Wobbling along X-ax...
 
Benachrichtigungen
Alles löschen

Wobbling along X-axis

5 Beiträge
3 Benutzer
0 Reactions
3,107 Ansichten
Giulio Buccini
(@julius)
Beiträge: 212
Reputable Member
Themenstarter
 

Hello,

I'm facing a quite frustrating problem with my little SC300 (first series).

By using an indicator and a rectilinear guide I discovered that the milling point is oscillating/tilting while jogging/moving the spindle along the X-axis.

It seems like a sinusoidal path in the Y-direction: initially the tip of the indicator deviates of (around) +0.1 mm, then return to the correct position, then deviates again of -0.1 mm.
This "cycle" is repeated along all of the X-axis every around 54,5 mm (if I remember correctly).

See the attached image.

In this conditions it is almost impossible to mill a long rectilinear side without obtaining some humps along it. Nerving... grrrrr....

I was thinking that the cause was some dirty stuff "glued" on one surface on one or more wheel. So I completed unmounted the portal, unmounted all of the wheels on the X axis, cleaned them with attention, checked the surfaces and the bearings for damages and then I remounted all back (was a huge work!).
But the wobbling cycle is still there. Grrrrrrr!

Could be a defect in the extrusion of the aluminum profile? I wrote to Stepcraft for a spare part, but still no answer.
Someone has faced the same problem in the past? What could be a possible solution?

I could try to buy and replace all of the wheels and the bearing on the X-axis. But mine seems still good conditions (no bearing play and very good surface conditions since I have always oiled and greased the machine).

SC300 + Spindle HF500 + Portalerhöhung + LinuxCNC + gsimple

 
Veröffentlicht : 18/03/2021 1:23 pm
(@volker-bunge)
Beiträge: 257
Reputable Member
 

Hello Julius,

I'm going to assume that you have a straight line in the X-axis. If not then please make this die.
Otherwise there is also a small problem with the Y-axis

Since the D300 is already a very small machine, the threaded rods are of course quite short and therefore quite insensitive to bending.
But try the following:

What type of milling do you use on the axles? Reduce the gifts. Not that you're just too fast.
If that doesn't help, then, in my opinion, you will build the X-axis.

Put the X-threaded rod on a flat surface. If it is the threaded rod, then this has now moved up and down at the ends when turning.
Then only a substitute helps.

If this is the case, then check again whether the bracket for the threaded rod and the Z-axis is correct and secure.

Now that you have heard everything you can, you can also test the aluminum profile right away. Place it on a straight surface and generate a light source with a flashlight, which you then hold between the surface and the aluminum profile. If the profile is straight, there is no light shining through. This test from both sides shows.

Then just put the two bearings on the threaded rod and turn again. This has been changed, then one or both bearings are defective.

Finally, you can test the engine to see whether it runs smoothly.

All of these tips are only possible with the pure elimination process in order to find out about the error.

I heard that yours will help you find the bug. Unfortunately, these errors are very annoying and unfortunately hardly to be found or felt without dismantling.

greeting

Volker
(automatische Google-Übersetzung)

(Deutsch)
Hallo Julius,

ich gehe mal davon aus, dass Du eine gerade Linie in der X-Achse fräst zum testen, so dass tatsächlich nur die X-Achse arbeitet. Wenn nicht, dann mache dies bitte.
(Es könnte sonst auch ein kleines Problem mit der Y-Achse vorliegen)

Da die D300 je schon eine recht kleine Maschine ist, sind da natürlich auch die Gewindestangen recht kurz und somit gegen ein Verbiegen recht unempfindlich.
Aber versuche es einmal mit folgendem:

Mit welcher Geschwindigkeit fräst Du auf den Achsen? Reduziere mal die Geschwindigkeiten. Nicht das Du einfach zu schnell bist.
Wenn das nichts hilft, dann wirst Du, meiner Meinung nach, die X-Achse auseinanderbauen müssen.

Lege die X-Gewindestange mal auf eine gerade Unterlage. Wenn es die Gewindestange ist, dann sollte sich diese jetzt beim drehen an den Enden auf und ab bewegen.
Da hilft dann nur Ersatz.

Sollte das nicht der Fall sein, dann prüfe doch einmal, ob die Halterung für die Gewindestange an der Z-Achse richtig sitzt und fest ist.

Da Du ja jetzt alles auseinander gebaut hat, kannst du auch gleich mal das Aluprofil testen. Leg es auf eine gerade Unterlage und erzeuge mit eine Taschenlampe eine Lichtquelle, die Du dann zwischen Unterlage und Aluprofil hälst. Ist das Profil gerade, dann sollte kein Licht durchscheinen. Diesen Test von beiden Seiten durchführen.

Dann stecke doch einfach noch die beiden Lager auf die Gewindestange und drehe noch einmal. Sollte hier sich was verändern, dann ist ein oder beide Lager defekt.

Zum guten Schluss kannst Du noch den Motor testen, ob dieser Rund läuft.

All diese Tipps sind nur dem reinen Ausschlussverfahren entnommen um so dem Fehler auf die Schliche zu kommen.

Ich hoffe, dass Dir meine Angaben helfen, den Fehler zu finden. Leider sind diese Fehler sehr lästig und leider kaum ohne Demontage zu finden bzw. zu beheben.

Gruß

Volker

Stepcraft V2 420, WinPC-NC USB Vollversion + HR10, EstlCam 10+XBox-Controller, HF500, Schwebehalter, Schleppmesser, Heißschneidedraht

 
Veröffentlicht : 19/03/2021 10:36 am
Andreas Schulz
(@andys1957)
Beiträge: 764
Prominent Member
 

Hallo Julius,

hast du wirklich über die gesamte X-Achse etwas mehr als 5 gleichmäßige Sinuswellen mit einer Breite von jeweils 54,5mm und einer Amplitude von +- 0,1mm? Falls das wirklich so ist, kann es eigentlich nur das Profil sein, welches dann aber schon bei der Produktion diesen "Schlag" bekommen haben muß. Oder aber eine der Rollen (bzw. deren Kugellager) ist defekt oder verschmutzt. Ich kenne den Durchmesser der Rollen jetzt nicht genau, aber bei einem Wiederholungsabstand von 54,5mm entspäche das einem Rollendurchmesser von ca. 17mm. Das könnte hinkommen...
Wie sieht denn dein Meßaufbau aus?

Gruß
Andy

Fräse: SC 2/600 mit PerfKit, T-Nut, Vakuum, Kress 1050 FME-1
CAD, CAM, Steuerung: FreeCAD, Estlcam, Estlcam/Arduino ("Forums-Shield")
Laser, Software: Sculpfun S30 Pro, Eleksmaker A3 mit Atomstack R30 (IR), Lightburn
3D-Druck: FB Reborn 2, Art. Genius und Gen. Pro, CTC Bizer, Elegoo Mars

 
Veröffentlicht : 19/03/2021 4:34 pm
Giulio Buccini
(@julius)
Beiträge: 212
Reputable Member
Themenstarter
 

aber bei einem Wiederholungsabstand von 54,5mm entspäche das einem Rollendurchmesser von ca. 17mm. Das könnte hinkommen...

17 mm X 3.14 = 53,38 mm

that is very close to 54,5 mm.... So the "guilty" could be one of the wheels (or the bearing) as you pointed out.
So I think that there is no other option that buying a complete new set a replace all of the three wheels and bearings on the X-axis.

Wie sieht denn dein Meßaufbau aus?

I just placed the flat surface of an L-shaped "winkel" (precision steel angle-square tool - made in Germany) on the aluminium table. The short arm of L-shape was pushed against the side wall of the steppie, while the long arm was pointing along the X-axis.

Then I mounted the indicator on the spindle and make my measurements by jogging it along the X-axis.
I tested both the "external" and the "internal" sides of the L-shaped square obtaining the same results. So I think that the square-angle tool is not the source of the error I saw.

Note that the parallelism between the spindle and the side wall of the Steppie is actually very good! I.e. less than half of a tenth (0.04-0.05 mm)

SC300 + Spindle HF500 + Portalerhöhung + LinuxCNC + gsimple

 
Veröffentlicht : 20/03/2021 11:35 am
Giulio Buccini
(@julius)
Beiträge: 212
Reputable Member
Themenstarter
 

Hello Volker, thanks for your tips.
I already know how to check a surface for linearity/flatness, since I was trained to that when I was young at the "industrial" high school (before moving to "Informatik" as specialized course of study).

Checking the linearity of the big "chunk" of aluminum is not easy for two reasons:
1.
I have not a such long square-tool to check the surface in one. Alternatively, I could slide an indicator over a precision flat-surface made of granite. But the costs of such tools are far above the cost of the machine itself.

2.
It is very hard to make the light coming from the back side of my measurement tool because of the "C-shaped" section of the aluminum chunk (in particular, if you want to check the most internal surface on which the wheels are rolling on.

I'm using the Steppie to mill just POM 99,99% of the time and I'm not using crazy speed. I just follow the recommendation of the well-know table that you can download from the Stepcraft/Sorotec website both.
As an example (POM material): 800 mm/min feed rate for a 2-cut/2mm-diameter endmill.

Anyway, I think the speed of cut is not in discussion here. I was just freely jogging the spindle along the X-axis with no load at all. I can imagine that under load the situation can only be the same or worse.

I posted my problem here only to know if the problem could be the big aluminium "chunk" acting as a guide for the wheels, or I have to replace only the wheels.

FINAL NOTE
I also thought that this problem is inherent to the way of the machine itself is build, therefore cannot be 100% eliminated.
Think of that: it is impossible for the wheels along the Y-axis to tilt while travelling just because they are rigidly connected with the wheels on the other side by the portal itself. Any tilt would result in a deformation of the portal, that is quite robust indeed.
On the X-axis things change a lot: the wheels are constrained to travel along two vertical guides embedded in that big chunk of aluminum. So, even a small imperfection can cause them to easily "rump-up" on one side of the guide. In this way the "sled" will be no more vertical...

This problem is solved on the Sorotec hobby-line machine by using two precision-shaft that serves as guide for the wheels (see attached image - number 13).

SC300 + Spindle HF500 + Portalerhöhung + LinuxCNC + gsimple

 
Veröffentlicht : 20/03/2021 12:24 pm
Teilen: