Filament diameter feedback

As I sit here cursing and swearing, filament that's got more curves than marilyn monroe...

I'm wondering - do any of the firmwares have support for a 'on the fly' measurement of the filament diameter - with capability to automagically adjust feed rate?

I was thinking along the lines of passing it through an IR beam just above the hot end and working back to a diameter, but there's probably some simpler options as well? I read in here it's been discussed more in terms of servo motors and encoders, Google's not find much other than some theoretical discussion...

Quote
hephaestus
I'm wondering - do any of the firmwares have support for a 'on the fly' measurement of the filament diameter - with capability to automagically adjust feed rate?

No (AFAIK) although this has been discussed more than once on the RepRap forums.

The general consensus (i.e. majority) are of the opinion that if you have filament that has an inconsistent diameter that it should be sent back to whoever you bought it from. As long as your filament manufacturer has proper tolerances then no such acrobatics are necessary.

Quote

Simpler is better.

Quote

Don't fix the blame, FIX THE PROBLEM!

Consider the following example:

Joe walks out of his house to pick up the paper, falls into a hole in his yard and injures his foot. He says the cause of the accident was that he was blinded by an oncoming auto. The next day, same thing, comes out of the house while it was raining and fell into the hole. He said it was wet and he slipped on the grass. On the next day same thing, this time it was the cat that distracted him.

His brother shakes his head and says: "Joe, you have a hole in your yard, FILL IN THE F**KING HOLE!".

Quote

Don't fix the blame, FIX THE PROBLEM!

---

Bob Morrison
Wörth am Rhein, Germany
"Luke, use the source!"
BLOG - PHOTOS - Thingiverse

I'd tend to agree, however for some of us who order cross border - it gets ugly... This rolls half done, and suddenly running into 15% variation along a foot - makes it hard to get tolerable prints. (And no, it's not a cheap roll, yet I've measured 2.78 to 3.09mm diameter - within a 300mm section)

We know there are manufacturing discrepancies, seems to happen with all suppliers, so is it better to make the machine adjust or to expect better QA from the manufacturers, I'd like to think both...

But not being a programmer - what would it take to make a feedback loop with a variable and adjustment for filament diameter?

You'd need some diameter measuring device and a firmware which adjusts E feedrate accordingly. Extra bonus if you take the distances between hotend entry, the feeder and the measuring device into account.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

A bigger problem is determining if your filament is even round. If your filament isn't perfectly round, just measuring diameter is not enough and it becomes a much harder problem to solve.

Are you sure your diameter discrepencies are not because your filament is slightly oval thus causing different measurements depending on the orientation of the calipers?

If we can't count on a consistent width, we would still be reliant on a consistent shape. It would be odd to find a supplier that is willing to do the QA for one of those and not the other.

Ideas for measuring the actual volume of a possibly oval filament are welcome. I am somewhat stumped.

John

Measurements at 90deg to each other and compare?

Hadn't pondered out of round, will have to go chase a hunk of that roll down and check.