So, according to this reprap article, the MXL belts the Ultimaker uses are
not actually well-suited to constantly reversing applications, as they are
designed for continuous motion in one direction and don't have an
anti-backlash profile:
http://mendelmax.com/tikiwiki/tiki-index.php?page=Choosing+Belts+and+...
They suggest GT2, which might be an interesting upgrade for the UM.
I'm building a Rostock and I'm using GT2 belts on that, so we'll see how
that goes. I've never had good layer to layer repeatability on my UM. If a
layer switches the direction its coming from (say, there's a hole in part
of the face), I always get a line due to backlash in the system - and I've
got belt tensioners that I think are working nicely (very stiff piano wire).
Does anyone with other belt tensioners actually have zero backlash?
Ya, I'm wondering about this too. I just spent the whole weekend rebuilding
my printer because I've got backlash in the x direction... So, ovals
instead of cirlces...
I keep tightening the belts but they keep stretching. I might give this a
go because I'm getting pretty tired of the backlash.
On Monday, September 17, 2012, Taylor Alexander wrote:
> So, according to this reprap article, the MXL belts the Ultimaker uses are
> not actually well-suited to constantly reversing applications, as they are
> designed for continuous motion in one direction and don't have an
> anti-backlash profile:
> They suggest GT2, which might be an interesting upgrade for the UM.
> I'm building a Rostock and I'm using GT2 belts on that, so we'll see how
> that goes. I've never had good layer to layer repeatability on my UM. If a
> layer switches the direction its coming from (say, there's a hole in part
> of the face), I always get a line due to backlash in the system - and I've
> got belt tensioners that I think are working nicely (very stiff piano wire).
> Does anyone with other belt tensioners actually have zero backlash?
> --
> You received this message because you are subscribed to the Google Groups
> "Ultimaker" group.
> There are also forums, which are the official gathering place for
> Ultimaker operators:
> http://forum.ultimaker.com/
> If you still want to post to this group, send email to
> ultimaker@googlegroups.com <javascript:_e({}, 'cvml',
> 'ultimaker@googlegroups.com');>
> To unsubscribe from this group, send email to
> ultimaker+unsubscribe@googlegroups.com <javascript:_e({}, 'cvml',
> 'ultimaker%2Bunsubscribe@googlegroups.com');>
> For more options, visit this group at
> http://groups.google.com/group/ultimaker?hl=en
On Monday, September 17, 2012 10:11:42 PM UTC-4, Kevin R wrote: > Ya, I'm wondering about this too. I just spent the whole weekend > rebuilding my printer because I've got backlash in the x direction... So, > ovals instead of cirlces...
> I keep tightening the belts but they keep stretching. I might give this a > go because I'm getting pretty tired of the backlash.
> On Monday, September 17, 2012, Taylor Alexander wrote:
>> So, according to this reprap article, the MXL belts the Ultimaker uses >> are not actually well-suited to constantly reversing applications, as they >> are designed for continuous motion in one direction and don't have an >> anti-backlash profile:
>> They suggest GT2, which might be an interesting upgrade for the UM.
>> I'm building a Rostock and I'm using GT2 belts on that, so we'll see how >> that goes. I've never had good layer to layer repeatability on my UM. If a >> layer switches the direction its coming from (say, there's a hole in part >> of the face), I always get a line due to backlash in the system - and I've >> got belt tensioners that I think are working nicely (very stiff piano wire).
>> Does anyone with other belt tensioners actually have zero backlash?
>> -- >> You received this message because you are subscribed to the Google Groups >> "Ultimaker" group.
>> There are also forums, which are the official gathering place for >> Ultimaker operators: >> http://forum.ultimaker.com/
>> If you still want to post to this group, send email to >> ultimaker@googlegroups.com >> To unsubscribe from this group, send email to >> ultimaker+unsubscribe@googlegroups.com >> For more options, visit this group at >> http://groups.google.com/group/ultimaker?hl=en
Ditto Dave's comment, I haven't tightened my long belts since I put them on
a year ago.
I get some vibration around hard corners on flat surfaces (the echo effect)
But that's because
I run it fast.
Usually ovals instead if circles for me was the short belts getting loose.
(Technically it was the X and Y
motors getting loose.) Same for infill not touching the edges. The "lash"
was in the short belt.
The motor screws tend to loosen. I put some lock washers on them and
haven't had a problem since.
On Mon, Sep 17, 2012 at 10:17 PM, ddurant <dduran...@gmail.com> wrote:
> Are you sure the belts are stretching? My original ones have been on there
> for over a year and still make a nice twang when I give them a flick..
> On Monday, September 17, 2012 10:11:42 PM UTC-4, Kevin R wrote:
>> Ya, I'm wondering about this too. I just spent the whole weekend
>> rebuilding my printer because I've got backlash in the x direction... So,
>> ovals instead of cirlces...
>> I keep tightening the belts but they keep stretching. I might give this a
>> go because I'm getting pretty tired of the backlash.
>> On Monday, September 17, 2012, Taylor Alexander wrote:
>>> They suggest GT2, which might be an interesting upgrade for the UM.
>>> I'm building a Rostock and I'm using GT2 belts on that, so we'll see how
>>> that goes. I've never had good layer to layer repeatability on my UM. If a
>>> layer switches the direction its coming from (say, there's a hole in part
>>> of the face), I always get a line due to backlash in the system - and I've
>>> got belt tensioners that I think are working nicely (very stiff piano wire).
>>> Does anyone with other belt tensioners actually have zero backlash?
>>> --
>>> You received this message because you are subscribed to the Google
>>> Groups "Ultimaker" group.
>>> There are also forums, which are the official gathering place for
>>> Ultimaker operators:
>>> http://forum.ultimaker.com/
>> PhD Candidate
>> Department of Chemical and Biological Engineering
>> University of British Columbia
>> P: 604.827.4123 C: 604.313.9149
>> --
> You received this message because you are subscribed to the Google Groups
> "Ultimaker" group.
> There are also forums, which are the official gathering place for
> Ultimaker operators:
> http://forum.ultimaker.com/
> If you still want to post to this group, send email to
> ultimaker@googlegroups.com
> To unsubscribe from this group, send email to
> ultimaker+unsubscribe@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/ultimaker?hl=en
It is easy with G-code to write a very little program to very slowly move a
short distance to activate a dial indicator. Then after stopping (pause) to
set the indicator to zero, continue to back up. for example, 0.1mm then
pause, 0.1mm then pause. See how much reverse movement is required before
the movement shows up on the dial indicator.
That would be static backlash. It is a different story when things are
moving fast and there are forces applied when quickly changing direction and
there are also mechanical resonances to take into account.
Bertho
From: Taylor Alexander Sent: Monday, September 17, 2012 19:52
So, according to this reprap article, the MXL belts the Ultimaker uses are
not actually well-suited to constantly reversing applications, as they are
designed for continuous motion in one direction and don't have an
anti-backlash profile:
They suggest GT2, which might be an interesting upgrade for the UM.
I'm building a Rostock and I'm using GT2 belts on that, so we'll see how
that goes. I've never had good layer to layer repeatability on my UM. If a
layer switches the direction its coming from (say, there's a hole in part of
the face), I always get a line due to backlash in the system - and I've got
belt tensioners that I think are working nicely (very stiff piano wire).
Does anyone with other belt tensioners actually have zero backlash?
That's interesting because GT2 belts is what the Makerbot Cupcake uses.
________________________________ From: Taylor Alexander <tlalexan...@gmail.com> To: ultimaker@googlegroups.com Sent: Monday, September 17, 2012 4:52 PM Subject: interesting note about the belts on the Ultimaker
So, according to this reprap article, the MXL belts the Ultimaker uses are not actually well-suited to constantly reversing applications, as they are designed for continuous motion in one direction and don't have an anti-backlash profile: http://mendelmax.com/tikiwiki/tiki-index.php?page=Choosing+Belts+and+...
They suggest GT2, which might be an interesting upgrade for the UM.
I'm building a Rostock and I'm using GT2 belts on that, so we'll see how that goes. I've never had good layer to layer repeatability on my UM. If a layer switches the direction its coming from (say, there's a hole in part of the face), I always get a line due to backlash in the system - and I've got belt tensioners that I think are working nicely (very stiff piano wire).
Does anyone with other belt tensioners actually have zero backlash? -- You received this message because you are subscribed to the Google Groups "Ultimaker" group.
There are also forums, which are the official gathering place for Ultimaker operators: http://forum.ultimaker.com/
If you still want to post to this group, send email to ultimaker@googlegroups.com To unsubscribe from this group, send email to ultimaker+unsubscribe@googlegroups.com For more options, visit this group at http://groups.google.com/group/ultimaker?hl=en
Another nice but less important thing is that GT2 belts are metric, and mxl
is not.
I've read a students report recently where one of the things they found was
that the repeatability was about 0.1mm. I suspect the belts played a large
part in that, as the theoretical resolution is a lot lower. One could
imagine a system with only one belt per axis too, as that will halve the
backlash due to the belts and pulleys.
On Tue, Sep 18, 2012 at 6:07 AM, Zip Zap <zzap...@yahoo.com> wrote:
> That's interesting because GT2 belts is what the Makerbot Cupcake uses.
> ------------------------------
> *From:* Taylor Alexander <tlalexan...@gmail.com>
> *To:* ultimaker@googlegroups.com
> *Sent:* Monday, September 17, 2012 4:52 PM
> *Subject:* interesting note about the belts on the Ultimaker
> So, according to this reprap article, the MXL belts the Ultimaker uses are
> not actually well-suited to constantly reversing applications, as they are
> designed for continuous motion in one direction and don't have an
> anti-backlash profile:
> They suggest GT2, which might be an interesting upgrade for the UM.
> I'm building a Rostock and I'm using GT2 belts on that, so we'll see how
> that goes. I've never had good layer to layer repeatability on my UM. If a
> layer switches the direction its coming from (say, there's a hole in part
> of the face), I always get a line due to backlash in the system - and I've
> got belt tensioners that I think are working nicely (very stiff piano wire).
> Does anyone with other belt tensioners actually have zero backlash?
> --
> You received this message because you are subscribed to the Google Groups
> "Ultimaker" group.
> There are also forums, which are the official gathering place for
> Ultimaker operators:
> http://forum.ultimaker.com/
> If you still want to post to this group, send email to
> ultimaker@googlegroups.com
> To unsubscribe from this group, send email to
> ultimaker+unsubscribe@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/ultimaker?hl=en
> --
> You received this message because you are subscribed to the Google Groups
> "Ultimaker" group.
> There are also forums, which are the official gathering place for
> Ultimaker operators:
> http://forum.ultimaker.com/
> If you still want to post to this group, send email to
> ultimaker@googlegroups.com
> To unsubscribe from this group, send email to
> ultimaker+unsubscribe@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/ultimaker?hl=en
Same for me - the original belts are holding up really well and I'm using the printer a lot. I organized a 3D printer meetup recently and while I don't know the details, the belts on all the 3 RepRap Mendels present seemed less precise? They had bigger "dips/gaps" in the belts and overall it seemed to me that they'd get less precision from that than what we have on the Ultimaker. They also has much longer belts, almost twice the length of what UM has.
Would love to see pictures of the Rostock build process Taylor?
On Tue, Sep 18, 2012 at 1:43 AM, Jensa <jens...@gmail.com> wrote:
> Same for me - the original belts are holding up really well and I'm using
> the printer a lot. I organized a 3D printer meetup recently and while I
> don't know the details, the belts on all the 3 RepRap Mendels present
> seemed less precise? They had bigger "dips/gaps" in the belts and overall
> it seemed to me that they'd get less precision from that than what we have
> on the Ultimaker. They also has much longer belts, almost twice the length
> of what UM has.
I mean, my belts are "holding up" fine, but I just still have crappy
prints, which is frustrating. I have slack somewhere and I wish it was
easier to eliminate.
> Would love to see pictures of the Rostock build process Taylor?
Awesome! Right now its just the printed parts, but I'm sure you guys will
be interested just as 3D printer nerds, so once the other mechanicals show
up I'll start an [off topic] thread about it for those who might be
interested. It sure is an interesting printer and I know you don't all have
your own communities full of people working on things like that! It seems
like I'll have most of the parts in two weeks or so, so expect something
more then. :)
> --
> You received this message because you are subscribed to the Google Groups
> "Ultimaker" group.
> There are also forums, which are the official gathering place for
> Ultimaker operators:
> http://forum.ultimaker.com/
> If you still want to post to this group, send email to
> ultimaker@googlegroups.com
> To unsubscribe from this group, send email to
> ultimaker+unsubscribe@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/ultimaker?hl=en
Yeah, I'm pretty sure. One belt in particular started tight then got
sloppy, so I added a tensioner. In about a month (during which time I only
did 3-4 prints) it got loose again. I've had to add a second tensioner to
it - even then it's not overly tight. I'm not convinced that's where 100%
of my backlash is coming from but it's certainly not helping.
My backlash is about 0.5 mm. Not huge but I'm trying to do cylindrical
models of engineered parts, so the printer hasn't been cutting it. I keep
having to send cylindrical parts out for RP. The amount I spent to print
parts last week because my UM couldn't print round circles could have
bought me another UM. I'm starting to feel a bit annoyed with the machine
considering the amount of time we've worked on it.
On Monday, September 17, 2012, Bill Culverhouse wrote:
> Ditto Dave's comment, I haven't tightened my long belts since I put them
> on a year ago.
> I get some vibration around hard corners on flat surfaces (the echo
> effect) But that's because
> I run it fast.
> Usually ovals instead if circles for me was the short belts getting loose.
> (Technically it was the X and Y
> motors getting loose.) Same for infill not touching the edges. The "lash"
> was in the short belt.
> The motor screws tend to loosen. I put some lock washers on them and
> haven't had a problem since.
> b
> On Mon, Sep 17, 2012 at 10:17 PM, ddurant <dduran...@gmail.com> wrote:
> Are you sure the belts are stretching? My original ones have been on there
> for over a year and still make a nice twang when I give them a flick..
> On Monday, September 17, 2012 10:11:42 PM UTC-4, Kevin R wrote:
> Ya, I'm wondering about this too. I just spent the whole weekend
> rebuilding my printer because I've got backlash in the x direction... So,
> ovals instead of cirlces...
> I keep tightening the belts but they keep stretching. I might give this a
> go because I'm getting pretty tired of the backlash.
> On Monday, September 17, 2012, Taylor Alexander wrote:
> They suggest GT2, which might be an interesting upgrade for the UM.
> I'm building a Rostock and I'm using GT2 belts on that, so we'll see how
> that goes. I've never had good layer to layer repeatability on my UM. If a
> layer switches the direction its coming from (say, there's a hole in part
> of the face), I always get a line due to backlash in the system - and I've
> got belt tensioners that I think are working nicely (very stiff piano wire).
> Does anyone with other belt tensioners actually have zero backlash?
> --
> You received this message because you are subscribed to the Google Groups
> "Ultimaker" group.
> There are also forums, which are the official gathering place for
> Ultimaker operators:
> http://forum.ultimaker.com/
> PhD Candidate
> Department of Chemical and Biological Engineering
> University of British Columbia
> P: 604.827.4123 C: 604.313.9149
> --
> You received this message because you are subscribed to the Google Groups
> "Ultimaker" group.
> There are also forums, which are the official gathering place for
> Ultimaker operators:
> http://forum.ultimaker.com/
> If you still want to post to this group, send email to
> ultimaker@googlegroups.com
> To unsubscribe from this group, send email to
> ultimaker+unsubscribe@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/ultimaker?hl=en
On Tue, Sep 18, 2012 at 4:43 AM, Jensa <jens...@gmail.com> wrote:
> Same for me - the original belts are holding up really well and I'm using
> the printer a lot. I organized a 3D printer meetup recently and while I
> don't know the details, the belts on all the 3 RepRap Mendels present
> seemed less precise? They had bigger "dips/gaps" in the belts and overall
> it seemed to me that they'd get less precision from that than what we have
> on the Ultimaker. They also has much longer belts, almost twice the length
> of what UM has.
> Would love to see pictures of the Rostock build process Taylor?
All 4 of my long belts (rev1 kit) have stretched. And I have to put more
tension on them.
Having said that, my prints are coming out nice and accurate at the moment.
Cheers,
-Joel
On Tue, Sep 18, 2012 at 11:16 PM, Bill Culverhouse <
bill.culverho...@gmail.com> wrote:
> I've got a Rostock project coming along. It's a bit further than this pic,
> I've borrowed the extruder drive of the UM
> to do some test prints with it.
> b
> On Tue, Sep 18, 2012 at 4:43 AM, Jensa <jens...@gmail.com> wrote:
>> Same for me - the original belts are holding up really well and I'm using
>> the printer a lot. I organized a 3D printer meetup recently and while I
>> don't know the details, the belts on all the 3 RepRap Mendels present
>> seemed less precise? They had bigger "dips/gaps" in the belts and overall
>> it seemed to me that they'd get less precision from that than what we have
>> on the Ultimaker. They also has much longer belts, almost twice the length
>> of what UM has.
>> Would love to see pictures of the Rostock build process Taylor?
>> J
>> --
> You received this message because you are subscribed to the Google Groups
> "Ultimaker" group.
> There are also forums, which are the official gathering place for
> Ultimaker operators:
> http://forum.ultimaker.com/
> If you still want to post to this group, send email to
> ultimaker@googlegroups.com
> To unsubscribe from this group, send email to
> ultimaker+unsubscribe@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/ultimaker?hl=en
I'll chime in here to say I see a LOT of belt stretching as well. To the point I've been considering cutting the belts and inserting a very stiff spring or something (stronger than the forces applied to the belt) because I can't tighten them any more with any of the belt tighteners on Thingiverse, and some are starting to get slack again. I had a nice twang when I first built the bot, but now the belt along the back panel just gives a dull thud when plucked.
On Monday, September 17, 2012 6:52:06 PM UTC-5, Taylor wrote:
> So, according to this reprap article, the MXL belts the Ultimaker uses are > not actually well-suited to constantly reversing applications, as they are > designed for continuous motion in one direction and don't have an > anti-backlash profile:
> They suggest GT2, which might be an interesting upgrade for the UM.
> I'm building a Rostock and I'm using GT2 belts on that, so we'll see how > that goes. I've never had good layer to layer repeatability on my UM. If a > layer switches the direction its coming from (say, there's a hole in part > of the face), I always get a line due to backlash in the system - and I've > got belt tensioners that I think are working nicely (very stiff piano wire).
> Does anyone with other belt tensioners actually have zero backlash?
I just did a calibration print of a 62 mm circle and using my calipers I measured it to be within +/-0.2mm along all axes. I've not done anything special other than tightening the 4 upper belts and 2 shorter belts every now and then. Not super-tight, but tight enough for them not to sway and have a nice "ring" to them.
My Ultimaker arrived in April. I wonder if they've changed the supplier of belts since then?
> On Monday, September 17, 2012 6:52:06 PM UTC-5, Taylor wrote:
>> So, according to this reprap article, the MXL belts the Ultimaker uses >> are not actually well-suited to constantly reversing applications, as they >> are designed for continuous motion in one direction and don't have an >> anti-backlash profile:
>> They suggest GT2, which might be an interesting upgrade for the UM.
>> I'm building a Rostock and I'm using GT2 belts on that, so we'll see how >> that goes. I've never had good layer to layer repeatability on my UM. If a >> layer switches the direction its coming from (say, there's a hole in part >> of the face), I always get a line due to backlash in the system - and I've >> got belt tensioners that I think are working nicely (very stiff piano wire).
>> Does anyone with other belt tensioners actually have zero backlash?
> On Monday, September 17, 2012 6:52:06 PM UTC-5, Taylor wrote:
>> So, according to this reprap article, the MXL belts the Ultimaker uses >> are not actually well-suited to constantly reversing applications, as they >> are designed for continuous motion in one direction and don't have an >> anti-backlash profile:
>> They suggest GT2, which might be an interesting upgrade for the UM.
>> I'm building a Rostock and I'm using GT2 belts on that, so we'll see how >> that goes. I've never had good layer to layer repeatability on my UM. If a >> layer switches the direction its coming from (say, there's a hole in part >> of the face), I always get a line due to backlash in the system - and I've >> got belt tensioners that I think are working nicely (very stiff piano wire).
>> Does anyone with other belt tensioners actually have zero backlash?
I work with cnc machines, and usally backlash is adjusted with parameters in the control, maybe this could be fixed in the UM firmware ?
if using a dial gauge and your X-axis have 0.3mm backlash (gcode moves the axis 40mm but the hardware moves 39,7mm) the firmware could handle this with driving the stepper the desired extra steps to compensate for backlash.
That might reduce backlash but I think its a pretty dynamic thing on the UM
so you're not going to eliminate it that way. Best to get the mechanical
backlash as low as possible.
I did see that on the Reprap wiki there is some debate as to the accuracy
of the article I originally linked to. So now I'm not sure. I didn't read
the debate part.
On Sep 20, 2012 1:08 AM, "catohagen" <catoha...@gmail.com> wrote:
> I work with cnc machines, and usally backlash is adjusted with parameters
> in the control, maybe this could be fixed in the UM firmware ?
> if using a dial gauge and your X-axis have 0.3mm backlash (gcode moves the
> axis 40mm but the hardware moves 39,7mm) the firmware could handle this
> with driving the stepper the desired extra steps to compensate for
> backlash.
> --
> You received this message because you are subscribed to the Google Groups
> "Ultimaker" group.
> There are also forums, which are the official gathering place for
> Ultimaker operators:
> http://forum.ultimaker.com/
> If you still want to post to this group, send email to
> ultimaker@googlegroups.com
> To unsubscribe from this group, send email to
> ultimaker+unsubscribe@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/ultimaker?hl=en
While you could compensate for backlash, it would not remove the play from the system (which is also present if you have backlash) and the play causes problems no software fix can solve.
On Thursday, September 20, 2012 10:08:11 AM UTC+2, catohagen wrote:
> I work with cnc machines, and usally backlash is adjusted with parameters > in the control, maybe this could be fixed in the UM firmware ?
> if using a dial gauge and your X-axis have 0.3mm backlash (gcode moves the > axis 40mm but the hardware moves 39,7mm) the firmware could handle this > with driving the stepper the desired extra steps to compensate for > backlash.
Although it is easy to add or subtract steps on a static system, but once
you have advanced motion control software it adds lots of problems for the
trajectory planning software when it is moving quickly.
The steps cannot be instantaneously added so there are issues of
acceleration and deacceleration for adding those steps and that interferes
with the planned movement of the other axis. There have been long
discussions in the CNC groups about the problems with backlash compensation
and fast moving machines.
On top of that, I think most of the backlash is dynamic and varying with
print speed. Even if the X-Y moves correctly, the head might wobble because
of acceleration forces. So in summary, fix the hardware.
Bertho
From: catohagen Sent: Thursday, September 20, 2012 04:08
I work with cnc machines, and usally backlash is adjusted with parameters in
the control, maybe this could be fixed in the UM firmware ?
if using a dial gauge and your X-axis have 0.3mm backlash (gcode moves the
axis 40mm but the hardware moves 39,7mm) the firmware could handle this with
driving the stepper the desired extra steps to compensate for backlash.
@ Dave and Bill: So just a quick update on my previous post, after
repositioning the bearings and adding a second tensioner to the long x
belts, my backlash in the x went from over 1mm to less than 0.2 mm. All the
test objects I printed were less than 0.1 mm from spec. I discovered one of
my bearing on the x rods wasn't 100% perpendicular to the rod and was
rubbing a bit on the bearing cover, which was adding some resistance in the
x axis. Adding the second belt tensioner help TONS. This one belt IS
stretching. When I first assembled it was tight.'after adding the first
tensioner it was tight. Now I need two? Crazy...
Back on topic: I think we should be tackling backlash from both angles.
With my CNC mill I spent tons of time getting the backlash down by tuning
the mechanics of the machine (I think I got it to 0.15 mm or so), then
honed it in using software and got zero backlash (I was machining micro phi
if channels and needed the precision). I understand the issues with trying
to do this with fast moving machines but software correction did work for
me when I was running it at 100 - 150 mm/s (cutting thin carbonate and
kapton sheets).
How hard would it be to have an experimental version of Marlin where some
of us can experiment with this?
On Thursday, September 20, 2012, Boman33 <Boma...@vinland.com> wrote:
> Although it is easy to add or subtract steps on a static system, but once
you have advanced motion control software it adds lots of problems for the
trajectory planning software when it is moving quickly.
> The steps cannot be instantaneously added so there are issues of
acceleration and deacceleration for adding those steps and that interferes
with the planned movement of the other axis. There have been long
discussions in the CNC groups about the problems with backlash compensation
and fast moving machines.
> On top of that, I think most of the backlash is dynamic and varying with
print speed. Even if the X-Y moves correctly, the head might wobble
because of acceleration forces. So in summary, fix the hardware.
> Bertho
> From: catohagen Sent: Thursday, September 20, 2012 04:08
> I work with cnc machines, and usally backlash is adjusted with parameters
in the control, maybe this could be fixed in the UM firmware ?
> if using a dial gauge and your X-axis have 0.3mm backlash (gcode moves
the axis 40mm but the hardware moves 39,7mm) the firmware could handle this
with driving the stepper the desired extra steps to compensate for backlash.
> --
> --
> You received this message because you are subscribed to the Google Groups
"Ultimaker" group.
> There are also forums, which are the official gathering place for
In the 298MXL they only come in fiberglass Neoprene Part number 298MXL025G (which I got.... excellent quality... no stretching and no belt tentioners)
and if you have crazy slack pick up the 297MXL Part number 297MXL025G
also fiberglass only
If the 300MXL (which the Ultimaker comes with) was working fine then you have a couple more options:
Part number 300MXL025G fiberglass
Part number 300MXL025UP urethane polyester Part number 300MXL025UK Urethane Kevlar
On Tuesday, September 18, 2012 2:41:17 PM UTC-5, Jennigma wrote:
> Do you have exact part numbers?
> On Tuesday, September 18, 2012 11:12:42 AM UTC-7, Chuck Ernst wrote:
>> I eventually bought new belts... slightly shorter 298 and 297 MXL and >> eliminated the need for the tentioners... You can buy better quality belts.
>> they come in the following types...
>> - Neoprene
>> - Polyurethane
>> - Fiberglass, Kevlar® or steel reinforcement available
>> In the states I bought them here and had zero problems.
>> On Monday, September 17, 2012 6:52:06 PM UTC-5, Taylor wrote:
>>> So, according to this reprap article, the MXL belts the Ultimaker uses >>> are not actually well-suited to constantly reversing applications, as they >>> are designed for continuous motion in one direction and don't have an >>> anti-backlash profile:
>>> They suggest GT2, which might be an interesting upgrade for the UM.
>>> I'm building a Rostock and I'm using GT2 belts on that, so we'll see how >>> that goes. I've never had good layer to layer repeatability on my UM. If a >>> layer switches the direction its coming from (say, there's a hole in part >>> of the face), I always get a line due to backlash in the system - and I've >>> got belt tensioners that I think are working nicely (very stiff piano wire).
>>> Does anyone with other belt tensioners actually have zero backlash?
On Thursday, September 20, 2012 11:52:05 AM UTC-5, Chuck Ernst wrote:
> sorry... just saw the question
> In the 298MXL they only come in fiberglass Neoprene Part number 298MXL025G > (which I got.... excellent quality... no stretching and no belt tentioners)
> and if you have crazy slack pick up the 297MXL Part number 297MXL025G
> also fiberglass only
> If the 300MXL (which the Ultimaker comes with) was working fine then you > have a couple more options:
> Part number 300MXL025G fiberglass
> Part number 300MXL025UP urethane polyester > Part number 300MXL025UK Urethane Kevlar
> There is a minimum $35 order so I picked up a bunch to try the different > types.
> Chuck
> On Tuesday, September 18, 2012 2:41:17 PM UTC-5, Jennigma wrote:
>> Do you have exact part numbers?
>> On Tuesday, September 18, 2012 11:12:42 AM UTC-7, Chuck Ernst wrote:
>>> I eventually bought new belts... slightly shorter 298 and 297 MXL and >>> eliminated the need for the tentioners... You can buy better quality belts.
>>> they come in the following types...
>>> - Neoprene
>>> - Polyurethane
>>> - Fiberglass, Kevlar® or steel reinforcement available
>>> In the states I bought them here and had zero problems.
>>> On Monday, September 17, 2012 6:52:06 PM UTC-5, Taylor wrote:
>>>> So, according to this reprap article, the MXL belts the Ultimaker uses >>>> are not actually well-suited to constantly reversing applications, as they >>>> are designed for continuous motion in one direction and don't have an >>>> anti-backlash profile:
>>>> They suggest GT2, which might be an interesting upgrade for the UM.
>>>> I'm building a Rostock and I'm using GT2 belts on that, so we'll see >>>> how that goes. I've never had good layer to layer repeatability on my UM. >>>> If a layer switches the direction its coming from (say, there's a hole in >>>> part of the face), I always get a line due to backlash in the system - and >>>> I've got belt tensioners that I think are working nicely (very stiff piano >>>> wire).
>>>> Does anyone with other belt tensioners actually have zero backlash?
Of the folks fighting backlash, what speed are you typically printing at and have you lubed the various bearings?
I've been printing at 60mm/s (with non-printing motion at 200mm/s) and my backlash seems pretty minimal (there is some). I find that if I crank the speed too much, one of my axis rods will shift by a couple of millimeters (as it were slightly too short).
Also, I cleaned all the rods and have been using sewing machine oil. It seems to work quite well, though I need to apply it more often. Not saying it is any better than the grease, just that I ran out of grease and it was the only thing I could find.
