looking for outages..but not hopeful..

.

[jason]

OK
No outages or stalls on axis.
( extruder does sometimes stall but I think I know the cause of that)

Various ratios trying to get square square but no dice..can't do better than 51.2 49.9..
But odd thing high lighted....
If I print the above 50 x50 twice on same print..with one rotated 45 degrees......
I get two prints 51.2 by 49.9
ab bc cd ca sides match and ab cd are 51.2 etc...so say the first side of each square is 51 on both the un totated and the rotated
Very very odd..I blame slicer...at the moment..but the g codes check out ok.....
Today I have read that marlin examines the gcodes looking a head ( for arcs and such)....so maybe there ..
But surely some one else has this problem
..not even just with sumpod...

[jason]

Right..I think I know whats wrong...

Backlash as suggested. ..
In my case in the x axis...
Heres how it was hidden..
No measure able backlash rack to pinion..
But "play" in x carriage such that far end can move and consequently the nozzle ..in an arc..of about 1-2 mm
This particular mdf component was damaged on arrival and in any case seemed made from different stuff to the rest of the kit ( more like low density than medium density)...
I have had to fit a repair plate to hold the motor but now I will unship it and make a replacement ..
Just tightening the guides or rack cant make up for this. .
So off I go...maybe now would be the time to remake the whole x carriage and gain five mm of z axis travel...and some stiffness into the bargin..

[jason]

Ok . made new plate..4 mm mild steel should be good.
Wimped out on bining the mdf lower part of the carriage..and have retained the mdf upper carriage too ( well this piece seems much "looser" than the rest. .but retaining it ( was concerned about damping and mechanical resonance. ...)
The fluff type mdf has been saturated with c a glue and cured in a press.....now seems a bit tougher. ..
Earlier tests have shown these components don't seem to get above ambient during print so don't expect the c a to fume furing use...
Wish me luck

[jason]

Thats better..50 mm dia measures out of round 0.63 mm on dia..
If I want better I will turn it...

.racks will enviably have some backlash..and 0.3 mm on a approx 4 mm rack is not going to go away with out a deliberate anti backlash measure..either pinion or counter weight / spring

[bdureau]

Jason
I have the same problem. Did you eventually solved yours?

[jason]

In short ..no..
It appears to me to be back lash/ triangulation of the X axis. ..small moves changing from say plus to minus lag and don't happen..flattening the circle at the 9 and 3 o'clock positions..
I think after remaking the upper carriage for x that without significant redesign this can't be fixed.. ( part of cause is the rack being outside the guide rails not inside (poor mechanical design))..a system preventing the carriage triangulation will be needed
In my case the backlash produces .0.56mm error on diameter largely irrespective of circle diameter. ..
Since this is a diameter error that equates to 0.28 mm..
I would like to know if anyone ever got better performance from a basic with rack..
If so please share... ..