Filament Flow Rate Project

Hey there,

I am working on a tool for measuring the flow rate ( [en.wikipedia.org] ) for selfmade filament. I've got most of the mechanical part figured out.
Since I've got no idea about microcontrollers or anything related to the electronic parts I need, I hope I get some got help here.

I want to measure (and record) the amount (filament) in gram which flows through my nozzle in a specific time with a specific temperatur.

So I am asking myself how to control and record hotend, time and weight with a microcontroller. Is this with a cheap microcontroller (RAMP?) possible ?
Can I use a normale scale and record the signal-output? ( USB-Output is to expensive for my profile)

Greetings

[1]

Haven't yet done any work with microcontrollers either but I imagine for the first prototype at least it's best to leave the main RAMPs do its normal job and have a separate microcontroller do the monitoring. If you need hot end temps - maybe you could attach a second thermocouple.

[2]

I've been looking with great interest at this setup:

[airtripper.com]
[hackaday.io]

The guy managed to record the force it takes the extruder to push filament through the bowden tube.
He's been able to record the history and plot it on a chart - looks pretty useful for tuning the machine.
It's pretty similar to what you want to do right?
Maybe you'll find his blog and that other page useful.

[3]

BTW would you like to share the details of the mechanical setup you're planning? I've seen a rotating wheel measuring the length of filament which has passed. But how would you actually measure the weight? The diameter of filament can fluctuate a little right? I think somebody said somewhere that if we had all this data - weight of material passed, force needed - the machine could become a lot more plug-and-play e.g. it could self tune a lot of printing parameters.

An auto calibrating extruder on the printer itself, with feedback to keep it tuned would be a huge step forwards. It would remove one of the more frustrating elements of calibration and enable truly consumer 3d printers.

I do not know how you would measure the volume of filament passing through a nozzle. But if you can you can feedback to the printer controller a value for e-steps/mm to increase or decrease filament flow.

However the reasons extruders drift off calibration are not easily compensated for eg hotbed bolt/drive gear getting clogged, filament stripping, blockages due to heat creep. Perhaps when there is a perfect extruder, and a perfect hot end having filament flow feedback would complete the system.

Edited 1 time(s). Last edit at 04/07/2016 09:40AM by DjDemonD.

Why not put white and black sections on the idler in extruder.
Reflective optical sensor aimed at it could sense segments passing by.

Attach it to a pin on an arduino and have it count segments
calculate length of filament traveling per segment
convert to tons --- you have weight of filament used

You could compare this to commanded extrusion
and
detect feed problem

ah -- but you would have to code for filament retraction???

Someone is already a good way along doing this [forums.reprap.org]

But I am not sure this is the complete solution, knowing how much the idler wheel moves, then using encoder pulses from it would take us a long way towards an extruder with feedback but its not the same as actually measuring the filament extruded. There must be a way to do this (ultrasonics? pressure sensors?).

To make a self calibrating extruder with feedback, we need to be able to measure (automatically) the filament diameter, if the filament has run out and the amount of filament extruded.

There is a discussion about measuring filament extruded here [forums.reprap.org]

A novel method of measuring filament extruded is needed. This is especially important relating to a conversation going on here [forums.reprap.org] about making a truly consumer 3dprinter, so no calibration required by the end user at all, and having the ability to use whatever brand of filament you want. I can see a model where a consumer printer uses the manufacturers filament which is tightly controlled and has extruder settings encoded into the spool for the printer to read when it is installed but this is not really where we want things to go, is it?

Edited 1 time(s). Last edit at 04/10/2016 03:30PM by DjDemonD.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Well if filament input is NOT a good measurement, what is?

Where does the filament go, if not out the nozzle,
does it evaporate, changed to energy (E=mc2) wow some real heating there.

There is NO area between nozzle exit and printed parts to measure the output!
and it is at 200 deg C.

The way you calibrate an extruder is
command 100mm extrude and measure how much filament goes in
then adjust the extruder stepper steps/mm

Your asking for the $100,000 mechanism to measure hot polymer on a $1000 printer

Do the color inkjet printers measure the ink spurting out of each color nozzle
to print those photo quality prints in outputs????

No what I am trying to do is get people thinking about if there is a way to measure the filament flow. Your opinion is that there is not and it is not necessary, fair enough.

And no inkjet printers don't need to measure the amount of ink they spray, as they rely on having cartridges supplied by the manufacturer then they work well, or cheapy copied cartridges, then they often work badly.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

So
FDM printers, with manufacturer provided cartridges, don't need to
measure the amount of polymer they extrude,
as they rely on having cartridges supplied by the manufacturer then they work well,

A inkjet printer sells for $70 and the cartridge costs $80
so
are you willing for:

A FDM 3D printer costs $1500 and the cartridge costs $1600

I don't think a $200,000 straysys 3D printer measures the filament output

Wow despite suggesting this idea yourself, about 4 comments above you're now opposed to being able to achieve filament flow feedback. The idler wheel method is being used already by several different groups. I am interested to brainstorm if there is a novel way to detect filament flow or more correctly volume extruded. If you don't know then say you don't know.

As for the cartridges, no I don't want to see 3d printing going down this route. But if you want to make a truly easy to use consumer printer then cartridges are one way to solve the problem that filament varies in its properties - control the filament. Another way is to make a much more sophisticated extruder.

I am trying to encourage a discussion about how to make a consumer 3d printer that is easy to use. Not technical. No calibration required and the ability to use a wide range of filaments. Some type of auto calibrating extruder which can detect the volume of filament extruded and control it via feedback is theoretically one way to do this.

The stratasys machine is a professional/enterprise level piece of equipment which requires training and expertise to operate correctly, as do your and my 3d printers. But if I want my grandad to use a 3d printer he does not want to spend hundreds of hours learning the intricacies of the machines.

Edited 1 time(s). Last edit at 04/11/2016 02:23PM by DjDemonD.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Quote
cozmicray
I don't think a $200,000 straysys 3D printer measures the filament output

Nope, they make sure it has a very accurate diameter and charge $250 a roll.
I have a Uprint-SE Plus at work and although its very easy to use, it has its limitations in the software.

Quote
cozmicray
A inkjet printer sells for $70 and the cartridge costs $80
so
are you willing for:

A FDM 3D printer costs $1500 and the cartridge costs $1600

That's not the business model. Every inkjet is sold at a loss so that profit can be made selling expensive ink.

So, an FDM printer would cost $300 and the cartridge $100. Wait, I think I just described the XYZ Da Vinci.

DaVinci cartridge is a box around a spool of filament.

If the DaVinci cartridge had a extruder and hot end in it,
it would cost $600

Quote
frankvdh


That's not the business model. Every inkjet is sold at a loss so that profit can be made selling expensive ink.

So, an FDM printer would cost $300 and the cartridge $100. Wait, I think I just described the XYZ Da Vinci.

I'm sure it would but it doesn't need one, it just has to be manufactured well and tested to work in the printer it's intended for.

But this is the business model we ought to avoid. We need a printer that can use a wide range of filaments automatically.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions