How would you measure filament flow?

[Ryan Carlyle]

So, we all know extruders lose some flow efficiency at high back-pressures, right? Higher pushing force = less length of filament pushed per motor rev. What I've seen is the tooth bite marks get closer together as load increases. But the main way I've seen to accurately measure the flow efficiency is to extrude X meters at Y speed and then measure the extrudate with a good digital scale. Which is pretty tedious.

Is an encoder on a filament idler the only other good way to measure this? Repetier and RRF can both measure filament motion but is the measurement outputted somewhere accessible? Is there any firmware with off the shelf parts that will spit out the instantaneous filament velocity to the user? I'm thinking like a measuring wheel or speedometer or something. 

[Ryan Carlyle]

Y'all think a speedometer could be rigged up on filament moving a few mm/sec? 

[David Crocker]

RRF includes a filament monitor calibration mechanism, to calibrate the change in position reported by the filament monitor against the amount of extrusion commanded. This calibration constant is reported by the M591 command. You can run this procedure at a few different extrusion speeds in order to measure the reduced extrusion with increasing back pressure. After doing a few sums, this can be fed into the M592 command to configure appropriate compensation.

Maybe we should look at adding a firmware feature or writing a script to automate this.

David Crocker, Duet3D Limited

www.duet3d.com
duet3d.dozuki.com
forum.duet3d.com

On 29/11/2019 19:15, Ryan Carlyle wrote:

So, we all know extruders lose some flow efficiency at high back-pressures, right? Higher pushing force = less length of filament pushed per motor rev. What I've seen is the tooth bite marks get closer together as load increases. But the main way I've seen to accurately measure the flow efficiency is to extrude X meters at Y speed and then measure the extrudate with a good digital scale. Which is pretty tedious.

Is an encoder on a filament idler the only other good way to measure this? Repetier and RRF can both measure filament motion but is the measurement outputted somewhere accessible? Is there any firmware with off the shelf parts that will spit out the instantaneous filament velocity to the user? I'm thinking like a measuring wheel or speedometer or something. 

--
You received this message because you are subscribed to the Google Groups "3DP Ideas" group.
To unsubscribe from this group and stop receiving emails from it, send an email to 3dp-ideas+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/3dp-ideas/75b9d250-073b-425d-ac22-66d935d72a05%40googlegroups.com.

[tray]

An optical mouse senor downstream of the drive wheel could read motion, but that's a mighty inconvenient spot. Upstream, the filament might be too smooth to activate the sensor unless you pretextured it with a sand paper roller, or use filament with a natural texture to it.  Sounds too fiddly.

[Ryan Carlyle]

David, what is RRF doing with the monitor info right now?

My interest is extruder grip performance characterization, but there’s probably some useful auto-calibration that could be done around printing temps or max printing speeds.

[Wing Wong]

Food for thought: micro camera sensors are fairly inexpensive. If you get two at 90' to one another measuring the same section of filament, you would be able to track movement of the filament as well as the thickness from two different views.

From that, you would be able to measure the rate of movement, limited by the pixel resolution, and be able to measure the thickness variation for that segment. Tracking that you would be able to calculate a continual volume measurement, no?

W.

Wing Wong
winge...@gmail.com
http://about.me/wingtangwong

https://www.facebook.com/wingedpower

On Fri, Nov 29, 2019 at 8:46 PM Ryan Carlyle <temp...@gmail.com> wrote:

David, what is RRF doing with the monitor info right now?

My interest is extruder grip performance characterization, but there’s probably some useful auto-calibration that could be done around printing temps or max printing speeds.

--
You received this message because you are subscribed to the Google Groups "3DP Ideas" group.
To unsubscribe from this group and stop receiving emails from it, send an email to 3dp-ideas+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/3dp-ideas/e7eecafe-4dc6-4d9f-86cd-c455b45a825f%40googlegroups.com.

[David Crocker]

It compares measured vs. commanded performance. If the correlation is outside a configured window, it pauses the print with a message.

On Sat, 30 Nov 2019, 04:46 Ryan Carlyle, <temp...@gmail.com> wrote:

David, what is RRF doing with the monitor info right now?

My interest is extruder grip performance characterization, but there’s probably some useful auto-calibration that could be done around printing temps or max printing speeds.

--
You received this message because you are subscribed to the Google Groups "3DP Ideas" group.
To unsubscribe from this group and stop receiving emails from it, send an email to 3dp-ideas+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/3dp-ideas/e7eecafe-4dc6-4d9f-86cd-c455b45a825f%40googlegroups.com.

[Ryan Carlyle]

David, if the Mcode commands let you freely check and report the calibration factor from a terminal, then that should be close enough for what I’m wanting to do. Accurate within the range of filament diameter variation at least. I’ll order one.

Whatever happened with the RRF optical filament sensor? That worked like an optical mouse, right?

[David Crocker]

We still sell the laser filament monitor. It is mechanically much simpler than the rotating magnet one. Unfortunately it doesn't work with some types of filament.

On Sat, 30 Nov 2019, 14:13 Ryan Carlyle, <temp...@gmail.com> wrote:

David, if the Mcode commands let you freely check and report the calibration factor from a terminal, then that should be close enough for what I’m wanting to do. Accurate within the range of filament diameter variation at least. I’ll order one.

Whatever happened with the RRF optical filament sensor? That worked like an optical mouse, right?

--
You received this message because you are subscribed to the Google Groups "3DP Ideas" group.
To unsubscribe from this group and stop receiving emails from it, send an email to 3dp-ideas+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/3dp-ideas/d1234f9b-3432-4214-a684-e74df72cf32e%40googlegroups.com.