I do not have any idea if this is possible or if any one have thought /
tested. And im quite certain that it wouldn't bee cheap.
But! my idea: putting a sensor that (touchless)measures the cut area
of fillament then virtualy conect it to a pulse number ( known control
principle) and by this correct calculate feedrate by real available
feed volume in heatzone.
this would eliminate problems with unconsistant fillament.
Wich type of sensor? Ultra sonic? image treatment? or radiation type?
(radiation is used sometimes in farmers treshers for yeld metering, it
calculate mass flow by notice less received radiation from source)
How does this sound?
(I'm from Sweeden and not shure i explained my idea too be clear)
Some ideas in this area have been thrown around, but nothing has really been done about them. The most concrete idea was to mark the surface of an idler bearing OR attach a (printable) encoder ring so that its motion could be read by an optical encoder. This would provide feedback to tell you if your filament is slipping (or gone), and would be fairly straightforward to implement, but of questionable utility as the extruder should be designed to prevent this from happening at all.
I think someone has also thrown out the of possibility (and if they haven't, I have now) of using some kind of force sensor/load cell or something on the idler block's tensioner to measure variations in filament diameter, though that wouldn't help with non-circular filament, and making sure that the feedstock is consistently round and as a consistent diameter will always be a better solution anyway.
On Sunday, April 8, 2012 at 2:09 PM, ESN wrote: > I do not have any idea if this is possible or if any one have thought / > tested. And im quite certain that it wouldn't bee cheap. > But! my idea: putting a sensor that (touchless)measures the cut area > of fillament then virtualy conect it to a pulse number ( known control > principle) and by this correct calculate feedrate by real available > feed volume in heatzone.
> this would eliminate problems with unconsistant fillament.
> Wich type of sensor? Ultra sonic? image treatment? or radiation type? > (radiation is used sometimes in farmers treshers for yeld metering, it > calculate mass flow by notice less received radiation from source)
> How does this sound? > (I'm from Sweeden and not shure i explained my idea too be clear)
> You received this message because you are subscribed to the Google > Groups "thingiverse" group. > To post to this group, send email to thingiverse@googlegroups.com (mailto:thingiverse@googlegroups.com) > To unsubscribe from this group, send email to > thingiverse+unsubscribe@googlegroups.com (mailto:thingiverse+unsubscribe@googlegroups.com) > For more options, visit this group at > http://groups.google.com/group/thingiverse?hl=en
I agree. consistent fillament is best but a acurate mass flow metering
would reduce costs for fillament and thou make ability to extend print
quality. (and in some ways maybe make concistent fillament more
expensive?)
On 8 Apr, 23:18, Whosawhatsis <whosawhat...@gmail.com> wrote:
> Some ideas in this area have been thrown around, but nothing has really been done about them. The most concrete idea was to mark the surface of an idler bearing OR attach a (printable) encoder ring so that its motion could be read by an optical encoder. This would provide feedback to tell you if your filament is slipping (or gone), and would be fairly straightforward to implement, but of questionable utility as the extruder should be designed to prevent this from happening at all.
> I think someone has also thrown out the of possibility (and if they haven't, I have now) of using some kind of force sensor/load cell or something on the idler block's tensioner to measure variations in filament diameter, though that wouldn't help with non-circular filament, and making sure that the feedstock is consistently round and as a consistent diameter will always be a better solution anyway.
> On Sunday, April 8, 2012 at 2:09 PM, ESN wrote:
> > I do not have any idea if this is possible or if any one have thought /
> > tested. And im quite certain that it wouldn't bee cheap.
> > But! my idea: putting a sensor that (touchless)measures the cut area
> > of fillament then virtualy conect it to a pulse number ( known control
> > principle) and by this correct calculate feedrate by real available
> > feed volume in heatzone.
> > this would eliminate problems with unconsistant fillament.
> > Wich type of sensor? Ultra sonic? image treatment? or radiation type?
> > (radiation is used sometimes in farmers treshers for yeld metering, it
> > calculate mass flow by notice less received radiation from source)
> > How does this sound?
> > (I'm from Sweeden and not shure i explained my idea too be clear)
> > You received this message because you are subscribed to the Google
> > Groups "thingiverse" group.
> > To post to this group, send email to thingiverse@googlegroups.com (mailto:thingiverse@googlegroups.com)
> > To unsubscribe from this group, send email to
> > thingiverse+unsubscribe@googlegroups.com (mailto:thingiverse+unsubscribe@googlegroups.com)
> > For more options, visit this group at
> >http://groups.google.com/group/thingiverse?hl=en
Is the variability of the filament really an issue?
Are you having problems with the filament dimensions as you receive
the filament, or the consistency of the filament drive mechanism?
The diameter variability of 2.9 +- 0.1 is bad enough as it represents
a potential 13% variation from ideal (2.9).
But does a reel or spool of filament vary like that. I thought the
variation was primarily between reels, rather than within one reel.
I don't use 1.75, but if it is +- 0.1mm then it is a bigger
variation. My understanding was that the +-0.1mm was a manufacturing
measurement tolerance, in that it is not easy to measure below that
size in a repeatable, automated way.
If you wanted to measure the filament, you would need to measure its
cross-section to get its instantaneous area. The manufacturers don't
do that because it is too hard. I don't think there is an easy way to
do it.
In either case it would make extrusion more complex, as you would need
to adjust your extrusion distances, according to the current filament
size.
HI Since i got no printer my self yet ( i have a rep strap in slow
process), I refer to discussions here on thingiverse about consitency
problems with fillament.
For me its obvious that it is the cross section and not only the
diameter at one or two angles that need to be measured,. That is why i
sugested radiation loss type of measuring. And yes this generates alot
of parameters to ad to the print process. I am not able to do this my
self. I just threw this sugsetion out so some one capable and
interested in this can take it further.
On 19 Apr, 06:36, SteveMac <steveab...@gmail.com> wrote:
> Is the variability of the filament really an issue?
> Are you having problems with the filament dimensions as you receive
> the filament, or the consistency of the filament drive mechanism?
> The diameter variability of 2.9 +- 0.1 is bad enough as it represents
> a potential 13% variation from ideal (2.9).
> But does a reel or spool of filament vary like that. I thought the
> variation was primarily between reels, rather than within one reel.
> I don't use 1.75, but if it is +- 0.1mm then it is a bigger
> variation. My understanding was that the +-0.1mm was a manufacturing
> measurement tolerance, in that it is not easy to measure below that
> size in a repeatable, automated way.
> If you wanted to measure the filament, you would need to measure its
> cross-section to get its instantaneous area. The manufacturers don't
> do that because it is too hard. I don't think there is an easy way to
> do it.
> In either case it would make extrusion more complex, as you would need
> to adjust your extrusion distances, according to the current filament
> size.
