Bondtech BGM steps?

Chriss

Hi *,

I use a Bondtech BMG extruder upgrade kit fpr Prusa i3 Mosquito on my CoreXY. That was a shortcut for having a fully assembled tool without designing a lot of parts by myself.

I have problems with the setup since the beginning so I want to ask the crowd for guidance now.

I calibrated the steps for "E" over and over again, with up to 300mm to get as accurate values as possible. I came to "M92 E425.80" and I had over extrusion (with the same filament) so I reduced the extrusion factor to 95% in the DWC and that look better now, the print is technically ok. But the print still looks over extruding.

What do you guys use for the E steps with a bondtech?

Cheers, Chriss

DIY-O-Sphere

Bondtech recommends 415.
My value is 412. (Clone)

Veti

i tend to reduce my e steps by 10% over the measured value.
pla does not like to be over extruded, but can take under extrusion a lot better.

A Former User

Some ideas:

Do you use pressure-advance? And if yes what exactly?
Mosquito´s have very little (de-)retract about 1-2mm is enough, what is your value here?

-> 415(rec. e-stps)/426(your e-stps) is about 0,97 so your 95% is close -> If you have calibrated your retract and pressure-advance I would guess you might land close to the 415 or less e-steps that bondtech proposes (your 426*0.95=ca. 405)?

(over extrusion might also result in the tension-screw (the one on the side with the spring) TOO tight screwd down -> double-gear extruders do not need the amount of tension between the gears like single-gear extruders -> dial it in just strong enough - but not more! We had to learn here the same when double-gear extruders became available to the masses)

Regards

deckingman

@Chriss said in Bondtech BGM steps?:

................. I calibrated the steps for "E" over and over again, with up to 300mm to get as accurate values as possible.

Did you repeat the calibration with the same piece of filament? If so, that can lead to inaccurate results. Also, 300mm of filament is not representative of what happens when you run a print (unless you have print moves of around 6 metres in length). For that matter, even 100mm is not representative of any actual print moves. Did you calibrate with hot end connected or not? Again, that can have an impact on the result that you see. As can the extruder speed if it's too high. If you've read all that, then by now you'll begin to understand that calibrating the extruder will only get you close and you need to fine tune the result by running some prints. As a general rule, most parts are surprisingly tolerant of under extrusion - even as mush as 15 - 20% under extrusion can produce quite acceptable results but the smallest amount of over extrusion will have a negative effect on print quality. For that reason, I've become accustomed to using 95% extrusion multiplier by default.
Finally, and for info, I have six (genuine) Bondtech BMGs on my machine. Here are the steps per mm that I have mine calibrated at.
M92 E409:407:405:405:403:405

A Former User

This post is deleted!

Chriss

@DIY-O-Sphere
Is your 412 measured or "learned" by prints?

@LB
I do not (yet) use pressure-advance.
I have 1mm retraction @45mm/s
The tension screw is lets say "tight" I had the problems that the extruder became a saw when I had many retractions with some prints.
You and deckingman convinced me to test with 405 now.

@deckingman
Yes, it is the same spool (calibration and printing)
Well, I used the 300mm to calibrate the E steps. My theory is that I will get more accurate e steps when I extrude more and measure over 300mm instead of 100. (I have no longer steel lineal than 300mm)
Funny to see how much difference you have with your six extruders. I would not have expected a range of 6 between the extruders.

Btw, here is a picture of my last print with 425:
Yes, over extrusion, that is clear, but why do I have this gab between the walls of the hole and the roof?

The other site:

zapta

@Chriss said in Bondtech BGM steps?:

so I reduced the extrusion factor to 95% in the DWC and that look better now, the print is technically ok. But the print still looks over extruding.

Printing at below 100% after calibrating the esteps is pretty much common. If you are looking for a good extrusion multiplier calibration you may want to take a look at https://github.com/supermerill/SuperSlicer . It's a variant of PrusaSlicer that also generates calibration prints.

For the extrusion calibration, it prints a few layers at 100% infill and then print one round line that is used to evaluate the print.

(You can probably set manually something similar in PrusaSlicer using the configuration modifiers).

DIY-O-Sphere

@Chriss said in Bondtech BGM steps?:

Is your 412 measured or "learned" by prints?

It's a measured value.